repository_name
stringclasses
316 values
func_path_in_repository
stringlengths
6
223
func_name
stringlengths
1
134
language
stringclasses
1 value
func_code_string
stringlengths
57
65.5k
func_documentation_string
stringlengths
1
46.3k
split_name
stringclasses
1 value
func_code_url
stringlengths
91
315
called_functions
listlengths
1
156
enclosing_scope
stringlengths
2
1.48M
vilmibm/done
parsedatetime/parsedatetime.py
_parse_date_rfc822
python
def _parse_date_rfc822(dateString): '''Parse an RFC822, RFC1123, RFC2822, or asctime-style date''' data = dateString.split() if data[0][-1] in (',', '.') or data[0].lower() in rfc822._daynames: del data[0] if len(data) == 4: s = data[3] i = s.find('+') if i > 0: data[3:] = [s[:i], s[i+1:]] else: data.append('') dateString = " ".join(data) if len(data) < 5: dateString += ' 00:00:00 GMT' return rfc822.parsedate_tz(dateString)
Parse an RFC822, RFC1123, RFC2822, or asctime-style date
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L154-L169
null
#!/usr/bin/env python """ Parse human-readable date/time text. """ __license__ = """ Copyright (c) 2004-2008 Mike Taylor Copyright (c) 2006-2008 Darshana Chhajed All rights reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ _debug = False import re import time import datetime import rfc822 import parsedatetime_consts # Copied from feedparser.py # Universal Feedparser # Copyright (c) 2002-2006, Mark Pilgrim, All rights reserved. # Originally a def inside of _parse_date_w3dtf() def _extract_date(m): year = int(m.group('year')) if year < 100: year = 100 * int(time.gmtime()[0] / 100) + int(year) if year < 1000: return 0, 0, 0 julian = m.group('julian') if julian: julian = int(julian) month = julian / 30 + 1 day = julian % 30 + 1 jday = None while jday != julian: t = time.mktime((year, month, day, 0, 0, 0, 0, 0, 0)) jday = time.gmtime(t)[-2] diff = abs(jday - julian) if jday > julian: if diff < day: day = day - diff else: month = month - 1 day = 31 elif jday < julian: if day + diff < 28: day = day + diff else: month = month + 1 return year, month, day month = m.group('month') day = 1 if month is None: month = 1 else: month = int(month) day = m.group('day') if day: day = int(day) else: day = 1 return year, month, day # Copied from feedparser.py # Universal Feedparser # Copyright (c) 2002-2006, Mark Pilgrim, All rights reserved. # Originally a def inside of _parse_date_w3dtf() def _extract_time(m): if not m: return 0, 0, 0 hours = m.group('hours') if not hours: return 0, 0, 0 hours = int(hours) minutes = int(m.group('minutes')) seconds = m.group('seconds') if seconds: seconds = int(seconds) else: seconds = 0 return hours, minutes, seconds # Copied from feedparser.py # Universal Feedparser # Copyright (c) 2002-2006, Mark Pilgrim, All rights reserved. # Modified to return a tuple instead of mktime # # Original comment: # W3DTF-style date parsing adapted from PyXML xml.utils.iso8601, written by # Drake and licensed under the Python license. Removed all range checking # for month, day, hour, minute, and second, since mktime will normalize # these later def _parse_date_w3dtf(dateString): # the __extract_date and __extract_time methods were # copied-out so they could be used by my code --bear def __extract_tzd(m): '''Return the Time Zone Designator as an offset in seconds from UTC.''' if not m: return 0 tzd = m.group('tzd') if not tzd: return 0 if tzd == 'Z': return 0 hours = int(m.group('tzdhours')) minutes = m.group('tzdminutes') if minutes: minutes = int(minutes) else: minutes = 0 offset = (hours*60 + minutes) * 60 if tzd[0] == '+': return -offset return offset __date_re = ('(?P<year>\d\d\d\d)' '(?:(?P<dsep>-|)' '(?:(?P<julian>\d\d\d)' '|(?P<month>\d\d)(?:(?P=dsep)(?P<day>\d\d))?))?') __tzd_re = '(?P<tzd>[-+](?P<tzdhours>\d\d)(?::?(?P<tzdminutes>\d\d))|Z)' __tzd_rx = re.compile(__tzd_re) __time_re = ('(?P<hours>\d\d)(?P<tsep>:|)(?P<minutes>\d\d)' '(?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?' + __tzd_re) __datetime_re = '%s(?:T%s)?' % (__date_re, __time_re) __datetime_rx = re.compile(__datetime_re) m = __datetime_rx.match(dateString) if (m is None) or (m.group() != dateString): return return _extract_date(m) + _extract_time(m) + (0, 0, 0) # Copied from feedparser.py # Universal Feedparser # Copyright (c) 2002-2006, Mark Pilgrim, All rights reserved. # Modified to return a tuple instead of mktime # def _parse_date_rfc822(dateString): '''Parse an RFC822, RFC1123, RFC2822, or asctime-style date''' data = dateString.split() if data[0][-1] in (',', '.') or data[0].lower() in rfc822._daynames: del data[0] if len(data) == 4: s = data[3] i = s.find('+') if i > 0: data[3:] = [s[:i], s[i+1:]] else: data.append('') dateString = " ".join(data) if len(data) < 5: dateString += ' 00:00:00 GMT' return rfc822.parsedate_tz(dateString) # rfc822.py defines several time zones, but we define some extra ones. # 'ET' is equivalent to 'EST', etc. _additional_timezones = {'AT': -400, 'ET': -500, 'CT': -600, 'MT': -700, 'PT': -800} rfc822._timezones.update(_additional_timezones) class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar._buildTime
python
def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple()
Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L237-L309
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar.parseDate
python
def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime
Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L312-L386
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar.parseDateText
python
def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime
Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L389-L440
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar.evalRanges
python
def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0)
Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L443-L606
[ "def parse(self, datetimeString, sourceTime=None):\n \"\"\"\n Splits the given C{datetimeString} into tokens, finds the regex\n patterns that match and then calculates a C{struct_time} value from\n the chunks.\n\n If C{sourceTime} is given then the C{struct_time} value will be\n calculated from that value, otherwise from the current date/time.\n\n If the C{datetimeString} is parsed and date/time value found then\n the second item of the returned tuple will be a flag to let you know\n what kind of C{struct_time} value is being returned::\n\n 0 = not parsed at all\n 1 = parsed as a C{date}\n 2 = parsed as a C{time}\n 3 = parsed as a C{datetime}\n\n @type datetimeString: string\n @param datetimeString: date/time text to evaluate\n @type sourceTime: struct_time\n @param sourceTime: C{struct_time} value to use as the base\n\n @rtype: tuple\n @return: tuple of: modified C{sourceTime} and the result flag\n \"\"\"\n\n if sourceTime:\n if isinstance(sourceTime, datetime.datetime):\n if _debug:\n print 'coercing datetime to timetuple'\n sourceTime = sourceTime.timetuple()\n else:\n if not isinstance(sourceTime, time.struct_time) and \\\n not isinstance(sourceTime, tuple):\n raise Exception('sourceTime is not a struct_time')\n\n s = datetimeString.strip().lower()\n parseStr = ''\n totalTime = sourceTime\n\n if s == '' :\n if sourceTime is not None:\n return (sourceTime, self.dateFlag + self.timeFlag)\n else:\n return (time.localtime(), 0)\n\n self.timeFlag = 0\n self.dateFlag = 0\n\n while len(s) > 0:\n flag = False\n chunk1 = ''\n chunk2 = ''\n\n if _debug:\n print 'parse (top of loop): [%s][%s]' % (s, parseStr)\n\n if parseStr == '':\n # Modifier like next\\prev..\n m = self.ptc.CRE_MODIFIER.search(s)\n if m is not None:\n self.modifierFlag = True\n if (m.group('modifier') != s):\n # capture remaining string\n parseStr = m.group('modifier')\n chunk1 = s[:m.start('modifier')].strip()\n chunk2 = s[m.end('modifier'):].strip()\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Modifier like from\\after\\prior..\n m = self.ptc.CRE_MODIFIER2.search(s)\n if m is not None:\n self.modifier2Flag = True\n if (m.group('modifier') != s):\n # capture remaining string\n parseStr = m.group('modifier')\n chunk1 = s[:m.start('modifier')].strip()\n chunk2 = s[m.end('modifier'):].strip()\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n valid_date = False\n for match in self.ptc.CRE_DATE3.finditer(s):\n # to prevent \"HH:MM(:SS) time strings\" expressions from triggering\n # this regex, we checks if the month field exists in the searched \n # expression, if it doesn't exist, the date field is not valid\n if match.group('mthname'):\n m = self.ptc.CRE_DATE3.search(s, match.start())\n valid_date = True\n break\n\n # String date format\n if valid_date:\n self.dateStrFlag = True\n self.dateFlag = 1\n if (m.group('date') != s):\n # capture remaining string\n parseStr = m.group('date')\n chunk1 = s[:m.start('date')]\n chunk2 = s[m.end('date'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Standard date format\n m = self.ptc.CRE_DATE.search(s)\n if m is not None:\n self.dateStdFlag = True\n self.dateFlag = 1\n if (m.group('date') != s):\n # capture remaining string\n parseStr = m.group('date')\n chunk1 = s[:m.start('date')]\n chunk2 = s[m.end('date'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Natural language day strings\n m = self.ptc.CRE_DAY.search(s)\n if m is not None:\n self.dayStrFlag = True\n self.dateFlag = 1\n if (m.group('day') != s):\n # capture remaining string\n parseStr = m.group('day')\n chunk1 = s[:m.start('day')]\n chunk2 = s[m.end('day'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Quantity + Units\n m = self.ptc.CRE_UNITS.search(s)\n if m is not None:\n self.unitsFlag = True\n if (m.group('qty') != s):\n # capture remaining string\n parseStr = m.group('qty')\n chunk1 = s[:m.start('qty')].strip()\n chunk2 = s[m.end('qty'):].strip()\n\n if chunk1[-1:] == '-':\n parseStr = '-%s' % parseStr\n chunk1 = chunk1[:-1]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Quantity + Units\n m = self.ptc.CRE_QUNITS.search(s)\n if m is not None:\n self.qunitsFlag = True\n\n if (m.group('qty') != s):\n # capture remaining string\n parseStr = m.group('qty')\n chunk1 = s[:m.start('qty')].strip()\n chunk2 = s[m.end('qty'):].strip()\n\n if chunk1[-1:] == '-':\n parseStr = '-%s' % parseStr\n chunk1 = chunk1[:-1]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s \n\n if parseStr == '':\n # Weekday\n m = self.ptc.CRE_WEEKDAY.search(s)\n if m is not None:\n gv = m.group('weekday')\n if s not in self.ptc.dayOffsets:\n self.weekdyFlag = True\n self.dateFlag = 1\n if (gv != s):\n # capture remaining string\n parseStr = gv\n chunk1 = s[:m.start('weekday')]\n chunk2 = s[m.end('weekday'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Natural language time strings\n m = self.ptc.CRE_TIME.search(s)\n if m is not None:\n self.timeStrFlag = True\n self.timeFlag = 2\n if (m.group('time') != s):\n # capture remaining string\n parseStr = m.group('time')\n chunk1 = s[:m.start('time')]\n chunk2 = s[m.end('time'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # HH:MM(:SS) am/pm time strings\n m = self.ptc.CRE_TIMEHMS2.search(s)\n if m is not None:\n self.meridianFlag = True\n self.timeFlag = 2\n if m.group('minutes') is not None:\n if m.group('seconds') is not None:\n parseStr = '%s:%s:%s %s' % (m.group('hours'),\n m.group('minutes'),\n m.group('seconds'),\n m.group('meridian'))\n else:\n parseStr = '%s:%s %s' % (m.group('hours'),\n m.group('minutes'),\n m.group('meridian'))\n else:\n parseStr = '%s %s' % (m.group('hours'),\n m.group('meridian'))\n\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('meridian'):]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n\n if parseStr == '':\n # HH:MM(:SS) time strings\n m = self.ptc.CRE_TIMEHMS.search(s)\n if m is not None:\n self.timeStdFlag = True\n self.timeFlag = 2\n if m.group('seconds') is not None:\n parseStr = '%s:%s:%s' % (m.group('hours'),\n m.group('minutes'),\n m.group('seconds'))\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('seconds'):]\n else:\n parseStr = '%s:%s' % (m.group('hours'),\n m.group('minutes'))\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('minutes'):]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n\n # if string does not match any regex, empty string to\n # come out of the while loop\n if not flag:\n s = ''\n\n if _debug:\n print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2)\n print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \\\n (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag)\n print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \\\n (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag)\n\n # evaluate the matched string\n if parseStr != '':\n if self.modifierFlag == True:\n t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime)\n # t is the unparsed part of the chunks.\n # If it is not date/time, return current\n # totalTime as it is; else return the output\n # after parsing t.\n if (t != '') and (t != None):\n tempDateFlag = self.dateFlag\n tempTimeFlag = self.timeFlag\n (totalTime2, flag) = self.parse(t, totalTime)\n\n if flag == 0 and totalTime is not None:\n self.timeFlag = tempTimeFlag\n self.dateFlag = tempDateFlag\n\n return (totalTime, self.dateFlag + self.timeFlag)\n else:\n return (totalTime2, self.dateFlag + self.timeFlag)\n\n elif self.modifier2Flag == True:\n totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime)\n\n if invalidFlag == True:\n self.dateFlag = 0\n self.timeFlag = 0\n\n else:\n totalTime = self._evalString(parseStr, totalTime)\n parseStr = ''\n\n # String is not parsed at all\n if totalTime is None or totalTime == sourceTime:\n totalTime = time.localtime()\n self.dateFlag = 0\n self.timeFlag = 0\n\n return (totalTime, self.dateFlag + self.timeFlag)\n" ]
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar._CalculateDOWDelta
python
def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff
Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L609-L678
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar._evalModifier
python
def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime
Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L681-L870
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar._evalModifier2
python
def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2)
Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L872-L943
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar._evalString
python
def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime
Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L946-L1158
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar.parse
python
def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag)
Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L1161-L1476
[ "def _evalModifier(self, modifier, chunk1, chunk2, sourceTime):\n \"\"\"\n Evaluate the C{modifier} string and following text (passed in\n as C{chunk1} and C{chunk2}) and if they match any known modifiers\n calculate the delta and apply it to C{sourceTime}.\n\n @type modifier: string\n @param modifier: modifier text to apply to sourceTime\n @type chunk1: string\n @param chunk1: first text chunk that followed modifier (if any)\n @type chunk2: string\n @param chunk2: second text chunk that followed modifier (if any)\n @type sourceTime: struct_time\n @param sourceTime: C{struct_time} value to use as the base\n\n @rtype: tuple\n @return: tuple of: remaining text and the modified sourceTime\n \"\"\"\n offset = self.ptc.Modifiers[modifier]\n\n if sourceTime is not None:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime\n else:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime()\n\n # capture the units after the modifier and the remaining\n # string after the unit\n m = self.ptc.CRE_REMAINING.search(chunk2)\n if m is not None:\n index = m.start() + 1\n unit = chunk2[:m.start()]\n chunk2 = chunk2[index:]\n else:\n unit = chunk2\n chunk2 = ''\n\n flag = False\n\n if unit == 'month' or \\\n unit == 'mth' or \\\n unit == 'm':\n if offset == 0:\n dy = self.ptc.daysInMonth(mth, yr)\n sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst)\n elif offset == 2:\n # if day is the last day of the month, calculate the last day\n # of the next month\n if dy == self.ptc.daysInMonth(mth, yr):\n dy = self.ptc.daysInMonth(mth + 1, yr)\n\n start = datetime.datetime(yr, mth, dy, 9, 0, 0)\n target = self.inc(start, month=1)\n sourceTime = target.timetuple()\n else:\n start = datetime.datetime(yr, mth, 1, 9, 0, 0)\n target = self.inc(start, month=offset)\n sourceTime = target.timetuple()\n\n flag = True\n self.dateFlag = 1\n\n if unit == 'week' or \\\n unit == 'wk' or \\\n unit == 'w':\n if offset == 0:\n start = datetime.datetime(yr, mth, dy, 17, 0, 0)\n target = start + datetime.timedelta(days=(4 - wd))\n sourceTime = target.timetuple()\n elif offset == 2:\n start = datetime.datetime(yr, mth, dy, 9, 0, 0)\n target = start + datetime.timedelta(days=7)\n sourceTime = target.timetuple()\n else:\n return self._evalModifier(modifier, chunk1, \"monday \" + chunk2, sourceTime)\n\n flag = True\n self.dateFlag = 1\n\n if unit == 'day' or \\\n unit == 'dy' or \\\n unit == 'd':\n if offset == 0:\n sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst)\n self.timeFlag = 2\n elif offset == 2:\n start = datetime.datetime(yr, mth, dy, hr, mn, sec)\n target = start + datetime.timedelta(days=1)\n sourceTime = target.timetuple()\n else:\n start = datetime.datetime(yr, mth, dy, 9, 0, 0)\n target = start + datetime.timedelta(days=offset)\n sourceTime = target.timetuple()\n\n flag = True\n self.dateFlag = 1\n\n if unit == 'hour' or \\\n unit == 'hr':\n if offset == 0:\n sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst)\n else:\n start = datetime.datetime(yr, mth, dy, hr, 0, 0)\n target = start + datetime.timedelta(hours=offset)\n sourceTime = target.timetuple()\n\n flag = True\n self.timeFlag = 2\n\n if unit == 'year' or \\\n unit == 'yr' or \\\n unit == 'y':\n if offset == 0:\n sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst)\n elif offset == 2:\n sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst)\n else:\n sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst)\n\n flag = True\n self.dateFlag = 1\n\n if flag == False:\n m = self.ptc.CRE_WEEKDAY.match(unit)\n if m is not None:\n wkdy = m.group()\n self.dateFlag = 1\n\n if modifier == 'eod':\n # Calculate the upcoming weekday\n self.modifierFlag = False\n (sourceTime, _) = self.parse(wkdy, sourceTime)\n sources = self.ptc.buildSources(sourceTime)\n self.timeFlag = 2\n\n if modifier in sources:\n sourceTime = sources[modifier]\n\n else:\n wkdy = self.ptc.WeekdayOffsets[wkdy]\n diff = self._CalculateDOWDelta(wd, wkdy, offset,\n self.ptc.DOWParseStyle,\n self.ptc.CurrentDOWParseStyle)\n start = datetime.datetime(yr, mth, dy, 9, 0, 0)\n target = start + datetime.timedelta(days=diff)\n sourceTime = target.timetuple()\n\n flag = True\n self.dateFlag = 1\n\n if not flag:\n m = self.ptc.CRE_TIME.match(unit)\n if m is not None:\n self.modifierFlag = False\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit)\n\n start = datetime.datetime(yr, mth, dy, hr, mn, sec)\n target = start + datetime.timedelta(days=offset)\n sourceTime = target.timetuple()\n flag = True\n else:\n self.modifierFlag = False\n\n # check if the remaining text is parsable and if so,\n # use it as the base time for the modifier source time\n t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime)\n\n if flag2 != 0:\n sourceTime = t\n\n sources = self.ptc.buildSources(sourceTime)\n\n if modifier in sources:\n sourceTime = sources[modifier]\n flag = True\n self.timeFlag = 2\n\n # if the word after next is a number, the string is more than likely\n # to be \"next 4 hrs\" which we will have to combine the units with the\n # rest of the string\n if not flag:\n if offset < 0:\n # if offset is negative, the unit has to be made negative\n unit = '-%s' % unit\n\n chunk2 = '%s %s' % (unit, chunk2)\n\n self.modifierFlag = False\n\n #return '%s %s' % (chunk1, chunk2), sourceTime\n return '%s' % chunk2, sourceTime\n", "def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime):\n \"\"\"\n Evaluate the C{modifier} string and following text (passed in\n as C{chunk1} and C{chunk2}) and if they match any known modifiers\n calculate the delta and apply it to C{sourceTime}.\n\n @type modifier: string\n @param modifier: modifier text to apply to C{sourceTime}\n @type chunk1: string\n @param chunk1: first text chunk that followed modifier (if any)\n @type chunk2: string\n @param chunk2: second text chunk that followed modifier (if any)\n @type sourceTime: struct_time\n @param sourceTime: C{struct_time} value to use as the base\n\n @rtype: tuple\n @return: tuple of: remaining text and the modified sourceTime\n \"\"\"\n offset = self.ptc.Modifiers[modifier]\n digit = r'\\d+'\n\n self.modifier2Flag = False\n\n # If the string after the negative modifier starts with digits,\n # then it is likely that the string is similar to ' before 3 days'\n # or 'evening prior to 3 days'.\n # In this case, the total time is calculated by subtracting '3 days'\n # from the current date.\n # So, we have to identify the quantity and negate it before parsing\n # the string.\n # This is not required for strings not starting with digits since the\n # string is enough to calculate the sourceTime\n if chunk2 != '':\n if offset < 0:\n m = re.match(digit, chunk2.strip())\n if m is not None:\n qty = int(m.group()) * -1\n chunk2 = chunk2[m.end():]\n chunk2 = '%d%s' % (qty, chunk2)\n\n sourceTime, flag1 = self.parse(chunk2, sourceTime)\n if flag1 == 0:\n flag1 = True\n else:\n flag1 = False\n flag2 = False\n else:\n flag1 = False\n\n if chunk1 != '':\n if offset < 0:\n m = re.search(digit, chunk1.strip())\n if m is not None:\n qty = int(m.group()) * -1\n chunk1 = chunk1[m.end():]\n chunk1 = '%d%s' % (qty, chunk1)\n\n tempDateFlag = self.dateFlag\n tempTimeFlag = self.timeFlag\n sourceTime2, flag2 = self.parse(chunk1, sourceTime)\n else:\n return sourceTime, (flag1 and flag2)\n\n # if chunk1 is not a datetime and chunk2 is then do not use datetime\n # value returned by parsing chunk1\n if not (flag1 == False and flag2 == 0):\n sourceTime = sourceTime2\n else:\n self.timeFlag = tempTimeFlag\n self.dateFlag = tempDateFlag\n\n return sourceTime, (flag1 and flag2)\n", "def _evalString(self, datetimeString, sourceTime=None):\n \"\"\"\n Calculate the datetime based on flags set by the L{parse()} routine\n\n Examples handled::\n RFC822, W3CDTF formatted dates\n HH:MM[:SS][ am/pm]\n MM/DD/YYYY\n DD MMMM YYYY\n\n @type datetimeString: string\n @param datetimeString: text to try and parse as more \"traditional\"\n date/time text\n @type sourceTime: struct_time\n @param sourceTime: C{struct_time} value to use as the base\n\n @rtype: datetime\n @return: calculated C{struct_time} value or current C{struct_time}\n if not parsed\n \"\"\"\n s = datetimeString.strip()\n now = time.localtime()\n\n # Given string date is a RFC822 date\n if sourceTime is None:\n sourceTime = _parse_date_rfc822(s)\n\n if sourceTime is not None:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime\n self.dateFlag = 1\n\n if (hr != 0) and (mn != 0) and (sec != 0):\n self.timeFlag = 2\n\n sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst)\n\n # Given string date is a W3CDTF date\n if sourceTime is None:\n sourceTime = _parse_date_w3dtf(s)\n\n if sourceTime is not None:\n self.dateFlag = 1\n self.timeFlag = 2\n\n if sourceTime is None:\n s = s.lower()\n\n # Given string is in the format HH:MM(:SS)(am/pm)\n if self.meridianFlag:\n if sourceTime is None:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now\n else:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime\n\n m = self.ptc.CRE_TIMEHMS2.search(s)\n if m is not None:\n dt = s[:m.start('meridian')].strip()\n if len(dt) <= 2:\n hr = int(dt)\n mn = 0\n sec = 0\n else:\n hr, mn, sec = _extract_time(m)\n\n if hr == 24:\n hr = 0\n\n sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst)\n meridian = m.group('meridian').lower()\n\n # if 'am' found and hour is 12 - force hour to 0 (midnight)\n if (meridian in self.ptc.am) and hr == 12:\n sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst)\n\n # if 'pm' found and hour < 12, add 12 to shift to evening\n if (meridian in self.ptc.pm) and hr < 12:\n sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst)\n\n # invalid time\n if hr > 24 or mn > 59 or sec > 59:\n sourceTime = now\n self.dateFlag = 0\n self.timeFlag = 0\n\n self.meridianFlag = False\n\n # Given string is in the format HH:MM(:SS)\n if self.timeStdFlag:\n if sourceTime is None:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now\n else:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime\n\n m = self.ptc.CRE_TIMEHMS.search(s)\n if m is not None:\n hr, mn, sec = _extract_time(m)\n if hr == 24:\n hr = 0\n\n if hr > 24 or mn > 59 or sec > 59:\n # invalid time\n sourceTime = now\n self.dateFlag = 0\n self.timeFlag = 0\n else:\n sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst)\n\n self.timeStdFlag = False\n\n # Given string is in the format 07/21/2006\n if self.dateStdFlag:\n sourceTime = self.parseDate(s)\n self.dateStdFlag = False\n\n # Given string is in the format \"May 23rd, 2005\"\n if self.dateStrFlag:\n sourceTime = self.parseDateText(s)\n self.dateStrFlag = False\n\n # Given string is a weekday\n if self.weekdyFlag:\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now\n\n start = datetime.datetime(yr, mth, dy, hr, mn, sec)\n wkdy = self.ptc.WeekdayOffsets[s]\n\n if wkdy > wd:\n qty = self._CalculateDOWDelta(wd, wkdy, 2,\n self.ptc.DOWParseStyle,\n self.ptc.CurrentDOWParseStyle)\n else:\n qty = self._CalculateDOWDelta(wd, wkdy, 2,\n self.ptc.DOWParseStyle,\n self.ptc.CurrentDOWParseStyle)\n\n target = start + datetime.timedelta(days=qty)\n wd = wkdy\n\n sourceTime = target.timetuple()\n self.weekdyFlag = False\n\n # Given string is a natural language time string like\n # lunch, midnight, etc\n if self.timeStrFlag:\n if s in self.ptc.re_values['now']:\n sourceTime = now\n else:\n sources = self.ptc.buildSources(sourceTime)\n\n if s in sources:\n sourceTime = sources[s]\n else:\n sourceTime = now\n self.dateFlag = 0\n self.timeFlag = 0\n\n self.timeStrFlag = False\n\n # Given string is a natural language date string like today, tomorrow..\n if self.dayStrFlag:\n if sourceTime is None:\n sourceTime = now\n\n (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime\n\n if s in self.ptc.dayOffsets:\n offset = self.ptc.dayOffsets[s]\n else:\n offset = 0\n\n start = datetime.datetime(yr, mth, dy, 9, 0, 0)\n target = start + datetime.timedelta(days=offset)\n sourceTime = target.timetuple()\n\n self.dayStrFlag = False\n\n # Given string is a time string with units like \"5 hrs 30 min\"\n if self.unitsFlag:\n modifier = '' # TODO\n\n if sourceTime is None:\n sourceTime = now\n\n m = self.ptc.CRE_UNITS.search(s)\n if m is not None:\n units = m.group('units')\n quantity = s[:m.start('units')]\n\n sourceTime = self._buildTime(sourceTime, quantity, modifier, units)\n self.unitsFlag = False\n\n # Given string is a time string with single char units like \"5 h 30 m\"\n if self.qunitsFlag:\n modifier = '' # TODO\n\n if sourceTime is None:\n sourceTime = now\n\n m = self.ptc.CRE_QUNITS.search(s)\n if m is not None:\n units = m.group('qunits')\n quantity = s[:m.start('qunits')]\n\n sourceTime = self._buildTime(sourceTime, quantity, modifier, units)\n self.qunitsFlag = False\n\n # Given string does not match anything\n if sourceTime is None:\n sourceTime = now\n self.dateFlag = 0\n self.timeFlag = 0\n\n return sourceTime\n", "def parse(self, datetimeString, sourceTime=None):\n \"\"\"\n Splits the given C{datetimeString} into tokens, finds the regex\n patterns that match and then calculates a C{struct_time} value from\n the chunks.\n\n If C{sourceTime} is given then the C{struct_time} value will be\n calculated from that value, otherwise from the current date/time.\n\n If the C{datetimeString} is parsed and date/time value found then\n the second item of the returned tuple will be a flag to let you know\n what kind of C{struct_time} value is being returned::\n\n 0 = not parsed at all\n 1 = parsed as a C{date}\n 2 = parsed as a C{time}\n 3 = parsed as a C{datetime}\n\n @type datetimeString: string\n @param datetimeString: date/time text to evaluate\n @type sourceTime: struct_time\n @param sourceTime: C{struct_time} value to use as the base\n\n @rtype: tuple\n @return: tuple of: modified C{sourceTime} and the result flag\n \"\"\"\n\n if sourceTime:\n if isinstance(sourceTime, datetime.datetime):\n if _debug:\n print 'coercing datetime to timetuple'\n sourceTime = sourceTime.timetuple()\n else:\n if not isinstance(sourceTime, time.struct_time) and \\\n not isinstance(sourceTime, tuple):\n raise Exception('sourceTime is not a struct_time')\n\n s = datetimeString.strip().lower()\n parseStr = ''\n totalTime = sourceTime\n\n if s == '' :\n if sourceTime is not None:\n return (sourceTime, self.dateFlag + self.timeFlag)\n else:\n return (time.localtime(), 0)\n\n self.timeFlag = 0\n self.dateFlag = 0\n\n while len(s) > 0:\n flag = False\n chunk1 = ''\n chunk2 = ''\n\n if _debug:\n print 'parse (top of loop): [%s][%s]' % (s, parseStr)\n\n if parseStr == '':\n # Modifier like next\\prev..\n m = self.ptc.CRE_MODIFIER.search(s)\n if m is not None:\n self.modifierFlag = True\n if (m.group('modifier') != s):\n # capture remaining string\n parseStr = m.group('modifier')\n chunk1 = s[:m.start('modifier')].strip()\n chunk2 = s[m.end('modifier'):].strip()\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Modifier like from\\after\\prior..\n m = self.ptc.CRE_MODIFIER2.search(s)\n if m is not None:\n self.modifier2Flag = True\n if (m.group('modifier') != s):\n # capture remaining string\n parseStr = m.group('modifier')\n chunk1 = s[:m.start('modifier')].strip()\n chunk2 = s[m.end('modifier'):].strip()\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n valid_date = False\n for match in self.ptc.CRE_DATE3.finditer(s):\n # to prevent \"HH:MM(:SS) time strings\" expressions from triggering\n # this regex, we checks if the month field exists in the searched \n # expression, if it doesn't exist, the date field is not valid\n if match.group('mthname'):\n m = self.ptc.CRE_DATE3.search(s, match.start())\n valid_date = True\n break\n\n # String date format\n if valid_date:\n self.dateStrFlag = True\n self.dateFlag = 1\n if (m.group('date') != s):\n # capture remaining string\n parseStr = m.group('date')\n chunk1 = s[:m.start('date')]\n chunk2 = s[m.end('date'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Standard date format\n m = self.ptc.CRE_DATE.search(s)\n if m is not None:\n self.dateStdFlag = True\n self.dateFlag = 1\n if (m.group('date') != s):\n # capture remaining string\n parseStr = m.group('date')\n chunk1 = s[:m.start('date')]\n chunk2 = s[m.end('date'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Natural language day strings\n m = self.ptc.CRE_DAY.search(s)\n if m is not None:\n self.dayStrFlag = True\n self.dateFlag = 1\n if (m.group('day') != s):\n # capture remaining string\n parseStr = m.group('day')\n chunk1 = s[:m.start('day')]\n chunk2 = s[m.end('day'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Quantity + Units\n m = self.ptc.CRE_UNITS.search(s)\n if m is not None:\n self.unitsFlag = True\n if (m.group('qty') != s):\n # capture remaining string\n parseStr = m.group('qty')\n chunk1 = s[:m.start('qty')].strip()\n chunk2 = s[m.end('qty'):].strip()\n\n if chunk1[-1:] == '-':\n parseStr = '-%s' % parseStr\n chunk1 = chunk1[:-1]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Quantity + Units\n m = self.ptc.CRE_QUNITS.search(s)\n if m is not None:\n self.qunitsFlag = True\n\n if (m.group('qty') != s):\n # capture remaining string\n parseStr = m.group('qty')\n chunk1 = s[:m.start('qty')].strip()\n chunk2 = s[m.end('qty'):].strip()\n\n if chunk1[-1:] == '-':\n parseStr = '-%s' % parseStr\n chunk1 = chunk1[:-1]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s \n\n if parseStr == '':\n # Weekday\n m = self.ptc.CRE_WEEKDAY.search(s)\n if m is not None:\n gv = m.group('weekday')\n if s not in self.ptc.dayOffsets:\n self.weekdyFlag = True\n self.dateFlag = 1\n if (gv != s):\n # capture remaining string\n parseStr = gv\n chunk1 = s[:m.start('weekday')]\n chunk2 = s[m.end('weekday'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # Natural language time strings\n m = self.ptc.CRE_TIME.search(s)\n if m is not None:\n self.timeStrFlag = True\n self.timeFlag = 2\n if (m.group('time') != s):\n # capture remaining string\n parseStr = m.group('time')\n chunk1 = s[:m.start('time')]\n chunk2 = s[m.end('time'):]\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n else:\n parseStr = s\n\n if parseStr == '':\n # HH:MM(:SS) am/pm time strings\n m = self.ptc.CRE_TIMEHMS2.search(s)\n if m is not None:\n self.meridianFlag = True\n self.timeFlag = 2\n if m.group('minutes') is not None:\n if m.group('seconds') is not None:\n parseStr = '%s:%s:%s %s' % (m.group('hours'),\n m.group('minutes'),\n m.group('seconds'),\n m.group('meridian'))\n else:\n parseStr = '%s:%s %s' % (m.group('hours'),\n m.group('minutes'),\n m.group('meridian'))\n else:\n parseStr = '%s %s' % (m.group('hours'),\n m.group('meridian'))\n\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('meridian'):]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n\n if parseStr == '':\n # HH:MM(:SS) time strings\n m = self.ptc.CRE_TIMEHMS.search(s)\n if m is not None:\n self.timeStdFlag = True\n self.timeFlag = 2\n if m.group('seconds') is not None:\n parseStr = '%s:%s:%s' % (m.group('hours'),\n m.group('minutes'),\n m.group('seconds'))\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('seconds'):]\n else:\n parseStr = '%s:%s' % (m.group('hours'),\n m.group('minutes'))\n chunk1 = s[:m.start('hours')]\n chunk2 = s[m.end('minutes'):]\n\n s = '%s %s' % (chunk1, chunk2)\n flag = True\n\n # if string does not match any regex, empty string to\n # come out of the while loop\n if not flag:\n s = ''\n\n if _debug:\n print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2)\n print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \\\n (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag)\n print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \\\n (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag)\n\n # evaluate the matched string\n if parseStr != '':\n if self.modifierFlag == True:\n t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime)\n # t is the unparsed part of the chunks.\n # If it is not date/time, return current\n # totalTime as it is; else return the output\n # after parsing t.\n if (t != '') and (t != None):\n tempDateFlag = self.dateFlag\n tempTimeFlag = self.timeFlag\n (totalTime2, flag) = self.parse(t, totalTime)\n\n if flag == 0 and totalTime is not None:\n self.timeFlag = tempTimeFlag\n self.dateFlag = tempDateFlag\n\n return (totalTime, self.dateFlag + self.timeFlag)\n else:\n return (totalTime2, self.dateFlag + self.timeFlag)\n\n elif self.modifier2Flag == True:\n totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime)\n\n if invalidFlag == True:\n self.dateFlag = 0\n self.timeFlag = 0\n\n else:\n totalTime = self._evalString(parseStr, totalTime)\n parseStr = ''\n\n # String is not parsed at all\n if totalTime is None or totalTime == sourceTime:\n totalTime = time.localtime()\n self.dateFlag = 0\n self.timeFlag = 0\n\n return (totalTime, self.dateFlag + self.timeFlag)\n" ]
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime.py
Calendar.inc
python
def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime.py#L1479-L1540
null
class Calendar: """ A collection of routines to input, parse and manipulate date and times. The text can either be 'normal' date values or it can be human readable. """ def __init__(self, constants=None): """ Default constructor for the L{Calendar} class. @type constants: object @param constants: Instance of the class L{parsedatetime_consts.Constants} @rtype: object @return: L{Calendar} instance """ # if a constants reference is not included, use default if constants is None: self.ptc = parsedatetime_consts.Constants() else: self.ptc = constants self.weekdyFlag = False # monday/tuesday/... self.dateStdFlag = False # 07/21/06 self.dateStrFlag = False # July 21st, 2006 self.timeStdFlag = False # 5:50 self.meridianFlag = False # am/pm self.dayStrFlag = False # tomorrow/yesterday/today/.. self.timeStrFlag = False # lunch/noon/breakfast/... self.modifierFlag = False # after/before/prev/next/.. self.modifier2Flag = False # after/before/prev/next/.. self.unitsFlag = False # hrs/weeks/yrs/min/.. self.qunitsFlag = False # h/m/t/d.. self.timeFlag = 0 self.dateFlag = 0 def _convertUnitAsWords(self, unitText): """ Converts text units into their number value Five = 5 Twenty Five = 25 Two hundred twenty five = 225 Two thousand and twenty five = 2025 Two thousand twenty five = 2025 @type unitText: string @param unitText: number text to convert @rtype: integer @return: numerical value of unitText """ # TODO: implement this pass def _buildTime(self, source, quantity, modifier, units): """ Take C{quantity}, C{modifier} and C{unit} strings and convert them into values. After converting, calcuate the time and return the adjusted sourceTime. @type source: time @param source: time to use as the base (or source) @type quantity: string @param quantity: quantity string @type modifier: string @param modifier: how quantity and units modify the source time @type units: string @param units: unit of the quantity (i.e. hours, days, months, etc) @rtype: struct_time @return: C{struct_time} of the calculated time """ if _debug: print '_buildTime: [%s][%s][%s]' % (quantity, modifier, units) if source is None: source = time.localtime() if quantity is None: quantity = '' else: quantity = quantity.strip() if len(quantity) == 0: qty = 1 else: try: qty = int(quantity) except ValueError: qty = 0 if modifier in self.ptc.Modifiers: qty = qty * self.ptc.Modifiers[modifier] if units is None or units == '': units = 'dy' # plurals are handled by regex's (could be a bug tho) (yr, mth, dy, hr, mn, sec, _, _, _) = source start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start if units.startswith('y'): target = self.inc(start, year=qty) self.dateFlag = 1 elif units.endswith('th') or units.endswith('ths'): target = self.inc(start, month=qty) self.dateFlag = 1 else: if units.startswith('d'): target = start + datetime.timedelta(days=qty) self.dateFlag = 1 elif units.startswith('h'): target = start + datetime.timedelta(hours=qty) self.timeFlag = 2 elif units.startswith('m'): target = start + datetime.timedelta(minutes=qty) self.timeFlag = 2 elif units.startswith('s'): target = start + datetime.timedelta(seconds=qty) self.timeFlag = 2 elif units.startswith('w'): target = start + datetime.timedelta(weeks=qty) self.dateFlag = 1 return target.timetuple() def parseDate(self, dateString): """ Parse short-form date strings:: '05/28/2006' or '04.21' @type dateString: string @param dateString: text to convert to a C{datetime} @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() # values pulled from regex's will be stored here and later # assigned to mth, dy, yr based on information from the locale # -1 is used as the marker value because we want zero values # to be passed thru so they can be flagged as errors later v1 = -1 v2 = -1 v3 = -1 s = dateString m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v1 = int(s[:index]) s = s[index + 1:] m = self.ptc.CRE_DATE2.search(s) if m is not None: index = m.start() v2 = int(s[:index]) v3 = int(s[index + 1:]) else: v2 = int(s.strip()) v = [ v1, v2, v3 ] d = { 'm': mth, 'd': dy, 'y': yr } for i in range(0, 3): n = v[i] c = self.ptc.dp_order[i] if n >= 0: d[c] = n # if the year is not specified and the date has already # passed, increment the year if v3 == -1 and ((mth > d['m']) or (mth == d['m'] and dy > d['d'])): yr = d['y'] + 1 else: yr = d['y'] mth = d['m'] dy = d['d'] # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 if _debug: print 'parseDate: ', yr, mth, dy, self.ptc.daysInMonth(mth, yr) if (mth > 0 and mth <= 12) and \ (dy > 0 and dy <= self.ptc.daysInMonth(mth, yr)): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() # return current time if date # string is invalid return sourceTime def parseDateText(self, dateString): """ Parse long-form date strings:: 'May 31st, 2006' 'Jan 1st' 'July 2006' @type dateString: string @param dateString: text to convert to a datetime @rtype: struct_time @return: calculated C{struct_time} value of dateString """ yr, mth, dy, hr, mn, sec, wd, yd, isdst = time.localtime() currentMth = mth currentDy = dy s = dateString.lower() m = self.ptc.CRE_DATE3.search(s) mth = m.group('mthname') mth = self.ptc.MonthOffsets[mth] if m.group('day') != None: dy = int(m.group('day')) else: dy = 1 if m.group('year') != None: yr = int(m.group('year')) # birthday epoch constraint if yr < self.ptc.BirthdayEpoch: yr += 2000 elif yr < 100: yr += 1900 elif (mth < currentMth) or (mth == currentMth and dy < currentDy): # if that day and month have already passed in this year, # then increment the year by 1 yr += 1 if dy > 0 and dy <= self.ptc.daysInMonth(mth, yr): sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) else: # Return current time if date string is invalid self.dateFlag = 0 self.timeFlag = 0 sourceTime = time.localtime() return sourceTime def evalRanges(self, datetimeString, sourceTime=None): """ Evaluate the C{datetimeString} text and determine if it represents a date or time range. @type datetimeString: string @param datetimeString: datetime text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: start datetime, end datetime and the invalid flag """ startTime = '' endTime = '' startDate = '' endDate = '' rangeFlag = 0 s = datetimeString.strip().lower() if self.ptc.rangeSep in s: s = s.replace(self.ptc.rangeSep, ' %s ' % self.ptc.rangeSep) s = s.replace(' ', ' ') m = self.ptc.CRE_TIMERNG1.search(s) if m is not None: rangeFlag = 1 else: m = self.ptc.CRE_TIMERNG2.search(s) if m is not None: rangeFlag = 2 else: m = self.ptc.CRE_TIMERNG4.search(s) if m is not None: rangeFlag = 7 else: m = self.ptc.CRE_TIMERNG3.search(s) if m is not None: rangeFlag = 3 else: m = self.ptc.CRE_DATERNG1.search(s) if m is not None: rangeFlag = 4 else: m = self.ptc.CRE_DATERNG2.search(s) if m is not None: rangeFlag = 5 else: m = self.ptc.CRE_DATERNG3.search(s) if m is not None: rangeFlag = 6 if _debug: print 'evalRanges: rangeFlag =', rangeFlag, '[%s]' % s if m is not None: if (m.group() != s): # capture remaining string parseStr = m.group() chunk1 = s[:m.start()] chunk2 = s[m.end():] s = '%s %s' % (chunk1, chunk2) flag = 1 sourceTime, flag = self.parse(s, sourceTime) if flag == 0: sourceTime = None else: parseStr = s if rangeFlag == 1: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 2: m = re.search(self.ptc.rangeSep, parseStr) startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 3 or rangeFlag == 7: m = re.search(self.ptc.rangeSep, parseStr) # capturing the meridian from the end time if self.ptc.usesMeridian: ampm = re.search(self.ptc.am[0], parseStr) # appending the meridian to the start time if ampm is not None: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[0]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()] + self.ptc.meridian[1]), sourceTime) else: startTime, sflag = self.parse((parseStr[:m.start()]), sourceTime) endTime, eflag = self.parse(parseStr[(m.start() + 1):], sourceTime) if (eflag != 0) and (sflag != 0): return (startTime, endTime, 2) elif rangeFlag == 4: m = re.search(self.ptc.rangeSep, parseStr) startDate, sflag = self.parse((parseStr[:m.start()]), sourceTime) endDate, eflag = self.parse((parseStr[(m.start() + 1):]), sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 5: m = re.search(self.ptc.rangeSep, parseStr) endDate = parseStr[(m.start() + 1):] # capturing the year from the end date date = self.ptc.CRE_DATE3.search(endDate) endYear = date.group('year') # appending the year to the start date if the start date # does not have year information and the end date does. # eg : "Aug 21 - Sep 4, 2007" if endYear is not None: startDate = (parseStr[:m.start()]).strip() date = self.ptc.CRE_DATE3.search(startDate) startYear = date.group('year') if startYear is None: startDate = startDate + ', ' + endYear else: startDate = parseStr[:m.start()] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) elif rangeFlag == 6: m = re.search(self.ptc.rangeSep, parseStr) startDate = parseStr[:m.start()] # capturing the month from the start date mth = self.ptc.CRE_DATE3.search(startDate) mth = mth.group('mthname') # appending the month name to the end date endDate = mth + parseStr[(m.start() + 1):] startDate, sflag = self.parse(startDate, sourceTime) endDate, eflag = self.parse(endDate, sourceTime) if (eflag != 0) and (sflag != 0): return (startDate, endDate, 1) else: # if range is not found sourceTime = time.localtime() return (sourceTime, sourceTime, 0) def _CalculateDOWDelta(self, wd, wkdy, offset, style, currentDayStyle): """ Based on the C{style} and C{currentDayStyle} determine what day-of-week value is to be returned. @type wd: integer @param wd: day-of-week value for the current day @type wkdy: integer @param wkdy: day-of-week value for the parsed day @type offset: integer @param offset: offset direction for any modifiers (-1, 0, 1) @type style: integer @param style: normally the value set in C{Constants.DOWParseStyle} @type currentDayStyle: integer @param currentDayStyle: normally the value set in C{Constants.CurrentDOWParseStyle} @rtype: integer @return: calculated day-of-week """ if offset == 1: # modifier is indicating future week eg: "next". # DOW is calculated as DOW of next week diff = 7 - wd + wkdy elif offset == -1: # modifier is indicating past week eg: "last","previous" # DOW is calculated as DOW of previous week diff = wkdy - wd - 7 elif offset == 0: # modifier is indiacting current week eg: "this" # DOW is calculated as DOW of this week diff = wkdy - wd elif offset == 2: # no modifier is present. # i.e. string to be parsed is just DOW if style == 1: # next occurance of the DOW is calculated if currentDayStyle == True: if wkdy >= wd: diff = wkdy - wd else: diff = 7 - wd + wkdy else: if wkdy > wd: diff = wkdy - wd else: diff = 7 - wd + wkdy elif style == -1: # last occurance of the DOW is calculated if currentDayStyle == True: if wkdy <= wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: if wkdy < wd: diff = wkdy - wd else: diff = wkdy - wd - 7 else: # occurance of the DOW in the current week is calculated diff = wkdy - wd if _debug: print "wd %s, wkdy %s, offset %d, style %d\n" % (wd, wkdy, offset, style) return diff def _evalModifier(self, modifier, chunk1, chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to sourceTime @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() # capture the units after the modifier and the remaining # string after the unit m = self.ptc.CRE_REMAINING.search(chunk2) if m is not None: index = m.start() + 1 unit = chunk2[:m.start()] chunk2 = chunk2[index:] else: unit = chunk2 chunk2 = '' flag = False if unit == 'month' or \ unit == 'mth' or \ unit == 'm': if offset == 0: dy = self.ptc.daysInMonth(mth, yr) sourceTime = (yr, mth, dy, 9, 0, 0, wd, yd, isdst) elif offset == 2: # if day is the last day of the month, calculate the last day # of the next month if dy == self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth + 1, yr) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = self.inc(start, month=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, 1, 9, 0, 0) target = self.inc(start, month=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'week' or \ unit == 'wk' or \ unit == 'w': if offset == 0: start = datetime.datetime(yr, mth, dy, 17, 0, 0) target = start + datetime.timedelta(days=(4 - wd)) sourceTime = target.timetuple() elif offset == 2: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=7) sourceTime = target.timetuple() else: return self._evalModifier(modifier, chunk1, "monday " + chunk2, sourceTime) flag = True self.dateFlag = 1 if unit == 'day' or \ unit == 'dy' or \ unit == 'd': if offset == 0: sourceTime = (yr, mth, dy, 17, 0, 0, wd, yd, isdst) self.timeFlag = 2 elif offset == 2: start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=1) sourceTime = target.timetuple() else: start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if unit == 'hour' or \ unit == 'hr': if offset == 0: sourceTime = (yr, mth, dy, hr, 0, 0, wd, yd, isdst) else: start = datetime.datetime(yr, mth, dy, hr, 0, 0) target = start + datetime.timedelta(hours=offset) sourceTime = target.timetuple() flag = True self.timeFlag = 2 if unit == 'year' or \ unit == 'yr' or \ unit == 'y': if offset == 0: sourceTime = (yr, 12, 31, hr, mn, sec, wd, yd, isdst) elif offset == 2: sourceTime = (yr + 1, mth, dy, hr, mn, sec, wd, yd, isdst) else: sourceTime = (yr + offset, 1, 1, 9, 0, 0, wd, yd, isdst) flag = True self.dateFlag = 1 if flag == False: m = self.ptc.CRE_WEEKDAY.match(unit) if m is not None: wkdy = m.group() self.dateFlag = 1 if modifier == 'eod': # Calculate the upcoming weekday self.modifierFlag = False (sourceTime, _) = self.parse(wkdy, sourceTime) sources = self.ptc.buildSources(sourceTime) self.timeFlag = 2 if modifier in sources: sourceTime = sources[modifier] else: wkdy = self.ptc.WeekdayOffsets[wkdy] diff = self._CalculateDOWDelta(wd, wkdy, offset, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=diff) sourceTime = target.timetuple() flag = True self.dateFlag = 1 if not flag: m = self.ptc.CRE_TIME.match(unit) if m is not None: self.modifierFlag = False (yr, mth, dy, hr, mn, sec, wd, yd, isdst), _ = self.parse(unit) start = datetime.datetime(yr, mth, dy, hr, mn, sec) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() flag = True else: self.modifierFlag = False # check if the remaining text is parsable and if so, # use it as the base time for the modifier source time t, flag2 = self.parse('%s %s' % (chunk1, unit), sourceTime) if flag2 != 0: sourceTime = t sources = self.ptc.buildSources(sourceTime) if modifier in sources: sourceTime = sources[modifier] flag = True self.timeFlag = 2 # if the word after next is a number, the string is more than likely # to be "next 4 hrs" which we will have to combine the units with the # rest of the string if not flag: if offset < 0: # if offset is negative, the unit has to be made negative unit = '-%s' % unit chunk2 = '%s %s' % (unit, chunk2) self.modifierFlag = False #return '%s %s' % (chunk1, chunk2), sourceTime return '%s' % chunk2, sourceTime def _evalModifier2(self, modifier, chunk1 , chunk2, sourceTime): """ Evaluate the C{modifier} string and following text (passed in as C{chunk1} and C{chunk2}) and if they match any known modifiers calculate the delta and apply it to C{sourceTime}. @type modifier: string @param modifier: modifier text to apply to C{sourceTime} @type chunk1: string @param chunk1: first text chunk that followed modifier (if any) @type chunk2: string @param chunk2: second text chunk that followed modifier (if any) @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: remaining text and the modified sourceTime """ offset = self.ptc.Modifiers[modifier] digit = r'\d+' self.modifier2Flag = False # If the string after the negative modifier starts with digits, # then it is likely that the string is similar to ' before 3 days' # or 'evening prior to 3 days'. # In this case, the total time is calculated by subtracting '3 days' # from the current date. # So, we have to identify the quantity and negate it before parsing # the string. # This is not required for strings not starting with digits since the # string is enough to calculate the sourceTime if chunk2 != '': if offset < 0: m = re.match(digit, chunk2.strip()) if m is not None: qty = int(m.group()) * -1 chunk2 = chunk2[m.end():] chunk2 = '%d%s' % (qty, chunk2) sourceTime, flag1 = self.parse(chunk2, sourceTime) if flag1 == 0: flag1 = True else: flag1 = False flag2 = False else: flag1 = False if chunk1 != '': if offset < 0: m = re.search(digit, chunk1.strip()) if m is not None: qty = int(m.group()) * -1 chunk1 = chunk1[m.end():] chunk1 = '%d%s' % (qty, chunk1) tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag sourceTime2, flag2 = self.parse(chunk1, sourceTime) else: return sourceTime, (flag1 and flag2) # if chunk1 is not a datetime and chunk2 is then do not use datetime # value returned by parsing chunk1 if not (flag1 == False and flag2 == 0): sourceTime = sourceTime2 else: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return sourceTime, (flag1 and flag2) def _evalString(self, datetimeString, sourceTime=None): """ Calculate the datetime based on flags set by the L{parse()} routine Examples handled:: RFC822, W3CDTF formatted dates HH:MM[:SS][ am/pm] MM/DD/YYYY DD MMMM YYYY @type datetimeString: string @param datetimeString: text to try and parse as more "traditional" date/time text @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: datetime @return: calculated C{struct_time} value or current C{struct_time} if not parsed """ s = datetimeString.strip() now = time.localtime() # Given string date is a RFC822 date if sourceTime is None: sourceTime = _parse_date_rfc822(s) if sourceTime is not None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst, _) = sourceTime self.dateFlag = 1 if (hr != 0) and (mn != 0) and (sec != 0): self.timeFlag = 2 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) # Given string date is a W3CDTF date if sourceTime is None: sourceTime = _parse_date_w3dtf(s) if sourceTime is not None: self.dateFlag = 1 self.timeFlag = 2 if sourceTime is None: s = s.lower() # Given string is in the format HH:MM(:SS)(am/pm) if self.meridianFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: dt = s[:m.start('meridian')].strip() if len(dt) <= 2: hr = int(dt) mn = 0 sec = 0 else: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) meridian = m.group('meridian').lower() # if 'am' found and hour is 12 - force hour to 0 (midnight) if (meridian in self.ptc.am) and hr == 12: sourceTime = (yr, mth, dy, 0, mn, sec, wd, yd, isdst) # if 'pm' found and hour < 12, add 12 to shift to evening if (meridian in self.ptc.pm) and hr < 12: sourceTime = (yr, mth, dy, hr + 12, mn, sec, wd, yd, isdst) # invalid time if hr > 24 or mn > 59 or sec > 59: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.meridianFlag = False # Given string is in the format HH:MM(:SS) if self.timeStdFlag: if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: hr, mn, sec = _extract_time(m) if hr == 24: hr = 0 if hr > 24 or mn > 59 or sec > 59: # invalid time sourceTime = now self.dateFlag = 0 self.timeFlag = 0 else: sourceTime = (yr, mth, dy, hr, mn, sec, wd, yd, isdst) self.timeStdFlag = False # Given string is in the format 07/21/2006 if self.dateStdFlag: sourceTime = self.parseDate(s) self.dateStdFlag = False # Given string is in the format "May 23rd, 2005" if self.dateStrFlag: sourceTime = self.parseDateText(s) self.dateStrFlag = False # Given string is a weekday if self.weekdyFlag: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = now start = datetime.datetime(yr, mth, dy, hr, mn, sec) wkdy = self.ptc.WeekdayOffsets[s] if wkdy > wd: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) else: qty = self._CalculateDOWDelta(wd, wkdy, 2, self.ptc.DOWParseStyle, self.ptc.CurrentDOWParseStyle) target = start + datetime.timedelta(days=qty) wd = wkdy sourceTime = target.timetuple() self.weekdyFlag = False # Given string is a natural language time string like # lunch, midnight, etc if self.timeStrFlag: if s in self.ptc.re_values['now']: sourceTime = now else: sources = self.ptc.buildSources(sourceTime) if s in sources: sourceTime = sources[s] else: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 self.timeStrFlag = False # Given string is a natural language date string like today, tomorrow.. if self.dayStrFlag: if sourceTime is None: sourceTime = now (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime if s in self.ptc.dayOffsets: offset = self.ptc.dayOffsets[s] else: offset = 0 start = datetime.datetime(yr, mth, dy, 9, 0, 0) target = start + datetime.timedelta(days=offset) sourceTime = target.timetuple() self.dayStrFlag = False # Given string is a time string with units like "5 hrs 30 min" if self.unitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_UNITS.search(s) if m is not None: units = m.group('units') quantity = s[:m.start('units')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.unitsFlag = False # Given string is a time string with single char units like "5 h 30 m" if self.qunitsFlag: modifier = '' # TODO if sourceTime is None: sourceTime = now m = self.ptc.CRE_QUNITS.search(s) if m is not None: units = m.group('qunits') quantity = s[:m.start('qunits')] sourceTime = self._buildTime(sourceTime, quantity, modifier, units) self.qunitsFlag = False # Given string does not match anything if sourceTime is None: sourceTime = now self.dateFlag = 0 self.timeFlag = 0 return sourceTime def parse(self, datetimeString, sourceTime=None): """ Splits the given C{datetimeString} into tokens, finds the regex patterns that match and then calculates a C{struct_time} value from the chunks. If C{sourceTime} is given then the C{struct_time} value will be calculated from that value, otherwise from the current date/time. If the C{datetimeString} is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of C{struct_time} value is being returned:: 0 = not parsed at all 1 = parsed as a C{date} 2 = parsed as a C{time} 3 = parsed as a C{datetime} @type datetimeString: string @param datetimeString: date/time text to evaluate @type sourceTime: struct_time @param sourceTime: C{struct_time} value to use as the base @rtype: tuple @return: tuple of: modified C{sourceTime} and the result flag """ if sourceTime: if isinstance(sourceTime, datetime.datetime): if _debug: print 'coercing datetime to timetuple' sourceTime = sourceTime.timetuple() else: if not isinstance(sourceTime, time.struct_time) and \ not isinstance(sourceTime, tuple): raise Exception('sourceTime is not a struct_time') s = datetimeString.strip().lower() parseStr = '' totalTime = sourceTime if s == '' : if sourceTime is not None: return (sourceTime, self.dateFlag + self.timeFlag) else: return (time.localtime(), 0) self.timeFlag = 0 self.dateFlag = 0 while len(s) > 0: flag = False chunk1 = '' chunk2 = '' if _debug: print 'parse (top of loop): [%s][%s]' % (s, parseStr) if parseStr == '': # Modifier like next\prev.. m = self.ptc.CRE_MODIFIER.search(s) if m is not None: self.modifierFlag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': # Modifier like from\after\prior.. m = self.ptc.CRE_MODIFIER2.search(s) if m is not None: self.modifier2Flag = True if (m.group('modifier') != s): # capture remaining string parseStr = m.group('modifier') chunk1 = s[:m.start('modifier')].strip() chunk2 = s[m.end('modifier'):].strip() flag = True else: parseStr = s if parseStr == '': valid_date = False for match in self.ptc.CRE_DATE3.finditer(s): # to prevent "HH:MM(:SS) time strings" expressions from triggering # this regex, we checks if the month field exists in the searched # expression, if it doesn't exist, the date field is not valid if match.group('mthname'): m = self.ptc.CRE_DATE3.search(s, match.start()) valid_date = True break # String date format if valid_date: self.dateStrFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Standard date format m = self.ptc.CRE_DATE.search(s) if m is not None: self.dateStdFlag = True self.dateFlag = 1 if (m.group('date') != s): # capture remaining string parseStr = m.group('date') chunk1 = s[:m.start('date')] chunk2 = s[m.end('date'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language day strings m = self.ptc.CRE_DAY.search(s) if m is not None: self.dayStrFlag = True self.dateFlag = 1 if (m.group('day') != s): # capture remaining string parseStr = m.group('day') chunk1 = s[:m.start('day')] chunk2 = s[m.end('day'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_UNITS.search(s) if m is not None: self.unitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Quantity + Units m = self.ptc.CRE_QUNITS.search(s) if m is not None: self.qunitsFlag = True if (m.group('qty') != s): # capture remaining string parseStr = m.group('qty') chunk1 = s[:m.start('qty')].strip() chunk2 = s[m.end('qty'):].strip() if chunk1[-1:] == '-': parseStr = '-%s' % parseStr chunk1 = chunk1[:-1] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Weekday m = self.ptc.CRE_WEEKDAY.search(s) if m is not None: gv = m.group('weekday') if s not in self.ptc.dayOffsets: self.weekdyFlag = True self.dateFlag = 1 if (gv != s): # capture remaining string parseStr = gv chunk1 = s[:m.start('weekday')] chunk2 = s[m.end('weekday'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # Natural language time strings m = self.ptc.CRE_TIME.search(s) if m is not None: self.timeStrFlag = True self.timeFlag = 2 if (m.group('time') != s): # capture remaining string parseStr = m.group('time') chunk1 = s[:m.start('time')] chunk2 = s[m.end('time'):] s = '%s %s' % (chunk1, chunk2) flag = True else: parseStr = s if parseStr == '': # HH:MM(:SS) am/pm time strings m = self.ptc.CRE_TIMEHMS2.search(s) if m is not None: self.meridianFlag = True self.timeFlag = 2 if m.group('minutes') is not None: if m.group('seconds') is not None: parseStr = '%s:%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('seconds'), m.group('meridian')) else: parseStr = '%s:%s %s' % (m.group('hours'), m.group('minutes'), m.group('meridian')) else: parseStr = '%s %s' % (m.group('hours'), m.group('meridian')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('meridian'):] s = '%s %s' % (chunk1, chunk2) flag = True if parseStr == '': # HH:MM(:SS) time strings m = self.ptc.CRE_TIMEHMS.search(s) if m is not None: self.timeStdFlag = True self.timeFlag = 2 if m.group('seconds') is not None: parseStr = '%s:%s:%s' % (m.group('hours'), m.group('minutes'), m.group('seconds')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('seconds'):] else: parseStr = '%s:%s' % (m.group('hours'), m.group('minutes')) chunk1 = s[:m.start('hours')] chunk2 = s[m.end('minutes'):] s = '%s %s' % (chunk1, chunk2) flag = True # if string does not match any regex, empty string to # come out of the while loop if not flag: s = '' if _debug: print 'parse (bottom) [%s][%s][%s][%s]' % (s, parseStr, chunk1, chunk2) print 'weekday %s, dateStd %s, dateStr %s, time %s, timeStr %s, meridian %s' % \ (self.weekdyFlag, self.dateStdFlag, self.dateStrFlag, self.timeStdFlag, self.timeStrFlag, self.meridianFlag) print 'dayStr %s, modifier %s, modifier2 %s, units %s, qunits %s' % \ (self.dayStrFlag, self.modifierFlag, self.modifier2Flag, self.unitsFlag, self.qunitsFlag) # evaluate the matched string if parseStr != '': if self.modifierFlag == True: t, totalTime = self._evalModifier(parseStr, chunk1, chunk2, totalTime) # t is the unparsed part of the chunks. # If it is not date/time, return current # totalTime as it is; else return the output # after parsing t. if (t != '') and (t != None): tempDateFlag = self.dateFlag tempTimeFlag = self.timeFlag (totalTime2, flag) = self.parse(t, totalTime) if flag == 0 and totalTime is not None: self.timeFlag = tempTimeFlag self.dateFlag = tempDateFlag return (totalTime, self.dateFlag + self.timeFlag) else: return (totalTime2, self.dateFlag + self.timeFlag) elif self.modifier2Flag == True: totalTime, invalidFlag = self._evalModifier2(parseStr, chunk1, chunk2, totalTime) if invalidFlag == True: self.dateFlag = 0 self.timeFlag = 0 else: totalTime = self._evalString(parseStr, totalTime) parseStr = '' # String is not parsed at all if totalTime is None or totalTime == sourceTime: totalTime = time.localtime() self.dateFlag = 0 self.timeFlag = 0 return (totalTime, self.dateFlag + self.timeFlag) def inc(self, source, month=None, year=None): """ Takes the given C{source} date, or current date if none is passed, and increments it according to the values passed in by month and/or year. This routine is needed because Python's C{timedelta()} function does not allow for month or year increments. @type source: struct_time @param source: C{struct_time} value to increment @type month: integer @param month: optional number of months to increment @type year: integer @param year: optional number of years to increment @rtype: datetime @return: C{source} incremented by the number of months and/or years """ yr = source.year mth = source.month dy = source.day if year: try: yi = int(year) except ValueError: yi = 0 yr += yi if month: try: mi = int(month) except ValueError: mi = 0 m = abs(mi) y = m / 12 # how many years are in month increment m = m % 12 # get remaining months if mi < 0: mth = mth - m # sub months from start month if mth < 1: # cross start-of-year? y -= 1 # yes - decrement year mth += 12 # and fix month else: mth = mth + m # add months to start month if mth > 12: # cross end-of-year? y += 1 # yes - increment year mth -= 12 # and fix month yr += y # if the day ends up past the last day of # the new month, set it to the last day if dy > self.ptc.daysInMonth(mth, yr): dy = self.ptc.daysInMonth(mth, yr) d = source.replace(year=yr, month=mth, day=dy) return source + (d - source)
vilmibm/done
parsedatetime/parsedatetime_consts.py
_initLocale
python
def _initLocale(ptc): def lcase(x): return x.lower() if pyicu and ptc.usePyICU: ptc.icuLocale = None if ptc.localeID is not None: ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is None: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is not None: break ptc.icuSymbols = pyicu.DateFormatSymbols(ptc.icuLocale) # grab ICU list of weekdays, skipping first entry which # is always blank wd = map(lcase, ptc.icuSymbols.getWeekdays()[1:]) swd = map(lcase, ptc.icuSymbols.getShortWeekdays()[1:]) # store them in our list with Monday first (ICU puts Sunday first) ptc.Weekdays = wd[1:] + wd[0:1] ptc.shortWeekdays = swd[1:] + swd[0:1] ptc.Months = map(lcase, ptc.icuSymbols.getMonths()) ptc.shortMonths = map(lcase, ptc.icuSymbols.getShortMonths()) # not quite sure how to init this so for now # set it to none so it will be set to the en_US defaults for now ptc.re_consts = None ptc.icu_df = { 'full': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.icu_tf = { 'full': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.dateFormats = { 'full': ptc.icu_df['full'].toPattern(), 'long': ptc.icu_df['long'].toPattern(), 'medium': ptc.icu_df['medium'].toPattern(), 'short': ptc.icu_df['short'].toPattern(), } ptc.timeFormats = { 'full': ptc.icu_tf['full'].toPattern(), 'long': ptc.icu_tf['long'].toPattern(), 'medium': ptc.icu_tf['medium'].toPattern(), 'short': ptc.icu_tf['short'].toPattern(), } else: if not ptc.localeID in pdtLocales: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] if ptc.localeID in pdtLocales: break ptc.locale = pdtLocales[ptc.localeID] ptc.usePyICU = False ptc.Weekdays = ptc.locale.Weekdays ptc.shortWeekdays = ptc.locale.shortWeekdays ptc.Months = ptc.locale.Months ptc.shortMonths = ptc.locale.shortMonths ptc.dateFormats = ptc.locale.dateFormats ptc.timeFormats = ptc.locale.timeFormats # these values are used to setup the various bits # of the regex values used to parse # # check if a local set of constants has been # provided, if not use en_US as the default if ptc.localeID in pdtLocales: ptc.re_sources = pdtLocales[ptc.localeID].re_sources ptc.re_values = pdtLocales[ptc.localeID].re_consts units = pdtLocales[ptc.localeID].units ptc.Modifiers = pdtLocales[ptc.localeID].modifiers ptc.dayOffsets = pdtLocales[ptc.localeID].dayoffsets # for now, pull over any missing keys from the US set for key in pdtLocales['en_US'].re_consts: if not key in ptc.re_values: ptc.re_values[key] = pdtLocales['en_US'].re_consts[key] else: ptc.re_sources = pdtLocales['en_US'].re_sources ptc.re_values = pdtLocales['en_US'].re_consts ptc.Modifiers = pdtLocales['en_US'].modifiers ptc.dayOffsets = pdtLocales['en_US'].dayoffsets units = pdtLocales['en_US'].units # escape any regex special characters that may be found wd = tuple(map(re.escape, ptc.Weekdays)) swd = tuple(map(re.escape, ptc.shortWeekdays)) mth = tuple(map(re.escape, ptc.Months)) smth = tuple(map(re.escape, ptc.shortMonths)) ptc.re_values['months'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % mth ptc.re_values['shortmonths'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % smth ptc.re_values['days'] = '%s|%s|%s|%s|%s|%s|%s' % wd ptc.re_values['shortdays'] = '%s|%s|%s|%s|%s|%s|%s' % swd l = [] for unit in units: l.append('|'.join(units[unit])) ptc.re_values['units'] = '|'.join(l) ptc.Units = ptc.re_values['units'].split('|')
Helper function to initialize the different lists and strings from either PyICU or one of the internal pdt Locales and store them into ptc.
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime_consts.py#L479-L598
null
#!/usr/bin/env python """ parsedatetime constants and helper functions to determine regex values from Locale information if present. Also contains the internal Locale classes to give some sane defaults if PyICU is not found. """ __license__ = """ Copyright (c) 2004-2008 Mike Taylor Copyright (c) 2006-2008 Darshana Chhajed Copyright (c) 2007 Bernd Zeimetz <bzed@debian.org> All rights reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ try: import PyICU as pyicu except: pyicu = None import datetime import calendar import time import re class pdtLocale_en: """ en_US Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for USA """ localeID = 'en_US' # don't use a unicode string dateSep = [ u'/', u'.' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, MMMM d, yyyy', 'long': 'MMMM d, yyyy', 'medium': 'MMM d, yyyy', 'short': 'M/d/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a z', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'm', u'd', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': -1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eod': 0, 'eo': 0 } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_au: """ en_AU Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for Australia """ localeID = 'en_AU' # don't use a unicode string dateSep = [ u'-', u'/' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, d MMMM yyyy', 'long': 'd MMMM yyyy', 'medium': 'dd/MM/yyyy', 'short': 'd/MM/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_es: """ es Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings in Spanish Note that I don't speak Spanish so many of the items below are still in English """ localeID = 'es' # don't use a unicode string dateSep = [ u'/' ] timeSep = [ u':' ] meridian = [] usesMeridian = False uses24 = True Weekdays = [ u'lunes', u'martes', u'mi\xe9rcoles', u'jueves', u'viernes', u's\xe1bado', u'domingo', ] shortWeekdays = [ u'lun', u'mar', u'mi\xe9', u'jue', u'vie', u's\xe1b', u'dom', ] Months = [ u'enero', u'febrero', u'marzo', u'abril', u'mayo', u'junio', u'julio', u'agosto', u'septiembre', u'octubre', u'noviembre', u'diciembre' ] shortMonths = [ u'ene', u'feb', u'mar', u'abr', u'may', u'jun', u'jul', u'ago', u'sep', u'oct', u'nov', u'dic' ] dateFormats = { 'full': "EEEE d' de 'MMMM' de 'yyyy", 'long': "d' de 'MMMM' de 'yyyy", 'medium': "dd-MMM-yy", 'short': "d/MM/yy", } timeFormats = { 'full': "HH'H'mm' 'ss z", 'long': "HH:mm:ss z", 'medium': "HH:mm:ss", 'short': "HH:mm", } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': timeSep, 'dateseperator': dateSep, 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_de: """ de_DE Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Contributed by Debian parsedatetime package maintainer Bernd Zeimetz <bzed@debian.org> Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for German """ localeID = 'de_DE' # don't use a unicode string dateSep = [ u'.' ] timeSep = [ u':' ] meridian = [ ] usesMeridian = False uses24 = True Weekdays = [ u'montag', u'dienstag', u'mittwoch', u'donnerstag', u'freitag', u'samstag', u'sonntag', ] shortWeekdays = [ u'mo', u'di', u'mi', u'do', u'fr', u'sa', u'so', ] Months = [ u'januar', u'februar', u'm\xe4rz', u'april', u'mai', u'juni', u'juli', u'august', u'september', u'oktober', u'november', u'dezember', ] shortMonths = [ u'jan', u'feb', u'mrz', u'apr', u'mai', u'jun', u'jul', u'aug', u'sep', u'okt', u'nov', u'dez', ] dateFormats = { 'full': u'EEEE, d. MMMM yyyy', 'long': u'd. MMMM yyyy', 'medium': u'dd.MM.yyyy', 'short': u'dd.MM.yy' } timeFormats = { 'full': u'HH:mm:ss v', 'long': u'HH:mm:ss z', 'medium': u'HH:mm:ss', 'short': u'HH:mm' } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'sekunden', 'sek', 's' ], 'minutes': [ 'minuten', 'min' , 'm' ], 'hours': [ 'stunden', 'std', 'h' ], 'days': [ 'tage', 't' ], 'weeks': [ 'wochen', 'w' ], 'months': [ 'monate' ], #the short version would be a capital M, #as I understand it we can't distinguis #between m for minutes and M for months. 'years': [ 'jahre', 'j' ], } # text constants to be used by regex's later re_consts = { 'specials': 'am|dem|der|im|in|den|zum', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': '', 'qunits': 'h|m|s|t|w|m|j', 'now': [ 'jetzt' ], } # Used to adjust the returned date before/after the source #still looking for insight on how to translate all of them to german. modifiers = { u'from': 1, u'before': -1, u'after': 1, u'vergangener': -1, u'vorheriger': -1, u'prev': -1, u'letzter': -1, u'n\xe4chster': 1, u'dieser': 0, u'previous': -1, u'in a': 2, u'end of': 0, u'eod': 0, u'eo': 0, } #morgen/abermorgen does not work, see http://code.google.com/p/parsedatetime/issues/detail?id=19 dayoffsets = { u'morgen': 1, u'heute': 0, u'gestern': -1, u'vorgestern': -2, u'\xfcbermorgen': 2, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { u'mittag': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittags': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittagessen': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'morgen': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'morgens': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'fr\e4hst\xe4ck': { 'hr': 8, 'mn': 0, 'sec': 0 }, u'abendessen': { 'hr': 19, 'mn': 0, 'sec': 0 }, u'abend': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'abends': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'mitternacht': { 'hr': 0, 'mn': 0, 'sec': 0 }, u'nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'nachts': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute abend': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'feierabend': { 'hr': 17, 'mn': 0, 'sec': 0 }, } pdtLocales = { 'en_US': pdtLocale_en, 'en_AU': pdtLocale_au, 'es_ES': pdtLocale_es, 'de_DE': pdtLocale_de, } def _initSymbols(ptc): """ Helper function to initialize the single character constants and other symbols needed. """ ptc.timeSep = [ u':' ] ptc.dateSep = [ u'/' ] ptc.meridian = [ u'AM', u'PM' ] ptc.usesMeridian = True ptc.uses24 = False if pyicu and ptc.usePyICU: am = u'' pm = u'' ts = '' # ICU doesn't seem to provide directly the # date or time seperator - so we have to # figure it out o = ptc.icu_tf['short'] s = ptc.timeFormats['short'] ptc.usesMeridian = u'a' in s ptc.uses24 = u'H' in s # '11:45 AM' or '11:45' s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) # ': AM' or ':' s = s.replace('11', '').replace('45', '') if len(s) > 0: ts = s[0] if ptc.usesMeridian: # '23:45 AM' or '23:45' am = s[1:].strip() s = o.format(datetime.datetime(2003, 10, 30, 23, 45)) if ptc.uses24: s = s.replace('23', '') else: s = s.replace('11', '') # 'PM' or '' pm = s.replace('45', '').replace(ts, '').strip() ptc.timeSep = [ ts ] ptc.meridian = [ am, pm ] o = ptc.icu_df['short'] s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) s = s.replace('10', '').replace('30', '').replace('03', '').replace('2003', '') if len(s) > 0: ds = s[0] else: ds = '/' ptc.dateSep = [ ds ] s = ptc.dateFormats['short'] l = s.lower().split(ds) dp_order = [] for s in l: if len(s) > 0: dp_order.append(s[:1]) ptc.dp_order = dp_order else: ptc.timeSep = ptc.locale.timeSep ptc.dateSep = ptc.locale.dateSep ptc.meridian = ptc.locale.meridian ptc.usesMeridian = ptc.locale.usesMeridian ptc.uses24 = ptc.locale.uses24 ptc.dp_order = ptc.locale.dp_order # build am and pm lists to contain # original case, lowercase and first-char # versions of the meridian text if len(ptc.meridian) > 0: am = ptc.meridian[0] ptc.am = [ am ] if len(am) > 0: ptc.am.append(am[0]) am = am.lower() ptc.am.append(am) ptc.am.append(am[0]) else: am = '' ptc.am = [ '', '' ] if len(ptc.meridian) > 1: pm = ptc.meridian[1] ptc.pm = [ pm ] if len(pm) > 0: ptc.pm.append(pm[0]) pm = pm.lower() ptc.pm.append(pm) ptc.pm.append(pm[0]) else: pm = '' ptc.pm = [ '', '' ] def _initPatterns(ptc): """ Helper function to take the different localized bits from ptc and create the regex strings. """ # TODO add code to parse the date formats and build the regexes up from sub-parts # TODO find all hard-coded uses of date/time seperators ptc.RE_DATE4 = r'''(?P<date>(((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?(,)?(\s)?) (?P<mthname>(%(months)s|%(shortmonths)s))\s? (?P<year>\d\d(\d\d)?)? ) )''' % ptc.re_values # I refactored DATE3 to fix Issue 16 http://code.google.com/p/parsedatetime/issues/detail?id=16 # I suspect the final line was for a trailing time - but testing shows it's not needed # ptc.RE_DATE3 = r'''(?P<date>((?P<mthname>(%(months)s|%(shortmonths)s))\s? # ((?P<day>\d\d?)(\s?|%(daysuffix)s|$)+)? # (,\s?(?P<year>\d\d(\d\d)?))?)) # (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DATE3 = r'''(?P<date>( (((?P<mthname>(%(months)s|%(shortmonths)s))| ((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?))(\s)?){1,2} ((,)?(\s)?(?P<year>\d\d(\d\d)?))? ) )''' % ptc.re_values ptc.RE_MONTH = r'''(\s?|^) (?P<month>( (?P<mthname>(%(months)s|%(shortmonths)s)) (\s?(?P<year>(\d\d\d\d)))? )) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_WEEKDAY = r'''(\s?|^) (?P<weekday>(%(days)s|%(shortdays)s)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_SPECIAL = r'(?P<special>^[%(specials)s]+)\s+' % ptc.re_values ptc.RE_UNITS = r'''(?P<qty>(-?\d+\s* (?P<units>((%(units)s)s?)) ))''' % ptc.re_values ptc.RE_QUNITS = r'''(?P<qty>(-?\d+\s? (?P<qunits>%(qunits)s) (\s?|,|$) ))''' % ptc.re_values ptc.RE_MODIFIER = r'''(\s?|^) (?P<modifier> (previous|prev|last|next|eod|eo|(end\sof)|(in\sa)))''' % ptc.re_values ptc.RE_MODIFIER2 = r'''(\s?|^) (?P<modifier> (from|before|after|ago|prior)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_TIMEHMS = r'''(\s?|^) (?P<hours>\d\d?) (?P<tsep>%(timeseperator)s|) (?P<minutes>\d\d) (?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?''' % ptc.re_values ptc.RE_TIMEHMS2 = r'''(?P<hours>(\d\d?)) ((?P<tsep>%(timeseperator)s|) (?P<minutes>(\d\d?)) (?:(?P=tsep) (?P<seconds>\d\d? (?:[.,]\d+)?))?)?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_TIMEHMS2 += r'\s?(?P<meridian>(%(meridian)s))' % ptc.re_values dateSeps = ''.join(ptc.dateSep) + '.' ptc.RE_DATE = r'''(\s?|^) (?P<date>(\d\d?[%s]\d\d?([%s]\d\d(\d\d)?)?)) (\s?|$|[^0-9a-zA-Z])''' % (dateSeps, dateSeps) ptc.RE_DATE2 = r'[%s]' % dateSeps ptc.RE_DAY = r'''(\s?|^) (?P<day>(today|tomorrow|yesterday)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DAY2 = r'''(?P<day>\d\d?)|(?P<suffix>%(daysuffix)s) ''' % ptc.re_values ptc.RE_TIME = r'''(\s?|^) (?P<time>(morning|breakfast|noon|lunch|evening|midnight|tonight|dinner|night|now)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_REMAINING = r'\s+' # Regex for date/time ranges ptc.RE_RTIMEHMS = r'''(\s?|^) (\d\d?)%(timeseperator)s (\d\d) (%(timeseperator)s(\d\d))? (\s?|$)''' % ptc.re_values ptc.RE_RTIMEHMS2 = r'''(\s?|^) (\d\d?) (%(timeseperator)s(\d\d?))? (%(timeseperator)s(\d\d?))?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_RTIMEHMS2 += r'\s?(%(meridian)s)' % ptc.re_values ptc.RE_RDATE = r'(\d+([%s]\d+)+)' % dateSeps ptc.RE_RDATE3 = r'''((((%(months)s))\s? ((\d\d?) (\s?|%(daysuffix)s|$)+)? (,\s?\d\d\d\d)?))''' % ptc.re_values # "06/07/06 - 08/09/06" ptc.DATERNG1 = ptc.RE_RDATE + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE ptc.DATERNG1 = ptc.DATERNG1 % ptc.re_values # "march 31 - june 1st, 2006" ptc.DATERNG2 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE3 ptc.DATERNG2 = ptc.DATERNG2 % ptc.re_values # "march 1rd -13th" ptc.DATERNG3 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?(\d\d?)\s?(rd|st|nd|th)?' ptc.DATERNG3 = ptc.DATERNG3 % ptc.re_values # "4:00:55 pm - 5:90:44 am", '4p-5p' ptc.TIMERNG1 = ptc.RE_RTIMEHMS2 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG1 = ptc.TIMERNG1 % ptc.re_values # "4:00 - 5:90 ", "4:55:55-3:44:55" ptc.TIMERNG2 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS ptc.TIMERNG2 = ptc.TIMERNG2 % ptc.re_values # "4-5pm " ptc.TIMERNG3 = r'\d\d?\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG3 = ptc.TIMERNG3 % ptc.re_values # "4:30-5pm " ptc.TIMERNG4 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG4 = ptc.TIMERNG4 % ptc.re_values def _initConstants(ptc): """ Create localized versions of the units, week and month names """ # build weekday offsets - yes, it assumes the Weekday and shortWeekday # lists are in the same order and Mon..Sun (Python style) ptc.WeekdayOffsets = {} o = 0 for key in ptc.Weekdays: ptc.WeekdayOffsets[key] = o o += 1 o = 0 for key in ptc.shortWeekdays: ptc.WeekdayOffsets[key] = o o += 1 # build month offsets - yes, it assumes the Months and shortMonths # lists are in the same order and Jan..Dec ptc.MonthOffsets = {} o = 1 for key in ptc.Months: ptc.MonthOffsets[key] = o o += 1 o = 1 for key in ptc.shortMonths: ptc.MonthOffsets[key] = o o += 1 # ptc.DaySuffixes = ptc.re_consts['daysuffix'].split('|') class Constants: """ Default set of constants for parsedatetime. If PyICU is present, then the class will first try to get PyICU to return a locale specified by C{localeID}. If either C{localeID} is None or if the locale does not exist within PyICU, then each of the locales defined in C{fallbackLocales} is tried in order. If PyICU is not present or none of the specified locales can be used, then the class will initialize itself to the en_US locale. if PyICU is not present or not requested, only the locales defined by C{pdtLocales} will be searched. """ def __init__(self, localeID=None, usePyICU=True, fallbackLocales=['en_US']): self.localeID = localeID self.fallbackLocales = fallbackLocales if 'en_US' not in self.fallbackLocales: self.fallbackLocales.append('en_US') # define non-locale specific constants self.locale = None self.usePyICU = usePyICU # starting cache of leap years # daysInMonth will add to this if during # runtime it gets a request for a year not found self._leapYears = [ 1904, 1908, 1912, 1916, 1920, 1924, 1928, 1932, 1936, 1940, 1944, 1948, 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024, 2028, 2032, 2036, 2040, 2044, 2048, 2052, 2056, 2060, 2064, 2068, 2072, 2076, 2080, 2084, 2088, 2092, 2096 ] self.Second = 1 self.Minute = 60 * self.Second self.Hour = 60 * self.Minute self.Day = 24 * self.Hour self.Week = 7 * self.Day self.Month = 30 * self.Day self.Year = 365 * self.Day self.rangeSep = u'-' self._DaysInMonthList = (31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) self.BirthdayEpoch = 50 # DOWParseStyle controls how we parse "Tuesday" # If the current day was Thursday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Current day marked as *** # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current -1,0 *** # week +1 +1 # # If the current day was Monday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 -1 # current *** 0,+1 # week +1 self.DOWParseStyle = 1 # CurrentDOWParseStyle controls how we parse "Friday" # If the current day was Friday and the text to parse is "Friday" # then the following table shows how each style would be returned # True/False. This also depends on DOWParseStyle. # # Current day marked as *** # # DOWParseStyle = 0 # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T,F # week +1 # # DOWParseStyle = -1 # Sun Mon Tue Wed Thu Fri Sat # week -1 F # current T # week +1 # # DOWParseStyle = +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T # week +1 F self.CurrentDOWParseStyle = False # initalize attributes to empty values to ensure # they are defined self.re_sources = None self.re_values = None self.Modifiers = None self.dayOffsets = None self.WeekdayOffsets = None self.MonthOffsets = None self.dateSep = None self.timeSep = None self.am = None self.pm = None self.meridian = None self.usesMeridian = None self.uses24 = None self.dp_order = None self.RE_DATE4 = r'' self.RE_DATE3 = r'' self.RE_MONTH = r'' self.RE_WEEKDAY = r'' self.RE_SPECIAL = r'' self.RE_UNITS = r'' self.RE_QUNITS = r'' self.RE_MODIFIER = r'' self.RE_MODIFIER2 = r'' self.RE_TIMEHMS = r'' self.RE_TIMEHMS2 = r'' self.RE_DATE = r'' self.RE_DATE2 = r'' self.RE_DAY = r'' self.RE_DAY2 = r'' self.RE_TIME = r'' self.RE_REMAINING = r'' self.RE_RTIMEHMS = r'' self.RE_RTIMEHMS2 = r'' self.RE_RDATE = r'' self.RE_RDATE3 = r'' self.DATERNG1 = r'' self.DATERNG2 = r'' self.DATERNG3 = r'' self.TIMERNG1 = r'' self.TIMERNG2 = r'' self.TIMERNG3 = r'' self.TIMERNG4 = r'' _initLocale(self) _initConstants(self) _initSymbols(self) _initPatterns(self) self.re_option = re.IGNORECASE + re.VERBOSE self.cre_source = { 'CRE_SPECIAL': self.RE_SPECIAL, 'CRE_UNITS': self.RE_UNITS, 'CRE_QUNITS': self.RE_QUNITS, 'CRE_MODIFIER': self.RE_MODIFIER, 'CRE_MODIFIER2': self.RE_MODIFIER2, 'CRE_TIMEHMS': self.RE_TIMEHMS, 'CRE_TIMEHMS2': self.RE_TIMEHMS2, 'CRE_DATE': self.RE_DATE, 'CRE_DATE2': self.RE_DATE2, 'CRE_DATE3': self.RE_DATE3, 'CRE_DATE4': self.RE_DATE4, 'CRE_MONTH': self.RE_MONTH, 'CRE_WEEKDAY': self.RE_WEEKDAY, 'CRE_DAY': self.RE_DAY, 'CRE_DAY2': self.RE_DAY2, 'CRE_TIME': self.RE_TIME, 'CRE_REMAINING': self.RE_REMAINING, 'CRE_RTIMEHMS': self.RE_RTIMEHMS, 'CRE_RTIMEHMS2': self.RE_RTIMEHMS2, 'CRE_RDATE': self.RE_RDATE, 'CRE_RDATE3': self.RE_RDATE3, 'CRE_TIMERNG1': self.TIMERNG1, 'CRE_TIMERNG2': self.TIMERNG2, 'CRE_TIMERNG3': self.TIMERNG3, 'CRE_TIMERNG4': self.TIMERNG4, 'CRE_DATERNG1': self.DATERNG1, 'CRE_DATERNG2': self.DATERNG2, 'CRE_DATERNG3': self.DATERNG3, } self.cre_keys = self.cre_source.keys() def __getattr__(self, name): if name in self.cre_keys: value = re.compile(self.cre_source[name], self.re_option) setattr(self, name, value) return value else: raise AttributeError, name def daysInMonth(self, month, year): """ Take the given month (1-12) and a given year (4 digit) return the number of days in the month adjusting for leap year as needed """ result = None if month > 0 and month <= 12: result = self._DaysInMonthList[month - 1] if month == 2: if year in self._leapYears: result += 1 else: if calendar.isleap(year): self._leapYears.append(year) result += 1 return result def buildSources(self, sourceTime=None): """ Return a dictionary of date/time tuples based on the keys found in self.re_sources. The current time is used as the default and any specified item found in self.re_sources is inserted into the value and the generated dictionary is returned. """ if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime sources = {} defaults = { 'yr': yr, 'mth': mth, 'dy': dy, 'hr': hr, 'mn': mn, 'sec': sec, } for item in self.re_sources: values = {} source = self.re_sources[item] for key in defaults.keys(): if key in source: values[key] = source[key] else: values[key] = defaults[key] sources[item] = ( values['yr'], values['mth'], values['dy'], values['hr'], values['mn'], values['sec'], wd, yd, isdst ) return sources
vilmibm/done
parsedatetime/parsedatetime_consts.py
_initSymbols
python
def _initSymbols(ptc): ptc.timeSep = [ u':' ] ptc.dateSep = [ u'/' ] ptc.meridian = [ u'AM', u'PM' ] ptc.usesMeridian = True ptc.uses24 = False if pyicu and ptc.usePyICU: am = u'' pm = u'' ts = '' # ICU doesn't seem to provide directly the # date or time seperator - so we have to # figure it out o = ptc.icu_tf['short'] s = ptc.timeFormats['short'] ptc.usesMeridian = u'a' in s ptc.uses24 = u'H' in s # '11:45 AM' or '11:45' s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) # ': AM' or ':' s = s.replace('11', '').replace('45', '') if len(s) > 0: ts = s[0] if ptc.usesMeridian: # '23:45 AM' or '23:45' am = s[1:].strip() s = o.format(datetime.datetime(2003, 10, 30, 23, 45)) if ptc.uses24: s = s.replace('23', '') else: s = s.replace('11', '') # 'PM' or '' pm = s.replace('45', '').replace(ts, '').strip() ptc.timeSep = [ ts ] ptc.meridian = [ am, pm ] o = ptc.icu_df['short'] s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) s = s.replace('10', '').replace('30', '').replace('03', '').replace('2003', '') if len(s) > 0: ds = s[0] else: ds = '/' ptc.dateSep = [ ds ] s = ptc.dateFormats['short'] l = s.lower().split(ds) dp_order = [] for s in l: if len(s) > 0: dp_order.append(s[:1]) ptc.dp_order = dp_order else: ptc.timeSep = ptc.locale.timeSep ptc.dateSep = ptc.locale.dateSep ptc.meridian = ptc.locale.meridian ptc.usesMeridian = ptc.locale.usesMeridian ptc.uses24 = ptc.locale.uses24 ptc.dp_order = ptc.locale.dp_order # build am and pm lists to contain # original case, lowercase and first-char # versions of the meridian text if len(ptc.meridian) > 0: am = ptc.meridian[0] ptc.am = [ am ] if len(am) > 0: ptc.am.append(am[0]) am = am.lower() ptc.am.append(am) ptc.am.append(am[0]) else: am = '' ptc.am = [ '', '' ] if len(ptc.meridian) > 1: pm = ptc.meridian[1] ptc.pm = [ pm ] if len(pm) > 0: ptc.pm.append(pm[0]) pm = pm.lower() ptc.pm.append(pm) ptc.pm.append(pm[0]) else: pm = '' ptc.pm = [ '', '' ]
Helper function to initialize the single character constants and other symbols needed.
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime_consts.py#L601-L707
null
#!/usr/bin/env python """ parsedatetime constants and helper functions to determine regex values from Locale information if present. Also contains the internal Locale classes to give some sane defaults if PyICU is not found. """ __license__ = """ Copyright (c) 2004-2008 Mike Taylor Copyright (c) 2006-2008 Darshana Chhajed Copyright (c) 2007 Bernd Zeimetz <bzed@debian.org> All rights reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ try: import PyICU as pyicu except: pyicu = None import datetime import calendar import time import re class pdtLocale_en: """ en_US Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for USA """ localeID = 'en_US' # don't use a unicode string dateSep = [ u'/', u'.' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, MMMM d, yyyy', 'long': 'MMMM d, yyyy', 'medium': 'MMM d, yyyy', 'short': 'M/d/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a z', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'm', u'd', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': -1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eod': 0, 'eo': 0 } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_au: """ en_AU Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for Australia """ localeID = 'en_AU' # don't use a unicode string dateSep = [ u'-', u'/' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, d MMMM yyyy', 'long': 'd MMMM yyyy', 'medium': 'dd/MM/yyyy', 'short': 'd/MM/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_es: """ es Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings in Spanish Note that I don't speak Spanish so many of the items below are still in English """ localeID = 'es' # don't use a unicode string dateSep = [ u'/' ] timeSep = [ u':' ] meridian = [] usesMeridian = False uses24 = True Weekdays = [ u'lunes', u'martes', u'mi\xe9rcoles', u'jueves', u'viernes', u's\xe1bado', u'domingo', ] shortWeekdays = [ u'lun', u'mar', u'mi\xe9', u'jue', u'vie', u's\xe1b', u'dom', ] Months = [ u'enero', u'febrero', u'marzo', u'abril', u'mayo', u'junio', u'julio', u'agosto', u'septiembre', u'octubre', u'noviembre', u'diciembre' ] shortMonths = [ u'ene', u'feb', u'mar', u'abr', u'may', u'jun', u'jul', u'ago', u'sep', u'oct', u'nov', u'dic' ] dateFormats = { 'full': "EEEE d' de 'MMMM' de 'yyyy", 'long': "d' de 'MMMM' de 'yyyy", 'medium': "dd-MMM-yy", 'short': "d/MM/yy", } timeFormats = { 'full': "HH'H'mm' 'ss z", 'long': "HH:mm:ss z", 'medium': "HH:mm:ss", 'short': "HH:mm", } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': timeSep, 'dateseperator': dateSep, 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_de: """ de_DE Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Contributed by Debian parsedatetime package maintainer Bernd Zeimetz <bzed@debian.org> Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for German """ localeID = 'de_DE' # don't use a unicode string dateSep = [ u'.' ] timeSep = [ u':' ] meridian = [ ] usesMeridian = False uses24 = True Weekdays = [ u'montag', u'dienstag', u'mittwoch', u'donnerstag', u'freitag', u'samstag', u'sonntag', ] shortWeekdays = [ u'mo', u'di', u'mi', u'do', u'fr', u'sa', u'so', ] Months = [ u'januar', u'februar', u'm\xe4rz', u'april', u'mai', u'juni', u'juli', u'august', u'september', u'oktober', u'november', u'dezember', ] shortMonths = [ u'jan', u'feb', u'mrz', u'apr', u'mai', u'jun', u'jul', u'aug', u'sep', u'okt', u'nov', u'dez', ] dateFormats = { 'full': u'EEEE, d. MMMM yyyy', 'long': u'd. MMMM yyyy', 'medium': u'dd.MM.yyyy', 'short': u'dd.MM.yy' } timeFormats = { 'full': u'HH:mm:ss v', 'long': u'HH:mm:ss z', 'medium': u'HH:mm:ss', 'short': u'HH:mm' } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'sekunden', 'sek', 's' ], 'minutes': [ 'minuten', 'min' , 'm' ], 'hours': [ 'stunden', 'std', 'h' ], 'days': [ 'tage', 't' ], 'weeks': [ 'wochen', 'w' ], 'months': [ 'monate' ], #the short version would be a capital M, #as I understand it we can't distinguis #between m for minutes and M for months. 'years': [ 'jahre', 'j' ], } # text constants to be used by regex's later re_consts = { 'specials': 'am|dem|der|im|in|den|zum', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': '', 'qunits': 'h|m|s|t|w|m|j', 'now': [ 'jetzt' ], } # Used to adjust the returned date before/after the source #still looking for insight on how to translate all of them to german. modifiers = { u'from': 1, u'before': -1, u'after': 1, u'vergangener': -1, u'vorheriger': -1, u'prev': -1, u'letzter': -1, u'n\xe4chster': 1, u'dieser': 0, u'previous': -1, u'in a': 2, u'end of': 0, u'eod': 0, u'eo': 0, } #morgen/abermorgen does not work, see http://code.google.com/p/parsedatetime/issues/detail?id=19 dayoffsets = { u'morgen': 1, u'heute': 0, u'gestern': -1, u'vorgestern': -2, u'\xfcbermorgen': 2, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { u'mittag': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittags': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittagessen': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'morgen': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'morgens': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'fr\e4hst\xe4ck': { 'hr': 8, 'mn': 0, 'sec': 0 }, u'abendessen': { 'hr': 19, 'mn': 0, 'sec': 0 }, u'abend': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'abends': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'mitternacht': { 'hr': 0, 'mn': 0, 'sec': 0 }, u'nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'nachts': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute abend': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'feierabend': { 'hr': 17, 'mn': 0, 'sec': 0 }, } pdtLocales = { 'en_US': pdtLocale_en, 'en_AU': pdtLocale_au, 'es_ES': pdtLocale_es, 'de_DE': pdtLocale_de, } def _initLocale(ptc): """ Helper function to initialize the different lists and strings from either PyICU or one of the internal pdt Locales and store them into ptc. """ def lcase(x): return x.lower() if pyicu and ptc.usePyICU: ptc.icuLocale = None if ptc.localeID is not None: ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is None: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is not None: break ptc.icuSymbols = pyicu.DateFormatSymbols(ptc.icuLocale) # grab ICU list of weekdays, skipping first entry which # is always blank wd = map(lcase, ptc.icuSymbols.getWeekdays()[1:]) swd = map(lcase, ptc.icuSymbols.getShortWeekdays()[1:]) # store them in our list with Monday first (ICU puts Sunday first) ptc.Weekdays = wd[1:] + wd[0:1] ptc.shortWeekdays = swd[1:] + swd[0:1] ptc.Months = map(lcase, ptc.icuSymbols.getMonths()) ptc.shortMonths = map(lcase, ptc.icuSymbols.getShortMonths()) # not quite sure how to init this so for now # set it to none so it will be set to the en_US defaults for now ptc.re_consts = None ptc.icu_df = { 'full': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.icu_tf = { 'full': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.dateFormats = { 'full': ptc.icu_df['full'].toPattern(), 'long': ptc.icu_df['long'].toPattern(), 'medium': ptc.icu_df['medium'].toPattern(), 'short': ptc.icu_df['short'].toPattern(), } ptc.timeFormats = { 'full': ptc.icu_tf['full'].toPattern(), 'long': ptc.icu_tf['long'].toPattern(), 'medium': ptc.icu_tf['medium'].toPattern(), 'short': ptc.icu_tf['short'].toPattern(), } else: if not ptc.localeID in pdtLocales: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] if ptc.localeID in pdtLocales: break ptc.locale = pdtLocales[ptc.localeID] ptc.usePyICU = False ptc.Weekdays = ptc.locale.Weekdays ptc.shortWeekdays = ptc.locale.shortWeekdays ptc.Months = ptc.locale.Months ptc.shortMonths = ptc.locale.shortMonths ptc.dateFormats = ptc.locale.dateFormats ptc.timeFormats = ptc.locale.timeFormats # these values are used to setup the various bits # of the regex values used to parse # # check if a local set of constants has been # provided, if not use en_US as the default if ptc.localeID in pdtLocales: ptc.re_sources = pdtLocales[ptc.localeID].re_sources ptc.re_values = pdtLocales[ptc.localeID].re_consts units = pdtLocales[ptc.localeID].units ptc.Modifiers = pdtLocales[ptc.localeID].modifiers ptc.dayOffsets = pdtLocales[ptc.localeID].dayoffsets # for now, pull over any missing keys from the US set for key in pdtLocales['en_US'].re_consts: if not key in ptc.re_values: ptc.re_values[key] = pdtLocales['en_US'].re_consts[key] else: ptc.re_sources = pdtLocales['en_US'].re_sources ptc.re_values = pdtLocales['en_US'].re_consts ptc.Modifiers = pdtLocales['en_US'].modifiers ptc.dayOffsets = pdtLocales['en_US'].dayoffsets units = pdtLocales['en_US'].units # escape any regex special characters that may be found wd = tuple(map(re.escape, ptc.Weekdays)) swd = tuple(map(re.escape, ptc.shortWeekdays)) mth = tuple(map(re.escape, ptc.Months)) smth = tuple(map(re.escape, ptc.shortMonths)) ptc.re_values['months'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % mth ptc.re_values['shortmonths'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % smth ptc.re_values['days'] = '%s|%s|%s|%s|%s|%s|%s' % wd ptc.re_values['shortdays'] = '%s|%s|%s|%s|%s|%s|%s' % swd l = [] for unit in units: l.append('|'.join(units[unit])) ptc.re_values['units'] = '|'.join(l) ptc.Units = ptc.re_values['units'].split('|') def _initPatterns(ptc): """ Helper function to take the different localized bits from ptc and create the regex strings. """ # TODO add code to parse the date formats and build the regexes up from sub-parts # TODO find all hard-coded uses of date/time seperators ptc.RE_DATE4 = r'''(?P<date>(((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?(,)?(\s)?) (?P<mthname>(%(months)s|%(shortmonths)s))\s? (?P<year>\d\d(\d\d)?)? ) )''' % ptc.re_values # I refactored DATE3 to fix Issue 16 http://code.google.com/p/parsedatetime/issues/detail?id=16 # I suspect the final line was for a trailing time - but testing shows it's not needed # ptc.RE_DATE3 = r'''(?P<date>((?P<mthname>(%(months)s|%(shortmonths)s))\s? # ((?P<day>\d\d?)(\s?|%(daysuffix)s|$)+)? # (,\s?(?P<year>\d\d(\d\d)?))?)) # (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DATE3 = r'''(?P<date>( (((?P<mthname>(%(months)s|%(shortmonths)s))| ((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?))(\s)?){1,2} ((,)?(\s)?(?P<year>\d\d(\d\d)?))? ) )''' % ptc.re_values ptc.RE_MONTH = r'''(\s?|^) (?P<month>( (?P<mthname>(%(months)s|%(shortmonths)s)) (\s?(?P<year>(\d\d\d\d)))? )) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_WEEKDAY = r'''(\s?|^) (?P<weekday>(%(days)s|%(shortdays)s)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_SPECIAL = r'(?P<special>^[%(specials)s]+)\s+' % ptc.re_values ptc.RE_UNITS = r'''(?P<qty>(-?\d+\s* (?P<units>((%(units)s)s?)) ))''' % ptc.re_values ptc.RE_QUNITS = r'''(?P<qty>(-?\d+\s? (?P<qunits>%(qunits)s) (\s?|,|$) ))''' % ptc.re_values ptc.RE_MODIFIER = r'''(\s?|^) (?P<modifier> (previous|prev|last|next|eod|eo|(end\sof)|(in\sa)))''' % ptc.re_values ptc.RE_MODIFIER2 = r'''(\s?|^) (?P<modifier> (from|before|after|ago|prior)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_TIMEHMS = r'''(\s?|^) (?P<hours>\d\d?) (?P<tsep>%(timeseperator)s|) (?P<minutes>\d\d) (?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?''' % ptc.re_values ptc.RE_TIMEHMS2 = r'''(?P<hours>(\d\d?)) ((?P<tsep>%(timeseperator)s|) (?P<minutes>(\d\d?)) (?:(?P=tsep) (?P<seconds>\d\d? (?:[.,]\d+)?))?)?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_TIMEHMS2 += r'\s?(?P<meridian>(%(meridian)s))' % ptc.re_values dateSeps = ''.join(ptc.dateSep) + '.' ptc.RE_DATE = r'''(\s?|^) (?P<date>(\d\d?[%s]\d\d?([%s]\d\d(\d\d)?)?)) (\s?|$|[^0-9a-zA-Z])''' % (dateSeps, dateSeps) ptc.RE_DATE2 = r'[%s]' % dateSeps ptc.RE_DAY = r'''(\s?|^) (?P<day>(today|tomorrow|yesterday)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DAY2 = r'''(?P<day>\d\d?)|(?P<suffix>%(daysuffix)s) ''' % ptc.re_values ptc.RE_TIME = r'''(\s?|^) (?P<time>(morning|breakfast|noon|lunch|evening|midnight|tonight|dinner|night|now)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_REMAINING = r'\s+' # Regex for date/time ranges ptc.RE_RTIMEHMS = r'''(\s?|^) (\d\d?)%(timeseperator)s (\d\d) (%(timeseperator)s(\d\d))? (\s?|$)''' % ptc.re_values ptc.RE_RTIMEHMS2 = r'''(\s?|^) (\d\d?) (%(timeseperator)s(\d\d?))? (%(timeseperator)s(\d\d?))?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_RTIMEHMS2 += r'\s?(%(meridian)s)' % ptc.re_values ptc.RE_RDATE = r'(\d+([%s]\d+)+)' % dateSeps ptc.RE_RDATE3 = r'''((((%(months)s))\s? ((\d\d?) (\s?|%(daysuffix)s|$)+)? (,\s?\d\d\d\d)?))''' % ptc.re_values # "06/07/06 - 08/09/06" ptc.DATERNG1 = ptc.RE_RDATE + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE ptc.DATERNG1 = ptc.DATERNG1 % ptc.re_values # "march 31 - june 1st, 2006" ptc.DATERNG2 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE3 ptc.DATERNG2 = ptc.DATERNG2 % ptc.re_values # "march 1rd -13th" ptc.DATERNG3 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?(\d\d?)\s?(rd|st|nd|th)?' ptc.DATERNG3 = ptc.DATERNG3 % ptc.re_values # "4:00:55 pm - 5:90:44 am", '4p-5p' ptc.TIMERNG1 = ptc.RE_RTIMEHMS2 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG1 = ptc.TIMERNG1 % ptc.re_values # "4:00 - 5:90 ", "4:55:55-3:44:55" ptc.TIMERNG2 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS ptc.TIMERNG2 = ptc.TIMERNG2 % ptc.re_values # "4-5pm " ptc.TIMERNG3 = r'\d\d?\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG3 = ptc.TIMERNG3 % ptc.re_values # "4:30-5pm " ptc.TIMERNG4 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG4 = ptc.TIMERNG4 % ptc.re_values def _initConstants(ptc): """ Create localized versions of the units, week and month names """ # build weekday offsets - yes, it assumes the Weekday and shortWeekday # lists are in the same order and Mon..Sun (Python style) ptc.WeekdayOffsets = {} o = 0 for key in ptc.Weekdays: ptc.WeekdayOffsets[key] = o o += 1 o = 0 for key in ptc.shortWeekdays: ptc.WeekdayOffsets[key] = o o += 1 # build month offsets - yes, it assumes the Months and shortMonths # lists are in the same order and Jan..Dec ptc.MonthOffsets = {} o = 1 for key in ptc.Months: ptc.MonthOffsets[key] = o o += 1 o = 1 for key in ptc.shortMonths: ptc.MonthOffsets[key] = o o += 1 # ptc.DaySuffixes = ptc.re_consts['daysuffix'].split('|') class Constants: """ Default set of constants for parsedatetime. If PyICU is present, then the class will first try to get PyICU to return a locale specified by C{localeID}. If either C{localeID} is None or if the locale does not exist within PyICU, then each of the locales defined in C{fallbackLocales} is tried in order. If PyICU is not present or none of the specified locales can be used, then the class will initialize itself to the en_US locale. if PyICU is not present or not requested, only the locales defined by C{pdtLocales} will be searched. """ def __init__(self, localeID=None, usePyICU=True, fallbackLocales=['en_US']): self.localeID = localeID self.fallbackLocales = fallbackLocales if 'en_US' not in self.fallbackLocales: self.fallbackLocales.append('en_US') # define non-locale specific constants self.locale = None self.usePyICU = usePyICU # starting cache of leap years # daysInMonth will add to this if during # runtime it gets a request for a year not found self._leapYears = [ 1904, 1908, 1912, 1916, 1920, 1924, 1928, 1932, 1936, 1940, 1944, 1948, 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024, 2028, 2032, 2036, 2040, 2044, 2048, 2052, 2056, 2060, 2064, 2068, 2072, 2076, 2080, 2084, 2088, 2092, 2096 ] self.Second = 1 self.Minute = 60 * self.Second self.Hour = 60 * self.Minute self.Day = 24 * self.Hour self.Week = 7 * self.Day self.Month = 30 * self.Day self.Year = 365 * self.Day self.rangeSep = u'-' self._DaysInMonthList = (31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) self.BirthdayEpoch = 50 # DOWParseStyle controls how we parse "Tuesday" # If the current day was Thursday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Current day marked as *** # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current -1,0 *** # week +1 +1 # # If the current day was Monday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 -1 # current *** 0,+1 # week +1 self.DOWParseStyle = 1 # CurrentDOWParseStyle controls how we parse "Friday" # If the current day was Friday and the text to parse is "Friday" # then the following table shows how each style would be returned # True/False. This also depends on DOWParseStyle. # # Current day marked as *** # # DOWParseStyle = 0 # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T,F # week +1 # # DOWParseStyle = -1 # Sun Mon Tue Wed Thu Fri Sat # week -1 F # current T # week +1 # # DOWParseStyle = +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T # week +1 F self.CurrentDOWParseStyle = False # initalize attributes to empty values to ensure # they are defined self.re_sources = None self.re_values = None self.Modifiers = None self.dayOffsets = None self.WeekdayOffsets = None self.MonthOffsets = None self.dateSep = None self.timeSep = None self.am = None self.pm = None self.meridian = None self.usesMeridian = None self.uses24 = None self.dp_order = None self.RE_DATE4 = r'' self.RE_DATE3 = r'' self.RE_MONTH = r'' self.RE_WEEKDAY = r'' self.RE_SPECIAL = r'' self.RE_UNITS = r'' self.RE_QUNITS = r'' self.RE_MODIFIER = r'' self.RE_MODIFIER2 = r'' self.RE_TIMEHMS = r'' self.RE_TIMEHMS2 = r'' self.RE_DATE = r'' self.RE_DATE2 = r'' self.RE_DAY = r'' self.RE_DAY2 = r'' self.RE_TIME = r'' self.RE_REMAINING = r'' self.RE_RTIMEHMS = r'' self.RE_RTIMEHMS2 = r'' self.RE_RDATE = r'' self.RE_RDATE3 = r'' self.DATERNG1 = r'' self.DATERNG2 = r'' self.DATERNG3 = r'' self.TIMERNG1 = r'' self.TIMERNG2 = r'' self.TIMERNG3 = r'' self.TIMERNG4 = r'' _initLocale(self) _initConstants(self) _initSymbols(self) _initPatterns(self) self.re_option = re.IGNORECASE + re.VERBOSE self.cre_source = { 'CRE_SPECIAL': self.RE_SPECIAL, 'CRE_UNITS': self.RE_UNITS, 'CRE_QUNITS': self.RE_QUNITS, 'CRE_MODIFIER': self.RE_MODIFIER, 'CRE_MODIFIER2': self.RE_MODIFIER2, 'CRE_TIMEHMS': self.RE_TIMEHMS, 'CRE_TIMEHMS2': self.RE_TIMEHMS2, 'CRE_DATE': self.RE_DATE, 'CRE_DATE2': self.RE_DATE2, 'CRE_DATE3': self.RE_DATE3, 'CRE_DATE4': self.RE_DATE4, 'CRE_MONTH': self.RE_MONTH, 'CRE_WEEKDAY': self.RE_WEEKDAY, 'CRE_DAY': self.RE_DAY, 'CRE_DAY2': self.RE_DAY2, 'CRE_TIME': self.RE_TIME, 'CRE_REMAINING': self.RE_REMAINING, 'CRE_RTIMEHMS': self.RE_RTIMEHMS, 'CRE_RTIMEHMS2': self.RE_RTIMEHMS2, 'CRE_RDATE': self.RE_RDATE, 'CRE_RDATE3': self.RE_RDATE3, 'CRE_TIMERNG1': self.TIMERNG1, 'CRE_TIMERNG2': self.TIMERNG2, 'CRE_TIMERNG3': self.TIMERNG3, 'CRE_TIMERNG4': self.TIMERNG4, 'CRE_DATERNG1': self.DATERNG1, 'CRE_DATERNG2': self.DATERNG2, 'CRE_DATERNG3': self.DATERNG3, } self.cre_keys = self.cre_source.keys() def __getattr__(self, name): if name in self.cre_keys: value = re.compile(self.cre_source[name], self.re_option) setattr(self, name, value) return value else: raise AttributeError, name def daysInMonth(self, month, year): """ Take the given month (1-12) and a given year (4 digit) return the number of days in the month adjusting for leap year as needed """ result = None if month > 0 and month <= 12: result = self._DaysInMonthList[month - 1] if month == 2: if year in self._leapYears: result += 1 else: if calendar.isleap(year): self._leapYears.append(year) result += 1 return result def buildSources(self, sourceTime=None): """ Return a dictionary of date/time tuples based on the keys found in self.re_sources. The current time is used as the default and any specified item found in self.re_sources is inserted into the value and the generated dictionary is returned. """ if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime sources = {} defaults = { 'yr': yr, 'mth': mth, 'dy': dy, 'hr': hr, 'mn': mn, 'sec': sec, } for item in self.re_sources: values = {} source = self.re_sources[item] for key in defaults.keys(): if key in source: values[key] = source[key] else: values[key] = defaults[key] sources[item] = ( values['yr'], values['mth'], values['dy'], values['hr'], values['mn'], values['sec'], wd, yd, isdst ) return sources
vilmibm/done
parsedatetime/parsedatetime_consts.py
_initPatterns
python
def _initPatterns(ptc): # TODO add code to parse the date formats and build the regexes up from sub-parts # TODO find all hard-coded uses of date/time seperators ptc.RE_DATE4 = r'''(?P<date>(((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?(,)?(\s)?) (?P<mthname>(%(months)s|%(shortmonths)s))\s? (?P<year>\d\d(\d\d)?)? ) )''' % ptc.re_values # I refactored DATE3 to fix Issue 16 http://code.google.com/p/parsedatetime/issues/detail?id=16 # I suspect the final line was for a trailing time - but testing shows it's not needed # ptc.RE_DATE3 = r'''(?P<date>((?P<mthname>(%(months)s|%(shortmonths)s))\s? # ((?P<day>\d\d?)(\s?|%(daysuffix)s|$)+)? # (,\s?(?P<year>\d\d(\d\d)?))?)) # (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DATE3 = r'''(?P<date>( (((?P<mthname>(%(months)s|%(shortmonths)s))| ((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?))(\s)?){1,2} ((,)?(\s)?(?P<year>\d\d(\d\d)?))? ) )''' % ptc.re_values ptc.RE_MONTH = r'''(\s?|^) (?P<month>( (?P<mthname>(%(months)s|%(shortmonths)s)) (\s?(?P<year>(\d\d\d\d)))? )) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_WEEKDAY = r'''(\s?|^) (?P<weekday>(%(days)s|%(shortdays)s)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_SPECIAL = r'(?P<special>^[%(specials)s]+)\s+' % ptc.re_values ptc.RE_UNITS = r'''(?P<qty>(-?\d+\s* (?P<units>((%(units)s)s?)) ))''' % ptc.re_values ptc.RE_QUNITS = r'''(?P<qty>(-?\d+\s? (?P<qunits>%(qunits)s) (\s?|,|$) ))''' % ptc.re_values ptc.RE_MODIFIER = r'''(\s?|^) (?P<modifier> (previous|prev|last|next|eod|eo|(end\sof)|(in\sa)))''' % ptc.re_values ptc.RE_MODIFIER2 = r'''(\s?|^) (?P<modifier> (from|before|after|ago|prior)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_TIMEHMS = r'''(\s?|^) (?P<hours>\d\d?) (?P<tsep>%(timeseperator)s|) (?P<minutes>\d\d) (?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?''' % ptc.re_values ptc.RE_TIMEHMS2 = r'''(?P<hours>(\d\d?)) ((?P<tsep>%(timeseperator)s|) (?P<minutes>(\d\d?)) (?:(?P=tsep) (?P<seconds>\d\d? (?:[.,]\d+)?))?)?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_TIMEHMS2 += r'\s?(?P<meridian>(%(meridian)s))' % ptc.re_values dateSeps = ''.join(ptc.dateSep) + '.' ptc.RE_DATE = r'''(\s?|^) (?P<date>(\d\d?[%s]\d\d?([%s]\d\d(\d\d)?)?)) (\s?|$|[^0-9a-zA-Z])''' % (dateSeps, dateSeps) ptc.RE_DATE2 = r'[%s]' % dateSeps ptc.RE_DAY = r'''(\s?|^) (?P<day>(today|tomorrow|yesterday)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DAY2 = r'''(?P<day>\d\d?)|(?P<suffix>%(daysuffix)s) ''' % ptc.re_values ptc.RE_TIME = r'''(\s?|^) (?P<time>(morning|breakfast|noon|lunch|evening|midnight|tonight|dinner|night|now)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_REMAINING = r'\s+' # Regex for date/time ranges ptc.RE_RTIMEHMS = r'''(\s?|^) (\d\d?)%(timeseperator)s (\d\d) (%(timeseperator)s(\d\d))? (\s?|$)''' % ptc.re_values ptc.RE_RTIMEHMS2 = r'''(\s?|^) (\d\d?) (%(timeseperator)s(\d\d?))? (%(timeseperator)s(\d\d?))?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_RTIMEHMS2 += r'\s?(%(meridian)s)' % ptc.re_values ptc.RE_RDATE = r'(\d+([%s]\d+)+)' % dateSeps ptc.RE_RDATE3 = r'''((((%(months)s))\s? ((\d\d?) (\s?|%(daysuffix)s|$)+)? (,\s?\d\d\d\d)?))''' % ptc.re_values # "06/07/06 - 08/09/06" ptc.DATERNG1 = ptc.RE_RDATE + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE ptc.DATERNG1 = ptc.DATERNG1 % ptc.re_values # "march 31 - june 1st, 2006" ptc.DATERNG2 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE3 ptc.DATERNG2 = ptc.DATERNG2 % ptc.re_values # "march 1rd -13th" ptc.DATERNG3 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?(\d\d?)\s?(rd|st|nd|th)?' ptc.DATERNG3 = ptc.DATERNG3 % ptc.re_values # "4:00:55 pm - 5:90:44 am", '4p-5p' ptc.TIMERNG1 = ptc.RE_RTIMEHMS2 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG1 = ptc.TIMERNG1 % ptc.re_values # "4:00 - 5:90 ", "4:55:55-3:44:55" ptc.TIMERNG2 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS ptc.TIMERNG2 = ptc.TIMERNG2 % ptc.re_values # "4-5pm " ptc.TIMERNG3 = r'\d\d?\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG3 = ptc.TIMERNG3 % ptc.re_values # "4:30-5pm " ptc.TIMERNG4 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG4 = ptc.TIMERNG4 % ptc.re_values
Helper function to take the different localized bits from ptc and create the regex strings.
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime_consts.py#L710-L838
null
#!/usr/bin/env python """ parsedatetime constants and helper functions to determine regex values from Locale information if present. Also contains the internal Locale classes to give some sane defaults if PyICU is not found. """ __license__ = """ Copyright (c) 2004-2008 Mike Taylor Copyright (c) 2006-2008 Darshana Chhajed Copyright (c) 2007 Bernd Zeimetz <bzed@debian.org> All rights reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ try: import PyICU as pyicu except: pyicu = None import datetime import calendar import time import re class pdtLocale_en: """ en_US Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for USA """ localeID = 'en_US' # don't use a unicode string dateSep = [ u'/', u'.' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, MMMM d, yyyy', 'long': 'MMMM d, yyyy', 'medium': 'MMM d, yyyy', 'short': 'M/d/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a z', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'm', u'd', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': -1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eod': 0, 'eo': 0 } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_au: """ en_AU Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for Australia """ localeID = 'en_AU' # don't use a unicode string dateSep = [ u'-', u'/' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, d MMMM yyyy', 'long': 'd MMMM yyyy', 'medium': 'dd/MM/yyyy', 'short': 'd/MM/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_es: """ es Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings in Spanish Note that I don't speak Spanish so many of the items below are still in English """ localeID = 'es' # don't use a unicode string dateSep = [ u'/' ] timeSep = [ u':' ] meridian = [] usesMeridian = False uses24 = True Weekdays = [ u'lunes', u'martes', u'mi\xe9rcoles', u'jueves', u'viernes', u's\xe1bado', u'domingo', ] shortWeekdays = [ u'lun', u'mar', u'mi\xe9', u'jue', u'vie', u's\xe1b', u'dom', ] Months = [ u'enero', u'febrero', u'marzo', u'abril', u'mayo', u'junio', u'julio', u'agosto', u'septiembre', u'octubre', u'noviembre', u'diciembre' ] shortMonths = [ u'ene', u'feb', u'mar', u'abr', u'may', u'jun', u'jul', u'ago', u'sep', u'oct', u'nov', u'dic' ] dateFormats = { 'full': "EEEE d' de 'MMMM' de 'yyyy", 'long': "d' de 'MMMM' de 'yyyy", 'medium': "dd-MMM-yy", 'short': "d/MM/yy", } timeFormats = { 'full': "HH'H'mm' 'ss z", 'long': "HH:mm:ss z", 'medium': "HH:mm:ss", 'short': "HH:mm", } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': timeSep, 'dateseperator': dateSep, 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_de: """ de_DE Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Contributed by Debian parsedatetime package maintainer Bernd Zeimetz <bzed@debian.org> Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for German """ localeID = 'de_DE' # don't use a unicode string dateSep = [ u'.' ] timeSep = [ u':' ] meridian = [ ] usesMeridian = False uses24 = True Weekdays = [ u'montag', u'dienstag', u'mittwoch', u'donnerstag', u'freitag', u'samstag', u'sonntag', ] shortWeekdays = [ u'mo', u'di', u'mi', u'do', u'fr', u'sa', u'so', ] Months = [ u'januar', u'februar', u'm\xe4rz', u'april', u'mai', u'juni', u'juli', u'august', u'september', u'oktober', u'november', u'dezember', ] shortMonths = [ u'jan', u'feb', u'mrz', u'apr', u'mai', u'jun', u'jul', u'aug', u'sep', u'okt', u'nov', u'dez', ] dateFormats = { 'full': u'EEEE, d. MMMM yyyy', 'long': u'd. MMMM yyyy', 'medium': u'dd.MM.yyyy', 'short': u'dd.MM.yy' } timeFormats = { 'full': u'HH:mm:ss v', 'long': u'HH:mm:ss z', 'medium': u'HH:mm:ss', 'short': u'HH:mm' } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'sekunden', 'sek', 's' ], 'minutes': [ 'minuten', 'min' , 'm' ], 'hours': [ 'stunden', 'std', 'h' ], 'days': [ 'tage', 't' ], 'weeks': [ 'wochen', 'w' ], 'months': [ 'monate' ], #the short version would be a capital M, #as I understand it we can't distinguis #between m for minutes and M for months. 'years': [ 'jahre', 'j' ], } # text constants to be used by regex's later re_consts = { 'specials': 'am|dem|der|im|in|den|zum', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': '', 'qunits': 'h|m|s|t|w|m|j', 'now': [ 'jetzt' ], } # Used to adjust the returned date before/after the source #still looking for insight on how to translate all of them to german. modifiers = { u'from': 1, u'before': -1, u'after': 1, u'vergangener': -1, u'vorheriger': -1, u'prev': -1, u'letzter': -1, u'n\xe4chster': 1, u'dieser': 0, u'previous': -1, u'in a': 2, u'end of': 0, u'eod': 0, u'eo': 0, } #morgen/abermorgen does not work, see http://code.google.com/p/parsedatetime/issues/detail?id=19 dayoffsets = { u'morgen': 1, u'heute': 0, u'gestern': -1, u'vorgestern': -2, u'\xfcbermorgen': 2, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { u'mittag': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittags': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittagessen': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'morgen': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'morgens': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'fr\e4hst\xe4ck': { 'hr': 8, 'mn': 0, 'sec': 0 }, u'abendessen': { 'hr': 19, 'mn': 0, 'sec': 0 }, u'abend': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'abends': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'mitternacht': { 'hr': 0, 'mn': 0, 'sec': 0 }, u'nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'nachts': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute abend': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'feierabend': { 'hr': 17, 'mn': 0, 'sec': 0 }, } pdtLocales = { 'en_US': pdtLocale_en, 'en_AU': pdtLocale_au, 'es_ES': pdtLocale_es, 'de_DE': pdtLocale_de, } def _initLocale(ptc): """ Helper function to initialize the different lists and strings from either PyICU or one of the internal pdt Locales and store them into ptc. """ def lcase(x): return x.lower() if pyicu and ptc.usePyICU: ptc.icuLocale = None if ptc.localeID is not None: ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is None: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is not None: break ptc.icuSymbols = pyicu.DateFormatSymbols(ptc.icuLocale) # grab ICU list of weekdays, skipping first entry which # is always blank wd = map(lcase, ptc.icuSymbols.getWeekdays()[1:]) swd = map(lcase, ptc.icuSymbols.getShortWeekdays()[1:]) # store them in our list with Monday first (ICU puts Sunday first) ptc.Weekdays = wd[1:] + wd[0:1] ptc.shortWeekdays = swd[1:] + swd[0:1] ptc.Months = map(lcase, ptc.icuSymbols.getMonths()) ptc.shortMonths = map(lcase, ptc.icuSymbols.getShortMonths()) # not quite sure how to init this so for now # set it to none so it will be set to the en_US defaults for now ptc.re_consts = None ptc.icu_df = { 'full': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.icu_tf = { 'full': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.dateFormats = { 'full': ptc.icu_df['full'].toPattern(), 'long': ptc.icu_df['long'].toPattern(), 'medium': ptc.icu_df['medium'].toPattern(), 'short': ptc.icu_df['short'].toPattern(), } ptc.timeFormats = { 'full': ptc.icu_tf['full'].toPattern(), 'long': ptc.icu_tf['long'].toPattern(), 'medium': ptc.icu_tf['medium'].toPattern(), 'short': ptc.icu_tf['short'].toPattern(), } else: if not ptc.localeID in pdtLocales: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] if ptc.localeID in pdtLocales: break ptc.locale = pdtLocales[ptc.localeID] ptc.usePyICU = False ptc.Weekdays = ptc.locale.Weekdays ptc.shortWeekdays = ptc.locale.shortWeekdays ptc.Months = ptc.locale.Months ptc.shortMonths = ptc.locale.shortMonths ptc.dateFormats = ptc.locale.dateFormats ptc.timeFormats = ptc.locale.timeFormats # these values are used to setup the various bits # of the regex values used to parse # # check if a local set of constants has been # provided, if not use en_US as the default if ptc.localeID in pdtLocales: ptc.re_sources = pdtLocales[ptc.localeID].re_sources ptc.re_values = pdtLocales[ptc.localeID].re_consts units = pdtLocales[ptc.localeID].units ptc.Modifiers = pdtLocales[ptc.localeID].modifiers ptc.dayOffsets = pdtLocales[ptc.localeID].dayoffsets # for now, pull over any missing keys from the US set for key in pdtLocales['en_US'].re_consts: if not key in ptc.re_values: ptc.re_values[key] = pdtLocales['en_US'].re_consts[key] else: ptc.re_sources = pdtLocales['en_US'].re_sources ptc.re_values = pdtLocales['en_US'].re_consts ptc.Modifiers = pdtLocales['en_US'].modifiers ptc.dayOffsets = pdtLocales['en_US'].dayoffsets units = pdtLocales['en_US'].units # escape any regex special characters that may be found wd = tuple(map(re.escape, ptc.Weekdays)) swd = tuple(map(re.escape, ptc.shortWeekdays)) mth = tuple(map(re.escape, ptc.Months)) smth = tuple(map(re.escape, ptc.shortMonths)) ptc.re_values['months'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % mth ptc.re_values['shortmonths'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % smth ptc.re_values['days'] = '%s|%s|%s|%s|%s|%s|%s' % wd ptc.re_values['shortdays'] = '%s|%s|%s|%s|%s|%s|%s' % swd l = [] for unit in units: l.append('|'.join(units[unit])) ptc.re_values['units'] = '|'.join(l) ptc.Units = ptc.re_values['units'].split('|') def _initSymbols(ptc): """ Helper function to initialize the single character constants and other symbols needed. """ ptc.timeSep = [ u':' ] ptc.dateSep = [ u'/' ] ptc.meridian = [ u'AM', u'PM' ] ptc.usesMeridian = True ptc.uses24 = False if pyicu and ptc.usePyICU: am = u'' pm = u'' ts = '' # ICU doesn't seem to provide directly the # date or time seperator - so we have to # figure it out o = ptc.icu_tf['short'] s = ptc.timeFormats['short'] ptc.usesMeridian = u'a' in s ptc.uses24 = u'H' in s # '11:45 AM' or '11:45' s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) # ': AM' or ':' s = s.replace('11', '').replace('45', '') if len(s) > 0: ts = s[0] if ptc.usesMeridian: # '23:45 AM' or '23:45' am = s[1:].strip() s = o.format(datetime.datetime(2003, 10, 30, 23, 45)) if ptc.uses24: s = s.replace('23', '') else: s = s.replace('11', '') # 'PM' or '' pm = s.replace('45', '').replace(ts, '').strip() ptc.timeSep = [ ts ] ptc.meridian = [ am, pm ] o = ptc.icu_df['short'] s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) s = s.replace('10', '').replace('30', '').replace('03', '').replace('2003', '') if len(s) > 0: ds = s[0] else: ds = '/' ptc.dateSep = [ ds ] s = ptc.dateFormats['short'] l = s.lower().split(ds) dp_order = [] for s in l: if len(s) > 0: dp_order.append(s[:1]) ptc.dp_order = dp_order else: ptc.timeSep = ptc.locale.timeSep ptc.dateSep = ptc.locale.dateSep ptc.meridian = ptc.locale.meridian ptc.usesMeridian = ptc.locale.usesMeridian ptc.uses24 = ptc.locale.uses24 ptc.dp_order = ptc.locale.dp_order # build am and pm lists to contain # original case, lowercase and first-char # versions of the meridian text if len(ptc.meridian) > 0: am = ptc.meridian[0] ptc.am = [ am ] if len(am) > 0: ptc.am.append(am[0]) am = am.lower() ptc.am.append(am) ptc.am.append(am[0]) else: am = '' ptc.am = [ '', '' ] if len(ptc.meridian) > 1: pm = ptc.meridian[1] ptc.pm = [ pm ] if len(pm) > 0: ptc.pm.append(pm[0]) pm = pm.lower() ptc.pm.append(pm) ptc.pm.append(pm[0]) else: pm = '' ptc.pm = [ '', '' ] def _initConstants(ptc): """ Create localized versions of the units, week and month names """ # build weekday offsets - yes, it assumes the Weekday and shortWeekday # lists are in the same order and Mon..Sun (Python style) ptc.WeekdayOffsets = {} o = 0 for key in ptc.Weekdays: ptc.WeekdayOffsets[key] = o o += 1 o = 0 for key in ptc.shortWeekdays: ptc.WeekdayOffsets[key] = o o += 1 # build month offsets - yes, it assumes the Months and shortMonths # lists are in the same order and Jan..Dec ptc.MonthOffsets = {} o = 1 for key in ptc.Months: ptc.MonthOffsets[key] = o o += 1 o = 1 for key in ptc.shortMonths: ptc.MonthOffsets[key] = o o += 1 # ptc.DaySuffixes = ptc.re_consts['daysuffix'].split('|') class Constants: """ Default set of constants for parsedatetime. If PyICU is present, then the class will first try to get PyICU to return a locale specified by C{localeID}. If either C{localeID} is None or if the locale does not exist within PyICU, then each of the locales defined in C{fallbackLocales} is tried in order. If PyICU is not present or none of the specified locales can be used, then the class will initialize itself to the en_US locale. if PyICU is not present or not requested, only the locales defined by C{pdtLocales} will be searched. """ def __init__(self, localeID=None, usePyICU=True, fallbackLocales=['en_US']): self.localeID = localeID self.fallbackLocales = fallbackLocales if 'en_US' not in self.fallbackLocales: self.fallbackLocales.append('en_US') # define non-locale specific constants self.locale = None self.usePyICU = usePyICU # starting cache of leap years # daysInMonth will add to this if during # runtime it gets a request for a year not found self._leapYears = [ 1904, 1908, 1912, 1916, 1920, 1924, 1928, 1932, 1936, 1940, 1944, 1948, 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024, 2028, 2032, 2036, 2040, 2044, 2048, 2052, 2056, 2060, 2064, 2068, 2072, 2076, 2080, 2084, 2088, 2092, 2096 ] self.Second = 1 self.Minute = 60 * self.Second self.Hour = 60 * self.Minute self.Day = 24 * self.Hour self.Week = 7 * self.Day self.Month = 30 * self.Day self.Year = 365 * self.Day self.rangeSep = u'-' self._DaysInMonthList = (31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) self.BirthdayEpoch = 50 # DOWParseStyle controls how we parse "Tuesday" # If the current day was Thursday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Current day marked as *** # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current -1,0 *** # week +1 +1 # # If the current day was Monday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 -1 # current *** 0,+1 # week +1 self.DOWParseStyle = 1 # CurrentDOWParseStyle controls how we parse "Friday" # If the current day was Friday and the text to parse is "Friday" # then the following table shows how each style would be returned # True/False. This also depends on DOWParseStyle. # # Current day marked as *** # # DOWParseStyle = 0 # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T,F # week +1 # # DOWParseStyle = -1 # Sun Mon Tue Wed Thu Fri Sat # week -1 F # current T # week +1 # # DOWParseStyle = +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T # week +1 F self.CurrentDOWParseStyle = False # initalize attributes to empty values to ensure # they are defined self.re_sources = None self.re_values = None self.Modifiers = None self.dayOffsets = None self.WeekdayOffsets = None self.MonthOffsets = None self.dateSep = None self.timeSep = None self.am = None self.pm = None self.meridian = None self.usesMeridian = None self.uses24 = None self.dp_order = None self.RE_DATE4 = r'' self.RE_DATE3 = r'' self.RE_MONTH = r'' self.RE_WEEKDAY = r'' self.RE_SPECIAL = r'' self.RE_UNITS = r'' self.RE_QUNITS = r'' self.RE_MODIFIER = r'' self.RE_MODIFIER2 = r'' self.RE_TIMEHMS = r'' self.RE_TIMEHMS2 = r'' self.RE_DATE = r'' self.RE_DATE2 = r'' self.RE_DAY = r'' self.RE_DAY2 = r'' self.RE_TIME = r'' self.RE_REMAINING = r'' self.RE_RTIMEHMS = r'' self.RE_RTIMEHMS2 = r'' self.RE_RDATE = r'' self.RE_RDATE3 = r'' self.DATERNG1 = r'' self.DATERNG2 = r'' self.DATERNG3 = r'' self.TIMERNG1 = r'' self.TIMERNG2 = r'' self.TIMERNG3 = r'' self.TIMERNG4 = r'' _initLocale(self) _initConstants(self) _initSymbols(self) _initPatterns(self) self.re_option = re.IGNORECASE + re.VERBOSE self.cre_source = { 'CRE_SPECIAL': self.RE_SPECIAL, 'CRE_UNITS': self.RE_UNITS, 'CRE_QUNITS': self.RE_QUNITS, 'CRE_MODIFIER': self.RE_MODIFIER, 'CRE_MODIFIER2': self.RE_MODIFIER2, 'CRE_TIMEHMS': self.RE_TIMEHMS, 'CRE_TIMEHMS2': self.RE_TIMEHMS2, 'CRE_DATE': self.RE_DATE, 'CRE_DATE2': self.RE_DATE2, 'CRE_DATE3': self.RE_DATE3, 'CRE_DATE4': self.RE_DATE4, 'CRE_MONTH': self.RE_MONTH, 'CRE_WEEKDAY': self.RE_WEEKDAY, 'CRE_DAY': self.RE_DAY, 'CRE_DAY2': self.RE_DAY2, 'CRE_TIME': self.RE_TIME, 'CRE_REMAINING': self.RE_REMAINING, 'CRE_RTIMEHMS': self.RE_RTIMEHMS, 'CRE_RTIMEHMS2': self.RE_RTIMEHMS2, 'CRE_RDATE': self.RE_RDATE, 'CRE_RDATE3': self.RE_RDATE3, 'CRE_TIMERNG1': self.TIMERNG1, 'CRE_TIMERNG2': self.TIMERNG2, 'CRE_TIMERNG3': self.TIMERNG3, 'CRE_TIMERNG4': self.TIMERNG4, 'CRE_DATERNG1': self.DATERNG1, 'CRE_DATERNG2': self.DATERNG2, 'CRE_DATERNG3': self.DATERNG3, } self.cre_keys = self.cre_source.keys() def __getattr__(self, name): if name in self.cre_keys: value = re.compile(self.cre_source[name], self.re_option) setattr(self, name, value) return value else: raise AttributeError, name def daysInMonth(self, month, year): """ Take the given month (1-12) and a given year (4 digit) return the number of days in the month adjusting for leap year as needed """ result = None if month > 0 and month <= 12: result = self._DaysInMonthList[month - 1] if month == 2: if year in self._leapYears: result += 1 else: if calendar.isleap(year): self._leapYears.append(year) result += 1 return result def buildSources(self, sourceTime=None): """ Return a dictionary of date/time tuples based on the keys found in self.re_sources. The current time is used as the default and any specified item found in self.re_sources is inserted into the value and the generated dictionary is returned. """ if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime sources = {} defaults = { 'yr': yr, 'mth': mth, 'dy': dy, 'hr': hr, 'mn': mn, 'sec': sec, } for item in self.re_sources: values = {} source = self.re_sources[item] for key in defaults.keys(): if key in source: values[key] = source[key] else: values[key] = defaults[key] sources[item] = ( values['yr'], values['mth'], values['dy'], values['hr'], values['mn'], values['sec'], wd, yd, isdst ) return sources
vilmibm/done
parsedatetime/parsedatetime_consts.py
_initConstants
python
def _initConstants(ptc): # build weekday offsets - yes, it assumes the Weekday and shortWeekday # lists are in the same order and Mon..Sun (Python style) ptc.WeekdayOffsets = {} o = 0 for key in ptc.Weekdays: ptc.WeekdayOffsets[key] = o o += 1 o = 0 for key in ptc.shortWeekdays: ptc.WeekdayOffsets[key] = o o += 1 # build month offsets - yes, it assumes the Months and shortMonths # lists are in the same order and Jan..Dec ptc.MonthOffsets = {} o = 1 for key in ptc.Months: ptc.MonthOffsets[key] = o o += 1 o = 1 for key in ptc.shortMonths: ptc.MonthOffsets[key] = o o += 1
Create localized versions of the units, week and month names
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime_consts.py#L841-L869
null
#!/usr/bin/env python """ parsedatetime constants and helper functions to determine regex values from Locale information if present. Also contains the internal Locale classes to give some sane defaults if PyICU is not found. """ __license__ = """ Copyright (c) 2004-2008 Mike Taylor Copyright (c) 2006-2008 Darshana Chhajed Copyright (c) 2007 Bernd Zeimetz <bzed@debian.org> All rights reserved. Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License. """ try: import PyICU as pyicu except: pyicu = None import datetime import calendar import time import re class pdtLocale_en: """ en_US Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for USA """ localeID = 'en_US' # don't use a unicode string dateSep = [ u'/', u'.' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, MMMM d, yyyy', 'long': 'MMMM d, yyyy', 'medium': 'MMM d, yyyy', 'short': 'M/d/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a z', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'm', u'd', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': -1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eod': 0, 'eo': 0 } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_au: """ en_AU Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for Australia """ localeID = 'en_AU' # don't use a unicode string dateSep = [ u'-', u'/' ] timeSep = [ u':' ] meridian = [ u'AM', u'PM' ] usesMeridian = True uses24 = False Weekdays = [ u'monday', u'tuesday', u'wednesday', u'thursday', u'friday', u'saturday', u'sunday', ] shortWeekdays = [ u'mon', u'tues', u'wed', u'thu', u'fri', u'sat', u'sun', ] Months = [ u'january', u'february', u'march', u'april', u'may', u'june', u'july', u'august', u'september', u'october', u'november', u'december', ] shortMonths = [ u'jan', u'feb', u'mar', u'apr', u'may', u'jun', u'jul', u'aug', u'sep', u'oct', u'nov', u'dec', ] dateFormats = { 'full': 'EEEE, d MMMM yyyy', 'long': 'd MMMM yyyy', 'medium': 'dd/MM/yyyy', 'short': 'd/MM/yy', } timeFormats = { 'full': 'h:mm:ss a z', 'long': 'h:mm:ss a', 'medium': 'h:mm:ss a', 'short': 'h:mm a', } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'meridian': 'am|pm|a.m.|p.m.|a|p', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_es: """ es Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings in Spanish Note that I don't speak Spanish so many of the items below are still in English """ localeID = 'es' # don't use a unicode string dateSep = [ u'/' ] timeSep = [ u':' ] meridian = [] usesMeridian = False uses24 = True Weekdays = [ u'lunes', u'martes', u'mi\xe9rcoles', u'jueves', u'viernes', u's\xe1bado', u'domingo', ] shortWeekdays = [ u'lun', u'mar', u'mi\xe9', u'jue', u'vie', u's\xe1b', u'dom', ] Months = [ u'enero', u'febrero', u'marzo', u'abril', u'mayo', u'junio', u'julio', u'agosto', u'septiembre', u'octubre', u'noviembre', u'diciembre' ] shortMonths = [ u'ene', u'feb', u'mar', u'abr', u'may', u'jun', u'jul', u'ago', u'sep', u'oct', u'nov', u'dic' ] dateFormats = { 'full': "EEEE d' de 'MMMM' de 'yyyy", 'long': "d' de 'MMMM' de 'yyyy", 'medium': "dd-MMM-yy", 'short': "d/MM/yy", } timeFormats = { 'full': "HH'H'mm' 'ss z", 'long': "HH:mm:ss z", 'medium': "HH:mm:ss", 'short': "HH:mm", } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'second', 'sec' ], 'minutes': [ 'minute', 'min' ], 'hours': [ 'hour', 'hr' ], 'days': [ 'day', 'dy' ], 'weeks': [ 'week', 'wk' ], 'months': [ 'month', 'mth' ], 'years': [ 'year', 'yr' ], } # text constants to be used by regex's later re_consts = { 'specials': 'in|on|of|at', 'timeseperator': timeSep, 'dateseperator': dateSep, 'rangeseperator': '-', 'daysuffix': 'rd|st|nd|th', 'qunits': 'h|m|s|d|w|m|y', 'now': [ 'now' ], } # Used to adjust the returned date before/after the source modifiers = { 'from': 1, 'before': -1, 'after': 1, 'ago': 1, 'prior': -1, 'prev': -1, 'last': -1, 'next': 1, 'previous': -1, 'in a': 2, 'end of': 0, 'eo': 0, } dayoffsets = { 'tomorrow': 1, 'today': 0, 'yesterday': -1, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { 'noon': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'lunch': { 'hr': 12, 'mn': 0, 'sec': 0 }, 'morning': { 'hr': 6, 'mn': 0, 'sec': 0 }, 'breakfast': { 'hr': 8, 'mn': 0, 'sec': 0 }, 'dinner': { 'hr': 19, 'mn': 0, 'sec': 0 }, 'evening': { 'hr': 18, 'mn': 0, 'sec': 0 }, 'midnight': { 'hr': 0, 'mn': 0, 'sec': 0 }, 'night': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'tonight': { 'hr': 21, 'mn': 0, 'sec': 0 }, 'eod': { 'hr': 17, 'mn': 0, 'sec': 0 }, } class pdtLocale_de: """ de_DE Locale constants This class will be used to initialize L{Constants} if PyICU is not located. Contributed by Debian parsedatetime package maintainer Bernd Zeimetz <bzed@debian.org> Defined as class variables are the lists and strings needed by parsedatetime to evaluate strings for German """ localeID = 'de_DE' # don't use a unicode string dateSep = [ u'.' ] timeSep = [ u':' ] meridian = [ ] usesMeridian = False uses24 = True Weekdays = [ u'montag', u'dienstag', u'mittwoch', u'donnerstag', u'freitag', u'samstag', u'sonntag', ] shortWeekdays = [ u'mo', u'di', u'mi', u'do', u'fr', u'sa', u'so', ] Months = [ u'januar', u'februar', u'm\xe4rz', u'april', u'mai', u'juni', u'juli', u'august', u'september', u'oktober', u'november', u'dezember', ] shortMonths = [ u'jan', u'feb', u'mrz', u'apr', u'mai', u'jun', u'jul', u'aug', u'sep', u'okt', u'nov', u'dez', ] dateFormats = { 'full': u'EEEE, d. MMMM yyyy', 'long': u'd. MMMM yyyy', 'medium': u'dd.MM.yyyy', 'short': u'dd.MM.yy' } timeFormats = { 'full': u'HH:mm:ss v', 'long': u'HH:mm:ss z', 'medium': u'HH:mm:ss', 'short': u'HH:mm' } dp_order = [ u'd', u'm', u'y' ] # this will be added to re_consts later units = { 'seconds': [ 'sekunden', 'sek', 's' ], 'minutes': [ 'minuten', 'min' , 'm' ], 'hours': [ 'stunden', 'std', 'h' ], 'days': [ 'tage', 't' ], 'weeks': [ 'wochen', 'w' ], 'months': [ 'monate' ], #the short version would be a capital M, #as I understand it we can't distinguis #between m for minutes and M for months. 'years': [ 'jahre', 'j' ], } # text constants to be used by regex's later re_consts = { 'specials': 'am|dem|der|im|in|den|zum', 'timeseperator': ':', 'rangeseperator': '-', 'daysuffix': '', 'qunits': 'h|m|s|t|w|m|j', 'now': [ 'jetzt' ], } # Used to adjust the returned date before/after the source #still looking for insight on how to translate all of them to german. modifiers = { u'from': 1, u'before': -1, u'after': 1, u'vergangener': -1, u'vorheriger': -1, u'prev': -1, u'letzter': -1, u'n\xe4chster': 1, u'dieser': 0, u'previous': -1, u'in a': 2, u'end of': 0, u'eod': 0, u'eo': 0, } #morgen/abermorgen does not work, see http://code.google.com/p/parsedatetime/issues/detail?id=19 dayoffsets = { u'morgen': 1, u'heute': 0, u'gestern': -1, u'vorgestern': -2, u'\xfcbermorgen': 2, } # special day and/or times, i.e. lunch, noon, evening # each element in the dictionary is a dictionary that is used # to fill in any value to be replace - the current date/time will # already have been populated by the method buildSources re_sources = { u'mittag': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittags': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'mittagessen': { 'hr': 12, 'mn': 0, 'sec': 0 }, u'morgen': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'morgens': { 'hr': 6, 'mn': 0, 'sec': 0 }, u'fr\e4hst\xe4ck': { 'hr': 8, 'mn': 0, 'sec': 0 }, u'abendessen': { 'hr': 19, 'mn': 0, 'sec': 0 }, u'abend': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'abends': { 'hr': 18, 'mn': 0, 'sec': 0 }, u'mitternacht': { 'hr': 0, 'mn': 0, 'sec': 0 }, u'nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'nachts': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute abend': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'heute nacht': { 'hr': 21, 'mn': 0, 'sec': 0 }, u'feierabend': { 'hr': 17, 'mn': 0, 'sec': 0 }, } pdtLocales = { 'en_US': pdtLocale_en, 'en_AU': pdtLocale_au, 'es_ES': pdtLocale_es, 'de_DE': pdtLocale_de, } def _initLocale(ptc): """ Helper function to initialize the different lists and strings from either PyICU or one of the internal pdt Locales and store them into ptc. """ def lcase(x): return x.lower() if pyicu and ptc.usePyICU: ptc.icuLocale = None if ptc.localeID is not None: ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is None: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] ptc.icuLocale = pyicu.Locale(ptc.localeID) if ptc.icuLocale is not None: break ptc.icuSymbols = pyicu.DateFormatSymbols(ptc.icuLocale) # grab ICU list of weekdays, skipping first entry which # is always blank wd = map(lcase, ptc.icuSymbols.getWeekdays()[1:]) swd = map(lcase, ptc.icuSymbols.getShortWeekdays()[1:]) # store them in our list with Monday first (ICU puts Sunday first) ptc.Weekdays = wd[1:] + wd[0:1] ptc.shortWeekdays = swd[1:] + swd[0:1] ptc.Months = map(lcase, ptc.icuSymbols.getMonths()) ptc.shortMonths = map(lcase, ptc.icuSymbols.getShortMonths()) # not quite sure how to init this so for now # set it to none so it will be set to the en_US defaults for now ptc.re_consts = None ptc.icu_df = { 'full': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createDateInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.icu_tf = { 'full': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kFull, ptc.icuLocale), 'long': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kLong, ptc.icuLocale), 'medium': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kMedium, ptc.icuLocale), 'short': pyicu.DateFormat.createTimeInstance(pyicu.DateFormat.kShort, ptc.icuLocale), } ptc.dateFormats = { 'full': ptc.icu_df['full'].toPattern(), 'long': ptc.icu_df['long'].toPattern(), 'medium': ptc.icu_df['medium'].toPattern(), 'short': ptc.icu_df['short'].toPattern(), } ptc.timeFormats = { 'full': ptc.icu_tf['full'].toPattern(), 'long': ptc.icu_tf['long'].toPattern(), 'medium': ptc.icu_tf['medium'].toPattern(), 'short': ptc.icu_tf['short'].toPattern(), } else: if not ptc.localeID in pdtLocales: for id in range(0, len(ptc.fallbackLocales)): ptc.localeID = ptc.fallbackLocales[id] if ptc.localeID in pdtLocales: break ptc.locale = pdtLocales[ptc.localeID] ptc.usePyICU = False ptc.Weekdays = ptc.locale.Weekdays ptc.shortWeekdays = ptc.locale.shortWeekdays ptc.Months = ptc.locale.Months ptc.shortMonths = ptc.locale.shortMonths ptc.dateFormats = ptc.locale.dateFormats ptc.timeFormats = ptc.locale.timeFormats # these values are used to setup the various bits # of the regex values used to parse # # check if a local set of constants has been # provided, if not use en_US as the default if ptc.localeID in pdtLocales: ptc.re_sources = pdtLocales[ptc.localeID].re_sources ptc.re_values = pdtLocales[ptc.localeID].re_consts units = pdtLocales[ptc.localeID].units ptc.Modifiers = pdtLocales[ptc.localeID].modifiers ptc.dayOffsets = pdtLocales[ptc.localeID].dayoffsets # for now, pull over any missing keys from the US set for key in pdtLocales['en_US'].re_consts: if not key in ptc.re_values: ptc.re_values[key] = pdtLocales['en_US'].re_consts[key] else: ptc.re_sources = pdtLocales['en_US'].re_sources ptc.re_values = pdtLocales['en_US'].re_consts ptc.Modifiers = pdtLocales['en_US'].modifiers ptc.dayOffsets = pdtLocales['en_US'].dayoffsets units = pdtLocales['en_US'].units # escape any regex special characters that may be found wd = tuple(map(re.escape, ptc.Weekdays)) swd = tuple(map(re.escape, ptc.shortWeekdays)) mth = tuple(map(re.escape, ptc.Months)) smth = tuple(map(re.escape, ptc.shortMonths)) ptc.re_values['months'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % mth ptc.re_values['shortmonths'] = '%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s|%s' % smth ptc.re_values['days'] = '%s|%s|%s|%s|%s|%s|%s' % wd ptc.re_values['shortdays'] = '%s|%s|%s|%s|%s|%s|%s' % swd l = [] for unit in units: l.append('|'.join(units[unit])) ptc.re_values['units'] = '|'.join(l) ptc.Units = ptc.re_values['units'].split('|') def _initSymbols(ptc): """ Helper function to initialize the single character constants and other symbols needed. """ ptc.timeSep = [ u':' ] ptc.dateSep = [ u'/' ] ptc.meridian = [ u'AM', u'PM' ] ptc.usesMeridian = True ptc.uses24 = False if pyicu and ptc.usePyICU: am = u'' pm = u'' ts = '' # ICU doesn't seem to provide directly the # date or time seperator - so we have to # figure it out o = ptc.icu_tf['short'] s = ptc.timeFormats['short'] ptc.usesMeridian = u'a' in s ptc.uses24 = u'H' in s # '11:45 AM' or '11:45' s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) # ': AM' or ':' s = s.replace('11', '').replace('45', '') if len(s) > 0: ts = s[0] if ptc.usesMeridian: # '23:45 AM' or '23:45' am = s[1:].strip() s = o.format(datetime.datetime(2003, 10, 30, 23, 45)) if ptc.uses24: s = s.replace('23', '') else: s = s.replace('11', '') # 'PM' or '' pm = s.replace('45', '').replace(ts, '').strip() ptc.timeSep = [ ts ] ptc.meridian = [ am, pm ] o = ptc.icu_df['short'] s = o.format(datetime.datetime(2003, 10, 30, 11, 45)) s = s.replace('10', '').replace('30', '').replace('03', '').replace('2003', '') if len(s) > 0: ds = s[0] else: ds = '/' ptc.dateSep = [ ds ] s = ptc.dateFormats['short'] l = s.lower().split(ds) dp_order = [] for s in l: if len(s) > 0: dp_order.append(s[:1]) ptc.dp_order = dp_order else: ptc.timeSep = ptc.locale.timeSep ptc.dateSep = ptc.locale.dateSep ptc.meridian = ptc.locale.meridian ptc.usesMeridian = ptc.locale.usesMeridian ptc.uses24 = ptc.locale.uses24 ptc.dp_order = ptc.locale.dp_order # build am and pm lists to contain # original case, lowercase and first-char # versions of the meridian text if len(ptc.meridian) > 0: am = ptc.meridian[0] ptc.am = [ am ] if len(am) > 0: ptc.am.append(am[0]) am = am.lower() ptc.am.append(am) ptc.am.append(am[0]) else: am = '' ptc.am = [ '', '' ] if len(ptc.meridian) > 1: pm = ptc.meridian[1] ptc.pm = [ pm ] if len(pm) > 0: ptc.pm.append(pm[0]) pm = pm.lower() ptc.pm.append(pm) ptc.pm.append(pm[0]) else: pm = '' ptc.pm = [ '', '' ] def _initPatterns(ptc): """ Helper function to take the different localized bits from ptc and create the regex strings. """ # TODO add code to parse the date formats and build the regexes up from sub-parts # TODO find all hard-coded uses of date/time seperators ptc.RE_DATE4 = r'''(?P<date>(((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?(,)?(\s)?) (?P<mthname>(%(months)s|%(shortmonths)s))\s? (?P<year>\d\d(\d\d)?)? ) )''' % ptc.re_values # I refactored DATE3 to fix Issue 16 http://code.google.com/p/parsedatetime/issues/detail?id=16 # I suspect the final line was for a trailing time - but testing shows it's not needed # ptc.RE_DATE3 = r'''(?P<date>((?P<mthname>(%(months)s|%(shortmonths)s))\s? # ((?P<day>\d\d?)(\s?|%(daysuffix)s|$)+)? # (,\s?(?P<year>\d\d(\d\d)?))?)) # (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DATE3 = r'''(?P<date>( (((?P<mthname>(%(months)s|%(shortmonths)s))| ((?P<day>\d\d?)(?P<suffix>%(daysuffix)s)?))(\s)?){1,2} ((,)?(\s)?(?P<year>\d\d(\d\d)?))? ) )''' % ptc.re_values ptc.RE_MONTH = r'''(\s?|^) (?P<month>( (?P<mthname>(%(months)s|%(shortmonths)s)) (\s?(?P<year>(\d\d\d\d)))? )) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_WEEKDAY = r'''(\s?|^) (?P<weekday>(%(days)s|%(shortdays)s)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_SPECIAL = r'(?P<special>^[%(specials)s]+)\s+' % ptc.re_values ptc.RE_UNITS = r'''(?P<qty>(-?\d+\s* (?P<units>((%(units)s)s?)) ))''' % ptc.re_values ptc.RE_QUNITS = r'''(?P<qty>(-?\d+\s? (?P<qunits>%(qunits)s) (\s?|,|$) ))''' % ptc.re_values ptc.RE_MODIFIER = r'''(\s?|^) (?P<modifier> (previous|prev|last|next|eod|eo|(end\sof)|(in\sa)))''' % ptc.re_values ptc.RE_MODIFIER2 = r'''(\s?|^) (?P<modifier> (from|before|after|ago|prior)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_TIMEHMS = r'''(\s?|^) (?P<hours>\d\d?) (?P<tsep>%(timeseperator)s|) (?P<minutes>\d\d) (?:(?P=tsep)(?P<seconds>\d\d(?:[.,]\d+)?))?''' % ptc.re_values ptc.RE_TIMEHMS2 = r'''(?P<hours>(\d\d?)) ((?P<tsep>%(timeseperator)s|) (?P<minutes>(\d\d?)) (?:(?P=tsep) (?P<seconds>\d\d? (?:[.,]\d+)?))?)?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_TIMEHMS2 += r'\s?(?P<meridian>(%(meridian)s))' % ptc.re_values dateSeps = ''.join(ptc.dateSep) + '.' ptc.RE_DATE = r'''(\s?|^) (?P<date>(\d\d?[%s]\d\d?([%s]\d\d(\d\d)?)?)) (\s?|$|[^0-9a-zA-Z])''' % (dateSeps, dateSeps) ptc.RE_DATE2 = r'[%s]' % dateSeps ptc.RE_DAY = r'''(\s?|^) (?P<day>(today|tomorrow|yesterday)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_DAY2 = r'''(?P<day>\d\d?)|(?P<suffix>%(daysuffix)s) ''' % ptc.re_values ptc.RE_TIME = r'''(\s?|^) (?P<time>(morning|breakfast|noon|lunch|evening|midnight|tonight|dinner|night|now)) (\s?|$|[^0-9a-zA-Z])''' % ptc.re_values ptc.RE_REMAINING = r'\s+' # Regex for date/time ranges ptc.RE_RTIMEHMS = r'''(\s?|^) (\d\d?)%(timeseperator)s (\d\d) (%(timeseperator)s(\d\d))? (\s?|$)''' % ptc.re_values ptc.RE_RTIMEHMS2 = r'''(\s?|^) (\d\d?) (%(timeseperator)s(\d\d?))? (%(timeseperator)s(\d\d?))?''' % ptc.re_values if 'meridian' in ptc.re_values: ptc.RE_RTIMEHMS2 += r'\s?(%(meridian)s)' % ptc.re_values ptc.RE_RDATE = r'(\d+([%s]\d+)+)' % dateSeps ptc.RE_RDATE3 = r'''((((%(months)s))\s? ((\d\d?) (\s?|%(daysuffix)s|$)+)? (,\s?\d\d\d\d)?))''' % ptc.re_values # "06/07/06 - 08/09/06" ptc.DATERNG1 = ptc.RE_RDATE + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE ptc.DATERNG1 = ptc.DATERNG1 % ptc.re_values # "march 31 - june 1st, 2006" ptc.DATERNG2 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RDATE3 ptc.DATERNG2 = ptc.DATERNG2 % ptc.re_values # "march 1rd -13th" ptc.DATERNG3 = ptc.RE_RDATE3 + r'\s?%(rangeseperator)s\s?(\d\d?)\s?(rd|st|nd|th)?' ptc.DATERNG3 = ptc.DATERNG3 % ptc.re_values # "4:00:55 pm - 5:90:44 am", '4p-5p' ptc.TIMERNG1 = ptc.RE_RTIMEHMS2 + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG1 = ptc.TIMERNG1 % ptc.re_values # "4:00 - 5:90 ", "4:55:55-3:44:55" ptc.TIMERNG2 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS ptc.TIMERNG2 = ptc.TIMERNG2 % ptc.re_values # "4-5pm " ptc.TIMERNG3 = r'\d\d?\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG3 = ptc.TIMERNG3 % ptc.re_values # "4:30-5pm " ptc.TIMERNG4 = ptc.RE_RTIMEHMS + r'\s?%(rangeseperator)s\s?' + ptc.RE_RTIMEHMS2 ptc.TIMERNG4 = ptc.TIMERNG4 % ptc.re_values # ptc.DaySuffixes = ptc.re_consts['daysuffix'].split('|') class Constants: """ Default set of constants for parsedatetime. If PyICU is present, then the class will first try to get PyICU to return a locale specified by C{localeID}. If either C{localeID} is None or if the locale does not exist within PyICU, then each of the locales defined in C{fallbackLocales} is tried in order. If PyICU is not present or none of the specified locales can be used, then the class will initialize itself to the en_US locale. if PyICU is not present or not requested, only the locales defined by C{pdtLocales} will be searched. """ def __init__(self, localeID=None, usePyICU=True, fallbackLocales=['en_US']): self.localeID = localeID self.fallbackLocales = fallbackLocales if 'en_US' not in self.fallbackLocales: self.fallbackLocales.append('en_US') # define non-locale specific constants self.locale = None self.usePyICU = usePyICU # starting cache of leap years # daysInMonth will add to this if during # runtime it gets a request for a year not found self._leapYears = [ 1904, 1908, 1912, 1916, 1920, 1924, 1928, 1932, 1936, 1940, 1944, 1948, 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024, 2028, 2032, 2036, 2040, 2044, 2048, 2052, 2056, 2060, 2064, 2068, 2072, 2076, 2080, 2084, 2088, 2092, 2096 ] self.Second = 1 self.Minute = 60 * self.Second self.Hour = 60 * self.Minute self.Day = 24 * self.Hour self.Week = 7 * self.Day self.Month = 30 * self.Day self.Year = 365 * self.Day self.rangeSep = u'-' self._DaysInMonthList = (31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) self.BirthdayEpoch = 50 # DOWParseStyle controls how we parse "Tuesday" # If the current day was Thursday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Current day marked as *** # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current -1,0 *** # week +1 +1 # # If the current day was Monday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 -1 # current *** 0,+1 # week +1 self.DOWParseStyle = 1 # CurrentDOWParseStyle controls how we parse "Friday" # If the current day was Friday and the text to parse is "Friday" # then the following table shows how each style would be returned # True/False. This also depends on DOWParseStyle. # # Current day marked as *** # # DOWParseStyle = 0 # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T,F # week +1 # # DOWParseStyle = -1 # Sun Mon Tue Wed Thu Fri Sat # week -1 F # current T # week +1 # # DOWParseStyle = +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T # week +1 F self.CurrentDOWParseStyle = False # initalize attributes to empty values to ensure # they are defined self.re_sources = None self.re_values = None self.Modifiers = None self.dayOffsets = None self.WeekdayOffsets = None self.MonthOffsets = None self.dateSep = None self.timeSep = None self.am = None self.pm = None self.meridian = None self.usesMeridian = None self.uses24 = None self.dp_order = None self.RE_DATE4 = r'' self.RE_DATE3 = r'' self.RE_MONTH = r'' self.RE_WEEKDAY = r'' self.RE_SPECIAL = r'' self.RE_UNITS = r'' self.RE_QUNITS = r'' self.RE_MODIFIER = r'' self.RE_MODIFIER2 = r'' self.RE_TIMEHMS = r'' self.RE_TIMEHMS2 = r'' self.RE_DATE = r'' self.RE_DATE2 = r'' self.RE_DAY = r'' self.RE_DAY2 = r'' self.RE_TIME = r'' self.RE_REMAINING = r'' self.RE_RTIMEHMS = r'' self.RE_RTIMEHMS2 = r'' self.RE_RDATE = r'' self.RE_RDATE3 = r'' self.DATERNG1 = r'' self.DATERNG2 = r'' self.DATERNG3 = r'' self.TIMERNG1 = r'' self.TIMERNG2 = r'' self.TIMERNG3 = r'' self.TIMERNG4 = r'' _initLocale(self) _initConstants(self) _initSymbols(self) _initPatterns(self) self.re_option = re.IGNORECASE + re.VERBOSE self.cre_source = { 'CRE_SPECIAL': self.RE_SPECIAL, 'CRE_UNITS': self.RE_UNITS, 'CRE_QUNITS': self.RE_QUNITS, 'CRE_MODIFIER': self.RE_MODIFIER, 'CRE_MODIFIER2': self.RE_MODIFIER2, 'CRE_TIMEHMS': self.RE_TIMEHMS, 'CRE_TIMEHMS2': self.RE_TIMEHMS2, 'CRE_DATE': self.RE_DATE, 'CRE_DATE2': self.RE_DATE2, 'CRE_DATE3': self.RE_DATE3, 'CRE_DATE4': self.RE_DATE4, 'CRE_MONTH': self.RE_MONTH, 'CRE_WEEKDAY': self.RE_WEEKDAY, 'CRE_DAY': self.RE_DAY, 'CRE_DAY2': self.RE_DAY2, 'CRE_TIME': self.RE_TIME, 'CRE_REMAINING': self.RE_REMAINING, 'CRE_RTIMEHMS': self.RE_RTIMEHMS, 'CRE_RTIMEHMS2': self.RE_RTIMEHMS2, 'CRE_RDATE': self.RE_RDATE, 'CRE_RDATE3': self.RE_RDATE3, 'CRE_TIMERNG1': self.TIMERNG1, 'CRE_TIMERNG2': self.TIMERNG2, 'CRE_TIMERNG3': self.TIMERNG3, 'CRE_TIMERNG4': self.TIMERNG4, 'CRE_DATERNG1': self.DATERNG1, 'CRE_DATERNG2': self.DATERNG2, 'CRE_DATERNG3': self.DATERNG3, } self.cre_keys = self.cre_source.keys() def __getattr__(self, name): if name in self.cre_keys: value = re.compile(self.cre_source[name], self.re_option) setattr(self, name, value) return value else: raise AttributeError, name def daysInMonth(self, month, year): """ Take the given month (1-12) and a given year (4 digit) return the number of days in the month adjusting for leap year as needed """ result = None if month > 0 and month <= 12: result = self._DaysInMonthList[month - 1] if month == 2: if year in self._leapYears: result += 1 else: if calendar.isleap(year): self._leapYears.append(year) result += 1 return result def buildSources(self, sourceTime=None): """ Return a dictionary of date/time tuples based on the keys found in self.re_sources. The current time is used as the default and any specified item found in self.re_sources is inserted into the value and the generated dictionary is returned. """ if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime sources = {} defaults = { 'yr': yr, 'mth': mth, 'dy': dy, 'hr': hr, 'mn': mn, 'sec': sec, } for item in self.re_sources: values = {} source = self.re_sources[item] for key in defaults.keys(): if key in source: values[key] = source[key] else: values[key] = defaults[key] sources[item] = ( values['yr'], values['mth'], values['dy'], values['hr'], values['mn'], values['sec'], wd, yd, isdst ) return sources
vilmibm/done
parsedatetime/parsedatetime_consts.py
Constants.buildSources
python
def buildSources(self, sourceTime=None): if sourceTime is None: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = time.localtime() else: (yr, mth, dy, hr, mn, sec, wd, yd, isdst) = sourceTime sources = {} defaults = { 'yr': yr, 'mth': mth, 'dy': dy, 'hr': hr, 'mn': mn, 'sec': sec, } for item in self.re_sources: values = {} source = self.re_sources[item] for key in defaults.keys(): if key in source: values[key] = source[key] else: values[key] = defaults[key] sources[item] = ( values['yr'], values['mth'], values['dy'], values['hr'], values['mn'], values['sec'], wd, yd, isdst ) return sources
Return a dictionary of date/time tuples based on the keys found in self.re_sources. The current time is used as the default and any specified item found in self.re_sources is inserted into the value and the generated dictionary is returned.
train
https://github.com/vilmibm/done/blob/7e5b60d2900ceddefa49de352a19b794199b51a8/parsedatetime/parsedatetime_consts.py#L1087-L1118
null
class Constants: """ Default set of constants for parsedatetime. If PyICU is present, then the class will first try to get PyICU to return a locale specified by C{localeID}. If either C{localeID} is None or if the locale does not exist within PyICU, then each of the locales defined in C{fallbackLocales} is tried in order. If PyICU is not present or none of the specified locales can be used, then the class will initialize itself to the en_US locale. if PyICU is not present or not requested, only the locales defined by C{pdtLocales} will be searched. """ def __init__(self, localeID=None, usePyICU=True, fallbackLocales=['en_US']): self.localeID = localeID self.fallbackLocales = fallbackLocales if 'en_US' not in self.fallbackLocales: self.fallbackLocales.append('en_US') # define non-locale specific constants self.locale = None self.usePyICU = usePyICU # starting cache of leap years # daysInMonth will add to this if during # runtime it gets a request for a year not found self._leapYears = [ 1904, 1908, 1912, 1916, 1920, 1924, 1928, 1932, 1936, 1940, 1944, 1948, 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004, 2008, 2012, 2016, 2020, 2024, 2028, 2032, 2036, 2040, 2044, 2048, 2052, 2056, 2060, 2064, 2068, 2072, 2076, 2080, 2084, 2088, 2092, 2096 ] self.Second = 1 self.Minute = 60 * self.Second self.Hour = 60 * self.Minute self.Day = 24 * self.Hour self.Week = 7 * self.Day self.Month = 30 * self.Day self.Year = 365 * self.Day self.rangeSep = u'-' self._DaysInMonthList = (31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31) self.BirthdayEpoch = 50 # DOWParseStyle controls how we parse "Tuesday" # If the current day was Thursday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Current day marked as *** # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current -1,0 *** # week +1 +1 # # If the current day was Monday and the text to parse is "Tuesday" # then the following table shows how each style would be returned # -1, 0, +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 -1 # current *** 0,+1 # week +1 self.DOWParseStyle = 1 # CurrentDOWParseStyle controls how we parse "Friday" # If the current day was Friday and the text to parse is "Friday" # then the following table shows how each style would be returned # True/False. This also depends on DOWParseStyle. # # Current day marked as *** # # DOWParseStyle = 0 # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T,F # week +1 # # DOWParseStyle = -1 # Sun Mon Tue Wed Thu Fri Sat # week -1 F # current T # week +1 # # DOWParseStyle = +1 # # Sun Mon Tue Wed Thu Fri Sat # week -1 # current T # week +1 F self.CurrentDOWParseStyle = False # initalize attributes to empty values to ensure # they are defined self.re_sources = None self.re_values = None self.Modifiers = None self.dayOffsets = None self.WeekdayOffsets = None self.MonthOffsets = None self.dateSep = None self.timeSep = None self.am = None self.pm = None self.meridian = None self.usesMeridian = None self.uses24 = None self.dp_order = None self.RE_DATE4 = r'' self.RE_DATE3 = r'' self.RE_MONTH = r'' self.RE_WEEKDAY = r'' self.RE_SPECIAL = r'' self.RE_UNITS = r'' self.RE_QUNITS = r'' self.RE_MODIFIER = r'' self.RE_MODIFIER2 = r'' self.RE_TIMEHMS = r'' self.RE_TIMEHMS2 = r'' self.RE_DATE = r'' self.RE_DATE2 = r'' self.RE_DAY = r'' self.RE_DAY2 = r'' self.RE_TIME = r'' self.RE_REMAINING = r'' self.RE_RTIMEHMS = r'' self.RE_RTIMEHMS2 = r'' self.RE_RDATE = r'' self.RE_RDATE3 = r'' self.DATERNG1 = r'' self.DATERNG2 = r'' self.DATERNG3 = r'' self.TIMERNG1 = r'' self.TIMERNG2 = r'' self.TIMERNG3 = r'' self.TIMERNG4 = r'' _initLocale(self) _initConstants(self) _initSymbols(self) _initPatterns(self) self.re_option = re.IGNORECASE + re.VERBOSE self.cre_source = { 'CRE_SPECIAL': self.RE_SPECIAL, 'CRE_UNITS': self.RE_UNITS, 'CRE_QUNITS': self.RE_QUNITS, 'CRE_MODIFIER': self.RE_MODIFIER, 'CRE_MODIFIER2': self.RE_MODIFIER2, 'CRE_TIMEHMS': self.RE_TIMEHMS, 'CRE_TIMEHMS2': self.RE_TIMEHMS2, 'CRE_DATE': self.RE_DATE, 'CRE_DATE2': self.RE_DATE2, 'CRE_DATE3': self.RE_DATE3, 'CRE_DATE4': self.RE_DATE4, 'CRE_MONTH': self.RE_MONTH, 'CRE_WEEKDAY': self.RE_WEEKDAY, 'CRE_DAY': self.RE_DAY, 'CRE_DAY2': self.RE_DAY2, 'CRE_TIME': self.RE_TIME, 'CRE_REMAINING': self.RE_REMAINING, 'CRE_RTIMEHMS': self.RE_RTIMEHMS, 'CRE_RTIMEHMS2': self.RE_RTIMEHMS2, 'CRE_RDATE': self.RE_RDATE, 'CRE_RDATE3': self.RE_RDATE3, 'CRE_TIMERNG1': self.TIMERNG1, 'CRE_TIMERNG2': self.TIMERNG2, 'CRE_TIMERNG3': self.TIMERNG3, 'CRE_TIMERNG4': self.TIMERNG4, 'CRE_DATERNG1': self.DATERNG1, 'CRE_DATERNG2': self.DATERNG2, 'CRE_DATERNG3': self.DATERNG3, } self.cre_keys = self.cre_source.keys() def __getattr__(self, name): if name in self.cre_keys: value = re.compile(self.cre_source[name], self.re_option) setattr(self, name, value) return value else: raise AttributeError, name def daysInMonth(self, month, year): """ Take the given month (1-12) and a given year (4 digit) return the number of days in the month adjusting for leap year as needed """ result = None if month > 0 and month <= 12: result = self._DaysInMonthList[month - 1] if month == 2: if year in self._leapYears: result += 1 else: if calendar.isleap(year): self._leapYears.append(year) result += 1 return result
MacHu-GWU/pyknackhq-project
pyknackhq/schema.py
Object.get_field_key
python
def get_field_key(self, key, using_name=True): try: if using_name: return self.f_name[key].key else: return self.f[key].key except KeyError: raise ValueError("'%s' are not found!" % key)
Given a field key or name, return it's field key.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/schema.py#L80-L89
null
class Object(object): """Data object class. Object are used to define an abstract concept of thing. For example, an employee can be an object having attributes: name, date of birth, phone, email, etc... """ def __init__(self, **kwargs): for k, v in kwargs.items(): object.__setattr__(self, k, v) self.f = OrderedDict() # {field_key: Field instance} self.f_name = OrderedDict() # {field_name: Field instance} for d in self.fields: field = Field.from_dict(d) self.f.setdefault(d["key"], field) self.f_name.setdefault(d["name"], field) def __str__(self): return "Object('%s')" % self.name def __repr__(self): return "Object(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Object(**d) @staticmethod def from_json(abspath): return Object.from_dict(load_js(abspath, enable_verbose=False)) def __iter__(self): return iter(self.f.values()) @property def all_field_key(self): """Return all available field_key. """ return [f.key for f in self.f.values()] @property def all_field_name(self): """Return all available field_name. """ return [f.name for f in self.f.values()] def get_field(self, key, using_name=True): """Given a field key or name, return the Field instance. """ try: if using_name: return self.f_name[key] else: return self.f[key] except KeyError: raise ValueError("'%s' are not found!" % key)
MacHu-GWU/pyknackhq-project
pyknackhq/schema.py
Application.get_object_key
python
def get_object_key(self, key, using_name=True): try: if using_name: return self.o_name[key].key else: return self.o[key].key except KeyError: raise ValueError("'%s' are not found!" % key)
Given a object key or name, return it's object key.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/schema.py#L149-L158
null
class Application(object): """Application class that holding object and its fields information. """ def __init__(self, **kwargs): self.data = {"application": kwargs} for k, v in kwargs.items(): object.__setattr__(self, k, v) self.o = OrderedDict() # {field_key: Field instance} self.o_name = OrderedDict() # {field_name: Field instance} for d in self.objects: object_ = Object.from_dict(d) self.o.setdefault(d["key"], object_) self.o_name.setdefault(d["name"], object_) def __str__(self): return "Application('%s')" % self.name def __repr__(self): return "Application('%s')" % self.name @staticmethod def from_dict(d): return Application(**d["application"]) @staticmethod def from_json(abspath): return Application.from_dict(load_js(abspath, enable_verbose=False)) def to_json(self, abspath): safe_dump_js(self.data, abspath, enable_verbose=False) def __iter__(self): return iter(self.o.values()) @property def all_object_key(self): """Return all available object_key. """ return [o.key for o in self.o.values()] @property def all_object_name(self): """Return all available object_name. """ return [o.name for o in self.o.values()] def get_object(self, key, using_name=True): """Given a object key or name, return the Object instance. """ try: if using_name: return self.o_name[key] else: return self.o[key] except KeyError: raise ValueError("'%s' are not found!" % key)
MacHu-GWU/pyknackhq-project
pyknackhq/zzz_manual_install.py
install
python
def install(): import os, shutil _ROOT = os.getcwd() _PACKAGE_NAME = os.path.basename(_ROOT) print("Installing [%s] to all python version..." % _PACKAGE_NAME) # find all Python release installed on this windows computer installed_python_version = list() for root, folder_list, _ in os.walk(r"C:\\"): for folder in folder_list: if folder.startswith("Python"): if os.path.exists(os.path.join(root, folder, "pythonw.exe")): installed_python_version.append(folder) break print("\tYou have installed: {0}".format(", ".join(installed_python_version))) # remove __pycache__ folder and *.pyc file print("\tRemoving *.pyc file ...") pyc_folder_list = list() for root, folder_list, _ in os.walk(_ROOT): if os.path.basename(root) == "__pycache__": pyc_folder_list.append(root) for folder in pyc_folder_list: shutil.rmtree(folder) print("\t\tall *.pyc file has been removed.") # install this package to all python version for py_root in installed_python_version: dst = os.path.join(r"C:\\", py_root, r"Lib\site-packages", _PACKAGE_NAME) try: shutil.rmtree(dst) except: pass print("\tRemoved %s." % dst) shutil.copytree(_ROOT, dst) print("\tInstalled %s." % dst) print("Complete!")
Install your package to all Python version you have installed on Windows.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/zzz_manual_install.py#L100-L141
null
#!/usr/bin/env python # -*- coding: utf-8 -*- """ Install your own package in one seconds! (Windows System Only!) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Put this script in your package directory, for example:: |---mypackage |---subpackage1 |---__init__.py |---subpackage2 |---__init__.py |---__init__.py |---module1.py |---module2.py |---zzz_manual_install.py <=== put it here Run this script as the main script. Then your package is automatically installed and replace the old one for all Python versions on your ``WINDOWS`` computer. This feature is extremely useful when you need refresh your package over and over again. But if you want to make an official release, you should make a setup.py and build the distribution by yourself. Read the following instruction: - For Python2: - https://docs.python.org/2/distutils/setupscript.html - https://docs.python.org/2/distutils/builtdist.html - For Python3: - https://docs.python.org/3.3/distutils/setupscript.html - https://docs.python.org/3.3/distutils/builtdist.html **Warning**: with python2, the project directory cannot have non-ascil char. ------------------------------------------------------------------------------- **中文文档** 本脚用于在Windows系统下傻瓜一键安装用户自己写的扩展包, 纯python实现。 例如你有一个扩展包叫 mypackage, 文件目录形如: ``C:\project\mypackage`` 则只需要把该脚本拷贝到 mypackage 目录下: ``C:\project\mypackage\zzz_manual_install.py`` 然后将本脚本以主脚本运行。则会把package文件中所有的 .pyc 文件清除后, 安装你所有 已安装的Python版本下。例如你安装了Python27和Python33, 那么就会创建以下目录并将 包里的所有文件拷贝到该目录下:: C:\Python27\Lib\site-packages\mypackage C:\Python33\Lib\site-packages\mypackage 然后你就可以用 ``import mypackage`` 调用你写的库了。 这一功能在调试阶段非常方便, 但最终发布时还是要通过写 ``setup.py`` 文件来制作 package的安装包。这一部分可以参考: - Python2: - https://docs.python.org/2/distutils/setupscript.html - https://docs.python.org/2/distutils/builtdist.html - Python3: - https://docs.python.org/3.3/distutils/setupscript.html - https://docs.python.org/3.3/distutils/builtdist.html 注: 项目目录在python2中不允许有中文路径。 About ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ **Copyright (c) 2015 by Sanhe Hu** - Author: Sanhe Hu - Email: husanhe@gmail.com - Lisence: MIT **Compatibility** - Python2: Yes - Python3: Yes **Prerequisites** - None class, method, func, exception ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ """ from __future__ import print_function, unicode_literals def install(): """Install your package to all Python version you have installed on Windows. """ import os, shutil _ROOT = os.getcwd() _PACKAGE_NAME = os.path.basename(_ROOT) print("Installing [%s] to all python version..." % _PACKAGE_NAME) # find all Python release installed on this windows computer installed_python_version = list() for root, folder_list, _ in os.walk(r"C:\\"): for folder in folder_list: if folder.startswith("Python"): if os.path.exists(os.path.join(root, folder, "pythonw.exe")): installed_python_version.append(folder) break print("\tYou have installed: {0}".format(", ".join(installed_python_version))) # remove __pycache__ folder and *.pyc file print("\tRemoving *.pyc file ...") pyc_folder_list = list() for root, folder_list, _ in os.walk(_ROOT): if os.path.basename(root) == "__pycache__": pyc_folder_list.append(root) for folder in pyc_folder_list: shutil.rmtree(folder) print("\t\tall *.pyc file has been removed.") # install this package to all python version for py_root in installed_python_version: dst = os.path.join(r"C:\\", py_root, r"Lib\site-packages", _PACKAGE_NAME) try: shutil.rmtree(dst) except: pass print("\tRemoved %s." % dst) shutil.copytree(_ROOT, dst) print("\tInstalled %s." % dst) print("Complete!") if __name__ == "__main__": install()
MacHu-GWU/pyknackhq-project
pyknackhq/js.py
load_js
python
def load_js(abspath, default=dict(), compress=False, enable_verbose=True): abspath = str(abspath) # try stringlize if compress: # check extension name if os.path.splitext(abspath)[1] != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: if os.path.splitext(abspath)[1] != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nLoading from %s..." % abspath) st = time.clock() if os.path.exists(abspath): # exists, then load if compress: with gzip.open(abspath, "rb") as f: js = json.loads(f.read().decode("utf-8")) else: with open(abspath, "r") as f: js = json.load(f) if enable_verbose: print("\tComplete! Elapse %.6f sec." % (time.clock() - st) ) return js else: if enable_verbose: print("\t%s not exists! cannot load! Create an default object " "instead" % abspath) return default
Load Json from file. If file are not exists, returns ``default``. :param abspath: File path. Use absolute path as much as you can. File extension has to be ``.json`` or ``.gz``. (for compressed Json) :type abspath: string :param default: (default dict()) If ``abspath`` not exists, return the default Python object instead. :param compress: (default False) Load from a gzip compressed Json file. Check :func:`dump_js()<dump_js>` function for more information. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import load_js >>> load_js("test.json") # if you have a json file Loading from test.json... Complete! Elapse 0.000432 sec. {'a': 1, 'b': 2} **中文文档** 从Json文件中读取数据 参数列表 :param abspath: 文件路径, 扩展名需为 ``.json`` 或 ``.gz`` :type abspath: ``字符串`` :param default: (默认 dict()) 如果文件路径不存在, 则会返回一个默认的Python对象。 :param compress: (默认 False) 是否从一个gzip压缩过的Json文件中读取数据。 请 参考 :func:`dump_js()<dump_js>` 获得更多信息. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值``
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/js.py#L52-L123
null
#!/usr/bin/env python # -*- coding: utf-8 -*- """ Module description ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ This module is re-pack of some json utility functions. - :func:`load_js`: Load Json from file. If file are not exists, returns user defined ``default value``. - :func:`dump_js`: Dump Json serializable object to file. - :func:`safe_dump_js`: An atomic write version of dump_js, silently overwrite existing file. - :func:`js2str`: Encode js to nicely formatted human readable string. - :func:`prt_js`: Print Json in pretty format. Highlight ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - :func:`load_js`, :func:`dump_js`, :func:`safe_dump_js` support gzip compress, size is **10 - 20 times** smaller in average. Compatibility ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - Python2: Yes - Python3: Yes Prerequisites ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - None Class, method, function, exception ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ """ from __future__ import print_function, unicode_literals import json, gzip import os, shutil import time def load_js(abspath, default=dict(), compress=False, enable_verbose=True): """Load Json from file. If file are not exists, returns ``default``. :param abspath: File path. Use absolute path as much as you can. File extension has to be ``.json`` or ``.gz``. (for compressed Json) :type abspath: string :param default: (default dict()) If ``abspath`` not exists, return the default Python object instead. :param compress: (default False) Load from a gzip compressed Json file. Check :func:`dump_js()<dump_js>` function for more information. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import load_js >>> load_js("test.json") # if you have a json file Loading from test.json... Complete! Elapse 0.000432 sec. {'a': 1, 'b': 2} **中文文档** 从Json文件中读取数据 参数列表 :param abspath: 文件路径, 扩展名需为 ``.json`` 或 ``.gz`` :type abspath: ``字符串`` :param default: (默认 dict()) 如果文件路径不存在, 则会返回一个默认的Python对象。 :param compress: (默认 False) 是否从一个gzip压缩过的Json文件中读取数据。 请 参考 :func:`dump_js()<dump_js>` 获得更多信息. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize if compress: # check extension name if os.path.splitext(abspath)[1] != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: if os.path.splitext(abspath)[1] != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nLoading from %s..." % abspath) st = time.clock() if os.path.exists(abspath): # exists, then load if compress: with gzip.open(abspath, "rb") as f: js = json.loads(f.read().decode("utf-8")) else: with open(abspath, "r") as f: js = json.load(f) if enable_verbose: print("\tComplete! Elapse %.6f sec." % (time.clock() - st) ) return js else: if enable_verbose: print("\t%s not exists! cannot load! Create an default object " "instead" % abspath) return default def dump_js(js, abspath, fastmode=False, replace=False, compress=False, enable_verbose=True): """Dump Json serializable object to file. Provides multiple choice to customize the behavior. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param replace: (default False) If ``True``, when you dump json to a existing path, it silently overwrite it. If False, an exception will be raised. Default False setting is to prevent overwrite file by mistake. :type replace: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> dump_js(js, "test.json", replace=True) Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 将Python中可被序列化的"字典", "列表"以及他们的组合, 按照Json的编码方式写入文件 文件 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为``.json``或``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param fastmode: (默认 False) 当为``True``时, Json编码时不对Key进行排序, 也不 进行缩进排版。这样做写入的速度更快, 文件的大小也更小。 :type fastmode: "布尔值" :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: "布尔值" :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: "布尔值" :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: "布尔值" """ abspath = str(abspath) # try stringlize if compress: # check extension name root, ext = os.path.splitext(abspath) if ext != ".gz": if ext != ".tmp": raise Exception("compressed json has to use extension '.gz'!") else: _, ext = os.path.splitext(root) if ext != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: root, ext = os.path.splitext(abspath) if ext != ".json": if ext != ".tmp": raise Exception("file extension are not '.json'!") else: _, ext = os.path.splitext(root) if ext != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nDumping to %s..." % abspath) st = time.clock() if os.path.exists(abspath): # if exists, check replace option if replace: # replace existing file if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) else: # stop, print error message raise Exception("\tCANNOT WRITE to %s, it's already " "exists" % abspath) else: # if not exists, just write to it if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) if enable_verbose: print("\tComplete! Elapse %.6f sec" % (time.clock() - st) ) def safe_dump_js(js, abspath, fastmode=False, compress=False, enable_verbose=True): """A stable version of dump_js, silently overwrite existing file. When your program been interrupted, you lose nothing. Typically if your program is interrupted by any reason, it only leaves a incomplete file. If you use replace=True, then you also lose your old file. So a bettr way is to: 1. dump json to a temp file. 2. when it's done, rename it to #abspath, overwrite the old one. This way guarantee atomic write. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> safe_dump_js(js, "test.json") Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 在对文件进行写入时, 如果程序中断, 则会留下一个不完整的文件。如果你使用了覆盖式 写入, 则你同时也丢失了原文件。所以为了保证写操作的原子性(要么全部完成, 要么全部 都不完成), 更好的方法是: 首先将文件写入一个临时文件中, 完成后再讲文件重命名, 覆盖旧文件。这样即使中途程序被中断, 也仅仅是留下了一个未完成的临时文件而已, 不会 影响原文件。 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为 ``.json`` 或 ``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: ``布尔值`` :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize temp_abspath = "%s.tmp" % abspath dump_js(js, temp_abspath, fastmode=fastmode, replace=True, compress=compress, enable_verbose=enable_verbose) shutil.move(temp_abspath, abspath) def js2str(js, sort_keys=True, indent=4): """Encode js to nicely formatted human readable string. (utf-8 encoding) Usage:: >>> from weatherlab.lib.dataIO.js import js2str >>> s = js2str({"a": 1, "b": 2}) >>> print(s) { "a": 1, "b": 2 } **中文文档** 将可Json化的Python对象转化成格式化的字符串。 """ return json.dumps(js, sort_keys=sort_keys, indent=indent, separators=("," , ": ")) def prt_js(js, sort_keys=True, indent=4): """Print Json in pretty format. There's a standard module pprint, can pretty print python dict and list. But it doesn't support sorted key. That why we need this func. Usage:: >>> from weatherlab.lib.dataIO.js import prt_js >>> prt_js({"a": 1, "b": 2}) { "a": 1, "b": 2 } **中文文档** 以人类可读的方式打印可Json化的Python对象。 """ print(js2str(js, sort_keys, indent) ) ############ # Unittest # ############ if __name__ == "__main__": import unittest class JSUnittest(unittest.TestCase): def test_write_and_read(self): data = {"a": [1, 2], "b": ["是", "否"]} safe_dump_js(data, "data.json") data = load_js("data.json") self.assertEqual(data["a"][0], 1) self.assertEqual(data["b"][0], "是") def test_js2str(self): data = {"a": [1, 2], "b": ["是", "否"]} prt_js(data) def test_compress(self): data = {"a": list(range(32)), "b": list(range(32)),} safe_dump_js(data, "data.gz", compress=True) prt_js(load_js("data.gz", compress=True)) def tearDown(self): for path in ["data.json", "data.gz"]: try: os.remove(path) except: pass unittest.main()
MacHu-GWU/pyknackhq-project
pyknackhq/js.py
dump_js
python
def dump_js(js, abspath, fastmode=False, replace=False, compress=False, enable_verbose=True): abspath = str(abspath) # try stringlize if compress: # check extension name root, ext = os.path.splitext(abspath) if ext != ".gz": if ext != ".tmp": raise Exception("compressed json has to use extension '.gz'!") else: _, ext = os.path.splitext(root) if ext != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: root, ext = os.path.splitext(abspath) if ext != ".json": if ext != ".tmp": raise Exception("file extension are not '.json'!") else: _, ext = os.path.splitext(root) if ext != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nDumping to %s..." % abspath) st = time.clock() if os.path.exists(abspath): # if exists, check replace option if replace: # replace existing file if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) else: # stop, print error message raise Exception("\tCANNOT WRITE to %s, it's already " "exists" % abspath) else: # if not exists, just write to it if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) if enable_verbose: print("\tComplete! Elapse %.6f sec" % (time.clock() - st) )
Dump Json serializable object to file. Provides multiple choice to customize the behavior. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param replace: (default False) If ``True``, when you dump json to a existing path, it silently overwrite it. If False, an exception will be raised. Default False setting is to prevent overwrite file by mistake. :type replace: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> dump_js(js, "test.json", replace=True) Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 将Python中可被序列化的"字典", "列表"以及他们的组合, 按照Json的编码方式写入文件 文件 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为``.json``或``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param fastmode: (默认 False) 当为``True``时, Json编码时不对Key进行排序, 也不 进行缩进排版。这样做写入的速度更快, 文件的大小也更小。 :type fastmode: "布尔值" :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: "布尔值" :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: "布尔值" :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: "布尔值"
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/js.py#L125-L258
null
#!/usr/bin/env python # -*- coding: utf-8 -*- """ Module description ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ This module is re-pack of some json utility functions. - :func:`load_js`: Load Json from file. If file are not exists, returns user defined ``default value``. - :func:`dump_js`: Dump Json serializable object to file. - :func:`safe_dump_js`: An atomic write version of dump_js, silently overwrite existing file. - :func:`js2str`: Encode js to nicely formatted human readable string. - :func:`prt_js`: Print Json in pretty format. Highlight ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - :func:`load_js`, :func:`dump_js`, :func:`safe_dump_js` support gzip compress, size is **10 - 20 times** smaller in average. Compatibility ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - Python2: Yes - Python3: Yes Prerequisites ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - None Class, method, function, exception ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ """ from __future__ import print_function, unicode_literals import json, gzip import os, shutil import time def load_js(abspath, default=dict(), compress=False, enable_verbose=True): """Load Json from file. If file are not exists, returns ``default``. :param abspath: File path. Use absolute path as much as you can. File extension has to be ``.json`` or ``.gz``. (for compressed Json) :type abspath: string :param default: (default dict()) If ``abspath`` not exists, return the default Python object instead. :param compress: (default False) Load from a gzip compressed Json file. Check :func:`dump_js()<dump_js>` function for more information. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import load_js >>> load_js("test.json") # if you have a json file Loading from test.json... Complete! Elapse 0.000432 sec. {'a': 1, 'b': 2} **中文文档** 从Json文件中读取数据 参数列表 :param abspath: 文件路径, 扩展名需为 ``.json`` 或 ``.gz`` :type abspath: ``字符串`` :param default: (默认 dict()) 如果文件路径不存在, 则会返回一个默认的Python对象。 :param compress: (默认 False) 是否从一个gzip压缩过的Json文件中读取数据。 请 参考 :func:`dump_js()<dump_js>` 获得更多信息. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize if compress: # check extension name if os.path.splitext(abspath)[1] != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: if os.path.splitext(abspath)[1] != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nLoading from %s..." % abspath) st = time.clock() if os.path.exists(abspath): # exists, then load if compress: with gzip.open(abspath, "rb") as f: js = json.loads(f.read().decode("utf-8")) else: with open(abspath, "r") as f: js = json.load(f) if enable_verbose: print("\tComplete! Elapse %.6f sec." % (time.clock() - st) ) return js else: if enable_verbose: print("\t%s not exists! cannot load! Create an default object " "instead" % abspath) return default def dump_js(js, abspath, fastmode=False, replace=False, compress=False, enable_verbose=True): """Dump Json serializable object to file. Provides multiple choice to customize the behavior. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param replace: (default False) If ``True``, when you dump json to a existing path, it silently overwrite it. If False, an exception will be raised. Default False setting is to prevent overwrite file by mistake. :type replace: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> dump_js(js, "test.json", replace=True) Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 将Python中可被序列化的"字典", "列表"以及他们的组合, 按照Json的编码方式写入文件 文件 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为``.json``或``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param fastmode: (默认 False) 当为``True``时, Json编码时不对Key进行排序, 也不 进行缩进排版。这样做写入的速度更快, 文件的大小也更小。 :type fastmode: "布尔值" :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: "布尔值" :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: "布尔值" :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: "布尔值" """ abspath = str(abspath) # try stringlize if compress: # check extension name root, ext = os.path.splitext(abspath) if ext != ".gz": if ext != ".tmp": raise Exception("compressed json has to use extension '.gz'!") else: _, ext = os.path.splitext(root) if ext != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: root, ext = os.path.splitext(abspath) if ext != ".json": if ext != ".tmp": raise Exception("file extension are not '.json'!") else: _, ext = os.path.splitext(root) if ext != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nDumping to %s..." % abspath) st = time.clock() if os.path.exists(abspath): # if exists, check replace option if replace: # replace existing file if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) else: # stop, print error message raise Exception("\tCANNOT WRITE to %s, it's already " "exists" % abspath) else: # if not exists, just write to it if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) if enable_verbose: print("\tComplete! Elapse %.6f sec" % (time.clock() - st) ) def safe_dump_js(js, abspath, fastmode=False, compress=False, enable_verbose=True): """A stable version of dump_js, silently overwrite existing file. When your program been interrupted, you lose nothing. Typically if your program is interrupted by any reason, it only leaves a incomplete file. If you use replace=True, then you also lose your old file. So a bettr way is to: 1. dump json to a temp file. 2. when it's done, rename it to #abspath, overwrite the old one. This way guarantee atomic write. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> safe_dump_js(js, "test.json") Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 在对文件进行写入时, 如果程序中断, 则会留下一个不完整的文件。如果你使用了覆盖式 写入, 则你同时也丢失了原文件。所以为了保证写操作的原子性(要么全部完成, 要么全部 都不完成), 更好的方法是: 首先将文件写入一个临时文件中, 完成后再讲文件重命名, 覆盖旧文件。这样即使中途程序被中断, 也仅仅是留下了一个未完成的临时文件而已, 不会 影响原文件。 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为 ``.json`` 或 ``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: ``布尔值`` :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize temp_abspath = "%s.tmp" % abspath dump_js(js, temp_abspath, fastmode=fastmode, replace=True, compress=compress, enable_verbose=enable_verbose) shutil.move(temp_abspath, abspath) def js2str(js, sort_keys=True, indent=4): """Encode js to nicely formatted human readable string. (utf-8 encoding) Usage:: >>> from weatherlab.lib.dataIO.js import js2str >>> s = js2str({"a": 1, "b": 2}) >>> print(s) { "a": 1, "b": 2 } **中文文档** 将可Json化的Python对象转化成格式化的字符串。 """ return json.dumps(js, sort_keys=sort_keys, indent=indent, separators=("," , ": ")) def prt_js(js, sort_keys=True, indent=4): """Print Json in pretty format. There's a standard module pprint, can pretty print python dict and list. But it doesn't support sorted key. That why we need this func. Usage:: >>> from weatherlab.lib.dataIO.js import prt_js >>> prt_js({"a": 1, "b": 2}) { "a": 1, "b": 2 } **中文文档** 以人类可读的方式打印可Json化的Python对象。 """ print(js2str(js, sort_keys, indent) ) ############ # Unittest # ############ if __name__ == "__main__": import unittest class JSUnittest(unittest.TestCase): def test_write_and_read(self): data = {"a": [1, 2], "b": ["是", "否"]} safe_dump_js(data, "data.json") data = load_js("data.json") self.assertEqual(data["a"][0], 1) self.assertEqual(data["b"][0], "是") def test_js2str(self): data = {"a": [1, 2], "b": ["是", "否"]} prt_js(data) def test_compress(self): data = {"a": list(range(32)), "b": list(range(32)),} safe_dump_js(data, "data.gz", compress=True) prt_js(load_js("data.gz", compress=True)) def tearDown(self): for path in ["data.json", "data.gz"]: try: os.remove(path) except: pass unittest.main()
MacHu-GWU/pyknackhq-project
pyknackhq/js.py
safe_dump_js
python
def safe_dump_js(js, abspath, fastmode=False, compress=False, enable_verbose=True): abspath = str(abspath) # try stringlize temp_abspath = "%s.tmp" % abspath dump_js(js, temp_abspath, fastmode=fastmode, replace=True, compress=compress, enable_verbose=enable_verbose) shutil.move(temp_abspath, abspath)
A stable version of dump_js, silently overwrite existing file. When your program been interrupted, you lose nothing. Typically if your program is interrupted by any reason, it only leaves a incomplete file. If you use replace=True, then you also lose your old file. So a bettr way is to: 1. dump json to a temp file. 2. when it's done, rename it to #abspath, overwrite the old one. This way guarantee atomic write. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> safe_dump_js(js, "test.json") Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 在对文件进行写入时, 如果程序中断, 则会留下一个不完整的文件。如果你使用了覆盖式 写入, 则你同时也丢失了原文件。所以为了保证写操作的原子性(要么全部完成, 要么全部 都不完成), 更好的方法是: 首先将文件写入一个临时文件中, 完成后再讲文件重命名, 覆盖旧文件。这样即使中途程序被中断, 也仅仅是留下了一个未完成的临时文件而已, 不会 影响原文件。 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为 ``.json`` 或 ``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: ``布尔值`` :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值``
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/js.py#L260-L335
[ "def dump_js(js, abspath, \n fastmode=False, replace=False, compress=False, enable_verbose=True):\n \"\"\"Dump Json serializable object to file.\n Provides multiple choice to customize the behavior.\n\n :param js: Serializable python object.\n :type js: dict or list\n\n :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` \n (for compressed json).\n :type abspath: string\n\n :param fastmode: (default False) If ``True``, then dumping json without \n sorted keys and pretty indent, and it's faster and smaller in size.\n :type fastmode: boolean\n\n :param replace: (default False) If ``True``, when you dump json to a existing \n path, it silently overwrite it. If False, an exception will be raised.\n Default False setting is to prevent overwrite file by mistake.\n :type replace: boolean\n\n :param compress: (default False) If ``True``, use GNU program gzip to \n compress the json file. Disk usage can be greatly reduced. But you have \n to use :func:`load_js(abspath, compress=True)<load_js>` in loading.\n :type compress: boolean\n\n :param enable_verbose: (default True) Trigger for message.\n :type enable_verbose: boolean\n\n Usage::\n\n >>> from weatherlab.lib.dataIO.js import dump_js\n >>> js = {\"a\": 1, \"b\": 2}\n >>> dump_js(js, \"test.json\", replace=True)\n Dumping to test.json...\n Complete! Elapse 0.002432 sec\n\n **中文文档**\n\n 将Python中可被序列化的\"字典\", \"列表\"以及他们的组合, 按照Json的编码方式写入文件\n 文件\n\n 参数列表\n\n :param js: 可Json化的Python对象\n :type js: ``字典`` 或 ``列表``\n\n :param abspath: 写入文件的路径。扩展名必须为``.json``或``.gz``, 其中gz用于被压\n 缩的Json\n :type abspath: ``字符串``\n\n :param fastmode: (默认 False) 当为``True``时, Json编码时不对Key进行排序, 也不\n 进行缩进排版。这样做写入的速度更快, 文件的大小也更小。\n :type fastmode: \"布尔值\"\n\n :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖\n 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。\n :type replace: \"布尔值\"\n\n :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。\n 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数\n :func:`load_js(abspath, compress=True)<load_js>`.\n :type compress: \"布尔值\"\n\n :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭.\n :type enable_verbose: \"布尔值\"\n \"\"\"\n abspath = str(abspath) # try stringlize\n\n if compress: # check extension name\n root, ext = os.path.splitext(abspath)\n if ext != \".gz\":\n if ext != \".tmp\":\n raise Exception(\"compressed json has to use extension '.gz'!\")\n else:\n _, ext = os.path.splitext(root)\n if ext != \".gz\":\n raise Exception(\"compressed json has to use extension '.gz'!\")\n else:\n root, ext = os.path.splitext(abspath)\n if ext != \".json\":\n if ext != \".tmp\":\n raise Exception(\"file extension are not '.json'!\")\n else:\n _, ext = os.path.splitext(root)\n if ext != \".json\":\n raise Exception(\"file extension are not '.json'!\")\n\n if enable_verbose:\n print(\"\\nDumping to %s...\" % abspath)\n st = time.clock()\n\n if os.path.exists(abspath): # if exists, check replace option\n if replace: # replace existing file\n if fastmode: # no sort and indent, do the fastest dumping\n if compress:\n with gzip.open(abspath, \"wb\") as f:\n f.write(json.dumps(js).encode(\"utf-8\"))\n else:\n with open(abspath, \"w\") as f:\n json.dump(js, f)\n else:\n if compress:\n with gzip.open(abspath, \"wb\") as f:\n f.write(json.dumps(js, sort_keys=True,\n indent=4, separators=(\",\" , \": \")).encode(\"utf-8\"))\n else:\n with open(abspath, \"w\") as f:\n json.dump(js, f, sort_keys=True, \n indent=4, separators=(\",\" , \": \") )\n else: # stop, print error message\n raise Exception(\"\\tCANNOT WRITE to %s, it's already \"\n \"exists\" % abspath)\n\n else: # if not exists, just write to it\n if fastmode: # no sort and indent, do the fastest dumping\n if compress:\n with gzip.open(abspath, \"wb\") as f:\n f.write(json.dumps(js).encode(\"utf-8\"))\n else:\n with open(abspath, \"w\") as f:\n json.dump(js, f)\n else:\n if compress:\n with gzip.open(abspath, \"wb\") as f:\n f.write(json.dumps(js, sort_keys=True,\n indent=4, separators=(\",\" , \": \")).encode(\"utf-8\"))\n else:\n with open(abspath, \"w\") as f:\n json.dump(js, f, sort_keys=True, \n indent=4, separators=(\",\" , \": \") )\n\n if enable_verbose:\n print(\"\\tComplete! Elapse %.6f sec\" % (time.clock() - st) )\n" ]
#!/usr/bin/env python # -*- coding: utf-8 -*- """ Module description ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ This module is re-pack of some json utility functions. - :func:`load_js`: Load Json from file. If file are not exists, returns user defined ``default value``. - :func:`dump_js`: Dump Json serializable object to file. - :func:`safe_dump_js`: An atomic write version of dump_js, silently overwrite existing file. - :func:`js2str`: Encode js to nicely formatted human readable string. - :func:`prt_js`: Print Json in pretty format. Highlight ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - :func:`load_js`, :func:`dump_js`, :func:`safe_dump_js` support gzip compress, size is **10 - 20 times** smaller in average. Compatibility ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - Python2: Yes - Python3: Yes Prerequisites ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - None Class, method, function, exception ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ """ from __future__ import print_function, unicode_literals import json, gzip import os, shutil import time def load_js(abspath, default=dict(), compress=False, enable_verbose=True): """Load Json from file. If file are not exists, returns ``default``. :param abspath: File path. Use absolute path as much as you can. File extension has to be ``.json`` or ``.gz``. (for compressed Json) :type abspath: string :param default: (default dict()) If ``abspath`` not exists, return the default Python object instead. :param compress: (default False) Load from a gzip compressed Json file. Check :func:`dump_js()<dump_js>` function for more information. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import load_js >>> load_js("test.json") # if you have a json file Loading from test.json... Complete! Elapse 0.000432 sec. {'a': 1, 'b': 2} **中文文档** 从Json文件中读取数据 参数列表 :param abspath: 文件路径, 扩展名需为 ``.json`` 或 ``.gz`` :type abspath: ``字符串`` :param default: (默认 dict()) 如果文件路径不存在, 则会返回一个默认的Python对象。 :param compress: (默认 False) 是否从一个gzip压缩过的Json文件中读取数据。 请 参考 :func:`dump_js()<dump_js>` 获得更多信息. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize if compress: # check extension name if os.path.splitext(abspath)[1] != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: if os.path.splitext(abspath)[1] != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nLoading from %s..." % abspath) st = time.clock() if os.path.exists(abspath): # exists, then load if compress: with gzip.open(abspath, "rb") as f: js = json.loads(f.read().decode("utf-8")) else: with open(abspath, "r") as f: js = json.load(f) if enable_verbose: print("\tComplete! Elapse %.6f sec." % (time.clock() - st) ) return js else: if enable_verbose: print("\t%s not exists! cannot load! Create an default object " "instead" % abspath) return default def dump_js(js, abspath, fastmode=False, replace=False, compress=False, enable_verbose=True): """Dump Json serializable object to file. Provides multiple choice to customize the behavior. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param replace: (default False) If ``True``, when you dump json to a existing path, it silently overwrite it. If False, an exception will be raised. Default False setting is to prevent overwrite file by mistake. :type replace: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> dump_js(js, "test.json", replace=True) Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 将Python中可被序列化的"字典", "列表"以及他们的组合, 按照Json的编码方式写入文件 文件 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为``.json``或``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param fastmode: (默认 False) 当为``True``时, Json编码时不对Key进行排序, 也不 进行缩进排版。这样做写入的速度更快, 文件的大小也更小。 :type fastmode: "布尔值" :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: "布尔值" :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: "布尔值" :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: "布尔值" """ abspath = str(abspath) # try stringlize if compress: # check extension name root, ext = os.path.splitext(abspath) if ext != ".gz": if ext != ".tmp": raise Exception("compressed json has to use extension '.gz'!") else: _, ext = os.path.splitext(root) if ext != ".gz": raise Exception("compressed json has to use extension '.gz'!") else: root, ext = os.path.splitext(abspath) if ext != ".json": if ext != ".tmp": raise Exception("file extension are not '.json'!") else: _, ext = os.path.splitext(root) if ext != ".json": raise Exception("file extension are not '.json'!") if enable_verbose: print("\nDumping to %s..." % abspath) st = time.clock() if os.path.exists(abspath): # if exists, check replace option if replace: # replace existing file if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) else: # stop, print error message raise Exception("\tCANNOT WRITE to %s, it's already " "exists" % abspath) else: # if not exists, just write to it if fastmode: # no sort and indent, do the fastest dumping if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f) else: if compress: with gzip.open(abspath, "wb") as f: f.write(json.dumps(js, sort_keys=True, indent=4, separators=("," , ": ")).encode("utf-8")) else: with open(abspath, "w") as f: json.dump(js, f, sort_keys=True, indent=4, separators=("," , ": ") ) if enable_verbose: print("\tComplete! Elapse %.6f sec" % (time.clock() - st) ) def safe_dump_js(js, abspath, fastmode=False, compress=False, enable_verbose=True): """A stable version of dump_js, silently overwrite existing file. When your program been interrupted, you lose nothing. Typically if your program is interrupted by any reason, it only leaves a incomplete file. If you use replace=True, then you also lose your old file. So a bettr way is to: 1. dump json to a temp file. 2. when it's done, rename it to #abspath, overwrite the old one. This way guarantee atomic write. :param js: Serializable python object. :type js: dict or list :param abspath: ``save as`` path, file extension has to be ``.json`` or ``.gz`` (for compressed json). :type abspath: string :param fastmode: (default False) If ``True``, then dumping json without sorted keys and pretty indent, and it's faster and smaller in size. :type fastmode: boolean :param compress: (default False) If ``True``, use GNU program gzip to compress the json file. Disk usage can be greatly reduced. But you have to use :func:`load_js(abspath, compress=True)<load_js>` in loading. :type compress: boolean :param enable_verbose: (default True) Trigger for message. :type enable_verbose: boolean Usage:: >>> from weatherlab.lib.dataIO.js import dump_js >>> js = {"a": 1, "b": 2} >>> safe_dump_js(js, "test.json") Dumping to test.json... Complete! Elapse 0.002432 sec **中文文档** 在对文件进行写入时, 如果程序中断, 则会留下一个不完整的文件。如果你使用了覆盖式 写入, 则你同时也丢失了原文件。所以为了保证写操作的原子性(要么全部完成, 要么全部 都不完成), 更好的方法是: 首先将文件写入一个临时文件中, 完成后再讲文件重命名, 覆盖旧文件。这样即使中途程序被中断, 也仅仅是留下了一个未完成的临时文件而已, 不会 影响原文件。 参数列表 :param js: 可Json化的Python对象 :type js: ``字典`` 或 ``列表`` :param abspath: 写入文件的路径。扩展名必须为 ``.json`` 或 ``.gz``, 其中gz用于被压 缩的Json :type abspath: ``字符串`` :param replace: (默认 False) 当为``True``时, 如果写入路径已经存在, 则会自动覆盖 原文件。而为``False``时, 则会抛出异常。防止误操作覆盖源文件。 :type replace: ``布尔值`` :param compress: (默认 False) 当为``True``时, 使用开源压缩标准gzip压缩Json文件。 通常能让文件大小缩小10-20倍不等。如要读取文件, 则需要使用函数 :func:`load_js(abspath, compress=True)<load_js>`. :type compress: ``布尔值`` :param enable_verbose: (默认 True) 是否打开信息提示开关, 批处理时建议关闭. :type enable_verbose: ``布尔值`` """ abspath = str(abspath) # try stringlize temp_abspath = "%s.tmp" % abspath dump_js(js, temp_abspath, fastmode=fastmode, replace=True, compress=compress, enable_verbose=enable_verbose) shutil.move(temp_abspath, abspath) def js2str(js, sort_keys=True, indent=4): """Encode js to nicely formatted human readable string. (utf-8 encoding) Usage:: >>> from weatherlab.lib.dataIO.js import js2str >>> s = js2str({"a": 1, "b": 2}) >>> print(s) { "a": 1, "b": 2 } **中文文档** 将可Json化的Python对象转化成格式化的字符串。 """ return json.dumps(js, sort_keys=sort_keys, indent=indent, separators=("," , ": ")) def prt_js(js, sort_keys=True, indent=4): """Print Json in pretty format. There's a standard module pprint, can pretty print python dict and list. But it doesn't support sorted key. That why we need this func. Usage:: >>> from weatherlab.lib.dataIO.js import prt_js >>> prt_js({"a": 1, "b": 2}) { "a": 1, "b": 2 } **中文文档** 以人类可读的方式打印可Json化的Python对象。 """ print(js2str(js, sort_keys, indent) ) ############ # Unittest # ############ if __name__ == "__main__": import unittest class JSUnittest(unittest.TestCase): def test_write_and_read(self): data = {"a": [1, 2], "b": ["是", "否"]} safe_dump_js(data, "data.json") data = load_js("data.json") self.assertEqual(data["a"][0], 1) self.assertEqual(data["b"][0], "是") def test_js2str(self): data = {"a": [1, 2], "b": ["是", "否"]} prt_js(data) def test_compress(self): data = {"a": list(range(32)), "b": list(range(32)),} safe_dump_js(data, "data.gz", compress=True) prt_js(load_js("data.gz", compress=True)) def tearDown(self): for path in ["data.json", "data.gz"]: try: os.remove(path) except: pass unittest.main()
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.convert_keys
python
def convert_keys(self, pydict): new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict
Convert field_name to field_key. {"field_name": value} => {"field_key": value}
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L40-L48
[ "def get_field_key(self, key, using_name=True):\n \"\"\"Given a field key or name, return it's field key.\n \"\"\"\n try:\n if using_name:\n return self.f_name[key].key\n else:\n return self.f[key].key\n except KeyError:\n raise ValueError(\"'%s' are not found!\" % key)\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.get_html_values
python
def get_html_values(self, pydict, recovery_name=True): new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict
Convert naive get response data to human readable field name format. using html data format.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L50-L62
null
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.get_raw_values
python
def get_raw_values(self, pydict, recovery_name=True): new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict
Convert naive get response data to human readable field name format. using raw data format.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L64-L77
null
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.convert_values
python
def convert_values(self, pydict): new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict
Convert knackhq data type instance to json friendly data.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L79-L88
null
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.insert_one
python
def insert_one(self, data, using_name=True): data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res
Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L93-L113
[ "def convert_keys(self, pydict):\n \"\"\"Convert field_name to field_key.\n\n {\"field_name\": value} => {\"field_key\": value}\n \"\"\"\n new_dict = dict()\n for key, value in pydict.items():\n new_dict[self.get_field_key(key)] = value\n return new_dict\n", "def convert_values(self, pydict):\n \"\"\"Convert knackhq data type instance to json friendly data.\n \"\"\"\n new_dict = dict()\n for key, value in pydict.items():\n try: # is it's BaseDataType Instance\n new_dict[key] = value._data\n except AttributeError:\n new_dict[key] = value\n return new_dict\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.insert
python
def insert(self, data, using_name=True): if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name)
Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L115-L130
[ "def insert_one(self, data, using_name=True):\n \"\"\"Insert one record.\n\n Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create\n\n For more information of the raw structure of all data type, read this:\n http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types\n\n :param data: dict type data\n :param using_name: if you are using field_name in data,\n please set using_name = True (it's the default), otherwise, False\n\n **中文文档**\n\n 插入一条记录\n \"\"\"\n data = self.convert_values(data)\n if using_name:\n data = self.convert_keys(data)\n res = self.post(self.post_url, data)\n return res\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.find_one
python
def find_one(self, id_, raw=True, recovery_name=True): url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res
Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L132-L164
[ "def get_html_values(self, pydict, recovery_name=True):\n \"\"\"Convert naive get response data to human readable field name format.\n\n using html data format.\n \"\"\"\n new_dict = {\"id\": pydict[\"id\"]}\n for field in self:\n if field.key in pydict:\n if recovery_name:\n new_dict[field.name] = pydict[field.key]\n else:\n new_dict[field.key] = pydict[field.key]\n return new_dict\n", "def get_raw_values(self, pydict, recovery_name=True):\n \"\"\"Convert naive get response data to human readable field name format.\n\n using raw data format.\n \"\"\"\n new_dict = {\"id\": pydict[\"id\"]}\n for field in self:\n raw_key = \"%s_raw\" % field.key\n if raw_key in pydict:\n if recovery_name:\n new_dict[field.name] = pydict[raw_key]\n else:\n new_dict[field.key] = pydict[raw_key]\n return new_dict\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.find
python
def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res
Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L166-L253
[ "def get_field_key(self, key, using_name=True):\n \"\"\"Given a field key or name, return it's field key.\n \"\"\"\n try:\n if using_name:\n return self.f_name[key].key\n else:\n return self.f[key].key\n except KeyError:\n raise ValueError(\"'%s' are not found!\" % key)\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.update_one
python
def update_one(self, id_, data, using_name=True): data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res
Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L255-L276
[ "def convert_keys(self, pydict):\n \"\"\"Convert field_name to field_key.\n\n {\"field_name\": value} => {\"field_key\": value}\n \"\"\"\n new_dict = dict()\n for key, value in pydict.items():\n new_dict[self.get_field_key(key)] = value\n return new_dict\n", "def convert_values(self, pydict):\n \"\"\"Convert knackhq data type instance to json friendly data.\n \"\"\"\n new_dict = dict()\n for key, value in pydict.items():\n try: # is it's BaseDataType Instance\n new_dict[key] = value._data\n except AttributeError:\n new_dict[key] = value\n return new_dict\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.delete_one
python
def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res
Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L278-L292
null
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
Collection.delete_all
python
def delete_all(self): for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L294-L302
[ "def find(self, filter=list(), \n sort_field=None, sort_order=None, \n page=None, rows_per_page=None,\n using_name=True, data_only=True, raw=True, recovery_name=True):\n \"\"\"Execute a find query.\n\n Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve\n\n :param filter: list of criterions. For more information: \n http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search\n :param sort_field: field_name or field_id, taking field_name by default.\n if using field_id, please set using_name = False.\n :param sort_order: -1 or 1, 1 means ascending, -1 means descending\n :param page and rows_per_page: skip first #page * #rows_per_page, \n returns #rows_per_page of records. For more information:\n http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination\n :param using_name: if you are using field_name in filter and sort_field, \n please set using_name = True (it's the default), otherwise, False\n :param data_only: set True you only need the data or the full api\n response\n :param raw: Default True, set True if you want the data in raw format. \n Otherwise, html format\n :param recovery_name: Default True, set True if you want field name\n instead of field key\n\n **中文文档**\n\n 返回多条记录\n \"\"\"\n if using_name: \n for criterion in filter:\n criterion[\"field\"] = self.get_field_key(criterion[\"field\"])\n\n if sort_field:\n sort_field = self.get_field_key(sort_field)\n\n if sort_order is None:\n pass\n elif sort_order == 1:\n sort_order = \"asc\"\n elif sort_order == -1:\n sort_order = \"desc\"\n else:\n raise ValueError\n\n params = dict()\n if len(filter) >= 1:\n params[\"filters\"] = json.dumps(filter)\n\n if sort_field:\n params[\"sort_field\"] = sort_field\n params[\"sort_order\"] = sort_order\n\n if (page is not None) \\\n and (rows_per_page is not None) \\\n and isinstance(page, int) \\\n and isinstance(rows_per_page, int) \\\n and (page >= 1) \\\n and (rows_per_page >= 1):\n params[\"page\"] = page\n params[\"rows_per_page\"] = rows_per_page\n\n res = self.get(self.get_url, params)\n\n # handle data_only and recovery\n if data_only:\n try:\n res = res[\"records\"]\n if raw:\n res = [self.get_raw_values(data, recovery_name) for data in res]\n else:\n res = [self.get_html_values(data, recovery_name) for data in res]\n except KeyError:\n pass\n else:\n if raw:\n try:\n res[\"records\"] = [\n self.get_raw_values(data, recovery_name) for data in res[\"records\"]]\n except KeyError:\n pass\n else:\n try:\n res[\"records\"] = [\n self.get_html_values(data, recovery_name) for data in res[\"records\"]]\n except KeyError:\n pass\n return res\n", "def delete_one(self, id_):\n \"\"\"Delete one record.\n\n Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete\n\n :param id_: record id_\n\n **中文文档**\n\n 删除一条记录\n \"\"\" \n url = \"https://api.knackhq.com/v1/objects/%s/records/%s\" % (\n self.key, id_)\n res = self.delete(url)\n return res\n" ]
class Collection(Object): """A collection is the equivalent of an RDBMS table, collection of MongoDB and object of Knackhq. Most of CRUD method can be executed using this. - :meth:`~Collection.insert_one` - :meth:`~Collection.insert` - :meth:`~Collection.find_one` - :meth:`~Collection.find` - :meth:`~Collection.update_one` - :meth:`~Collection.delete_one` - :meth:`~Collection.delete_all` """ def __str__(self): return "Collection('%s')" % self.name def __repr__(self): return "Collection(key='%s', name='%s')" % (self.key, self.name) @staticmethod def from_dict(d): return Collection(**d) @property def get_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key @property def post_url(self): return "https://api.knackhq.com/v1/objects/%s/records" % self.key def convert_keys(self, pydict): """Convert field_name to field_key. {"field_name": value} => {"field_key": value} """ new_dict = dict() for key, value in pydict.items(): new_dict[self.get_field_key(key)] = value return new_dict def get_html_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using html data format. """ new_dict = {"id": pydict["id"]} for field in self: if field.key in pydict: if recovery_name: new_dict[field.name] = pydict[field.key] else: new_dict[field.key] = pydict[field.key] return new_dict def get_raw_values(self, pydict, recovery_name=True): """Convert naive get response data to human readable field name format. using raw data format. """ new_dict = {"id": pydict["id"]} for field in self: raw_key = "%s_raw" % field.key if raw_key in pydict: if recovery_name: new_dict[field.name] = pydict[raw_key] else: new_dict[field.key] = pydict[raw_key] return new_dict def convert_values(self, pydict): """Convert knackhq data type instance to json friendly data. """ new_dict = dict() for key, value in pydict.items(): try: # is it's BaseDataType Instance new_dict[key] = value._data except AttributeError: new_dict[key] = value return new_dict #-------------------------------------------------------------------------# # CRUD method # #-------------------------------------------------------------------------# def insert_one(self, data, using_name=True): """Insert one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#create For more information of the raw structure of all data type, read this: http://helpdesk.knackhq.com/support/solutions/articles/5000446405-field-types :param data: dict type data :param using_name: if you are using field_name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入一条记录 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) res = self.post(self.post_url, data) return res def insert(self, data, using_name=True): """Insert one or many records. :param data: dict type data or list of dict :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 插入多条记录 """ if isinstance(data, list): # if iterable, insert one by one for d in data: self.insert_one(d, using_name=using_name) else: # not iterable, execute insert_one self.insert_one(data, using_name=using_name) def find_one(self, id_, raw=True, recovery_name=True): """Find one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param id_: record id_ :param using_name: if you are using field name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.get(url) if raw: try: res = self.get_raw_values(res, recovery_name=recovery_name) except: pass else: try: res = self.get_html_values(res, recovery_name=recovery_name) except: pass return res def find(self, filter=list(), sort_field=None, sort_order=None, page=None, rows_per_page=None, using_name=True, data_only=True, raw=True, recovery_name=True): """Execute a find query. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#retrieve :param filter: list of criterions. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000447623-api-reference-filters-search :param sort_field: field_name or field_id, taking field_name by default. if using field_id, please set using_name = False. :param sort_order: -1 or 1, 1 means ascending, -1 means descending :param page and rows_per_page: skip first #page * #rows_per_page, returns #rows_per_page of records. For more information: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#pagination :param using_name: if you are using field_name in filter and sort_field, please set using_name = True (it's the default), otherwise, False :param data_only: set True you only need the data or the full api response :param raw: Default True, set True if you want the data in raw format. Otherwise, html format :param recovery_name: Default True, set True if you want field name instead of field key **中文文档** 返回多条记录 """ if using_name: for criterion in filter: criterion["field"] = self.get_field_key(criterion["field"]) if sort_field: sort_field = self.get_field_key(sort_field) if sort_order is None: pass elif sort_order == 1: sort_order = "asc" elif sort_order == -1: sort_order = "desc" else: raise ValueError params = dict() if len(filter) >= 1: params["filters"] = json.dumps(filter) if sort_field: params["sort_field"] = sort_field params["sort_order"] = sort_order if (page is not None) \ and (rows_per_page is not None) \ and isinstance(page, int) \ and isinstance(rows_per_page, int) \ and (page >= 1) \ and (rows_per_page >= 1): params["page"] = page params["rows_per_page"] = rows_per_page res = self.get(self.get_url, params) # handle data_only and recovery if data_only: try: res = res["records"] if raw: res = [self.get_raw_values(data, recovery_name) for data in res] else: res = [self.get_html_values(data, recovery_name) for data in res] except KeyError: pass else: if raw: try: res["records"] = [ self.get_raw_values(data, recovery_name) for data in res["records"]] except KeyError: pass else: try: res["records"] = [ self.get_html_values(data, recovery_name) for data in res["records"]] except KeyError: pass return res def update_one(self, id_, data, using_name=True): """Update one record. Any fields you don't specify will remain unchanged. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#update :param id_: record id_ :param data: the new data fields and values :param using_name: if you are using field name in data, please set using_name = True (it's the default), otherwise, False **中文文档** 对一条记录进行更新 """ data = self.convert_values(data) if using_name: data = self.convert_keys(data) url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.put(url, data) return res def delete_one(self, id_): """Delete one record. Ref: http://helpdesk.knackhq.com/support/solutions/articles/5000446111-api-reference-root-access#delete :param id_: record id_ **中文文档** 删除一条记录 """ url = "https://api.knackhq.com/v1/objects/%s/records/%s" % ( self.key, id_) res = self.delete(url) return res def delete_all(self): """Delete all record in the table/collection of this object. **中文文档** 删除表中的所有记录 """ for record in self.find(using_name=False, data_only=True): res = self.delete_one(record["id"])
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
KnackhqAuth.get
python
def get(self, url, params=dict()): try: res = requests.get(url, headers=self.headers, params=params) return json.loads(res.text) except Exception as e: print(e) return "error"
Http get method wrapper, to support search.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L330-L338
null
class KnackhqAuth(object): """Knackhq API authentication class. :param application_id: str type, Application ID :param api_key: str type, API Key To get your Application ID and API Key, read this tutorial: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#key """ def __init__(self, application_id, api_key): self.application_id = application_id self.api_key = api_key self.headers = { "X-Knack-Application-Id": self.application_id, "X-Knack-REST-API-Key": self.api_key, "Content-Type": "application/json", } @staticmethod def from_dict(d): return KnackhqAuth(**d) @staticmethod def from_json(abspath): return KnackhqAuth.from_dict(load_js(abspath, enable_verbose=False)) def post(self, url, data): """Http post method wrapper, to support insert. """ try: res = requests.post( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error" def put(self, url, data): """Http put method wrapper, to support update. """ try: res = requests.put( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error" def delete(self, url): """Http delete method wrapper, to support delete. """ try: res = requests.delete(url, headers=self.headers) return json.loads(res.text) except Exception as e: print(e) return "error"
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
KnackhqAuth.post
python
def post(self, url, data): try: res = requests.post( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error"
Http post method wrapper, to support insert.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L340-L349
null
class KnackhqAuth(object): """Knackhq API authentication class. :param application_id: str type, Application ID :param api_key: str type, API Key To get your Application ID and API Key, read this tutorial: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#key """ def __init__(self, application_id, api_key): self.application_id = application_id self.api_key = api_key self.headers = { "X-Knack-Application-Id": self.application_id, "X-Knack-REST-API-Key": self.api_key, "Content-Type": "application/json", } @staticmethod def from_dict(d): return KnackhqAuth(**d) @staticmethod def from_json(abspath): return KnackhqAuth.from_dict(load_js(abspath, enable_verbose=False)) def get(self, url, params=dict()): """Http get method wrapper, to support search. """ try: res = requests.get(url, headers=self.headers, params=params) return json.loads(res.text) except Exception as e: print(e) return "error" def put(self, url, data): """Http put method wrapper, to support update. """ try: res = requests.put( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error" def delete(self, url): """Http delete method wrapper, to support delete. """ try: res = requests.delete(url, headers=self.headers) return json.loads(res.text) except Exception as e: print(e) return "error"
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
KnackhqAuth.delete
python
def delete(self, url): try: res = requests.delete(url, headers=self.headers) return json.loads(res.text) except Exception as e: print(e) return "error"
Http delete method wrapper, to support delete.
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L362-L370
null
class KnackhqAuth(object): """Knackhq API authentication class. :param application_id: str type, Application ID :param api_key: str type, API Key To get your Application ID and API Key, read this tutorial: http://helpdesk.knackhq.com/support/solutions/articles/5000444173-working-with-the-api#key """ def __init__(self, application_id, api_key): self.application_id = application_id self.api_key = api_key self.headers = { "X-Knack-Application-Id": self.application_id, "X-Knack-REST-API-Key": self.api_key, "Content-Type": "application/json", } @staticmethod def from_dict(d): return KnackhqAuth(**d) @staticmethod def from_json(abspath): return KnackhqAuth.from_dict(load_js(abspath, enable_verbose=False)) def get(self, url, params=dict()): """Http get method wrapper, to support search. """ try: res = requests.get(url, headers=self.headers, params=params) return json.loads(res.text) except Exception as e: print(e) return "error" def post(self, url, data): """Http post method wrapper, to support insert. """ try: res = requests.post( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error" def put(self, url, data): """Http put method wrapper, to support update. """ try: res = requests.put( url, headers=self.headers, data=json.dumps(data)) return json.loads(res.text) except Exception as e: print(e) return "error"
MacHu-GWU/pyknackhq-project
pyknackhq/client.py
KnackhqClient.get_collection
python
def get_collection(self, key, using_name=True): object_ = self.application.get_object(key, using_name=using_name) collection = Collection.from_dict(object_.__dict__) for http_cmd in ["get", "post", "put", "delete"]: collection.__setattr__(http_cmd, self.auth.__getattribute__(http_cmd)) return collection
Get :class:`Collection` instance. :param key: object_key or object_name :param using_name: True if getting object by object name
train
https://github.com/MacHu-GWU/pyknackhq-project/blob/dd937f24d7b0a351ba3818eb746c31b29a8cc341/pyknackhq/client.py#L410-L420
[ "def from_dict(d):\n return Collection(**d)\n" ]
class KnackhqClient(object): """Knackhq API client class. :param auth: A :class:`KnackAuth` instance. :param application: An :class:`~pyknackhq.schema.Application` instance. If it is not given, the client automatically pull it from knack server. How to construct a knackhq api client:: from pyknackhq import KnackhqClient, KnackhqAuth auth = KnackhqAuth(application_id="your app id", api_key="your api key") client = KnackClient(auth=auth) """ def __init__(self, auth, application=None): self.auth = auth if isinstance(application, Application): self.application = application else: # get the schema json, construct Application instance res = requests.get( "https://api.knackhq.com/v1/applications/%s" % self.auth.application_id) self.application = Application.from_dict(json.loads(res.text)) def __str__(self): return "KnackhqClient(application='%s')" % self.application def __repr__(self): return str(self) @property def all_object_key(self): return self.application.all_object_key @property def all_object_name(self): return self.application.all_object_name def get_collection(self, key, using_name=True): """Get :class:`Collection` instance. :param key: object_key or object_name :param using_name: True if getting object by object name """ object_ = self.application.get_object(key, using_name=using_name) collection = Collection.from_dict(object_.__dict__) for http_cmd in ["get", "post", "put", "delete"]: collection.__setattr__(http_cmd, self.auth.__getattribute__(http_cmd)) return collection def export_schema(self, abspath): """Export application detailed information to a nicely formatted json file. """ self.application.to_json(abspath)
ryanjdillon/pylleo
pylleo/utils_bokeh.py
create_bokeh_server
python
def create_bokeh_server(io_loop, files, argvs, host, port): '''Start bokeh server with applications paths''' from bokeh.server.server import Server from bokeh.command.util import build_single_handler_applications # Turn file paths into bokeh apps apps = build_single_handler_applications(files, argvs) # kwargs lifted from bokeh serve call to Server, with created io_loop kwargs = { 'io_loop':io_loop, 'generate_session_ids':True, 'redirect_root':True, 'use_x_headers':False, 'secret_key':None, 'num_procs':1, 'host': host, 'sign_sessions':False, 'develop':False, 'port':port, 'use_index':True } server = Server(apps,**kwargs) return server
Start bokeh server with applications paths
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils_bokeh.py#L2-L26
null
def run_server_to_disconnect(files, port=5000, new='tab'): def start_bokeh(io_loop): '''Start the `io_loop`''' io_loop.start() return None def launch_app(host, app_name, new): '''Lauch app in browser Ideally this would `bokeh.util.browser.view()`, but it doesn't work ''' import webbrowser # Map method strings to webbrowser method options = {'current':0, 'window':1, 'tab':2} # Concatenate url and open in browser, creating a session app_url = 'http://{}/{}'.format(host, app_name) print('Opening `{}` in browser'.format(app_url)) webbrowser.open(app_url, new=options[new]) return None def server_loop(server, io_loop): '''Check connections once session created and close on disconnect''' import time connected = [True,] session_loaded = False while any(connected): # Check if no session started on server sessions = server.get_sessions() if not session_loaded: if sessions: session_loaded = True # Once 1+ sessions started, check for no connections else: # List of bools for each session connected = [True,]*len(sessions) # Set `connected` item false no connections on session for i in range(len(sessions)): if sessions[i].connection_count == 0: connected[i] = False # Keep the pace down time.sleep(2) # Stop server once opened session connections closed io_loop.stop() return None import os import threading import tornado.ioloop import tornado.autoreload import time # Initialize some values, sanatize the paths to the bokeh plots argvs = {} app_names = [] for path in files: argvs[path] = None app_names.append(os.path.splitext(os.path.split(path)[1])[0]) # Concate hostname/port for creating handlers, launching apps host = 'localhost:{}'.format(port) # Initialize the tornado server io_loop = tornado.ioloop.IOLoop.instance() tornado.autoreload.start(io_loop) # Add the io_loop to the bokeh server server = create_bokeh_server(io_loop, files, argvs, host, port) print('Starting the server on {}'.format(host)) args = (io_loop,) th_startup = threading.Thread(target=start_bokeh, args=args) th_startup.start() # Launch each application in own tab or window th_launch = [None,]*len(app_names) for i in range(len(app_names)): args = (host, app_names[i], new) th_launch[i] = threading.Thread(target=launch_app, args=args) th_launch[i].start() # Delay to allow tabs to open in same browser window time.sleep(2) # Run session connection test, then stop `io_loop` args = (server, io_loop) th_shutdown = threading.Thread(target=server_loop, args=args) th_shutdown.start() return None
ryanjdillon/pylleo
pylleo/utils.py
predict_encoding
python
def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding']
Get file encoding of a text file
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L15-L24
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model)) def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot)) def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
ryanjdillon/pylleo
pylleo/utils.py
get_n_header
python
def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header
Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L27-L52
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding'] def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model)) def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot)) def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
ryanjdillon/pylleo
pylleo/utils.py
get_tag_params
python
def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model))
Load param strs and n_header based on model of tag model
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L55-L67
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding'] def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot)) def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
ryanjdillon/pylleo
pylleo/utils.py
find_file
python
def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path
Find path of file in directory containing the search string
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L70-L85
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding'] def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model)) def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot)) def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
ryanjdillon/pylleo
pylleo/utils.py
nearest
python
def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot))
Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot`
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L104-L119
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding'] def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model)) def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
ryanjdillon/pylleo
pylleo/utils.py
parse_experiment_params
python
def parse_experiment_params(name_exp): '''Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters ''' if ('/' in name_exp) or ('\\' in name_exp): raise ValueError("The path {} appears to be a path. Please pass " "only the data directory's name (i.e. the " "experiment name)".format(name_exp)) tag_params = dict() tag_params['experiment'] = name_exp tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','') tag_params['tag_id'] = name_exp.split('_')[2] tag_params['animal'] = name_exp.split('_')[3] tag_params['notes'] = name_exp.split('_')[4] return tag_params
Parse experiment parameters from the data directory name Args ---- name_exp: str Name of data directory with experiment parameters Returns ------- tag_params: dict of str Dictionary of parsed experiment parameters
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/utils.py#L122-L147
null
def get_testdata_path(tag_model): '''Get path to sample data directory for given tag model''' import os tag_model = tag_model.upper().replace('-','').replace('_','') sample_path = os.path.join('../datasets/{}'.format(tag_model)) if not os.path.isdir(sample_path): raise FileNotFoundError('No sample dataset found for tag ' '{}.'.format(tag_model)) return sample_path def predict_encoding(file_path, n_lines=20): '''Get file encoding of a text file''' import chardet # Open the file as binary data with open(file_path, 'rb') as f: # Join binary lines for specified number of lines rawdata = b''.join([f.readline() for _ in range(n_lines)]) return chardet.detect(rawdata)['encoding'] def get_n_header(f, header_char='"'): '''Get the nummber of header rows in a Little Leonardo data file Args ---- f : file stream File handle for the file from which header rows will be read header_char: str Character array at beginning of each header line Returns ------- n_header: int Number of header rows in Little Leonardo data file ''' n_header = 0 reading_headers = True while reading_headers: line = f.readline() if line.startswith(header_char): n_header += 1 else: reading_headers = False return n_header def get_tag_params(tag_model): '''Load param strs and n_header based on model of tag model''' tag_model = tag_model.replace('-', '') tags = dict() tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z', 'Depth', 'Propeller', 'Temperature'] # Return tag parameters if found, else raise error if tag_model in tags: return tags[tag_model] else: raise KeyError('{} not found in tag dictionary'.format(tag_model)) def find_file(path_dir, search_str, file_ext): '''Find path of file in directory containing the search string''' import os file_path = None for file_name in os.listdir(path_dir): if (search_str in file_name) and (file_name.endswith(file_ext)): file_path = os.path.join(path_dir, file_name) break if file_path == None: raise SystemError('No file found containing string: ' '{}.'.format(search_str)) return file_path def posix_string(s): '''Return string in lower case with spaces and dashes as underscores Args ---- s: str string to modify Returns ------- s_mod: str string with ` ` and `-` replaced with `_` ''' return s.lower().replace(' ','_').replace('-','_') def nearest(items, pivot): '''Find nearest value in array, including datetimes Args ---- items: iterable List of values from which to find nearest value to `pivot` pivot: int or float Value to find nearest of in `items` Returns ------- nearest: int or float Value in items nearest to `pivot` ''' return min(items, key=lambda x: abs(x - pivot))
ryanjdillon/pylleo
pylleo/lleoio.py
read_meta
python
def read_meta(path_dir, tag_model, tag_id): '''Read meta data from Little Leonardo data header rows Args ---- path_dir: str Parent directory containing lleo data files tag_model: str Little Leonardo tag model name tag_id: str, int Little Leonardo tag ID number Returns ------- meta: dict dictionary with meta data from header lines of lleo data files ''' from collections import OrderedDict import os import yamlord from . import utils def _parse_meta_line(line): '''Return key, value pair parsed from data header line''' # Parse the key and its value from the line key, val = line.replace(':', '').replace('"', '').split(',') return key.strip(), val.strip() def _read_meta_all(f, meta, n_header): '''Read all meta data from header rows of data file''' # Skip 'File name' line f.seek(0) _ = f.readline() # Create child dictionary for channel / file line = f.readline() key_ch, val_ch = _parse_meta_line(line) val_ch = utils.posix_string(val_ch) meta['parameters'][val_ch] = OrderedDict() # Write header values to channel dict for _ in range(n_header-2): line = f.readline() key, val = _parse_meta_line(line) meta['parameters'][val_ch][key] = val.strip() return meta def _create_meta(path_dir, tag_model, tag_id): '''Create meta data dictionary''' import datetime from . import utils param_strs = utils.get_tag_params(tag_model) # Create dictionary of meta data meta = OrderedDict() # Create fields for the parameters in data directory name exp_name = os.path.split(path_dir)[1] params_tag = utils.parse_experiment_params(exp_name) for key, value in params_tag.items(): meta[key] = value fmt = "%Y-%m-%d %H:%M:%S" meta['date_modified'] = datetime.datetime.now().strftime(fmt) meta['parameters'] = OrderedDict() for param_str in param_strs: print('Create meta entry for {}'.format(param_str)) path_file = utils.find_file(path_dir, param_str, '.TXT') # Get number of header rows enc = utils.predict_encoding(path_file, n_lines=20) with open(path_file, 'r', encoding=enc) as f: n_header = utils.get_n_header(f) f.seek(0) meta = _read_meta_all(f, meta, n_header=n_header) return meta # Load meta data from YAML file if it already exists meta_yaml_path = os.path.join(path_dir, 'meta.yml') # Load file if exists else create if os.path.isfile(meta_yaml_path): meta = yamlord.read_yaml(meta_yaml_path) # Else create meta dictionary and save to YAML else: meta = _create_meta(path_dir, tag_model, tag_id) yamlord.write_yaml(meta, meta_yaml_path) return meta
Read meta data from Little Leonardo data header rows Args ---- path_dir: str Parent directory containing lleo data files tag_model: str Little Leonardo tag model name tag_id: str, int Little Leonardo tag ID number Returns ------- meta: dict dictionary with meta data from header lines of lleo data files
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleoio.py#L2-L103
[ "def _create_meta(path_dir, tag_model, tag_id):\n '''Create meta data dictionary'''\n import datetime\n from . import utils\n\n param_strs = utils.get_tag_params(tag_model)\n\n # Create dictionary of meta data\n meta = OrderedDict()\n\n # Create fields for the parameters in data directory name\n exp_name = os.path.split(path_dir)[1]\n params_tag = utils.parse_experiment_params(exp_name)\n for key, value in params_tag.items():\n meta[key] = value\n\n fmt = \"%Y-%m-%d %H:%M:%S\"\n meta['date_modified'] = datetime.datetime.now().strftime(fmt)\n\n meta['parameters'] = OrderedDict()\n\n for param_str in param_strs:\n print('Create meta entry for {}'.format(param_str))\n\n path_file = utils.find_file(path_dir, param_str, '.TXT')\n # Get number of header rows\n enc = utils.predict_encoding(path_file, n_lines=20)\n with open(path_file, 'r', encoding=enc) as f:\n n_header = utils.get_n_header(f)\n f.seek(0)\n meta = _read_meta_all(f, meta, n_header=n_header)\n\n return meta\n" ]
def read_data(meta, path_dir, sample_f=1, decimate=False, overwrite=False): '''Read accelerometry data from leonardo txt files Args ---- meta: dict Dictionary of meta data from header lines of lleo data files path_dir: str Parent directory containing lleo data files sample_f: int Return every `sample_f` data points Returns ------- acc: pandas.DataFrame Dataframe containing accelerometry data on x, y, z axes [m/s^2] depth: pandas.DataFrame Dataframe containing depth data [m] prop: pandas.DataFrame Dataframe containing speed data from propeller temp: pandas.DataFrame Dataframe containing temperature data ''' import os import pandas from . import utils def _generate_datetimes(date, time, interval_s, n_timestamps): '''Generate list of datetimes from date/time with given interval''' from datetime import datetime, timedelta import pandas # TODO problematic if both m/d d/m options fmts = ['%Y/%m/%d %H%M%S', '%d/%m/%Y %H%M%S', '%m/%d/%Y %I%M%S %p', '%d/%m/%Y %I%M%S %p',] for fmt in fmts: try: start = pandas.to_datetime('{} {}'.format(date,time), format=fmt) except: print('Date format {:18} incorrect, ' 'trying next...'.format(fmt)) else: print('Date format {:18} correct.'.format(fmt)) break # Create datetime array datetimes = list() for i in range(n_timestamps): secs = interval_s*i datetimes.append(start + timedelta(seconds=secs)) return datetimes def _read_data_file(meta, path_dir, param_str): '''Read single Little Leonardo txt data file''' import numpy import os import pandas from . import utils # Get path of data file and associated pickle file path_file = utils.find_file(path_dir, param_str, '.TXT') col_name = utils.posix_string(param_str) # Get number of header rows in file enc = utils.predict_encoding(path_file, n_lines=20) with open(path_file, 'r', encoding=enc) as f: n_header = utils.get_n_header(f) print('\nReading: {}'.format(col_name)) data = numpy.genfromtxt(path_file, skip_header=n_header) interval_s = float(meta['parameters'][col_name]['Interval(Sec)']) date = meta['parameters'][col_name]['Start date'] time = meta['parameters'][col_name]['Start time'] # TODO review # Generate summed data if propeller sampling rate not 1 if (col_name == 'propeller') and (interval_s < 1): print('Too high sampling interval, taking sums') # Sampling rate fs = int(1/interval_s) print('data before', data.max()) # Drop elements to make divisible by fs for summing data = data[:-int(len(data)%fs)] # Reshape to 2D with columns `fs` in length to be summed data = data.reshape(fs, int(len(data)/fs)) data = numpy.sum(data, axis=0) interval_s = 1 print('data after', data.max()) datetimes = _generate_datetimes(date, time, interval_s, len(data)) data = numpy.vstack((datetimes, data)).T df = pandas.DataFrame(data, columns=['datetimes', col_name]) return df # Get list of string parameter names for tag model param_names = utils.get_tag_params(meta['tag_model']) # Load pickle file exists and code unchanged pickle_file = os.path.join(path_dir, 'pydata_'+meta['experiment']+'.p') # Load or create pandas DataFrame with parameters associated with tag model if (os.path.exists(pickle_file)) and (overwrite is not True): data_df = pandas.read_pickle(pickle_file) else: first_col = True for name in param_names: next_df = _read_data_file(meta, path_dir, name) if first_col == False: data_df = pandas.merge(data_df, next_df, on='datetimes', how='left') else: data_df = next_df first_col = False print('') # Covert columns to `datetime64` or `float64` types data_df = data_df.apply(lambda x: pandas.to_numeric(x, errors='ignore')) # Save file to pickle data_df.to_pickle(pickle_file) # Return DataFrame with ever `sample_f` values return data_df.iloc[::sample_f,:]
ryanjdillon/pylleo
pylleo/lleoio.py
read_data
python
def read_data(meta, path_dir, sample_f=1, decimate=False, overwrite=False): '''Read accelerometry data from leonardo txt files Args ---- meta: dict Dictionary of meta data from header lines of lleo data files path_dir: str Parent directory containing lleo data files sample_f: int Return every `sample_f` data points Returns ------- acc: pandas.DataFrame Dataframe containing accelerometry data on x, y, z axes [m/s^2] depth: pandas.DataFrame Dataframe containing depth data [m] prop: pandas.DataFrame Dataframe containing speed data from propeller temp: pandas.DataFrame Dataframe containing temperature data ''' import os import pandas from . import utils def _generate_datetimes(date, time, interval_s, n_timestamps): '''Generate list of datetimes from date/time with given interval''' from datetime import datetime, timedelta import pandas # TODO problematic if both m/d d/m options fmts = ['%Y/%m/%d %H%M%S', '%d/%m/%Y %H%M%S', '%m/%d/%Y %I%M%S %p', '%d/%m/%Y %I%M%S %p',] for fmt in fmts: try: start = pandas.to_datetime('{} {}'.format(date,time), format=fmt) except: print('Date format {:18} incorrect, ' 'trying next...'.format(fmt)) else: print('Date format {:18} correct.'.format(fmt)) break # Create datetime array datetimes = list() for i in range(n_timestamps): secs = interval_s*i datetimes.append(start + timedelta(seconds=secs)) return datetimes def _read_data_file(meta, path_dir, param_str): '''Read single Little Leonardo txt data file''' import numpy import os import pandas from . import utils # Get path of data file and associated pickle file path_file = utils.find_file(path_dir, param_str, '.TXT') col_name = utils.posix_string(param_str) # Get number of header rows in file enc = utils.predict_encoding(path_file, n_lines=20) with open(path_file, 'r', encoding=enc) as f: n_header = utils.get_n_header(f) print('\nReading: {}'.format(col_name)) data = numpy.genfromtxt(path_file, skip_header=n_header) interval_s = float(meta['parameters'][col_name]['Interval(Sec)']) date = meta['parameters'][col_name]['Start date'] time = meta['parameters'][col_name]['Start time'] # TODO review # Generate summed data if propeller sampling rate not 1 if (col_name == 'propeller') and (interval_s < 1): print('Too high sampling interval, taking sums') # Sampling rate fs = int(1/interval_s) print('data before', data.max()) # Drop elements to make divisible by fs for summing data = data[:-int(len(data)%fs)] # Reshape to 2D with columns `fs` in length to be summed data = data.reshape(fs, int(len(data)/fs)) data = numpy.sum(data, axis=0) interval_s = 1 print('data after', data.max()) datetimes = _generate_datetimes(date, time, interval_s, len(data)) data = numpy.vstack((datetimes, data)).T df = pandas.DataFrame(data, columns=['datetimes', col_name]) return df # Get list of string parameter names for tag model param_names = utils.get_tag_params(meta['tag_model']) # Load pickle file exists and code unchanged pickle_file = os.path.join(path_dir, 'pydata_'+meta['experiment']+'.p') # Load or create pandas DataFrame with parameters associated with tag model if (os.path.exists(pickle_file)) and (overwrite is not True): data_df = pandas.read_pickle(pickle_file) else: first_col = True for name in param_names: next_df = _read_data_file(meta, path_dir, name) if first_col == False: data_df = pandas.merge(data_df, next_df, on='datetimes', how='left') else: data_df = next_df first_col = False print('') # Covert columns to `datetime64` or `float64` types data_df = data_df.apply(lambda x: pandas.to_numeric(x, errors='ignore')) # Save file to pickle data_df.to_pickle(pickle_file) # Return DataFrame with ever `sample_f` values return data_df.iloc[::sample_f,:]
Read accelerometry data from leonardo txt files Args ---- meta: dict Dictionary of meta data from header lines of lleo data files path_dir: str Parent directory containing lleo data files sample_f: int Return every `sample_f` data points Returns ------- acc: pandas.DataFrame Dataframe containing accelerometry data on x, y, z axes [m/s^2] depth: pandas.DataFrame Dataframe containing depth data [m] prop: pandas.DataFrame Dataframe containing speed data from propeller temp: pandas.DataFrame Dataframe containing temperature data
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleoio.py#L106-L240
[ "def get_tag_params(tag_model):\n '''Load param strs and n_header based on model of tag model'''\n\n tag_model = tag_model.replace('-', '')\n tags = dict()\n tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z',\n 'Depth', 'Propeller', 'Temperature']\n\n # Return tag parameters if found, else raise error\n if tag_model in tags:\n return tags[tag_model]\n else:\n raise KeyError('{} not found in tag dictionary'.format(tag_model))\n", "def _read_data_file(meta, path_dir, param_str):\n '''Read single Little Leonardo txt data file'''\n import numpy\n import os\n import pandas\n\n from . import utils\n\n # Get path of data file and associated pickle file\n path_file = utils.find_file(path_dir, param_str, '.TXT')\n col_name = utils.posix_string(param_str)\n\n # Get number of header rows in file\n enc = utils.predict_encoding(path_file, n_lines=20)\n with open(path_file, 'r', encoding=enc) as f:\n n_header = utils.get_n_header(f)\n\n print('\\nReading: {}'.format(col_name))\n\n data = numpy.genfromtxt(path_file, skip_header=n_header)\n\n interval_s = float(meta['parameters'][col_name]['Interval(Sec)'])\n date = meta['parameters'][col_name]['Start date']\n time = meta['parameters'][col_name]['Start time']\n\n # TODO review\n # Generate summed data if propeller sampling rate not 1\n if (col_name == 'propeller') and (interval_s < 1):\n print('Too high sampling interval, taking sums')\n # Sampling rate\n fs = int(1/interval_s)\n\n print('data before', data.max())\n # Drop elements to make divisible by fs for summing\n data = data[:-int(len(data)%fs)]\n\n # Reshape to 2D with columns `fs` in length to be summed\n data = data.reshape(fs, int(len(data)/fs))\n data = numpy.sum(data, axis=0)\n interval_s = 1\n\n print('data after', data.max())\n\n datetimes = _generate_datetimes(date, time, interval_s, len(data))\n data = numpy.vstack((datetimes, data)).T\n df = pandas.DataFrame(data, columns=['datetimes', col_name])\n\n return df\n" ]
def read_meta(path_dir, tag_model, tag_id): '''Read meta data from Little Leonardo data header rows Args ---- path_dir: str Parent directory containing lleo data files tag_model: str Little Leonardo tag model name tag_id: str, int Little Leonardo tag ID number Returns ------- meta: dict dictionary with meta data from header lines of lleo data files ''' from collections import OrderedDict import os import yamlord from . import utils def _parse_meta_line(line): '''Return key, value pair parsed from data header line''' # Parse the key and its value from the line key, val = line.replace(':', '').replace('"', '').split(',') return key.strip(), val.strip() def _read_meta_all(f, meta, n_header): '''Read all meta data from header rows of data file''' # Skip 'File name' line f.seek(0) _ = f.readline() # Create child dictionary for channel / file line = f.readline() key_ch, val_ch = _parse_meta_line(line) val_ch = utils.posix_string(val_ch) meta['parameters'][val_ch] = OrderedDict() # Write header values to channel dict for _ in range(n_header-2): line = f.readline() key, val = _parse_meta_line(line) meta['parameters'][val_ch][key] = val.strip() return meta def _create_meta(path_dir, tag_model, tag_id): '''Create meta data dictionary''' import datetime from . import utils param_strs = utils.get_tag_params(tag_model) # Create dictionary of meta data meta = OrderedDict() # Create fields for the parameters in data directory name exp_name = os.path.split(path_dir)[1] params_tag = utils.parse_experiment_params(exp_name) for key, value in params_tag.items(): meta[key] = value fmt = "%Y-%m-%d %H:%M:%S" meta['date_modified'] = datetime.datetime.now().strftime(fmt) meta['parameters'] = OrderedDict() for param_str in param_strs: print('Create meta entry for {}'.format(param_str)) path_file = utils.find_file(path_dir, param_str, '.TXT') # Get number of header rows enc = utils.predict_encoding(path_file, n_lines=20) with open(path_file, 'r', encoding=enc) as f: n_header = utils.get_n_header(f) f.seek(0) meta = _read_meta_all(f, meta, n_header=n_header) return meta # Load meta data from YAML file if it already exists meta_yaml_path = os.path.join(path_dir, 'meta.yml') # Load file if exists else create if os.path.isfile(meta_yaml_path): meta = yamlord.read_yaml(meta_yaml_path) # Else create meta dictionary and save to YAML else: meta = _create_meta(path_dir, tag_model, tag_id) yamlord.write_yaml(meta, meta_yaml_path) return meta
ryanjdillon/pylleo
pylleo/lleocal.py
get_cal_data
python
def get_cal_data(data_df, cal_dict, param): '''Get data along specified axis during calibration intervals Args ---- data_df: pandas.DataFrame Pandas dataframe with lleo data cal_dict: dict Calibration dictionary Returns ------- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position See also -------- lleoio.read_data: creates pandas dataframe `data_df` read_cal: creates `cal_dict` and describes fields ''' param = param.lower().replace(' ','_').replace('-','_') idx_lower_start = cal_dict['parameters'][param]['lower']['start'] idx_lower_end = cal_dict['parameters'][param]['lower']['end'] idx_upper_start = cal_dict['parameters'][param]['upper']['start'] idx_upper_end = cal_dict['parameters'][param]['upper']['end'] idx_lower = (data_df.index >= idx_lower_start) & \ (data_df.index <= idx_lower_end) idx_upper = (data_df.index >= idx_upper_start) & \ (data_df.index <= idx_upper_end) return data_df[param][idx_lower], data_df[param][idx_upper]
Get data along specified axis during calibration intervals Args ---- data_df: pandas.DataFrame Pandas dataframe with lleo data cal_dict: dict Calibration dictionary Returns ------- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position See also -------- lleoio.read_data: creates pandas dataframe `data_df` read_cal: creates `cal_dict` and describes fields
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleocal.py#L2-L38
null
def read_cal(cal_yaml_path): '''Load calibration file if exists, else create Args ---- cal_yaml_path: str Path to calibration YAML file Returns ------- cal_dict: dict Key value pairs of calibration meta data ''' from collections import OrderedDict import datetime import os import warnings import yamlord from . import utils def __create_cal(cal_yaml_path): cal_dict = OrderedDict() # Add experiment name for calibration reference base_path, _ = os.path.split(cal_yaml_path) _, exp_name = os.path.split(base_path) cal_dict['experiment'] = exp_name return cal_dict # Try reading cal file, else create if os.path.isfile(cal_yaml_path): cal_dict = yamlord.read_yaml(cal_yaml_path) else: cal_dict = __create_cal(cal_yaml_path) cal_dict['parameters'] = OrderedDict() for key, val in utils.parse_experiment_params(cal_dict['experiment']).items(): cal_dict[key] = val fmt = "%Y-%m-%d %H:%M:%S" cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt) return cal_dict def update(data_df, cal_dict, param, bound, start, end): '''Update calibration times for give parameter and boundary''' from collections import OrderedDict if param not in cal_dict['parameters']: cal_dict['parameters'][param] = OrderedDict() if bound not in cal_dict['parameters'][param]: cal_dict['parameters'][param][bound] = OrderedDict() cal_dict['parameters'][param][bound]['start'] = start cal_dict['parameters'][param][bound]['end'] = end return cal_dict def fit1d(lower, upper): '''Fit acceleration data at lower and upper boundaries of gravity Args ---- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position Returns ------- p: ndarray Polynomial coefficients, highest power first. If y was 2-D, the coefficients for k-th data set are in p[:,k]. From `numpy.polyfit()`. NOTE ---- This method should be compared agaist alternate linalg method, which allows for 2d for 2d poly, see - http://stackoverflow.com/a/33966967/943773 A = numpy.vstack(lower, upper).transpose() y = A[:,1] m, c = numpy.linalg.lstsq(A, y)[0] ''' import numpy # Get smallest size as index position for slicing idx = min(len(lower), len(upper)) # Stack accelerometer count values for upper and lower bounds of curve x = numpy.hstack((lower[:idx].values, upper[:idx].values)) x = x.astype(float) # Make corresponding y array where all lower bound points equal -g # and all upper bound points equal +g y = numpy.zeros(len(x), dtype=float) y[:idx] = -1.0 # negative gravity y[idx:] = 1.0 # positive gravity return numpy.polyfit(x, y, deg=1) def calibrate_acc(data_df, cal_dict): def apply_poly(data_df, cal_dict, param): '''Apply poly fit to data array''' import numpy poly = cal_dict['parameters'][param]['poly'] a = numpy.polyval(poly, data_df[param]) return a.astype(float) # Apply calibration and add as a new column to the dataframe for ax in ['x', 'y', 'z']: col = 'A{}_g'.format(ax) col_cal = 'acceleration_{}'.format(ax) data_df[col] = apply_poly(data_df, cal_dict, col_cal) return data_df def create_speed_csv(cal_fname, data): import numpy # Get a mask of values which contain a sample, assuming the propeller was # not sampled at as high of a frequency as the accelerometer notnan = ~numpy.isnan(data['propeller']) # Read speed, start, and end times from csv cal = pandas.read_csv(cal_fname) # For each calibration in `speed_calibrations.csv` for i in range(len(cal)): start = cal.loc[i, 'start'] start = cal.loc[i, 'end'] dt0 = pylleo.utils.nearest(data['datetimes'][notnan], start) dt1 = pylleo.utils.nearest(data['datetimes'][notnan], end) cal_mask = (data['datetimes']>=dt0) & (data['datetimes']<=dt1) count_avg = data['propeller'][cal_mask].mean() cal.loc[i, 'count_average'] = count_avg cal.to_csv(cal_fname) return cal def calibrate_propeller(data_df, cal_fname, plot=False): def speed_calibration_average(cal_fname, plot): '''Cacluate the coefficients for the mean fit of calibrations Notes ----- `cal_fname` should contain three columns: date,est_speed,count_average 2014-04-18,2.012,30 ''' import datetime import matplotlib.pyplot as plt import numpy import pandas # Read calibration data calibs = pandas.read_csv(cal_fname) calibs['date'] = pandas.to_datetime(calibs['date']) # Get unique dates to process fits for udates = numpy.unique(calibs['date']) # Create x data for samples and output array for y n_samples = 1000 x = numpy.arange(n_samples) fits = numpy.zeros((len(udates), n_samples), dtype=float) # Calculate fit coefficients then store `n_samples number of samples # Force intercept through zero (i.e. zero counts = zero speed) # http://stackoverflow.com/a/9994484/943773 for i in range(len(udates)): cal = calibs[calibs['date']==udates[i]] xi = cal['count_average'].values[:, numpy.newaxis] yi = cal['est_speed'].values m, _, _, _ = numpy.linalg.lstsq(xi, yi) fits[i, :] = m*x # Add fit to plot if switch on if plot: plt.plot(x, fits[i,:], label='cal{}'.format(i)) # Calculate average of calibration samples y_avg = numpy.mean(fits, axis=0) # Add average fit to plot and show if switch on if plot: plt.plot(x, y_avg, label='avg') plt.legend() plt.show() # Calculate fit coefficients for average samples x_avg = x[:, numpy.newaxis] m_avg, _, _, _ = numpy.linalg.lstsq(x_avg, y_avg) return m_avg m_avg = speed_calibration_average(cal_fname, plot=plot) data_df['speed'] = m_avg * data_df['propeller'] return data_df
ryanjdillon/pylleo
pylleo/lleocal.py
read_cal
python
def read_cal(cal_yaml_path): '''Load calibration file if exists, else create Args ---- cal_yaml_path: str Path to calibration YAML file Returns ------- cal_dict: dict Key value pairs of calibration meta data ''' from collections import OrderedDict import datetime import os import warnings import yamlord from . import utils def __create_cal(cal_yaml_path): cal_dict = OrderedDict() # Add experiment name for calibration reference base_path, _ = os.path.split(cal_yaml_path) _, exp_name = os.path.split(base_path) cal_dict['experiment'] = exp_name return cal_dict # Try reading cal file, else create if os.path.isfile(cal_yaml_path): cal_dict = yamlord.read_yaml(cal_yaml_path) else: cal_dict = __create_cal(cal_yaml_path) cal_dict['parameters'] = OrderedDict() for key, val in utils.parse_experiment_params(cal_dict['experiment']).items(): cal_dict[key] = val fmt = "%Y-%m-%d %H:%M:%S" cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt) return cal_dict
Load calibration file if exists, else create Args ---- cal_yaml_path: str Path to calibration YAML file Returns ------- cal_dict: dict Key value pairs of calibration meta data
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleocal.py#L41-L85
[ "def parse_experiment_params(name_exp):\n '''Parse experiment parameters from the data directory name\n\n Args\n ----\n name_exp: str\n Name of data directory with experiment parameters\n\n Returns\n -------\n tag_params: dict of str\n Dictionary of parsed experiment parameters\n '''\n if ('/' in name_exp) or ('\\\\' in name_exp):\n raise ValueError(\"The path {} appears to be a path. Please pass \"\n \"only the data directory's name (i.e. the \"\n \"experiment name)\".format(name_exp))\n\n tag_params = dict()\n tag_params['experiment'] = name_exp\n tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','')\n tag_params['tag_id'] = name_exp.split('_')[2]\n tag_params['animal'] = name_exp.split('_')[3]\n tag_params['notes'] = name_exp.split('_')[4]\n\n return tag_params\n", "def __create_cal(cal_yaml_path):\n cal_dict = OrderedDict()\n\n # Add experiment name for calibration reference\n base_path, _ = os.path.split(cal_yaml_path)\n _, exp_name = os.path.split(base_path)\n cal_dict['experiment'] = exp_name\n\n return cal_dict\n" ]
def get_cal_data(data_df, cal_dict, param): '''Get data along specified axis during calibration intervals Args ---- data_df: pandas.DataFrame Pandas dataframe with lleo data cal_dict: dict Calibration dictionary Returns ------- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position See also -------- lleoio.read_data: creates pandas dataframe `data_df` read_cal: creates `cal_dict` and describes fields ''' param = param.lower().replace(' ','_').replace('-','_') idx_lower_start = cal_dict['parameters'][param]['lower']['start'] idx_lower_end = cal_dict['parameters'][param]['lower']['end'] idx_upper_start = cal_dict['parameters'][param]['upper']['start'] idx_upper_end = cal_dict['parameters'][param]['upper']['end'] idx_lower = (data_df.index >= idx_lower_start) & \ (data_df.index <= idx_lower_end) idx_upper = (data_df.index >= idx_upper_start) & \ (data_df.index <= idx_upper_end) return data_df[param][idx_lower], data_df[param][idx_upper] def update(data_df, cal_dict, param, bound, start, end): '''Update calibration times for give parameter and boundary''' from collections import OrderedDict if param not in cal_dict['parameters']: cal_dict['parameters'][param] = OrderedDict() if bound not in cal_dict['parameters'][param]: cal_dict['parameters'][param][bound] = OrderedDict() cal_dict['parameters'][param][bound]['start'] = start cal_dict['parameters'][param][bound]['end'] = end return cal_dict def fit1d(lower, upper): '''Fit acceleration data at lower and upper boundaries of gravity Args ---- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position Returns ------- p: ndarray Polynomial coefficients, highest power first. If y was 2-D, the coefficients for k-th data set are in p[:,k]. From `numpy.polyfit()`. NOTE ---- This method should be compared agaist alternate linalg method, which allows for 2d for 2d poly, see - http://stackoverflow.com/a/33966967/943773 A = numpy.vstack(lower, upper).transpose() y = A[:,1] m, c = numpy.linalg.lstsq(A, y)[0] ''' import numpy # Get smallest size as index position for slicing idx = min(len(lower), len(upper)) # Stack accelerometer count values for upper and lower bounds of curve x = numpy.hstack((lower[:idx].values, upper[:idx].values)) x = x.astype(float) # Make corresponding y array where all lower bound points equal -g # and all upper bound points equal +g y = numpy.zeros(len(x), dtype=float) y[:idx] = -1.0 # negative gravity y[idx:] = 1.0 # positive gravity return numpy.polyfit(x, y, deg=1) def calibrate_acc(data_df, cal_dict): def apply_poly(data_df, cal_dict, param): '''Apply poly fit to data array''' import numpy poly = cal_dict['parameters'][param]['poly'] a = numpy.polyval(poly, data_df[param]) return a.astype(float) # Apply calibration and add as a new column to the dataframe for ax in ['x', 'y', 'z']: col = 'A{}_g'.format(ax) col_cal = 'acceleration_{}'.format(ax) data_df[col] = apply_poly(data_df, cal_dict, col_cal) return data_df def create_speed_csv(cal_fname, data): import numpy # Get a mask of values which contain a sample, assuming the propeller was # not sampled at as high of a frequency as the accelerometer notnan = ~numpy.isnan(data['propeller']) # Read speed, start, and end times from csv cal = pandas.read_csv(cal_fname) # For each calibration in `speed_calibrations.csv` for i in range(len(cal)): start = cal.loc[i, 'start'] start = cal.loc[i, 'end'] dt0 = pylleo.utils.nearest(data['datetimes'][notnan], start) dt1 = pylleo.utils.nearest(data['datetimes'][notnan], end) cal_mask = (data['datetimes']>=dt0) & (data['datetimes']<=dt1) count_avg = data['propeller'][cal_mask].mean() cal.loc[i, 'count_average'] = count_avg cal.to_csv(cal_fname) return cal def calibrate_propeller(data_df, cal_fname, plot=False): def speed_calibration_average(cal_fname, plot): '''Cacluate the coefficients for the mean fit of calibrations Notes ----- `cal_fname` should contain three columns: date,est_speed,count_average 2014-04-18,2.012,30 ''' import datetime import matplotlib.pyplot as plt import numpy import pandas # Read calibration data calibs = pandas.read_csv(cal_fname) calibs['date'] = pandas.to_datetime(calibs['date']) # Get unique dates to process fits for udates = numpy.unique(calibs['date']) # Create x data for samples and output array for y n_samples = 1000 x = numpy.arange(n_samples) fits = numpy.zeros((len(udates), n_samples), dtype=float) # Calculate fit coefficients then store `n_samples number of samples # Force intercept through zero (i.e. zero counts = zero speed) # http://stackoverflow.com/a/9994484/943773 for i in range(len(udates)): cal = calibs[calibs['date']==udates[i]] xi = cal['count_average'].values[:, numpy.newaxis] yi = cal['est_speed'].values m, _, _, _ = numpy.linalg.lstsq(xi, yi) fits[i, :] = m*x # Add fit to plot if switch on if plot: plt.plot(x, fits[i,:], label='cal{}'.format(i)) # Calculate average of calibration samples y_avg = numpy.mean(fits, axis=0) # Add average fit to plot and show if switch on if plot: plt.plot(x, y_avg, label='avg') plt.legend() plt.show() # Calculate fit coefficients for average samples x_avg = x[:, numpy.newaxis] m_avg, _, _, _ = numpy.linalg.lstsq(x_avg, y_avg) return m_avg m_avg = speed_calibration_average(cal_fname, plot=plot) data_df['speed'] = m_avg * data_df['propeller'] return data_df
ryanjdillon/pylleo
pylleo/lleocal.py
update
python
def update(data_df, cal_dict, param, bound, start, end): '''Update calibration times for give parameter and boundary''' from collections import OrderedDict if param not in cal_dict['parameters']: cal_dict['parameters'][param] = OrderedDict() if bound not in cal_dict['parameters'][param]: cal_dict['parameters'][param][bound] = OrderedDict() cal_dict['parameters'][param][bound]['start'] = start cal_dict['parameters'][param][bound]['end'] = end return cal_dict
Update calibration times for give parameter and boundary
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleocal.py#L88-L100
null
def get_cal_data(data_df, cal_dict, param): '''Get data along specified axis during calibration intervals Args ---- data_df: pandas.DataFrame Pandas dataframe with lleo data cal_dict: dict Calibration dictionary Returns ------- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position See also -------- lleoio.read_data: creates pandas dataframe `data_df` read_cal: creates `cal_dict` and describes fields ''' param = param.lower().replace(' ','_').replace('-','_') idx_lower_start = cal_dict['parameters'][param]['lower']['start'] idx_lower_end = cal_dict['parameters'][param]['lower']['end'] idx_upper_start = cal_dict['parameters'][param]['upper']['start'] idx_upper_end = cal_dict['parameters'][param]['upper']['end'] idx_lower = (data_df.index >= idx_lower_start) & \ (data_df.index <= idx_lower_end) idx_upper = (data_df.index >= idx_upper_start) & \ (data_df.index <= idx_upper_end) return data_df[param][idx_lower], data_df[param][idx_upper] def read_cal(cal_yaml_path): '''Load calibration file if exists, else create Args ---- cal_yaml_path: str Path to calibration YAML file Returns ------- cal_dict: dict Key value pairs of calibration meta data ''' from collections import OrderedDict import datetime import os import warnings import yamlord from . import utils def __create_cal(cal_yaml_path): cal_dict = OrderedDict() # Add experiment name for calibration reference base_path, _ = os.path.split(cal_yaml_path) _, exp_name = os.path.split(base_path) cal_dict['experiment'] = exp_name return cal_dict # Try reading cal file, else create if os.path.isfile(cal_yaml_path): cal_dict = yamlord.read_yaml(cal_yaml_path) else: cal_dict = __create_cal(cal_yaml_path) cal_dict['parameters'] = OrderedDict() for key, val in utils.parse_experiment_params(cal_dict['experiment']).items(): cal_dict[key] = val fmt = "%Y-%m-%d %H:%M:%S" cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt) return cal_dict def fit1d(lower, upper): '''Fit acceleration data at lower and upper boundaries of gravity Args ---- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position Returns ------- p: ndarray Polynomial coefficients, highest power first. If y was 2-D, the coefficients for k-th data set are in p[:,k]. From `numpy.polyfit()`. NOTE ---- This method should be compared agaist alternate linalg method, which allows for 2d for 2d poly, see - http://stackoverflow.com/a/33966967/943773 A = numpy.vstack(lower, upper).transpose() y = A[:,1] m, c = numpy.linalg.lstsq(A, y)[0] ''' import numpy # Get smallest size as index position for slicing idx = min(len(lower), len(upper)) # Stack accelerometer count values for upper and lower bounds of curve x = numpy.hstack((lower[:idx].values, upper[:idx].values)) x = x.astype(float) # Make corresponding y array where all lower bound points equal -g # and all upper bound points equal +g y = numpy.zeros(len(x), dtype=float) y[:idx] = -1.0 # negative gravity y[idx:] = 1.0 # positive gravity return numpy.polyfit(x, y, deg=1) def calibrate_acc(data_df, cal_dict): def apply_poly(data_df, cal_dict, param): '''Apply poly fit to data array''' import numpy poly = cal_dict['parameters'][param]['poly'] a = numpy.polyval(poly, data_df[param]) return a.astype(float) # Apply calibration and add as a new column to the dataframe for ax in ['x', 'y', 'z']: col = 'A{}_g'.format(ax) col_cal = 'acceleration_{}'.format(ax) data_df[col] = apply_poly(data_df, cal_dict, col_cal) return data_df def create_speed_csv(cal_fname, data): import numpy # Get a mask of values which contain a sample, assuming the propeller was # not sampled at as high of a frequency as the accelerometer notnan = ~numpy.isnan(data['propeller']) # Read speed, start, and end times from csv cal = pandas.read_csv(cal_fname) # For each calibration in `speed_calibrations.csv` for i in range(len(cal)): start = cal.loc[i, 'start'] start = cal.loc[i, 'end'] dt0 = pylleo.utils.nearest(data['datetimes'][notnan], start) dt1 = pylleo.utils.nearest(data['datetimes'][notnan], end) cal_mask = (data['datetimes']>=dt0) & (data['datetimes']<=dt1) count_avg = data['propeller'][cal_mask].mean() cal.loc[i, 'count_average'] = count_avg cal.to_csv(cal_fname) return cal def calibrate_propeller(data_df, cal_fname, plot=False): def speed_calibration_average(cal_fname, plot): '''Cacluate the coefficients for the mean fit of calibrations Notes ----- `cal_fname` should contain three columns: date,est_speed,count_average 2014-04-18,2.012,30 ''' import datetime import matplotlib.pyplot as plt import numpy import pandas # Read calibration data calibs = pandas.read_csv(cal_fname) calibs['date'] = pandas.to_datetime(calibs['date']) # Get unique dates to process fits for udates = numpy.unique(calibs['date']) # Create x data for samples and output array for y n_samples = 1000 x = numpy.arange(n_samples) fits = numpy.zeros((len(udates), n_samples), dtype=float) # Calculate fit coefficients then store `n_samples number of samples # Force intercept through zero (i.e. zero counts = zero speed) # http://stackoverflow.com/a/9994484/943773 for i in range(len(udates)): cal = calibs[calibs['date']==udates[i]] xi = cal['count_average'].values[:, numpy.newaxis] yi = cal['est_speed'].values m, _, _, _ = numpy.linalg.lstsq(xi, yi) fits[i, :] = m*x # Add fit to plot if switch on if plot: plt.plot(x, fits[i,:], label='cal{}'.format(i)) # Calculate average of calibration samples y_avg = numpy.mean(fits, axis=0) # Add average fit to plot and show if switch on if plot: plt.plot(x, y_avg, label='avg') plt.legend() plt.show() # Calculate fit coefficients for average samples x_avg = x[:, numpy.newaxis] m_avg, _, _, _ = numpy.linalg.lstsq(x_avg, y_avg) return m_avg m_avg = speed_calibration_average(cal_fname, plot=plot) data_df['speed'] = m_avg * data_df['propeller'] return data_df
ryanjdillon/pylleo
pylleo/lleocal.py
fit1d
python
def fit1d(lower, upper): '''Fit acceleration data at lower and upper boundaries of gravity Args ---- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position Returns ------- p: ndarray Polynomial coefficients, highest power first. If y was 2-D, the coefficients for k-th data set are in p[:,k]. From `numpy.polyfit()`. NOTE ---- This method should be compared agaist alternate linalg method, which allows for 2d for 2d poly, see - http://stackoverflow.com/a/33966967/943773 A = numpy.vstack(lower, upper).transpose() y = A[:,1] m, c = numpy.linalg.lstsq(A, y)[0] ''' import numpy # Get smallest size as index position for slicing idx = min(len(lower), len(upper)) # Stack accelerometer count values for upper and lower bounds of curve x = numpy.hstack((lower[:idx].values, upper[:idx].values)) x = x.astype(float) # Make corresponding y array where all lower bound points equal -g # and all upper bound points equal +g y = numpy.zeros(len(x), dtype=float) y[:idx] = -1.0 # negative gravity y[idx:] = 1.0 # positive gravity return numpy.polyfit(x, y, deg=1)
Fit acceleration data at lower and upper boundaries of gravity Args ---- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position Returns ------- p: ndarray Polynomial coefficients, highest power first. If y was 2-D, the coefficients for k-th data set are in p[:,k]. From `numpy.polyfit()`. NOTE ---- This method should be compared agaist alternate linalg method, which allows for 2d for 2d poly, see - http://stackoverflow.com/a/33966967/943773 A = numpy.vstack(lower, upper).transpose() y = A[:,1] m, c = numpy.linalg.lstsq(A, y)[0]
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/lleocal.py#L103-L143
null
def get_cal_data(data_df, cal_dict, param): '''Get data along specified axis during calibration intervals Args ---- data_df: pandas.DataFrame Pandas dataframe with lleo data cal_dict: dict Calibration dictionary Returns ------- lower: pandas dataframe slice of lleo datafram containing points at -1g calibration position upper: pandas dataframe slice of lleo datafram containing points at -1g calibration position See also -------- lleoio.read_data: creates pandas dataframe `data_df` read_cal: creates `cal_dict` and describes fields ''' param = param.lower().replace(' ','_').replace('-','_') idx_lower_start = cal_dict['parameters'][param]['lower']['start'] idx_lower_end = cal_dict['parameters'][param]['lower']['end'] idx_upper_start = cal_dict['parameters'][param]['upper']['start'] idx_upper_end = cal_dict['parameters'][param]['upper']['end'] idx_lower = (data_df.index >= idx_lower_start) & \ (data_df.index <= idx_lower_end) idx_upper = (data_df.index >= idx_upper_start) & \ (data_df.index <= idx_upper_end) return data_df[param][idx_lower], data_df[param][idx_upper] def read_cal(cal_yaml_path): '''Load calibration file if exists, else create Args ---- cal_yaml_path: str Path to calibration YAML file Returns ------- cal_dict: dict Key value pairs of calibration meta data ''' from collections import OrderedDict import datetime import os import warnings import yamlord from . import utils def __create_cal(cal_yaml_path): cal_dict = OrderedDict() # Add experiment name for calibration reference base_path, _ = os.path.split(cal_yaml_path) _, exp_name = os.path.split(base_path) cal_dict['experiment'] = exp_name return cal_dict # Try reading cal file, else create if os.path.isfile(cal_yaml_path): cal_dict = yamlord.read_yaml(cal_yaml_path) else: cal_dict = __create_cal(cal_yaml_path) cal_dict['parameters'] = OrderedDict() for key, val in utils.parse_experiment_params(cal_dict['experiment']).items(): cal_dict[key] = val fmt = "%Y-%m-%d %H:%M:%S" cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt) return cal_dict def update(data_df, cal_dict, param, bound, start, end): '''Update calibration times for give parameter and boundary''' from collections import OrderedDict if param not in cal_dict['parameters']: cal_dict['parameters'][param] = OrderedDict() if bound not in cal_dict['parameters'][param]: cal_dict['parameters'][param][bound] = OrderedDict() cal_dict['parameters'][param][bound]['start'] = start cal_dict['parameters'][param][bound]['end'] = end return cal_dict def calibrate_acc(data_df, cal_dict): def apply_poly(data_df, cal_dict, param): '''Apply poly fit to data array''' import numpy poly = cal_dict['parameters'][param]['poly'] a = numpy.polyval(poly, data_df[param]) return a.astype(float) # Apply calibration and add as a new column to the dataframe for ax in ['x', 'y', 'z']: col = 'A{}_g'.format(ax) col_cal = 'acceleration_{}'.format(ax) data_df[col] = apply_poly(data_df, cal_dict, col_cal) return data_df def create_speed_csv(cal_fname, data): import numpy # Get a mask of values which contain a sample, assuming the propeller was # not sampled at as high of a frequency as the accelerometer notnan = ~numpy.isnan(data['propeller']) # Read speed, start, and end times from csv cal = pandas.read_csv(cal_fname) # For each calibration in `speed_calibrations.csv` for i in range(len(cal)): start = cal.loc[i, 'start'] start = cal.loc[i, 'end'] dt0 = pylleo.utils.nearest(data['datetimes'][notnan], start) dt1 = pylleo.utils.nearest(data['datetimes'][notnan], end) cal_mask = (data['datetimes']>=dt0) & (data['datetimes']<=dt1) count_avg = data['propeller'][cal_mask].mean() cal.loc[i, 'count_average'] = count_avg cal.to_csv(cal_fname) return cal def calibrate_propeller(data_df, cal_fname, plot=False): def speed_calibration_average(cal_fname, plot): '''Cacluate the coefficients for the mean fit of calibrations Notes ----- `cal_fname` should contain three columns: date,est_speed,count_average 2014-04-18,2.012,30 ''' import datetime import matplotlib.pyplot as plt import numpy import pandas # Read calibration data calibs = pandas.read_csv(cal_fname) calibs['date'] = pandas.to_datetime(calibs['date']) # Get unique dates to process fits for udates = numpy.unique(calibs['date']) # Create x data for samples and output array for y n_samples = 1000 x = numpy.arange(n_samples) fits = numpy.zeros((len(udates), n_samples), dtype=float) # Calculate fit coefficients then store `n_samples number of samples # Force intercept through zero (i.e. zero counts = zero speed) # http://stackoverflow.com/a/9994484/943773 for i in range(len(udates)): cal = calibs[calibs['date']==udates[i]] xi = cal['count_average'].values[:, numpy.newaxis] yi = cal['est_speed'].values m, _, _, _ = numpy.linalg.lstsq(xi, yi) fits[i, :] = m*x # Add fit to plot if switch on if plot: plt.plot(x, fits[i,:], label='cal{}'.format(i)) # Calculate average of calibration samples y_avg = numpy.mean(fits, axis=0) # Add average fit to plot and show if switch on if plot: plt.plot(x, y_avg, label='avg') plt.legend() plt.show() # Calculate fit coefficients for average samples x_avg = x[:, numpy.newaxis] m_avg, _, _, _ = numpy.linalg.lstsq(x_avg, y_avg) return m_avg m_avg = speed_calibration_average(cal_fname, plot=plot) data_df['speed'] = m_avg * data_df['propeller'] return data_df
ryanjdillon/pylleo
pylleo/calapp/main.py
plot_triaxial
python
def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats
Plot pandas dataframe containing an x, y, and z column
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L18-L43
null
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
load_data
python
def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data
Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L46-L79
[ "def parse_experiment_params(name_exp):\n '''Parse experiment parameters from the data directory name\n\n Args\n ----\n name_exp: str\n Name of data directory with experiment parameters\n\n Returns\n -------\n tag_params: dict of str\n Dictionary of parsed experiment parameters\n '''\n if ('/' in name_exp) or ('\\\\' in name_exp):\n raise ValueError(\"The path {} appears to be a path. Please pass \"\n \"only the data directory's name (i.e. the \"\n \"experiment name)\".format(name_exp))\n\n tag_params = dict()\n tag_params['experiment'] = name_exp\n tag_params['tag_model'] = (name_exp.split('_')[1]).replace('-','')\n tag_params['tag_id'] = name_exp.split('_')[2]\n tag_params['animal'] = name_exp.split('_')[3]\n tag_params['notes'] = name_exp.split('_')[4]\n\n return tag_params\n", "def read_meta(path_dir, tag_model, tag_id):\n '''Read meta data from Little Leonardo data header rows\n\n Args\n ----\n path_dir: str\n Parent directory containing lleo data files\n tag_model: str\n Little Leonardo tag model name\n tag_id: str, int\n Little Leonardo tag ID number\n\n Returns\n -------\n meta: dict\n dictionary with meta data from header lines of lleo data files\n '''\n from collections import OrderedDict\n import os\n import yamlord\n\n from . import utils\n\n def _parse_meta_line(line):\n '''Return key, value pair parsed from data header line'''\n\n # Parse the key and its value from the line\n key, val = line.replace(':', '').replace('\"', '').split(',')\n\n return key.strip(), val.strip()\n\n\n def _read_meta_all(f, meta, n_header):\n '''Read all meta data from header rows of data file'''\n\n # Skip 'File name' line\n f.seek(0)\n _ = f.readline()\n\n # Create child dictionary for channel / file\n line = f.readline()\n key_ch, val_ch = _parse_meta_line(line)\n val_ch = utils.posix_string(val_ch)\n meta['parameters'][val_ch] = OrderedDict()\n\n # Write header values to channel dict\n for _ in range(n_header-2):\n line = f.readline()\n key, val = _parse_meta_line(line)\n meta['parameters'][val_ch][key] = val.strip()\n\n return meta\n\n\n def _create_meta(path_dir, tag_model, tag_id):\n '''Create meta data dictionary'''\n import datetime\n from . import utils\n\n param_strs = utils.get_tag_params(tag_model)\n\n # Create dictionary of meta data\n meta = OrderedDict()\n\n # Create fields for the parameters in data directory name\n exp_name = os.path.split(path_dir)[1]\n params_tag = utils.parse_experiment_params(exp_name)\n for key, value in params_tag.items():\n meta[key] = value\n\n fmt = \"%Y-%m-%d %H:%M:%S\"\n meta['date_modified'] = datetime.datetime.now().strftime(fmt)\n\n meta['parameters'] = OrderedDict()\n\n for param_str in param_strs:\n print('Create meta entry for {}'.format(param_str))\n\n path_file = utils.find_file(path_dir, param_str, '.TXT')\n # Get number of header rows\n enc = utils.predict_encoding(path_file, n_lines=20)\n with open(path_file, 'r', encoding=enc) as f:\n n_header = utils.get_n_header(f)\n f.seek(0)\n meta = _read_meta_all(f, meta, n_header=n_header)\n\n return meta\n\n\n # Load meta data from YAML file if it already exists\n meta_yaml_path = os.path.join(path_dir, 'meta.yml')\n\n # Load file if exists else create\n if os.path.isfile(meta_yaml_path):\n meta = yamlord.read_yaml(meta_yaml_path)\n\n # Else create meta dictionary and save to YAML\n else:\n meta = _create_meta(path_dir, tag_model, tag_id)\n yamlord.write_yaml(meta, meta_yaml_path)\n\n return meta\n", "def get_tag_params(tag_model):\n '''Load param strs and n_header based on model of tag model'''\n\n tag_model = tag_model.replace('-', '')\n tags = dict()\n tags['W190PD3GT'] = ['Acceleration-X', 'Acceleration-Y', 'Acceleration-Z',\n 'Depth', 'Propeller', 'Temperature']\n\n # Return tag parameters if found, else raise error\n if tag_model in tags:\n return tags[tag_model]\n else:\n raise KeyError('{} not found in tag dictionary'.format(tag_model))\n", "def read_data(meta, path_dir, sample_f=1, decimate=False, overwrite=False):\n '''Read accelerometry data from leonardo txt files\n\n Args\n ----\n meta: dict\n Dictionary of meta data from header lines of lleo data files\n path_dir: str\n Parent directory containing lleo data files\n sample_f: int\n Return every `sample_f` data points\n\n Returns\n -------\n acc: pandas.DataFrame\n Dataframe containing accelerometry data on x, y, z axes [m/s^2]\n depth: pandas.DataFrame\n Dataframe containing depth data [m]\n prop: pandas.DataFrame\n Dataframe containing speed data from propeller\n temp: pandas.DataFrame\n Dataframe containing temperature data\n '''\n import os\n import pandas\n\n from . import utils\n\n def _generate_datetimes(date, time, interval_s, n_timestamps):\n '''Generate list of datetimes from date/time with given interval'''\n from datetime import datetime, timedelta\n import pandas\n\n # TODO problematic if both m/d d/m options\n fmts = ['%Y/%m/%d %H%M%S',\n '%d/%m/%Y %H%M%S',\n '%m/%d/%Y %I%M%S %p',\n '%d/%m/%Y %I%M%S %p',]\n\n for fmt in fmts:\n try:\n start = pandas.to_datetime('{} {}'.format(date,time), format=fmt)\n except:\n print('Date format {:18} incorrect, '\n 'trying next...'.format(fmt))\n else:\n print('Date format {:18} correct.'.format(fmt))\n break\n\n # Create datetime array\n datetimes = list()\n for i in range(n_timestamps):\n secs = interval_s*i\n datetimes.append(start + timedelta(seconds=secs))\n\n return datetimes\n\n\n def _read_data_file(meta, path_dir, param_str):\n '''Read single Little Leonardo txt data file'''\n import numpy\n import os\n import pandas\n\n from . import utils\n\n # Get path of data file and associated pickle file\n path_file = utils.find_file(path_dir, param_str, '.TXT')\n col_name = utils.posix_string(param_str)\n\n # Get number of header rows in file\n enc = utils.predict_encoding(path_file, n_lines=20)\n with open(path_file, 'r', encoding=enc) as f:\n n_header = utils.get_n_header(f)\n\n print('\\nReading: {}'.format(col_name))\n\n data = numpy.genfromtxt(path_file, skip_header=n_header)\n\n interval_s = float(meta['parameters'][col_name]['Interval(Sec)'])\n date = meta['parameters'][col_name]['Start date']\n time = meta['parameters'][col_name]['Start time']\n\n # TODO review\n # Generate summed data if propeller sampling rate not 1\n if (col_name == 'propeller') and (interval_s < 1):\n print('Too high sampling interval, taking sums')\n # Sampling rate\n fs = int(1/interval_s)\n\n print('data before', data.max())\n # Drop elements to make divisible by fs for summing\n data = data[:-int(len(data)%fs)]\n\n # Reshape to 2D with columns `fs` in length to be summed\n data = data.reshape(fs, int(len(data)/fs))\n data = numpy.sum(data, axis=0)\n interval_s = 1\n\n print('data after', data.max())\n\n datetimes = _generate_datetimes(date, time, interval_s, len(data))\n data = numpy.vstack((datetimes, data)).T\n df = pandas.DataFrame(data, columns=['datetimes', col_name])\n\n return df\n\n # Get list of string parameter names for tag model\n param_names = utils.get_tag_params(meta['tag_model'])\n\n # Load pickle file exists and code unchanged\n pickle_file = os.path.join(path_dir, 'pydata_'+meta['experiment']+'.p')\n\n # Load or create pandas DataFrame with parameters associated with tag model\n if (os.path.exists(pickle_file)) and (overwrite is not True):\n data_df = pandas.read_pickle(pickle_file)\n else:\n first_col = True\n for name in param_names:\n next_df = _read_data_file(meta, path_dir, name)\n if first_col == False:\n data_df = pandas.merge(data_df, next_df, on='datetimes', how='left')\n else:\n data_df = next_df\n first_col = False\n print('')\n\n # Covert columns to `datetime64` or `float64` types\n data_df = data_df.apply(lambda x: pandas.to_numeric(x, errors='ignore'))\n\n # Save file to pickle\n data_df.to_pickle(pickle_file)\n\n # Return DataFrame with ever `sample_f` values\n return data_df.iloc[::sample_f,:]\n" ]
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_parent
python
def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None
Update data directories drop down with new parent directory
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L82-L109
[ "def callback_datadirs(attr, old, new):\n '''Update source and controls with data loaded from selected directory'''\n import os\n\n global data\n\n try:\n # Load data from new data directory\n path_dir = os.path.join(parent_input.value, new)\n data, params_tag, params_data = load_data(path_dir)\n\n # Make title with new data directory\n p.title.text = 'Calibrating {}'.format(params_tag['experiment'])\n\n # Update `source` data fields from dataframe\n dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']]\n source.data = dict(x = list(data['acceleration_x']),\n y = list(data['acceleration_y']),\n z = list(data['acceleration_z']),\n ind = list(data.index),\n dt = list(data['datetimes']),\n dt_str = dt_str)\n\n # Update values for control widgets\n param_checkbox.active = [0, 1, 2]\n param_select.options = params_data\n param_select.value = params_data[0]\n regions = ['lower', 'upper']\n region_select.options = regions\n region_select.value = regions[0]\n start_input.value = str(data.index[0])\n end_input.value = str(data.index[-1])\n except Exception as e:\n msg = '''\n Problem loading data directory `{}`.\n\n Please check that data exists in that directory.\n\n Details:\n {}\n '''.format(new, e)\n output_window.text = output_template.format(msg)\n\n\n return None\n" ]
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_datadirs
python
def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None
Update source and controls with data loaded from selected directory
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L112-L156
[ "def load_data(path_dir):\n '''Load data, directory parameters, and accelerometer parameter names\n\n Args\n ----\n path_dir: str\n Path to the data directory\n\n Returns\n -------\n data: pandas.DataFrame\n Experiment data\n params_tag: dict\n A dictionary of parameters parsed from the directory name\n params_data: list\n A list of the accelerometer parameter names\n '''\n import os\n import pylleo\n\n exp_name = os.path.split(path_dir)[1]\n params_tag = pylleo.utils.parse_experiment_params(exp_name)\n\n # Load the Little Leonardo tag data\n meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'],\n params_tag['tag_id'])\n data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f)\n\n # Get and curate the parameter names of the loaded dataframe\n params_data = pylleo.utils.get_tag_params(params_tag['tag_model'])\n params_data = [pylleo.utils.posix_string(p) for p in params_data]\n params_data = [p for p in params_data if p.startswith('acc')]\n\n return data, params_tag, params_data\n" ]
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_box_select
python
def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None
Update TextInput start/end entries from BoxSelectTool selection
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L159-L176
null
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_checkbox
python
def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None
Update visible data from parameters selectin in the CheckboxSelect
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L179-L187
null
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_save_indices
python
def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None
Save index from bokeh textinput
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L190-L227
[ "def update(data_df, cal_dict, param, bound, start, end):\n '''Update calibration times for give parameter and boundary'''\n from collections import OrderedDict\n\n if param not in cal_dict['parameters']:\n cal_dict['parameters'][param] = OrderedDict()\n if bound not in cal_dict['parameters'][param]:\n cal_dict['parameters'][param][bound] = OrderedDict()\n\n cal_dict['parameters'][param][bound]['start'] = start\n cal_dict['parameters'][param][bound]['end'] = end\n\n return cal_dict\n", "def read_cal(cal_yaml_path):\n '''Load calibration file if exists, else create\n\n Args\n ----\n cal_yaml_path: str\n Path to calibration YAML file\n\n Returns\n -------\n cal_dict: dict\n Key value pairs of calibration meta data\n '''\n from collections import OrderedDict\n import datetime\n import os\n import warnings\n import yamlord\n\n from . import utils\n\n def __create_cal(cal_yaml_path):\n cal_dict = OrderedDict()\n\n # Add experiment name for calibration reference\n base_path, _ = os.path.split(cal_yaml_path)\n _, exp_name = os.path.split(base_path)\n cal_dict['experiment'] = exp_name\n\n return cal_dict\n\n # Try reading cal file, else create\n if os.path.isfile(cal_yaml_path):\n cal_dict = yamlord.read_yaml(cal_yaml_path)\n else:\n cal_dict = __create_cal(cal_yaml_path)\n cal_dict['parameters'] = OrderedDict()\n\n for key, val in utils.parse_experiment_params(cal_dict['experiment']).items():\n cal_dict[key] = val\n\n fmt = \"%Y-%m-%d %H:%M:%S\"\n cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt)\n\n return cal_dict\n" ]
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
ryanjdillon/pylleo
pylleo/calapp/main.py
callback_save_poly
python
def callback_save_poly(): '''Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration) ''' import datetime import pylleo import yamlord import itertools def _check_param_regions(param, regions, cal_dict): msg = ''' <b>{}</b> was not found in the calibration dictionary. Process that parameter and then try saving the polyfit again. '''.format(param) params_present = True if param not in cal_dict['parameters']: params_present = False msg.format(param) else: for region in regions: if region not in cal_dict['parameters'][param]: params_present = False msg.format('{}/{}'.format(param, region)) output_window.text = output_template.format(msg) return params_present def _check_index_order(param, regions, cal_dict): '''Check that index positions exist for each calibration region''' indices_present = True for region in regions: start = cal_dict['parameters'][param][region]['start'] end = cal_dict['parameters'][param][region]['end'] # Check if start comes after end if int(start) > int(end): indices_present = False msg = ''' The start index ({}) comes after the end index ({}). Please set new start/end indexes for <b>{}/{}</b> '''.format(start, end, param, region) msg.format(start, end, param, region) output_window.text = output_template.format(msg) return indices_present if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, cal_fname) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Get currently selected parameter param = param_select.value regions = region_select.options # Check that index positions have been recorded in `cal.yml` if not _check_index_order(param, regions, cal_dict): return None # Check that index positions are in sequence if not _check_index_order(param, regions, cal_dict): return None param = (param_select.value).lower().replace('-','_') try: msg = ''' Saved polyfit for <b>{}</b> to <b>{}</b>. '''.format(param, cal_fname) output_window.text = output_template.format(msg) lower, upper = pylleo.lleocal.get_cal_data(data, cal_dict, param) poly = list(pylleo.lleocal.fit1d(lower, upper)) poly = [float(str(i)) for i in poly] cal_dict['parameters'][param]['poly'] = poly yamlord.write_yaml(cal_dict, cal_yaml_path) except Exception as e: msg = 'Problem saving polyfit: {}'.format(e) output_window.text = output_template.format(msg) else: msg = ''' You must first load data and select indices for calibration regions before you can save to polyfit to `cal.yml` ''' output_window.text = output_template.format(msg) return None
Perform polyfit once regions selected Globals: cal_fname, data (read-only, so no declaration)
train
https://github.com/ryanjdillon/pylleo/blob/b9b999fef19eaeccce4f207ab1b6198287c1bfec/pylleo/calapp/main.py#L230-L323
[ "def read_cal(cal_yaml_path):\n '''Load calibration file if exists, else create\n\n Args\n ----\n cal_yaml_path: str\n Path to calibration YAML file\n\n Returns\n -------\n cal_dict: dict\n Key value pairs of calibration meta data\n '''\n from collections import OrderedDict\n import datetime\n import os\n import warnings\n import yamlord\n\n from . import utils\n\n def __create_cal(cal_yaml_path):\n cal_dict = OrderedDict()\n\n # Add experiment name for calibration reference\n base_path, _ = os.path.split(cal_yaml_path)\n _, exp_name = os.path.split(base_path)\n cal_dict['experiment'] = exp_name\n\n return cal_dict\n\n # Try reading cal file, else create\n if os.path.isfile(cal_yaml_path):\n cal_dict = yamlord.read_yaml(cal_yaml_path)\n else:\n cal_dict = __create_cal(cal_yaml_path)\n cal_dict['parameters'] = OrderedDict()\n\n for key, val in utils.parse_experiment_params(cal_dict['experiment']).items():\n cal_dict[key] = val\n\n fmt = \"%Y-%m-%d %H:%M:%S\"\n cal_dict['date_modified'] = datetime.datetime.now().strftime(fmt)\n\n return cal_dict\n", "def _check_index_order(param, regions, cal_dict):\n '''Check that index positions exist for each calibration region'''\n\n indices_present = True\n for region in regions:\n start = cal_dict['parameters'][param][region]['start']\n end = cal_dict['parameters'][param][region]['end']\n # Check if start comes after end\n if int(start) > int(end):\n indices_present = False\n msg = '''\n The start index ({}) comes after the end index ({}).\n\n Please set new start/end indexes for <b>{}/{}</b>\n '''.format(start, end, param, region)\n msg.format(start, end, param, region)\n output_window.text = output_template.format(msg)\n\n return indices_present\n" ]
''' LeoADCA Little Leonardo Accelerometer Data Calibration Application This app will launch an window in your default broweser to visually identify the times at which various axis of the lleo tag have been placed into +/-g orientations. Enter the start and end times of these orientation periods, then click 'save' to write those to a calibration YAML file (cal.yml) in the data directory Example ------- bokeh serve --show bokeh_calibration.py ''' def plot_triaxial(height, width, tools): '''Plot pandas dataframe containing an x, y, and z column''' import bokeh.plotting p = bokeh.plotting.figure(x_axis_type='datetime', plot_height=height, plot_width=width, title=' ', toolbar_sticky=False, tools=tools, active_drag=BoxZoomTool(), output_backend='webgl') p.yaxis.axis_label = 'Acceleration (count)' p.xaxis.axis_label = 'Time (timezone as programmed)' # Plot accelerometry data as lines and scatter (for BoxSelectTool) colors = ['#1b9e77', '#d95f02', '#7570b3'] axes = ['x', 'y', 'z'] lines = [None,]*3 scats = [None,]*3 for i, (ax, c) in enumerate(zip(axes, colors)): lines[i] = p.line(y=ax, x='dt', color=c, legend=False, source=source) scats[i] = p.scatter(y=ax, x='dt', color=c, legend=False, size=1, source=source) return p, lines, scats def load_data(path_dir): '''Load data, directory parameters, and accelerometer parameter names Args ---- path_dir: str Path to the data directory Returns ------- data: pandas.DataFrame Experiment data params_tag: dict A dictionary of parameters parsed from the directory name params_data: list A list of the accelerometer parameter names ''' import os import pylleo exp_name = os.path.split(path_dir)[1] params_tag = pylleo.utils.parse_experiment_params(exp_name) # Load the Little Leonardo tag data meta = pylleo.lleoio.read_meta(path_dir, params_tag['tag_model'], params_tag['tag_id']) data = pylleo.lleoio.read_data(meta, path_dir, sample_f=sample_f) # Get and curate the parameter names of the loaded dataframe params_data = pylleo.utils.get_tag_params(params_tag['tag_model']) params_data = [pylleo.utils.posix_string(p) for p in params_data] params_data = [p for p in params_data if p.startswith('acc')] return data, params_tag, params_data def callback_parent(attr, old, new): '''Update data directories drop down with new parent directory''' import os # Remove accidental white space if copy/pasted new = new.strip() parent_input.value = new # Verify new parent path exists and update `datadirs_select` widget if os.path.exists(new): # Create sorted list of data directories, ignore files joinisdir = lambda parent, d: os.path.isdir(os.path.join(parent, d)) options = sorted([d for d in os.listdir(new) if joinisdir(new, d)]) # Update dropdown list of available data directories and select first datadirs_select.options = options datadirs_select.value = options[0] callback_datadirs('value', options[0], options[0]) else: msg = ''' The parent path `{}` does not exist. Check that you have entered the absolute path. '''.format(new) output_window.text = output_template.format(msg) return None def callback_datadirs(attr, old, new): '''Update source and controls with data loaded from selected directory''' import os global data try: # Load data from new data directory path_dir = os.path.join(parent_input.value, new) data, params_tag, params_data = load_data(path_dir) # Make title with new data directory p.title.text = 'Calibrating {}'.format(params_tag['experiment']) # Update `source` data fields from dataframe dt_str = [dt.strftime(dt_fmt) for dt in data['datetimes']] source.data = dict(x = list(data['acceleration_x']), y = list(data['acceleration_y']), z = list(data['acceleration_z']), ind = list(data.index), dt = list(data['datetimes']), dt_str = dt_str) # Update values for control widgets param_checkbox.active = [0, 1, 2] param_select.options = params_data param_select.value = params_data[0] regions = ['lower', 'upper'] region_select.options = regions region_select.value = regions[0] start_input.value = str(data.index[0]) end_input.value = str(data.index[-1]) except Exception as e: msg = ''' Problem loading data directory `{}`. Please check that data exists in that directory. Details: {} '''.format(new, e) output_window.text = output_template.format(msg) return None def callback_box_select(attr, old, new): '''Update TextInput start/end entries from BoxSelectTool selection''' # Get indices of selection ind = sorted(new['1d']['indices']) if new is None: start_input.value = '0' end_input.value = '0' else: start_input.value = str(source.data['ind'][ind[0]]) end_input.value = str(source.data['ind'][ind[-1]]) msg = ''' New start and end index values set. ''' output_window.text = output_template.format(msg) return None def callback_checkbox(attr, old, new): '''Update visible data from parameters selectin in the CheckboxSelect''' import numpy for i in range(len(lines)): lines[i].visible = i in param_checkbox.active scats[i].visible = i in param_checkbox.active return None def callback_save_indices(): '''Save index from bokeh textinput''' import datetime import os import pylleo import yamlord if datadirs_select.value != 'None': path_dir = os.path.join(parent_input.value, datadirs_select.value) cal_yaml_path = os.path.join(path_dir, 'cal.yml') param = (param_select.value).lower().replace('-','_') region = region_select.value start = int(start_input.value) end = int(end_input.value) msg = ''' Updated calibration times for:<br> <b>{}/{}</b> <br> <br> star index: {}<br> end index: {}<br> '''.format(param, region, start, end) output_window.text = output_template.format(msg) cal_dict = pylleo.lleocal.read_cal(cal_yaml_path) # Generalize for Class-ifying cal_dict = pylleo.lleocal.update(data, cal_dict, param, region, start, end) yamlord.write_yaml(cal_dict, cal_yaml_path) else: msg = ''' You must first load data and select indices for calibration regions before you can save the indices to `cal.yml` ''' output_window.text = output_template.format(msg) return None import datetime import numpy import os import sys import subprocess from bokeh.layouts import widgetbox, column, row from bokeh.models import ColumnDataSource from bokeh.models import PanTool, WheelZoomTool, BoxZoomTool, HoverTool from bokeh.models import BoxSelectTool from bokeh.models.widgets import Div, PreText, CheckboxButtonGroup from bokeh.models.widgets import Select, TextInput, Button from bokeh.io import curdoc # DATA #------------------------------------------------------------------------------ cal_fname = 'cal.yml' sample_f = 30 dt_fmt = '%H:%M' # Create Column Data Source that will be used by the plot # use 6hr span to avoid straing xaxis labels t0 = datetime.datetime.now() t1 = t0 + datetime.timedelta(hours=6) source = ColumnDataSource(data = dict(x = [0, 0], y = [0, 0], z = [0, 0], ind = [0, 0], dt = [t0, t1], dt_str = [t0.strftime(dt_fmt), t1.strftime(dt_fmt)], )) # Input #------------------------------------------------------------------------------ # Path for entering the parent directory of data directories title = 'Parent directory:' css = ['widthfix'] parent_input = TextInput(value='', title=title, css_classes=css) parent_input.on_change('value', callback_parent) # Dropdown list of data directories in parent to load data from data_dirs = ['None'] title = 'Data directories:' datadirs_select = Select(title=title, value=data_dirs[0], options=data_dirs) datadirs_select.on_change('value', callback_datadirs) # Select which axes to select calibration start/end points param_checkbox_pre = PreText(text='Axes to display') labels_ax = ['x', 'y', 'z'] active_ax = [] param_checkbox = CheckboxButtonGroup(labels=labels_ax, active=active_ax) param_checkbox.on_change('active', callback_checkbox) # Select with parameter to collect start/end times for and perform a data fit params_data = ['None'] title = 'Parameter to calibrate:' param_select = Select(title=title, value=params_data[0], options=params_data) # Select upper or lower acceleration region to calibrate regions = ['None'] title = 'Bound (lower = -g; upper = +g):' region_select = Select(title=title, value=regions[0], options=regions) # User input start end times, save to cal start_input = TextInput(value='0', title='Start index:') end_input = TextInput(value='0', title='End index:') # Save the start end times selcted with BoxSelectTool (or manually entered) button_save = Button(label='Save Index Values', button_type='success') button_save.on_click(callback_save_indices) # Perform a polyfit on the data points occuring between the start/end points # for the parameter and region selected from the dropdown menus button_poly = Button(label='Perform Polyfit', button_type='success') button_poly.on_click(callback_save_poly) # Print text output from callback/button routines in styled div container output_template = ('<div style="display:inline-block; width:300px; ' 'height:150px; padding: 10px; background-color:#f2f2f2; ' 'border-radius:10px; overflow:scroll">{}</div>') output_window = Div(text=output_template.format('Status updates display here')) # Plotting #------------------------------------------------------------------------------ # Format data to display when HoverTool activated hover = HoverTool(tooltips=[('index', '@ind'), ('acc', '$y'), ('time', '@dt_str'), ]) # Define plots tools and create plot object and glyph objects tools = [PanTool(), WheelZoomTool(), BoxSelectTool(), BoxZoomTool(), hover] p, lines, scats = plot_triaxial(height=300, width=800, tools=tools) p.select(BoxSelectTool).select_every_mousemove = False # Force run of callback to make dummy line not visible at init callback_checkbox('active', active_ax, active_ax) # Update start/end input text boxes with BoxSelectTool for scat in scats: scat.data_source.on_change('selected', callback_box_select) # Rendering #------------------------------------------------------------------------------ # Bundle controls for inserting into the layout controls = (param_checkbox_pre, param_checkbox, param_select, region_select, start_input, end_input, button_save, button_poly) # Create layout row1 = row(column(widgetbox(parent_input, datadirs_select))) col1 = column(widgetbox(*controls), width=350) # See `output_template for css sizing of window vbuffer = row([], height=35) col2 = column(vbuffer, widgetbox(output_window)) row2 = row(col1, col2) layout = column(p, row1, row2, width=1100) # Generate document from layout curdoc().add_root(layout)
lwcook/horsetail-matching
horsetailmatching/hm.py
_extalg
python
def _extalg(xarr, alpha=100, axis=None): '''Given an array xarr of values, smoothly return the max/min''' return (np.sum(xarr * np.exp(alpha*xarr), axis=axis, keepdims=True)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True))
Given an array xarr of values, smoothly return the max/min
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L834-L837
null
import pdb import time import math import copy import warnings import numpy as np class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples ############################################################################## ## Private functions ############################################################################## def _extgrad(xarr, alpha=100, axis=None): '''Given an array xarr of values, return the gradient of the smooth min/max swith respect to each entry in the array''' term1 = (np.exp(alpha*xarr)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) term2 = 1 + alpha*(xarr - _extalg(xarr, alpha, axis=axis)) return term1*term2 def _ramp(x, width): return _minsmooth(1, _maxsmooth(0, (x - width/2)*(1/width))) def _trint(x, width): w = width/2. xb = _maxsmooth(-w, _minsmooth(x, w)) y1 = 0.5 + xb/w + xb**2/(2*w**2) y2 = xb/w - xb**2/(2*w**2) return _minsmooth(y1, 0.5) + _maxsmooth(y2, 0.0) def _minsmooth(a, b, eps=0.0000): return 0.5*(a + b - np.sqrt((a-b)**2 + eps**2)) def _maxsmooth(a, b, eps=0.0000): return 0.5*(a + b + np.sqrt((a-b)**2 + eps**2)) def _step(x): return 1 * (x > 0) def _erf(r): ## Numerical implementation of the error function for matrix comptibility # save the sign of x sign = np.sign(r) x = np.absolute(r) # constants a1 = 0.254829592 a2 = -0.284496736 a3 = 1.421413741 a4 = -1.453152027 a5 = 1.061405429 p = 0.3275911 # A&S formula 7.1.26 t = 1.0/(1.0 + p*x) y = 1.0 - (((((a5*t + a4)*t) + a3)*t + a2)*t + a1)*t*np.exp(-x*x) return sign*y # erf(-x) = -erf(x) def _kernel(points, M, bw, ktype='gauss', bGrad=False): if ktype == 'gauss' or ktype == 'gaussian': KernelMat = (1./M)*((1 + _erf((points/bw)/np.sqrt(2.)))/2.) # KernelMat = np.zeros(points.shape) # for ir in np.arange(points.shape[0]): # for ic in np.arange(points.shape[1]): # KernelMat[ir, ic] = (1./M)*((1. + # math.erf((points[ir, ic]/bw)/math.sqrt(2.)))/2.) elif ktype == 'uniform' or ktype == 'uni': KernelMat = (1./M)*_ramp(points, width=bw*np.sqrt(12)) elif ktype == 'triangle' or ktype == 'tri': KernelMat = (1./M)*_trint(points, width=bw*2.*np.sqrt(6)) if bGrad: if ktype == 'gauss' or ktype == 'gaussian': const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) KernelGradMat = const_term * np.exp(-(1./2.) * (points/bw)**2) elif ktype == 'uniform' or ktype == 'uni': width = bw*np.sqrt(12) const = (1./M)*(1./width) KernelGradMat = const*(_step(points+width/2) - _step(points-width/2)) elif ktype == 'triangle' or ktype == 'tri': width = bw*2.*np.sqrt(6) const = (1./M)*(2./width) KernelGradMat = const*(_ramp(points+width/4, width/2) - _ramp(points-width/4, width/2)) return KernelMat, KernelGradMat else: return KernelMat def _matrix_integration(q, h, t): ''' Returns the dp metric for a single horsetail curve at a given value of the epistemic uncertainties''' N = len(q) # correction if CDF has gone out of trapezium range if h[-1] < 0.9: h[-1] = 1.0 W = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) dp = (q - t).T.dot(W).dot(q - t) return dp def _matrix_grad(q, h, h_dx, t, t_prime): ''' Returns the gradient with respect to a single variable''' N = len(q) W = np.zeros([N, N]) Wprime = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) Wprime[i, i] = \ 0.5*(h_dx[min(i+1, N-1)] - h_dx[max(i-1, 0)]) tgrad = np.array([t_prime[i]*h_dx[i] for i in np.arange(N)]) grad = 2.0*(q - t).T.dot(W).dot(-1.0*tgrad) \ + (q - t).T.dot(Wprime).dot(q - t) return grad def _appendPlotArrays(q, h, integration_points): q = np.insert(q, 0, q[0]) h = np.insert(h, 0, 0) q = np.insert(q, 0, min(integration_points)) h = np.insert(h, 0, 0) q = np.append(q, q[-1]) h = np.append(h, 1) q = np.append(q, max(integration_points)) h = np.append(h, 1) return q, h def _finDiff(fobj, dv, f0=None, eps=10**-6): if f0 is None: f0 = fobj(dv) fbase = copy.copy(f0) fnew = fobj(dv + eps) return float((fnew - fbase)/eps) def _makeIter(x): try: iter(x) return [xi for xi in x] except: return [x] def _intervalSample(returned_samples, bounds): if len(returned_samples) < 1: return bounds[0] elif len(returned_samples) < 2: return bounds[1] else: return np.random.uniform(bounds[0], bounds[1])
lwcook/horsetail-matching
horsetailmatching/hm.py
_extgrad
python
def _extgrad(xarr, alpha=100, axis=None): '''Given an array xarr of values, return the gradient of the smooth min/max swith respect to each entry in the array''' term1 = (np.exp(alpha*xarr)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) term2 = 1 + alpha*(xarr - _extalg(xarr, alpha, axis=axis)) return term1*term2
Given an array xarr of values, return the gradient of the smooth min/max swith respect to each entry in the array
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L839-L846
null
import pdb import time import math import copy import warnings import numpy as np class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples ############################################################################## ## Private functions ############################################################################## def _extalg(xarr, alpha=100, axis=None): '''Given an array xarr of values, smoothly return the max/min''' return (np.sum(xarr * np.exp(alpha*xarr), axis=axis, keepdims=True)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) def _ramp(x, width): return _minsmooth(1, _maxsmooth(0, (x - width/2)*(1/width))) def _trint(x, width): w = width/2. xb = _maxsmooth(-w, _minsmooth(x, w)) y1 = 0.5 + xb/w + xb**2/(2*w**2) y2 = xb/w - xb**2/(2*w**2) return _minsmooth(y1, 0.5) + _maxsmooth(y2, 0.0) def _minsmooth(a, b, eps=0.0000): return 0.5*(a + b - np.sqrt((a-b)**2 + eps**2)) def _maxsmooth(a, b, eps=0.0000): return 0.5*(a + b + np.sqrt((a-b)**2 + eps**2)) def _step(x): return 1 * (x > 0) def _erf(r): ## Numerical implementation of the error function for matrix comptibility # save the sign of x sign = np.sign(r) x = np.absolute(r) # constants a1 = 0.254829592 a2 = -0.284496736 a3 = 1.421413741 a4 = -1.453152027 a5 = 1.061405429 p = 0.3275911 # A&S formula 7.1.26 t = 1.0/(1.0 + p*x) y = 1.0 - (((((a5*t + a4)*t) + a3)*t + a2)*t + a1)*t*np.exp(-x*x) return sign*y # erf(-x) = -erf(x) def _kernel(points, M, bw, ktype='gauss', bGrad=False): if ktype == 'gauss' or ktype == 'gaussian': KernelMat = (1./M)*((1 + _erf((points/bw)/np.sqrt(2.)))/2.) # KernelMat = np.zeros(points.shape) # for ir in np.arange(points.shape[0]): # for ic in np.arange(points.shape[1]): # KernelMat[ir, ic] = (1./M)*((1. + # math.erf((points[ir, ic]/bw)/math.sqrt(2.)))/2.) elif ktype == 'uniform' or ktype == 'uni': KernelMat = (1./M)*_ramp(points, width=bw*np.sqrt(12)) elif ktype == 'triangle' or ktype == 'tri': KernelMat = (1./M)*_trint(points, width=bw*2.*np.sqrt(6)) if bGrad: if ktype == 'gauss' or ktype == 'gaussian': const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) KernelGradMat = const_term * np.exp(-(1./2.) * (points/bw)**2) elif ktype == 'uniform' or ktype == 'uni': width = bw*np.sqrt(12) const = (1./M)*(1./width) KernelGradMat = const*(_step(points+width/2) - _step(points-width/2)) elif ktype == 'triangle' or ktype == 'tri': width = bw*2.*np.sqrt(6) const = (1./M)*(2./width) KernelGradMat = const*(_ramp(points+width/4, width/2) - _ramp(points-width/4, width/2)) return KernelMat, KernelGradMat else: return KernelMat def _matrix_integration(q, h, t): ''' Returns the dp metric for a single horsetail curve at a given value of the epistemic uncertainties''' N = len(q) # correction if CDF has gone out of trapezium range if h[-1] < 0.9: h[-1] = 1.0 W = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) dp = (q - t).T.dot(W).dot(q - t) return dp def _matrix_grad(q, h, h_dx, t, t_prime): ''' Returns the gradient with respect to a single variable''' N = len(q) W = np.zeros([N, N]) Wprime = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) Wprime[i, i] = \ 0.5*(h_dx[min(i+1, N-1)] - h_dx[max(i-1, 0)]) tgrad = np.array([t_prime[i]*h_dx[i] for i in np.arange(N)]) grad = 2.0*(q - t).T.dot(W).dot(-1.0*tgrad) \ + (q - t).T.dot(Wprime).dot(q - t) return grad def _appendPlotArrays(q, h, integration_points): q = np.insert(q, 0, q[0]) h = np.insert(h, 0, 0) q = np.insert(q, 0, min(integration_points)) h = np.insert(h, 0, 0) q = np.append(q, q[-1]) h = np.append(h, 1) q = np.append(q, max(integration_points)) h = np.append(h, 1) return q, h def _finDiff(fobj, dv, f0=None, eps=10**-6): if f0 is None: f0 = fobj(dv) fbase = copy.copy(f0) fnew = fobj(dv + eps) return float((fnew - fbase)/eps) def _makeIter(x): try: iter(x) return [xi for xi in x] except: return [x] def _intervalSample(returned_samples, bounds): if len(returned_samples) < 1: return bounds[0] elif len(returned_samples) < 2: return bounds[1] else: return np.random.uniform(bounds[0], bounds[1])
lwcook/horsetail-matching
horsetailmatching/hm.py
_matrix_integration
python
def _matrix_integration(q, h, t): ''' Returns the dp metric for a single horsetail curve at a given value of the epistemic uncertainties''' N = len(q) # correction if CDF has gone out of trapezium range if h[-1] < 0.9: h[-1] = 1.0 W = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) dp = (q - t).T.dot(W).dot(q - t) return dp
Returns the dp metric for a single horsetail curve at a given value of the epistemic uncertainties
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L921-L936
null
import pdb import time import math import copy import warnings import numpy as np class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples ############################################################################## ## Private functions ############################################################################## def _extalg(xarr, alpha=100, axis=None): '''Given an array xarr of values, smoothly return the max/min''' return (np.sum(xarr * np.exp(alpha*xarr), axis=axis, keepdims=True)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) def _extgrad(xarr, alpha=100, axis=None): '''Given an array xarr of values, return the gradient of the smooth min/max swith respect to each entry in the array''' term1 = (np.exp(alpha*xarr)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) term2 = 1 + alpha*(xarr - _extalg(xarr, alpha, axis=axis)) return term1*term2 def _ramp(x, width): return _minsmooth(1, _maxsmooth(0, (x - width/2)*(1/width))) def _trint(x, width): w = width/2. xb = _maxsmooth(-w, _minsmooth(x, w)) y1 = 0.5 + xb/w + xb**2/(2*w**2) y2 = xb/w - xb**2/(2*w**2) return _minsmooth(y1, 0.5) + _maxsmooth(y2, 0.0) def _minsmooth(a, b, eps=0.0000): return 0.5*(a + b - np.sqrt((a-b)**2 + eps**2)) def _maxsmooth(a, b, eps=0.0000): return 0.5*(a + b + np.sqrt((a-b)**2 + eps**2)) def _step(x): return 1 * (x > 0) def _erf(r): ## Numerical implementation of the error function for matrix comptibility # save the sign of x sign = np.sign(r) x = np.absolute(r) # constants a1 = 0.254829592 a2 = -0.284496736 a3 = 1.421413741 a4 = -1.453152027 a5 = 1.061405429 p = 0.3275911 # A&S formula 7.1.26 t = 1.0/(1.0 + p*x) y = 1.0 - (((((a5*t + a4)*t) + a3)*t + a2)*t + a1)*t*np.exp(-x*x) return sign*y # erf(-x) = -erf(x) def _kernel(points, M, bw, ktype='gauss', bGrad=False): if ktype == 'gauss' or ktype == 'gaussian': KernelMat = (1./M)*((1 + _erf((points/bw)/np.sqrt(2.)))/2.) # KernelMat = np.zeros(points.shape) # for ir in np.arange(points.shape[0]): # for ic in np.arange(points.shape[1]): # KernelMat[ir, ic] = (1./M)*((1. + # math.erf((points[ir, ic]/bw)/math.sqrt(2.)))/2.) elif ktype == 'uniform' or ktype == 'uni': KernelMat = (1./M)*_ramp(points, width=bw*np.sqrt(12)) elif ktype == 'triangle' or ktype == 'tri': KernelMat = (1./M)*_trint(points, width=bw*2.*np.sqrt(6)) if bGrad: if ktype == 'gauss' or ktype == 'gaussian': const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) KernelGradMat = const_term * np.exp(-(1./2.) * (points/bw)**2) elif ktype == 'uniform' or ktype == 'uni': width = bw*np.sqrt(12) const = (1./M)*(1./width) KernelGradMat = const*(_step(points+width/2) - _step(points-width/2)) elif ktype == 'triangle' or ktype == 'tri': width = bw*2.*np.sqrt(6) const = (1./M)*(2./width) KernelGradMat = const*(_ramp(points+width/4, width/2) - _ramp(points-width/4, width/2)) return KernelMat, KernelGradMat else: return KernelMat def _matrix_grad(q, h, h_dx, t, t_prime): ''' Returns the gradient with respect to a single variable''' N = len(q) W = np.zeros([N, N]) Wprime = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) Wprime[i, i] = \ 0.5*(h_dx[min(i+1, N-1)] - h_dx[max(i-1, 0)]) tgrad = np.array([t_prime[i]*h_dx[i] for i in np.arange(N)]) grad = 2.0*(q - t).T.dot(W).dot(-1.0*tgrad) \ + (q - t).T.dot(Wprime).dot(q - t) return grad def _appendPlotArrays(q, h, integration_points): q = np.insert(q, 0, q[0]) h = np.insert(h, 0, 0) q = np.insert(q, 0, min(integration_points)) h = np.insert(h, 0, 0) q = np.append(q, q[-1]) h = np.append(h, 1) q = np.append(q, max(integration_points)) h = np.append(h, 1) return q, h def _finDiff(fobj, dv, f0=None, eps=10**-6): if f0 is None: f0 = fobj(dv) fbase = copy.copy(f0) fnew = fobj(dv + eps) return float((fnew - fbase)/eps) def _makeIter(x): try: iter(x) return [xi for xi in x] except: return [x] def _intervalSample(returned_samples, bounds): if len(returned_samples) < 1: return bounds[0] elif len(returned_samples) < 2: return bounds[1] else: return np.random.uniform(bounds[0], bounds[1])
lwcook/horsetail-matching
horsetailmatching/hm.py
_matrix_grad
python
def _matrix_grad(q, h, h_dx, t, t_prime): ''' Returns the gradient with respect to a single variable''' N = len(q) W = np.zeros([N, N]) Wprime = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) Wprime[i, i] = \ 0.5*(h_dx[min(i+1, N-1)] - h_dx[max(i-1, 0)]) tgrad = np.array([t_prime[i]*h_dx[i] for i in np.arange(N)]) grad = 2.0*(q - t).T.dot(W).dot(-1.0*tgrad) \ + (q - t).T.dot(Wprime).dot(q - t) return grad
Returns the gradient with respect to a single variable
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L938-L954
null
import pdb import time import math import copy import warnings import numpy as np class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples ############################################################################## ## Private functions ############################################################################## def _extalg(xarr, alpha=100, axis=None): '''Given an array xarr of values, smoothly return the max/min''' return (np.sum(xarr * np.exp(alpha*xarr), axis=axis, keepdims=True)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) def _extgrad(xarr, alpha=100, axis=None): '''Given an array xarr of values, return the gradient of the smooth min/max swith respect to each entry in the array''' term1 = (np.exp(alpha*xarr)/ np.sum(np.exp(alpha*xarr), axis=axis, keepdims=True)) term2 = 1 + alpha*(xarr - _extalg(xarr, alpha, axis=axis)) return term1*term2 def _ramp(x, width): return _minsmooth(1, _maxsmooth(0, (x - width/2)*(1/width))) def _trint(x, width): w = width/2. xb = _maxsmooth(-w, _minsmooth(x, w)) y1 = 0.5 + xb/w + xb**2/(2*w**2) y2 = xb/w - xb**2/(2*w**2) return _minsmooth(y1, 0.5) + _maxsmooth(y2, 0.0) def _minsmooth(a, b, eps=0.0000): return 0.5*(a + b - np.sqrt((a-b)**2 + eps**2)) def _maxsmooth(a, b, eps=0.0000): return 0.5*(a + b + np.sqrt((a-b)**2 + eps**2)) def _step(x): return 1 * (x > 0) def _erf(r): ## Numerical implementation of the error function for matrix comptibility # save the sign of x sign = np.sign(r) x = np.absolute(r) # constants a1 = 0.254829592 a2 = -0.284496736 a3 = 1.421413741 a4 = -1.453152027 a5 = 1.061405429 p = 0.3275911 # A&S formula 7.1.26 t = 1.0/(1.0 + p*x) y = 1.0 - (((((a5*t + a4)*t) + a3)*t + a2)*t + a1)*t*np.exp(-x*x) return sign*y # erf(-x) = -erf(x) def _kernel(points, M, bw, ktype='gauss', bGrad=False): if ktype == 'gauss' or ktype == 'gaussian': KernelMat = (1./M)*((1 + _erf((points/bw)/np.sqrt(2.)))/2.) # KernelMat = np.zeros(points.shape) # for ir in np.arange(points.shape[0]): # for ic in np.arange(points.shape[1]): # KernelMat[ir, ic] = (1./M)*((1. + # math.erf((points[ir, ic]/bw)/math.sqrt(2.)))/2.) elif ktype == 'uniform' or ktype == 'uni': KernelMat = (1./M)*_ramp(points, width=bw*np.sqrt(12)) elif ktype == 'triangle' or ktype == 'tri': KernelMat = (1./M)*_trint(points, width=bw*2.*np.sqrt(6)) if bGrad: if ktype == 'gauss' or ktype == 'gaussian': const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) KernelGradMat = const_term * np.exp(-(1./2.) * (points/bw)**2) elif ktype == 'uniform' or ktype == 'uni': width = bw*np.sqrt(12) const = (1./M)*(1./width) KernelGradMat = const*(_step(points+width/2) - _step(points-width/2)) elif ktype == 'triangle' or ktype == 'tri': width = bw*2.*np.sqrt(6) const = (1./M)*(2./width) KernelGradMat = const*(_ramp(points+width/4, width/2) - _ramp(points-width/4, width/2)) return KernelMat, KernelGradMat else: return KernelMat def _matrix_integration(q, h, t): ''' Returns the dp metric for a single horsetail curve at a given value of the epistemic uncertainties''' N = len(q) # correction if CDF has gone out of trapezium range if h[-1] < 0.9: h[-1] = 1.0 W = np.zeros([N, N]) for i in range(N): W[i, i] = 0.5*(h[min(i+1, N-1)] - h[max(i-1, 0)]) dp = (q - t).T.dot(W).dot(q - t) return dp def _appendPlotArrays(q, h, integration_points): q = np.insert(q, 0, q[0]) h = np.insert(h, 0, 0) q = np.insert(q, 0, min(integration_points)) h = np.insert(h, 0, 0) q = np.append(q, q[-1]) h = np.append(h, 1) q = np.append(q, max(integration_points)) h = np.append(h, 1) return q, h def _finDiff(fobj, dv, f0=None, eps=10**-6): if f0 is None: f0 = fobj(dv) fbase = copy.copy(f0) fnew = fobj(dv + eps) return float((fnew - fbase)/eps) def _makeIter(x): try: iter(x) return [xi for xi in x] except: return [x] def _intervalSample(returned_samples, bounds): if len(returned_samples) < 1: return bounds[0] elif len(returned_samples) < 2: return bounds[1] else: return np.random.uniform(bounds[0], bounds[1])
lwcook/horsetail-matching
horsetailmatching/hm.py
HorsetailMatching.evalSamples
python
def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples
Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L292-L326
[ "def _makeIter(x):\n try:\n iter(x)\n return [xi for xi in x]\n except:\n return [x]\n", "def _makeSurrogates(self, x):\n\n # Get quadrature points\n if self.surrogate_points is None:\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)],\n copy=False)\n u_sparse = np.vstack([m.flatten() for m in mesh]).T\n else:\n u_sparse = self.surrogate_points\n\n N_sparse = u_sparse.shape[0]\n q_sparse = np.zeros(N_sparse)\n\n # Get surrogates in correct form\n if not self.jac:\n for iu, u in enumerate(u_sparse):\n q_sparse[iu] = self.fqoi(x, u)\n\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n\n def fqoi(u):\n return surr_qoi(u)\n fgrad = False\n surr_jac = False\n\n else:\n g_sparse = np.zeros([N_sparse, self._N_dv])\n for iu, u in enumerate(u_sparse):\n if isinstance(self.jac, bool) and self.jac:\n q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u)\n else:\n q_sparse[iu] = self.fqoi(x, u)\n g_sparse[iu, :] = self.jac(x, u)\n\n if not self.surrogate_jac:\n fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)]\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n for k in np.arange(self._N_dv):\n fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k])\n def surr_grad(u):\n return [f(u) for f in fpartial]\n else:\n if isinstance(self.surrogate_jac, bool) and self.surrogate_jac:\n surr_qoi, surr_grad = self.surrogate(\n u_sparse, q_sparse, g_sparse)\n else:\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n surr_grad = self.surrogate_jac(u_sparse, g_sparse)\n\n def fqoi(u):\n return(surr_qoi(u))\n def fgrad(u):\n return(surr_grad(u))\n surr_jac = fgrad\n\n return fqoi, fgrad, surr_jac\n", " def _getParameterSamples(self):\n\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n\n get_new = True\n if self.reuse_samples and self.u_samples is not None:\n if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u):\n if self.verbose:\n print('''Stored samples do not match current dimensions,\n getting new samples''')\n else:\n get_new = False\n\n if get_new:\n if self.verbose:\n print('Getting uncertain parameter samples')\n\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n N_prob = len(self.prob_uncertainties)\n N_int = len(self.int_uncertainties)\n# u_samples = np.zeros([self.samples_int, self.samples_prob, N_u])\n\n u_samples_prob = np.zeros([self.samples_int, self.samples_prob,\n len(self.prob_uncertainties)])\n u_samples_int = np.zeros([self.samples_int, self.samples_prob,\n len(self.int_uncertainties)])\n\n u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)])\n for kk, uk in enumerate(self.int_uncertainties):\n if callable(uk):\n samps = np.array(uk()).flatten()\n if len(samps) != self.samples_prob:\n raise Exception('Number of samples returned not equal ' +\n 'to specified number of samples: please set number of ' +\n 'samples with samples_prob attribute')\n else:\n u_ints[:, kk] = samps\n elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds\n lb, ub = uk[0], uk[1]\n u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int)\n u_ints[0, kk] = lb\n u_ints[-1, kk] = ub\n elif hasattr(uk, 'getSample'):\n for ii in np.arange(self.samples_int):\n u_ints[ii, kk] = uk.getSample()\n else:\n raise TypeError('Unsupported interval uncertainty type')\n\n u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1))\n\n u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)])\n for kk, uk in enumerate(self.prob_uncertainties):\n if callable(uk):\n samps = np.array(uk()).flatten()\n if len(samps) != self.samples_prob:\n raise Exception('Number of samples returned not equal ' +\n 'to specified number of samples: please set number of ' +\n 'samples with samples_prob attribute')\n else:\n u_probs[:, kk] = samps\n elif hasattr(uk, 'getSample'):\n for jj in np.arange(self.samples_prob):\n u_probs[jj, kk] = uk.getSample()\n else:\n raise TypeError('Unsupported probabilistic uncertainty type')\n\n u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1))\n\n u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2)\n\n self.u_samples = u_samples\n return u_samples\n else:\n if self.verbose:\n print('Re-using stored samples')\n return self.u_samples\n", "def _evalSamples(self, u_samples, fqoi, fgrad, jac):\n\n # Array of shape (M_int, M_prob)\n grad_samples = None\n q_samples = np.zeros([self.samples_int, self.samples_prob])\n if not jac:\n for ii in np.arange(q_samples.shape[0]):\n for jj in np.arange(q_samples.shape[1]):\n q_samples[ii, jj] = fqoi(u_samples[ii, jj])\n else:\n grad_samples = np.zeros([self.samples_int, self.samples_prob,\n self._N_dv])\n for ii in np.arange(q_samples.shape[0]):\n for jj in np.arange(q_samples.shape[1]):\n if isinstance(jac, bool) and jac:\n (q, grad) = fqoi(u_samples[ii, jj])\n q_samples[ii, jj] = float(q)\n grad_samples[ii, jj, :] = [_ for _ in grad]\n else:\n q_samples[ii, jj] = fqoi(u_samples[ii, jj])\n grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj])\n\n self.grad_samples = grad_samples\n\n self.q_samples = q_samples\n\n return q_samples, grad_samples\n" ]
class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples
lwcook/horsetail-matching
horsetailmatching/hm.py
HorsetailMatching.evalMetric
python
def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method)
Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0)
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L328-L363
[ "def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None):\n '''Evaluates the density matching metric from given samples of the quantity\n of interest and gradient instead of evaluating them at a design.\n\n :param np.ndarray q_samples: samples of the quantity of interest,\n size (M_int, M_prob)\n :param np.ndarray grad_samples: samples of the gradien,\n size (M_int, M_prob, n_x)\n\n :return: metric_value - value of the metric\n\n :rtype: float\n\n '''\n return self._evalDensityMetric(q_samples, grad_samples)\n", " def evalSamples(self, x):\n '''Evalautes the samples of quantity of interest and its gradient\n (if supplied) at the given values of the design variables\n\n :param iterable x: values of the design variables, this is passed as\n the first argument to the function fqoi\n\n :return: (values of the quantity of interest, values of the gradient)\n :rtype: Tuple\n '''\n\n # Make sure dimensions are correct\n# u_sample_dimensions = self._processDimensions()\n\n self._N_dv = len(_makeIter(x))\n\n if self.verbose:\n print('Evaluating surrogate')\n if self.surrogate is None:\n def fqoi(u):\n return self.fqoi(x, u)\n def fgrad(u):\n return self.jac(x, u)\n jac = self.jac\n else:\n fqoi, fgrad, surr_jac = self._makeSurrogates(x)\n jac = surr_jac\n\n u_samples = self._getParameterSamples()\n\n if self.verbose:\n print('Evaluating quantity of interest at samples')\n q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac)\n\n return q_samples, grad_samples\n", " def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None):\n '''Evaluates the horsetail matching metric from given samples of the quantity\n of interest and gradient instead of evaluating them at a design.\n\n :param np.ndarray q_samples: samples of the quantity of interest,\n size (M_int, M_prob)\n :param np.ndarray grad_samples: samples of the gradien,\n size (M_int, M_prob, n_x)\n\n :return: metric_value - value of the metric\n\n :rtype: float\n\n '''\n # Make sure dimensions are correct\n# u_sample_dimensions = self._processDimensions()\n\n q_samples = np.array(q_samples)\n if not (q_samples.shape[0] == self.samples_int and\n q_samples.shape[1] == self.samples_prob):\n raise ValueError('Shape of q_samples should be [M_int, M_prob]')\n\n if grad_samples is not None:\n grad_samples = np.array(grad_samples)\n if not (grad_samples.shape[0] == self.samples_int and\n grad_samples.shape[1] == self.samples_prob):\n raise ValueError('''Shape of grad_samples\n should be [M_int, M_prob, n_dv]''')\n\n if method is None:\n method = self.method\n\n if method.lower() == 'empirical':\n return self._evalMetricEmpirical(q_samples, grad_samples)\n elif method.lower() == 'kernel':\n return self._evalMetricKernel(q_samples, grad_samples)\n else:\n raise ValueError('Unsupported metric evalation method')\n" ]
class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples
lwcook/horsetail-matching
horsetailmatching/hm.py
HorsetailMatching.evalMetricFromSamples
python
def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method')
Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L365-L402
[ "def _evalMetricEmpirical(self, q_samples, grad_samples=None):\n\n M_prob = self.samples_prob\n M_int = self.samples_int\n\n if M_int > 1:\n alpha = self.alpha\n else:\n alpha = 1\n\n h_htail = np.zeros([M_int, M_prob])\n q_htail = np.zeros([M_int, M_prob])\n q_l = np.zeros(M_prob)\n q_u = np.zeros(M_prob)\n if grad_samples is not None:\n g_htail = np.zeros([M_int, M_prob, self._N_dv])\n g_l = np.zeros([M_prob, self._N_dv])\n g_u = np.zeros([M_prob, self._N_dv])\n Du_grad = np.zeros(self._N_dv)\n Dl_grad = np.zeros(self._N_dv)\n\n for ii in np.arange(M_int):\n # Get empirical CDF by sorting samples at each value of intervals\n sortinds = np.argsort(q_samples[ii, :])\n q_htail[ii, :] = q_samples[ii, sortinds]\n M = q_samples.shape[1]\n h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)]\n\n if grad_samples is not None:\n for ix in np.arange(self._N_dv):\n g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix]\n\n for jj in np.arange(M_prob):\n q_u[jj] = min(q_htail[:, jj])\n q_l[jj] = max(q_htail[:, jj])\n\n if grad_samples is not None:\n q_u[jj] = _extalg(q_htail[:, jj], -1*alpha)\n q_l[jj] = _extalg(q_htail[:, jj], alpha)\n for ix in np.arange(self._N_dv):\n gtemp = _extgrad(q_htail[:, jj], -1*alpha)\n g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix])\n gtemp = _extgrad(q_htail[:, jj], alpha)\n g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix])\n\n h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs\n t_u = [self._ftarg_u(hi) for hi in h_u]\n t_l = [self._ftarg_l(hi) for hi in h_u]\n\n self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u\n self._qh, self._hh = q_htail, h_htail\n self._tl, self._tu = t_l, t_u\n self._qis = None\n\n Du = (1./M_prob)*sum((q_u - t_u)**2)\n Dl = (1./M_prob)*sum((q_l - t_l)**2)\n dhat = np.sqrt(Du + Dl)\n\n if self.verbose:\n print('Metric: ' + str(dhat))\n\n if grad_samples is not None:\n for ix in np.arange(self._N_dv):\n Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix])\n Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix])\n\n dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad))\n if self.verbose:\n print('Gradient: ' + str([g for g in dhat_grad]))\n\n return dhat, dhat_grad\n\n else:\n return dhat\n" ]
class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples
lwcook/horsetail-matching
horsetailmatching/hm.py
HorsetailMatching.getHorsetail
python
def getHorsetail(self): '''Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show() ''' if hasattr(self, '_ql'): ql, qu, hl, hu = self._ql, self._qu, self._hl, self._hu qh, hh = self._qh, self._hh if self._qis is not None: ql, hl = _appendPlotArrays(ql, hl, self._qis) qu, hu = _appendPlotArrays(qu, hu, self._qis) CDFs = [] for qi, hi in zip(qh, hh): CDFs.append((qi, hi)) upper_target = [self._ftarg_u(h) for h in hu] upper_curve = (qu, hu, upper_target) lower_target = [self._ftarg_l(h) for h in hl] lower_curve = (ql, hl, lower_target) return upper_curve, lower_curve, CDFs else: raise ValueError('''The metric has not been evaluated at any design point so the horsetail does not exist''')
Function that gets vectors of the horsetail plot at the last design evaluated. :return: upper_curve, lower_curve, CDFs - returns three parameters, the first two are tuples containing pairs of x/y vectors of the upper and lower bounds on the CDFs (the horsetail plot). The third parameter is a list of x/y tuples for individual CDFs propagated at each sampled value of the interval uncertainties *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> (x1, y1, t1), (x2, y2, t2), CDFs = theHM.getHorsetail() >>> matplotlib.pyplot(x1, y1, 'b') >>> matplotlib.pyplot(x2, y2, 'b') >>> for (x, y) in CDFs: ... matplotlib.pyplot(x, y, 'k:') >>> matplotlib.pyplot.show()
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/hm.py#L405-L450
null
class HorsetailMatching(object): '''Class for using horsetail matching within an optimization. The main functionality is to evaluate the horsetail matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a HorsetailMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Each can be an instance of the UncertainParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be a function which returns sample(s) using whatever method is desired. :param list int_uncertainties: list of interval uncertainties [default []]. Each can be an instance of the IntervalParameter class, in which case they will be sampled using the getSample() method. Alternatiely each can be specified as a tuple/list of the bounds. :param function ftarget: function that returns the value of the target inverse CDF given a value in [0,1]. Can be a tuple that gives two target fuctions, one for the upper bound and one for the lower bound on the CDF under mixed uncertainties [default t(h) = 0] :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param str method: method with which to evaluate the horsetil matching metric, can be 'empirical' or 'kernel' [default 'empirical' if jac is False else default 'kernel']. :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param int samples_int: number of samples to take from the interval uncertainties. Note that under mixed uncertainties, a nested loop is used to evaluate the metric so the total number of samples will be samples_prob*samples_int (at each interval uncertainty sample samples_prob samples are taken from the probabilistic uncertainties). [default 50] :param list integration_points: Only for method='kernel'. The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: Only for method='kernel'. The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param str kernel_type: Only for method='kernel'. The type of kernel to use, can be 'gaussian', 'uniform', or 'triangle' [default 'gaussian']. :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. *Example Declarations*:: >>> from horsetailmatching import HorsetailMatching, UncertainParameter, PolySurrogate >>> def myFunc(x, u): return x[0]*x[1] + u >>> def myGrad(x, u): return [x[1], x[0]] >>> def myTarg1(h): return 1-h**3 >>> def myTarg2(h): return 2-h**3 >>> u1 = UniformParameter() >>> u2 = IntervalParameter() >>> U = [u1, u2] >>> poly = PolySurrogate(dimensions=2) >>> poly_points = poly.getQuadraturePoints() >>> theHM = HorsetailMatching(myFunc, U) >>> theHM = HorsetailMatching(myFunc, U, jac=myGrad, method='kernel') >>> theHM = HorsetailMatching(myFunc, U, ftarget=myTarg1) >>> theHM = HorsetailMatching(myFunc, U, ftarget=(myTarg1, myTarg2)) >>> theHM = HorsetailMatching(myFunc, U, samples_prob=500, samples_int = 50) >>> theHM = HorsetailMatching(myFunc, U, method='kernel', integration_points=numpy.linspace(0, 10, 100), kernel_bandwidth=0.01) >>> theHM = HorsetailMatching(myFunc, U, surrogate=poly.surrogate, surrogate_jac=False, surrogate_points=poly_points) >>> theHM = HorsetailMatching(myFunc, U, verbose=True, reuse_samples=True) ''' def __init__(self, fqoi, prob_uncertainties, int_uncertainties=[], ftarget=None, jac=False, method=None, samples_prob=100, samples_int=50, integration_points=None, kernel_bandwidth=None, kernel_type='gaussian', alpha=400, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi # self.uncertain_parameters = uncertain_parameters self.prob_uncertainties = prob_uncertainties self.int_uncertainties = int_uncertainties self.ftarget = ftarget self.jac = jac self.method = method # Must be done after setting jac self.samples_prob = samples_prob self.samples_int = samples_int self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.kernel_type = kernel_type self.alpha = alpha self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose ############################################################################### ## Properties with non-trivial setting behaviour ############################################################################### # @property # def uncertain_parameters(self): # return self._u_params # # @uncertain_parameters.setter # def uncertain_parameters(self, params): # self._u_params = _makeIter(params) # if len(self._u_params) == 0: # raise ValueError('No uncertain parameters provided') # # self._u_int, self._u_prob = [], [] # for ii, u in enumerate(self._u_params): # if u.is_interval_uncertainty: # self._u_int.append((ii, u)) # else: # self._u_prob.append((ii, u)) @property def prob_uncertainties(self): return self._prob_uncertainties @prob_uncertainties.setter def prob_uncertainties(self, params): self._prob_uncertainties = _makeIter(params) @property def int_uncertainties(self): return self._int_uncertainties @int_uncertainties.setter def int_uncertainties(self, params): self._int_uncertainties = _makeIter(params) @property def samples_prob(self): return self._samples_prob @samples_prob.setter def samples_prob(self, value): if len(self.prob_uncertainties) > 0: self._samples_prob = value else: self._samples_prob = 1 @property def samples_int(self): return self._samples_int @samples_int.setter def samples_int(self, value): if len(self.int_uncertainties) > 0: self._samples_int = value else: self._samples_int = 1 @property def method(self): return self._method @method.setter def method(self, value): if value is None: if self.jac is False: self._method = 'empirical' else: self._method = 'kernel' else: self._method = value @property def ftarget(self): return self._ftarget @ftarget.setter def ftarget(self, value): def standardTarget(h): return 0 try: iter(value) self._ftarg_u = value[0] self._ftarg_l = value[1] self._ftarget = value except: if value is None: self._ftarget = standardTarget else: self._ftarget = value self._ftarg_u = self._ftarget self._ftarg_l = self._ftarget @property def u_samples(self): return self._u_samples @u_samples.setter def u_samples(self, samples): if samples is not None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) if (not isinstance(samples, np.ndarray) or samples.shape != (self.samples_int, self.samples_prob, N_u)): raise TypeError('u_samples should be a np.array of size' '(samples_int, samples_prob, num_uncertanities)') self._u_samples = samples @property def kernel_type(self): return self._kernel_type @kernel_type.setter def kernel_type(self, value): allowed_types = ['gaussian', 'uniform', 'triangle'] if value not in allowed_types: raise ValueError('Kernel type must be one of'+ ', '.join([str(t) for t in allowed_types])) else: self._kernel_type = value ############################################################################## ## Public Methods ############################################################################## def evalSamples(self, x): '''Evalautes the samples of quantity of interest and its gradient (if supplied) at the given values of the design variables :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: (values of the quantity of interest, values of the gradient) :rtype: Tuple ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) return q_samples, grad_samples def evalMetric(self, x, method=None): '''Evaluates the horsetail matching metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param str method: method to use to evaluate the metric ('empirical' or 'kernel') :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theHM = HorsetailMatching(myFunc, u) >>> x0 = [1, 2] >>> theHM.evalMetric(x0) ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() if self.verbose: print('----------') print('At design: ' + str(x)) q_samples, grad_samples = self.evalSamples(x) if self.verbose: print('Evaluating metric') return self.evalMetricFromSamples(q_samples, grad_samples, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the horsetail matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' # Make sure dimensions are correct # u_sample_dimensions = self._processDimensions() q_samples = np.array(q_samples) if not (q_samples.shape[0] == self.samples_int and q_samples.shape[1] == self.samples_prob): raise ValueError('Shape of q_samples should be [M_int, M_prob]') if grad_samples is not None: grad_samples = np.array(grad_samples) if not (grad_samples.shape[0] == self.samples_int and grad_samples.shape[1] == self.samples_prob): raise ValueError('''Shape of grad_samples should be [M_int, M_prob, n_dv]''') if method is None: method = self.method if method.lower() == 'empirical': return self._evalMetricEmpirical(q_samples, grad_samples) elif method.lower() == 'kernel': return self._evalMetricKernel(q_samples, grad_samples) else: raise ValueError('Unsupported metric evalation method') ############################################################################## ## Private methods ## ############################################################################## def _evalMetricEmpirical(self, q_samples, grad_samples=None): M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 h_htail = np.zeros([M_int, M_prob]) q_htail = np.zeros([M_int, M_prob]) q_l = np.zeros(M_prob) q_u = np.zeros(M_prob) if grad_samples is not None: g_htail = np.zeros([M_int, M_prob, self._N_dv]) g_l = np.zeros([M_prob, self._N_dv]) g_u = np.zeros([M_prob, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) for ii in np.arange(M_int): # Get empirical CDF by sorting samples at each value of intervals sortinds = np.argsort(q_samples[ii, :]) q_htail[ii, :] = q_samples[ii, sortinds] M = q_samples.shape[1] h_htail[ii, :] = [(1./M)*(0.5 + j) for j in range(M)] if grad_samples is not None: for ix in np.arange(self._N_dv): g_htail[ii, :, ix] = grad_samples[ii, sortinds, ix] for jj in np.arange(M_prob): q_u[jj] = min(q_htail[:, jj]) q_l[jj] = max(q_htail[:, jj]) if grad_samples is not None: q_u[jj] = _extalg(q_htail[:, jj], -1*alpha) q_l[jj] = _extalg(q_htail[:, jj], alpha) for ix in np.arange(self._N_dv): gtemp = _extgrad(q_htail[:, jj], -1*alpha) g_u[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) gtemp = _extgrad(q_htail[:, jj], alpha) g_l[jj, ix] = gtemp.dot(g_htail[:, jj, ix]) h_u, h_l = h_htail[0], h_htail[0] # h is same for all ECDFs t_u = [self._ftarg_u(hi) for hi in h_u] t_l = [self._ftarg_l(hi) for hi in h_u] self._ql, self._qu, self._hl, self._hu = q_l, q_u, h_l, h_u self._qh, self._hh = q_htail, h_htail self._tl, self._tu = t_l, t_u self._qis = None Du = (1./M_prob)*sum((q_u - t_u)**2) Dl = (1./M_prob)*sum((q_l - t_l)**2) dhat = np.sqrt(Du + Dl) if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: for ix in np.arange(self._N_dv): Du_grad[ix] = (1./M_prob)*sum(2*(q_u - t_u)*g_u[:, ix]) Dl_grad[ix] = (1./M_prob)*sum(2*(q_l - t_l)*g_l[:, ix]) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _getKernelParameters(self, q_samples): # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if len(self.prob_uncertainties) > 0: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = 0.33*((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) else: bw = 1e-3 self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth ## Initalize arrays and prepare calculation q_min = np.amin(q_samples) q_max = np.amax(q_samples) if self.integration_points is None: q_range = q_max - q_min qis_full = np.linspace(q_min - q_range, q_max + q_range, 10000) self.integration_points = qis_full else: qis_full = np.array(self.integration_points) ii_low, ii_high = 0, len(qis_full) try: ii_high, qi_high = next((iq, qi) for iq, qi in enumerate(qis_full) if qi > q_max + 20*bw) except StopIteration: warnings.warn('Sample found higher than range of integration points') try: iiN_low, qi_low = next((iq, qi) for iq, qi in enumerate(qis_full[::-1]) if qi < q_min - 20*bw) ii_low = len(qis_full) - (iiN_low+1) except StopIteration: warnings.warn('Sample found lower than range of integration points') qis = qis_full[ii_low:ii_high+1] # Only evaluate over range of samples self._qis = qis return qis, bw def _evalMetricKernel(self, q_samples, grad_samples=None): qis, bw = self._getKernelParameters(q_samples) N_quad = len(qis) M_prob = self.samples_prob M_int = self.samples_int if M_int > 1: alpha = self.alpha else: alpha = 1 fhtail = np.zeros([N_quad, M_int]) qhtail = np.zeros([N_quad, M_int]) if grad_samples is not None: fht_grad = np.zeros([N_quad, M_int, self._N_dv]) hu_grad = np.zeros([N_quad, self._N_dv]) hl_grad = np.zeros([N_quad, self._N_dv]) Du_grad = np.zeros(self._N_dv) Dl_grad = np.zeros(self._N_dv) # ALGORITHM 1 from publication # Evaluate all individual CDFs and their gradients for mm in np.arange(M_int): qjs = q_samples[mm, :] rmat = qis.reshape([N_quad, 1])-qjs.reshape([1, M_prob]) if grad_samples is not None: Kcdf, Kprime = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=True) for ix in np.arange(self._N_dv): grad_js = grad_samples[mm, :, ix] fht_grad[:, mm, ix] = Kprime.dot(-1*grad_js) else: Kcdf = _kernel(rmat, M_prob, bw=bw, ktype=self.kernel_type, bGrad=False) fhtail[:, mm] = Kcdf.dot(np.ones([M_prob, 1])).flatten() qhtail[:, mm] = qis # ALGORITHM 2 from publication # Find horsetail curves - envelope of the CDFs and their gradients # In Matrix form if grad_samples is None: hu = np.max(fhtail, axis=1).flatten() hl = np.min(fhtail, axis=1).flatten() else: hu = _extalg(fhtail, alpha, axis=1).flatten() hl = _extalg(fhtail, -1*alpha, axis=1).flatten() Su_prime = _extgrad(fhtail, alpha, axis=1) Sl_prime = _extgrad(fhtail, -1*alpha, axis=1) for kx in np.arange(self._N_dv): fis_grad = fht_grad[:, :, kx] for ii in np.arange(N_quad): hu_grad[ii, kx] = Su_prime[ii, :].dot(fis_grad[ii, :]) hl_grad[ii, kx] = Sl_prime[ii, :].dot(fis_grad[ii, :]) # ALGORITHM 3 from publication # Evaluate overall metric and gradient using matrix multipliation tu = np.array([self._ftarg_u(hi) for hi in hu]) tl = np.array([self._ftarg_l(hi) for hi in hl]) Du = _matrix_integration(qis, hu, tu) Dl = _matrix_integration(qis, hl, tl) dhat = float(np.sqrt(Du + Dl)) self._ql, self._qu, self._hl, self._hu = qis, qis, hl, hu self._qh, self._hh = qhtail, fhtail self._tl, self._tu = tl, tu if self.verbose: print('Metric: ' + str(dhat)) if grad_samples is not None: tu_pr = np.array([_finDiff(self._ftarg_u, hi) for hi in hu]) tl_pr = np.array([_finDiff(self._ftarg_l, hi) for hi in hl]) for kx in np.arange(self._N_dv): Du_grad[kx] = _matrix_grad(qis, hu, hu_grad[:, kx], tu, tu_pr) Dl_grad[kx] = _matrix_grad(qis, hl, hl_grad[:, kx], tl, tl_pr) dhat_grad = (0.5*(Du+Dl)**(-0.5)*(Du_grad + Dl_grad)) if self.verbose: print('Gradient: ' + str([g for g in dhat_grad])) return dhat, dhat_grad else: return dhat def _makeSurrogates(self, x): # Get quadrature points if self.surrogate_points is None: N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)], copy=False) u_sparse = np.vstack([m.flatten() for m in mesh]).T else: u_sparse = self.surrogate_points N_sparse = u_sparse.shape[0] q_sparse = np.zeros(N_sparse) # Get surrogates in correct form if not self.jac: for iu, u in enumerate(u_sparse): q_sparse[iu] = self.fqoi(x, u) surr_qoi = self.surrogate(u_sparse, q_sparse) def fqoi(u): return surr_qoi(u) fgrad = False surr_jac = False else: g_sparse = np.zeros([N_sparse, self._N_dv]) for iu, u in enumerate(u_sparse): if isinstance(self.jac, bool) and self.jac: q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u) else: q_sparse[iu] = self.fqoi(x, u) g_sparse[iu, :] = self.jac(x, u) if not self.surrogate_jac: fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)] surr_qoi = self.surrogate(u_sparse, q_sparse) for k in np.arange(self._N_dv): fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k]) def surr_grad(u): return [f(u) for f in fpartial] else: if isinstance(self.surrogate_jac, bool) and self.surrogate_jac: surr_qoi, surr_grad = self.surrogate( u_sparse, q_sparse, g_sparse) else: surr_qoi = self.surrogate(u_sparse, q_sparse) surr_grad = self.surrogate_jac(u_sparse, g_sparse) def fqoi(u): return(surr_qoi(u)) def fgrad(u): return(surr_grad(u)) surr_jac = fgrad return fqoi, fgrad, surr_jac def _getParameterSamples(self): N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) get_new = True if self.reuse_samples and self.u_samples is not None: if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u): if self.verbose: print('''Stored samples do not match current dimensions, getting new samples''') else: get_new = False if get_new: if self.verbose: print('Getting uncertain parameter samples') N_u = len(self.prob_uncertainties) + len(self.int_uncertainties) N_prob = len(self.prob_uncertainties) N_int = len(self.int_uncertainties) # u_samples = np.zeros([self.samples_int, self.samples_prob, N_u]) u_samples_prob = np.zeros([self.samples_int, self.samples_prob, len(self.prob_uncertainties)]) u_samples_int = np.zeros([self.samples_int, self.samples_prob, len(self.int_uncertainties)]) u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)]) for kk, uk in enumerate(self.int_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_ints[:, kk] = samps elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds lb, ub = uk[0], uk[1] u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int) u_ints[0, kk] = lb u_ints[-1, kk] = ub elif hasattr(uk, 'getSample'): for ii in np.arange(self.samples_int): u_ints[ii, kk] = uk.getSample() else: raise TypeError('Unsupported interval uncertainty type') u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1)) u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)]) for kk, uk in enumerate(self.prob_uncertainties): if callable(uk): samps = np.array(uk()).flatten() if len(samps) != self.samples_prob: raise Exception('Number of samples returned not equal ' + 'to specified number of samples: please set number of ' + 'samples with samples_prob attribute') else: u_probs[:, kk] = samps elif hasattr(uk, 'getSample'): for jj in np.arange(self.samples_prob): u_probs[jj, kk] = uk.getSample() else: raise TypeError('Unsupported probabilistic uncertainty type') u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1)) u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2) self.u_samples = u_samples return u_samples else: if self.verbose: print('Re-using stored samples') return self.u_samples def _evalSamples(self, u_samples, fqoi, fgrad, jac): # Array of shape (M_int, M_prob) grad_samples = None q_samples = np.zeros([self.samples_int, self.samples_prob]) if not jac: for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): q_samples[ii, jj] = fqoi(u_samples[ii, jj]) else: grad_samples = np.zeros([self.samples_int, self.samples_prob, self._N_dv]) for ii in np.arange(q_samples.shape[0]): for jj in np.arange(q_samples.shape[1]): if isinstance(jac, bool) and jac: (q, grad) = fqoi(u_samples[ii, jj]) q_samples[ii, jj] = float(q) grad_samples[ii, jj, :] = [_ for _ in grad] else: q_samples[ii, jj] = fqoi(u_samples[ii, jj]) grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj]) self.grad_samples = grad_samples self.q_samples = q_samples return q_samples, grad_samples
lwcook/horsetail-matching
horsetailmatching/parameters.py
UncertainParameter.evalPDF
python
def evalPDF(self, u_values): '''Returns the PDF of the uncertain parameter evaluated at the values provided in u_values. :param iterable u_values: values of the uncertain parameter at which to evaluate the PDF *Example Usage* :: >>> u = UniformParameter() >>> X = numpy.linspace(-1, 1, 100) >>> Y = [u.evalPDF(x) for x in X] ''' if isinstance(u_values, np.ndarray): return self._evalPDF(u_values) else: try: iter(u_values) return [self._evalPDF(u) for u in u_values] except: return self._evalPDF(u_values)
Returns the PDF of the uncertain parameter evaluated at the values provided in u_values. :param iterable u_values: values of the uncertain parameter at which to evaluate the PDF *Example Usage* :: >>> u = UniformParameter() >>> X = numpy.linspace(-1, 1, 100) >>> Y = [u.evalPDF(x) for x in X]
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/parameters.py#L91-L113
null
class UncertainParameter(object): '''Base Class for handling uncertain parameters in optimization under uncertainty problems using horsetail matching. If this class is used, a custom distribution must be provided. Otherwise one of the child classes UniformParameter, IntervalParameter, or GaussianParameter should be used. All child classes use the methods getSample and evalPDF. :param function pdf: pdf function of distribution. Bounds on the distribution should also be provided via the lower_bound and upper_bound arguments. :param double lower_bound: lower bound of the distribution [default -1] :param double upper_bound: upper bound of the distribution [default 1] *Example Declaration* :: >>> def myPDF(q): if q > 2.5 or q < 1.5: return 0 else: return 1/(2.5 - 1.5) >>> u = UncertainParameter(pdf=myPDF, lower_bound=1.5, upper_bound=2.5) ''' default_lb = -1 default_ub = 1 def __init__(self, pdf=None, lower_bound=default_lb, upper_bound=default_ub): self.pdf = pdf self.lower_bound = lower_bound self.upper_bound = upper_bound self.is_interval_uncertainty = False self._max_pdf_val = None ############################################################################### ## Properties with non-trivail setting behaviour ############################################################################### @property def lower_bound(self): return self._lb @lower_bound.setter def lower_bound(self, value): if hasattr(self, '_ub') and value > self.upper_bound: raise ValueError('Lower bound cannot be greater than upper bound') self._lb = value @property def upper_bound(self): return self._ub @upper_bound.setter def upper_bound(self, value): if hasattr(self, '_lb') and value < self.lower_bound: raise ValueError('Lower bound cannot be greater than upper bound') self._ub = value ############################################################################### ## Public Methods ############################################################################### def getSample(self): '''Returns a random sample of the uncertain variable according to its distribution. *Example Usage* :: >>> u = UniformParameter() >>> u_sample = u.getSample() ''' ## _getSample is overwritten in child classes return self._getSample() ############################################################################### ## Private Methods ############################################################################### def _getSample(self): if self._max_pdf_val is None: self._max_pdf_val = self._getMaxPDFVal(self.evalPDF, self.lower_bound, self.upper_bound) while True: zscale = self._max_pdf_val*1.1 uval = (self.lower_bound + np.random.random()*(self.upper_bound-self.lower_bound)) zval = zscale*np.random.random() if zval < self.evalPDF(uval): return uval def _getMaxPDFVal(self, evalPDF, lower_bound, upper_bound): max_pdf_val = 0 for ui in np.linspace(lower_bound, upper_bound, 20): pdfi = evalPDF(ui) if pdfi > max_pdf_val: max_pdf_val = pdfi return max_pdf_val def _evalPDF(self, u): if u < self.lower_bound or u > self.upper_bound: return 0 else: return self.pdf(u)
lwcook/horsetail-matching
horsetailmatching/weightedsum.py
WeightedSum.evalMetric
python
def evalMetric(self, x, w1=None, w2=None): '''Evaluates the weighted sum metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param float w1: value to weight the mean by :param float w2: value to weight the std by :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float ''' if w1 is None: w1 = self.w1 if w2 is None: w2 = self.w2 if self.verbose: print('----------') print('At design: ' + str(x)) self._N_dv = len(_makeIter(x)) if self.verbose: print('Evaluating surrogate') if self.surrogate is None: def fqoi(u): return self.fqoi(x, u) def fgrad(u): return self.jac(x, u) jac = self.jac else: fqoi, fgrad, surr_jac = self._makeSurrogates(x) jac = surr_jac u_samples = self._getParameterSamples() if self.verbose: print('Evaluating quantity of interest at samples') q_samples, grad_samples = self._evalSamples(u_samples, fqoi, fgrad, jac) if self.verbose: print('Evaluating metric') return self._evalWeightedSumMetric(q_samples, grad_samples)
Evaluates the weighted sum metric at given values of the design variables. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :param float w1: value to weight the mean by :param float w2: value to weight the std by :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/weightedsum.py#L96-L143
[ "def _makeIter(x):\n try:\n iter(x)\n return [xi for xi in x]\n except:\n return [x]\n", "def _makeSurrogates(self, x):\n\n # Get quadrature points\n if self.surrogate_points is None:\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n mesh = np.meshgrid(*[np.linspace(-1, 1, 5) for n in np.arange(N_u)],\n copy=False)\n u_sparse = np.vstack([m.flatten() for m in mesh]).T\n else:\n u_sparse = self.surrogate_points\n\n N_sparse = u_sparse.shape[0]\n q_sparse = np.zeros(N_sparse)\n\n # Get surrogates in correct form\n if not self.jac:\n for iu, u in enumerate(u_sparse):\n q_sparse[iu] = self.fqoi(x, u)\n\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n\n def fqoi(u):\n return surr_qoi(u)\n fgrad = False\n surr_jac = False\n\n else:\n g_sparse = np.zeros([N_sparse, self._N_dv])\n for iu, u in enumerate(u_sparse):\n if isinstance(self.jac, bool) and self.jac:\n q_sparse[iu], g_sparse[iu, :] = self.fqoi(x, u)\n else:\n q_sparse[iu] = self.fqoi(x, u)\n g_sparse[iu, :] = self.jac(x, u)\n\n if not self.surrogate_jac:\n fpartial = [lambda u: 0 for _ in np.arange(self._N_dv)]\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n for k in np.arange(self._N_dv):\n fpartial[k] = self.surrogate(u_sparse, g_sparse[:, k])\n def surr_grad(u):\n return [f(u) for f in fpartial]\n else:\n if isinstance(self.surrogate_jac, bool) and self.surrogate_jac:\n surr_qoi, surr_grad = self.surrogate(\n u_sparse, q_sparse, g_sparse)\n else:\n surr_qoi = self.surrogate(u_sparse, q_sparse)\n surr_grad = self.surrogate_jac(u_sparse, g_sparse)\n\n def fqoi(u):\n return(surr_qoi(u))\n def fgrad(u):\n return(surr_grad(u))\n surr_jac = fgrad\n\n return fqoi, fgrad, surr_jac\n", " def _getParameterSamples(self):\n\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n\n get_new = True\n if self.reuse_samples and self.u_samples is not None:\n if self.u_samples.shape != (self.samples_int, self.samples_prob, N_u):\n if self.verbose:\n print('''Stored samples do not match current dimensions,\n getting new samples''')\n else:\n get_new = False\n\n if get_new:\n if self.verbose:\n print('Getting uncertain parameter samples')\n\n N_u = len(self.prob_uncertainties) + len(self.int_uncertainties)\n N_prob = len(self.prob_uncertainties)\n N_int = len(self.int_uncertainties)\n# u_samples = np.zeros([self.samples_int, self.samples_prob, N_u])\n\n u_samples_prob = np.zeros([self.samples_int, self.samples_prob,\n len(self.prob_uncertainties)])\n u_samples_int = np.zeros([self.samples_int, self.samples_prob,\n len(self.int_uncertainties)])\n\n u_ints = np.zeros([self.samples_int, len(self.int_uncertainties)])\n for kk, uk in enumerate(self.int_uncertainties):\n if callable(uk):\n samps = np.array(uk()).flatten()\n if len(samps) != self.samples_prob:\n raise Exception('Number of samples returned not equal ' +\n 'to specified number of samples: please set number of ' +\n 'samples with samples_prob attribute')\n else:\n u_ints[:, kk] = samps\n elif isinstance(uk, (tuple, list)): ## See if given as tuple/list of bounds\n lb, ub = uk[0], uk[1]\n u_ints[:, kk] = np.random.uniform(lb, ub, size=self.samples_int)\n u_ints[0, kk] = lb\n u_ints[-1, kk] = ub\n elif hasattr(uk, 'getSample'):\n for ii in np.arange(self.samples_int):\n u_ints[ii, kk] = uk.getSample()\n else:\n raise TypeError('Unsupported interval uncertainty type')\n\n u_samples_int = np.tile(u_ints[:, np.newaxis], (1, self.samples_prob, 1))\n\n u_probs = np.zeros([self.samples_prob, len(self.prob_uncertainties)])\n for kk, uk in enumerate(self.prob_uncertainties):\n if callable(uk):\n samps = np.array(uk()).flatten()\n if len(samps) != self.samples_prob:\n raise Exception('Number of samples returned not equal ' +\n 'to specified number of samples: please set number of ' +\n 'samples with samples_prob attribute')\n else:\n u_probs[:, kk] = samps\n elif hasattr(uk, 'getSample'):\n for jj in np.arange(self.samples_prob):\n u_probs[jj, kk] = uk.getSample()\n else:\n raise TypeError('Unsupported probabilistic uncertainty type')\n\n u_samples_prob = np.tile(u_probs[np.newaxis, :], (self.samples_int, 1, 1))\n\n u_samples = np.concatenate((u_samples_int, u_samples_prob), axis=2)\n\n self.u_samples = u_samples\n return u_samples\n else:\n if self.verbose:\n print('Re-using stored samples')\n return self.u_samples\n", "def _evalSamples(self, u_samples, fqoi, fgrad, jac):\n\n # Array of shape (M_int, M_prob)\n grad_samples = None\n q_samples = np.zeros([self.samples_int, self.samples_prob])\n if not jac:\n for ii in np.arange(q_samples.shape[0]):\n for jj in np.arange(q_samples.shape[1]):\n q_samples[ii, jj] = fqoi(u_samples[ii, jj])\n else:\n grad_samples = np.zeros([self.samples_int, self.samples_prob,\n self._N_dv])\n for ii in np.arange(q_samples.shape[0]):\n for jj in np.arange(q_samples.shape[1]):\n if isinstance(jac, bool) and jac:\n (q, grad) = fqoi(u_samples[ii, jj])\n q_samples[ii, jj] = float(q)\n grad_samples[ii, jj, :] = [_ for _ in grad]\n else:\n q_samples[ii, jj] = fqoi(u_samples[ii, jj])\n grad_samples[ii, jj, :] = fgrad(u_samples[ii, jj])\n\n self.grad_samples = grad_samples\n\n self.q_samples = q_samples\n\n return q_samples, grad_samples\n", "def _evalWeightedSumMetric(self, q_samples, grad_samples=None):\n\n fjs = np.array(q_samples).flatten()\n M = self.samples_prob\n\n mean = (1./M)*np.sum(fjs)\n var = (1./M)*np.sum([(fj - mean)**2 for fj in fjs])\n\n ws = self.w1*mean + self.w2*np.sqrt(var)\n\n if grad_samples is None:\n return ws\n else:\n ndv = grad_samples.shape[2]\n gradjs = grad_samples[0, :, :]\n\n gradient = np.zeros(ndv)\n for kdv in range(ndv):\n\n meang, varg = 0., 0.\n for j, fj in enumerate(fjs):\n meang += (1./M)*float(gradjs[j, kdv])\n varg += (1./M)*2*(fj - mean)*float(gradjs[j, kdv])\n\n gradient[kdv] = meang + 0.5*(var**-0.5)*varg\n\n return ws, gradient\n" ]
class WeightedSum(HorsetailMatching): '''Class for using weighted sum of moments within an optimization. The code is written such that all arguments that can be used at the initialization of a WeightedSum object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Is a list of UncertainParameter objects, or a list of functions that return samples of the each uncertainty. :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. ''' def __init__(self, fqoi, prob_uncertainties, jac=False, samples_prob=1000, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False, w1=1, w2=1): self.fqoi = fqoi self.prob_uncertainties = prob_uncertainties self.int_uncertainties = [] self.jac = jac self.samples_prob = samples_prob self.samples_int = 1 self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose self.w1 = w1 self.w2 = w2 ############################################################################## ## Public Methods ############################################################################## ############################################################################## ## Private methods ## ############################################################################## def _evalWeightedSumMetric(self, q_samples, grad_samples=None): fjs = np.array(q_samples).flatten() M = self.samples_prob mean = (1./M)*np.sum(fjs) var = (1./M)*np.sum([(fj - mean)**2 for fj in fjs]) ws = self.w1*mean + self.w2*np.sqrt(var) if grad_samples is None: return ws else: ndv = grad_samples.shape[2] gradjs = grad_samples[0, :, :] gradient = np.zeros(ndv) for kdv in range(ndv): meang, varg = 0., 0. for j, fj in enumerate(fjs): meang += (1./M)*float(gradjs[j, kdv]) varg += (1./M)*2*(fj - mean)*float(gradjs[j, kdv]) gradient[kdv] = meang + 0.5*(var**-0.5)*varg return ws, gradient def getHorsetail(self): return ([0], [0]), ([0], [0]), [([0], [0])]
lwcook/horsetail-matching
horsetailmatching/surrogates.py
eval_poly
python
def eval_poly(uvec, nvec, Jvec): '''Evaluate multi-dimensional polynomials through tensor multiplication. :param list uvec: vector value of the uncertain parameters at which to evaluate the polynomial :param list nvec: order in each dimension at which to evaluate the polynomial :param list Jvec: Jacobi matrix of each dimension's 1D polynomial :return: poly_value - value of the polynomial evaluated at uvec :rtype: float ''' us = _makeIter(uvec) ns = _makeIter(nvec) Js = _makeIter(Jvec) return np.prod([_eval_poly_1D(u, n, J) for u, n, J in zip(us, ns, Js)])
Evaluate multi-dimensional polynomials through tensor multiplication. :param list uvec: vector value of the uncertain parameters at which to evaluate the polynomial :param list nvec: order in each dimension at which to evaluate the polynomial :param list Jvec: Jacobi matrix of each dimension's 1D polynomial :return: poly_value - value of the polynomial evaluated at uvec :rtype: float
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/surrogates.py#L186-L204
[ "def _makeIter(x):\n try:\n iter(x)\n return [xi for xi in x]\n except:\n return [x]\n" ]
import numpy as np import math import pdb class PolySurrogate(object): '''Class for creating surrogate models using non-intrusive polynomial chaos. :param int dimensions: number of dimensions of the polynomial expansion :param int order: order of the polynomial expansion [default 3] :param str/list poly_type: string of the type of polynomials to use in the expansion, or list of strings where each entry in the list is the type of polynomial to use in the corresponding dimension. Supported polynomial types are legendre and gaussian. [default legendre] *Example Declaration*:: >>> thePC = PolySurrogate(dimensions=3) >>> thePC = PolySurrogate(dimensions=3, order=3) >>> thePC = PolySurrogate(dimensions=3, order=3, poly_type='legendre') ''' def __init__(self, dimensions, order=3, poly_type='legendre'): self.dims = dimensions self.P = int(order) + 1 if isinstance(poly_type, basestring): self.poly_types = [poly_type for _ in np.arange(self.dims)] else: self.poly_types = _makeIter(poly_type) self.J_list = [_define_poly_J(p, self.P) for p in self.poly_types] imesh = np.meshgrid(*[np.arange(self.P) for d in np.arange(self.dims)]) self.index_polys = np.vstack([m.flatten() for m in imesh]).T self.N_poly = len(self.index_polys) self.coeffs = np.zeros([self.P for __ in np.arange(self.dims)]) def surrogate(self, u_sparse, q_sparse): '''Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q) ''' self.train(q_sparse) def model(u): return self.predict(u) return model def predict(self, u): '''Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1]) ''' y, ysub = 0, np.zeros(self.N_poly) for ip in range(self.N_poly): inds = tuple(self.index_polys[ip]) ysub[ip] = self.coeffs[inds]*eval_poly(u, inds, self.J_list) y += ysub[ip] self.response_components = ysub return y def train(self, ftrain): '''Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1]) ''' self.coeffs = 0*self.coeffs upoints, wpoints = self.getQuadraturePointsAndWeights() try: fpoints = [ftrain(u) for u in upoints] except TypeError: fpoints = ftrain for ipoly in np.arange(self.N_poly): inds = tuple(self.index_polys[ipoly]) coeff = 0.0 for (u, q, w) in zip(upoints, fpoints, wpoints): coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w) self.coeffs[inds] = coeff return None def getQuadraturePointsAndWeights(self): '''Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray) ''' qw_list, qp_list = [], [] for ii in np.arange(len(self.J_list)): d, Q = np.linalg.eig(self.J_list[ii]) qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d) qw = (Q[0, qpi]**2).reshape([d.size, 1]) qw_list.append(qw) qp_list.append(qp) umesh = np.meshgrid(*qp_list) upoints = np.vstack([m.flatten() for m in umesh]).T wmesh = np.meshgrid(*qw_list) wpoints = np.vstack([m.flatten() for m in wmesh]).T return upoints, wpoints def getQuadraturePoints(self): '''Gets the quadrature points at which the output values must be found in order to train the polynomial expansion using gaussian quadrature. :return: upoints - a np.ndarray of size (num_polynomials, num_dimensions) :rtype: np.ndarray ''' upoints, _ = self.getQuadraturePointsAndWeights() return upoints ## -------------------------------------------------------------------------- ## Private funtions for polynomials ## -------------------------------------------------------------------------- def _eval_poly_1D(s, k, Jmat): if k == -1: return 0.0 elif k == 0: return 1.0 else: ki = k-1 beta_k = float(Jmat[ki+1, ki]) alpha_km1 = float(Jmat[ki, ki]) if k == 1: beta_km1 = 0. else: beta_km1 = float(Jmat[ki, ki-1]) return (1.0/float(beta_k))*( (s - alpha_km1)*_eval_poly_1D(s, k-1, Jmat) - beta_km1*_eval_poly_1D(s, k-2, Jmat)) def _define_poly_J(typestr, order, a=1, b=1): n = order # Define ab, the matrix of alpha and beta values # These are recurrence coefficients if typestr == 'legendre' or typestr == 'uniform': l, r = -1, 1 o = l + (r-l)/2.0 ab = np.zeros([n, 2],float) if n > 0: ab[0, 0], ab[0, 1] = o,1 for k in np.arange(2, n+1, 1): ik, ab[ik, 0] = k-1, o if k == 2: numer = float(((r-l)**2)*(k-1)*(k-1)*(k-1)) denom = float(((2*(k-1))**2)*(2*(k-1)+1)) else: numer = float(((r-l)**2)*(k-1)*(k-1)*(k-1)*(k-1)) denom = float(((2*(k-1))**2)*(2*(k-1)+1)*(2*(k-1)-1)) ab[ik, 1] = numer / denom elif typestr == 'hermite' or typestr == 'gaussian': mu = 0 mu0 = math.gamma(mu+0.5) if n==1: ab = np.array([[0, mu0]]) else: ab = np.zeros([n, 2]) nvechalf = np.array(range(1, n))*0.5 nvechalf[0::2] += mu ab[0, 1], ab[1::, 1] = mu0, nvechalf # Define J, the jacobi matrix from recurrence coefficients in ab J = np.zeros([n, n], float) if n == 1: J = np.array([[ab[0, 0]]]) else: J[0, 0] = ab[0, 0] J[0, 1] = math.sqrt(ab[1, 1]) for i in np.arange(2, n, 1): ii = i-1 J[ii, ii] = ab[ii,0] J[ii, ii-1] = math.sqrt(ab[ii, 1]) J[ii, ii+1] = math.sqrt(ab[ii+1, 1]) J[n-1, n-1] = ab[n-1, 0] J[n-1, n-2] = math.sqrt(ab[n-1, 1]) return J def _makeIter(x): try: iter(x) return [xi for xi in x] except: return [x]
lwcook/horsetail-matching
horsetailmatching/surrogates.py
PolySurrogate.surrogate
python
def surrogate(self, u_sparse, q_sparse): '''Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q) ''' self.train(q_sparse) def model(u): return self.predict(u) return model
Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q)
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/surrogates.py#L43-L70
[ "def train(self, ftrain):\n '''Trains the polynomial expansion.\n\n :param numpy.ndarray/function ftrain: output values corresponding to the\n quadrature points given by the getQuadraturePoints method to\n which the expansion should be trained. Or a function that should be evaluated\n at the quadrature points to give these output values.\n\n *Sample Usage*::\n\n >>> thePC = PolySurrogate(dimensions=2)\n >>> thePC.train(myFunc)\n >>> predicted_q = thePC.predict([0, 1])\n\n >>> thePC = PolySurrogate(dimensions=2)\n >>> U = thePC.getQuadraturePoints()\n >>> Q = [myFunc(u) for u in U]\n >>> thePC.train(Q)\n >>> predicted_q = thePC.predict([0, 1])\n\n '''\n self.coeffs = 0*self.coeffs\n\n upoints, wpoints = self.getQuadraturePointsAndWeights()\n\n try:\n fpoints = [ftrain(u) for u in upoints]\n except TypeError:\n fpoints = ftrain\n\n for ipoly in np.arange(self.N_poly):\n\n inds = tuple(self.index_polys[ipoly])\n coeff = 0.0\n for (u, q, w) in zip(upoints, fpoints, wpoints):\n coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w)\n\n self.coeffs[inds] = coeff\n return None\n" ]
class PolySurrogate(object): '''Class for creating surrogate models using non-intrusive polynomial chaos. :param int dimensions: number of dimensions of the polynomial expansion :param int order: order of the polynomial expansion [default 3] :param str/list poly_type: string of the type of polynomials to use in the expansion, or list of strings where each entry in the list is the type of polynomial to use in the corresponding dimension. Supported polynomial types are legendre and gaussian. [default legendre] *Example Declaration*:: >>> thePC = PolySurrogate(dimensions=3) >>> thePC = PolySurrogate(dimensions=3, order=3) >>> thePC = PolySurrogate(dimensions=3, order=3, poly_type='legendre') ''' def __init__(self, dimensions, order=3, poly_type='legendre'): self.dims = dimensions self.P = int(order) + 1 if isinstance(poly_type, basestring): self.poly_types = [poly_type for _ in np.arange(self.dims)] else: self.poly_types = _makeIter(poly_type) self.J_list = [_define_poly_J(p, self.P) for p in self.poly_types] imesh = np.meshgrid(*[np.arange(self.P) for d in np.arange(self.dims)]) self.index_polys = np.vstack([m.flatten() for m in imesh]).T self.N_poly = len(self.index_polys) self.coeffs = np.zeros([self.P for __ in np.arange(self.dims)]) def predict(self, u): '''Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1]) ''' y, ysub = 0, np.zeros(self.N_poly) for ip in range(self.N_poly): inds = tuple(self.index_polys[ip]) ysub[ip] = self.coeffs[inds]*eval_poly(u, inds, self.J_list) y += ysub[ip] self.response_components = ysub return y def train(self, ftrain): '''Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1]) ''' self.coeffs = 0*self.coeffs upoints, wpoints = self.getQuadraturePointsAndWeights() try: fpoints = [ftrain(u) for u in upoints] except TypeError: fpoints = ftrain for ipoly in np.arange(self.N_poly): inds = tuple(self.index_polys[ipoly]) coeff = 0.0 for (u, q, w) in zip(upoints, fpoints, wpoints): coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w) self.coeffs[inds] = coeff return None def getQuadraturePointsAndWeights(self): '''Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray) ''' qw_list, qp_list = [], [] for ii in np.arange(len(self.J_list)): d, Q = np.linalg.eig(self.J_list[ii]) qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d) qw = (Q[0, qpi]**2).reshape([d.size, 1]) qw_list.append(qw) qp_list.append(qp) umesh = np.meshgrid(*qp_list) upoints = np.vstack([m.flatten() for m in umesh]).T wmesh = np.meshgrid(*qw_list) wpoints = np.vstack([m.flatten() for m in wmesh]).T return upoints, wpoints def getQuadraturePoints(self): '''Gets the quadrature points at which the output values must be found in order to train the polynomial expansion using gaussian quadrature. :return: upoints - a np.ndarray of size (num_polynomials, num_dimensions) :rtype: np.ndarray ''' upoints, _ = self.getQuadraturePointsAndWeights() return upoints
lwcook/horsetail-matching
horsetailmatching/surrogates.py
PolySurrogate.predict
python
def predict(self, u): '''Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1]) ''' y, ysub = 0, np.zeros(self.N_poly) for ip in range(self.N_poly): inds = tuple(self.index_polys[ip]) ysub[ip] = self.coeffs[inds]*eval_poly(u, inds, self.J_list) y += ysub[ip] self.response_components = ysub return y
Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1])
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/surrogates.py#L72-L98
[ "def eval_poly(uvec, nvec, Jvec):\n '''Evaluate multi-dimensional polynomials through tensor multiplication.\n\n :param list uvec: vector value of the uncertain parameters at which to evaluate the\n polynomial\n\n :param list nvec: order in each dimension at which to evaluate the polynomial\n\n :param list Jvec: Jacobi matrix of each dimension's 1D polynomial\n\n :return: poly_value - value of the polynomial evaluated at uvec\n\n :rtype: float\n\n '''\n us = _makeIter(uvec)\n ns = _makeIter(nvec)\n Js = _makeIter(Jvec)\n return np.prod([_eval_poly_1D(u, n, J) for u, n, J in zip(us, ns, Js)])\n" ]
class PolySurrogate(object): '''Class for creating surrogate models using non-intrusive polynomial chaos. :param int dimensions: number of dimensions of the polynomial expansion :param int order: order of the polynomial expansion [default 3] :param str/list poly_type: string of the type of polynomials to use in the expansion, or list of strings where each entry in the list is the type of polynomial to use in the corresponding dimension. Supported polynomial types are legendre and gaussian. [default legendre] *Example Declaration*:: >>> thePC = PolySurrogate(dimensions=3) >>> thePC = PolySurrogate(dimensions=3, order=3) >>> thePC = PolySurrogate(dimensions=3, order=3, poly_type='legendre') ''' def __init__(self, dimensions, order=3, poly_type='legendre'): self.dims = dimensions self.P = int(order) + 1 if isinstance(poly_type, basestring): self.poly_types = [poly_type for _ in np.arange(self.dims)] else: self.poly_types = _makeIter(poly_type) self.J_list = [_define_poly_J(p, self.P) for p in self.poly_types] imesh = np.meshgrid(*[np.arange(self.P) for d in np.arange(self.dims)]) self.index_polys = np.vstack([m.flatten() for m in imesh]).T self.N_poly = len(self.index_polys) self.coeffs = np.zeros([self.P for __ in np.arange(self.dims)]) def surrogate(self, u_sparse, q_sparse): '''Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q) ''' self.train(q_sparse) def model(u): return self.predict(u) return model def train(self, ftrain): '''Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1]) ''' self.coeffs = 0*self.coeffs upoints, wpoints = self.getQuadraturePointsAndWeights() try: fpoints = [ftrain(u) for u in upoints] except TypeError: fpoints = ftrain for ipoly in np.arange(self.N_poly): inds = tuple(self.index_polys[ipoly]) coeff = 0.0 for (u, q, w) in zip(upoints, fpoints, wpoints): coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w) self.coeffs[inds] = coeff return None def getQuadraturePointsAndWeights(self): '''Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray) ''' qw_list, qp_list = [], [] for ii in np.arange(len(self.J_list)): d, Q = np.linalg.eig(self.J_list[ii]) qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d) qw = (Q[0, qpi]**2).reshape([d.size, 1]) qw_list.append(qw) qp_list.append(qp) umesh = np.meshgrid(*qp_list) upoints = np.vstack([m.flatten() for m in umesh]).T wmesh = np.meshgrid(*qw_list) wpoints = np.vstack([m.flatten() for m in wmesh]).T return upoints, wpoints def getQuadraturePoints(self): '''Gets the quadrature points at which the output values must be found in order to train the polynomial expansion using gaussian quadrature. :return: upoints - a np.ndarray of size (num_polynomials, num_dimensions) :rtype: np.ndarray ''' upoints, _ = self.getQuadraturePointsAndWeights() return upoints
lwcook/horsetail-matching
horsetailmatching/surrogates.py
PolySurrogate.train
python
def train(self, ftrain): '''Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1]) ''' self.coeffs = 0*self.coeffs upoints, wpoints = self.getQuadraturePointsAndWeights() try: fpoints = [ftrain(u) for u in upoints] except TypeError: fpoints = ftrain for ipoly in np.arange(self.N_poly): inds = tuple(self.index_polys[ipoly]) coeff = 0.0 for (u, q, w) in zip(upoints, fpoints, wpoints): coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w) self.coeffs[inds] = coeff return None
Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1])
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/surrogates.py#L100-L138
[ "def eval_poly(uvec, nvec, Jvec):\n '''Evaluate multi-dimensional polynomials through tensor multiplication.\n\n :param list uvec: vector value of the uncertain parameters at which to evaluate the\n polynomial\n\n :param list nvec: order in each dimension at which to evaluate the polynomial\n\n :param list Jvec: Jacobi matrix of each dimension's 1D polynomial\n\n :return: poly_value - value of the polynomial evaluated at uvec\n\n :rtype: float\n\n '''\n us = _makeIter(uvec)\n ns = _makeIter(nvec)\n Js = _makeIter(Jvec)\n return np.prod([_eval_poly_1D(u, n, J) for u, n, J in zip(us, ns, Js)])\n", "def getQuadraturePointsAndWeights(self):\n '''Gets the quadrature points and weights for gaussian quadrature\n integration of inner products from the definition of the polynomials in\n each dimension.\n\n\n :return: (u_points, w_points) - np.ndarray of shape\n (num_polynomials, num_dimensions) and a np.ndarray of size\n (num_polynomials)\n\n :rtype: (np.ndarray, np.ndarray)\n '''\n\n qw_list, qp_list = [], []\n for ii in np.arange(len(self.J_list)):\n\n d, Q = np.linalg.eig(self.J_list[ii])\n qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d)\n qw = (Q[0, qpi]**2).reshape([d.size, 1])\n\n qw_list.append(qw)\n qp_list.append(qp)\n\n umesh = np.meshgrid(*qp_list)\n upoints = np.vstack([m.flatten() for m in umesh]).T\n\n wmesh = np.meshgrid(*qw_list)\n wpoints = np.vstack([m.flatten() for m in wmesh]).T\n\n return upoints, wpoints\n" ]
class PolySurrogate(object): '''Class for creating surrogate models using non-intrusive polynomial chaos. :param int dimensions: number of dimensions of the polynomial expansion :param int order: order of the polynomial expansion [default 3] :param str/list poly_type: string of the type of polynomials to use in the expansion, or list of strings where each entry in the list is the type of polynomial to use in the corresponding dimension. Supported polynomial types are legendre and gaussian. [default legendre] *Example Declaration*:: >>> thePC = PolySurrogate(dimensions=3) >>> thePC = PolySurrogate(dimensions=3, order=3) >>> thePC = PolySurrogate(dimensions=3, order=3, poly_type='legendre') ''' def __init__(self, dimensions, order=3, poly_type='legendre'): self.dims = dimensions self.P = int(order) + 1 if isinstance(poly_type, basestring): self.poly_types = [poly_type for _ in np.arange(self.dims)] else: self.poly_types = _makeIter(poly_type) self.J_list = [_define_poly_J(p, self.P) for p in self.poly_types] imesh = np.meshgrid(*[np.arange(self.P) for d in np.arange(self.dims)]) self.index_polys = np.vstack([m.flatten() for m in imesh]).T self.N_poly = len(self.index_polys) self.coeffs = np.zeros([self.P for __ in np.arange(self.dims)]) def surrogate(self, u_sparse, q_sparse): '''Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q) ''' self.train(q_sparse) def model(u): return self.predict(u) return model def predict(self, u): '''Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1]) ''' y, ysub = 0, np.zeros(self.N_poly) for ip in range(self.N_poly): inds = tuple(self.index_polys[ip]) ysub[ip] = self.coeffs[inds]*eval_poly(u, inds, self.J_list) y += ysub[ip] self.response_components = ysub return y def getQuadraturePointsAndWeights(self): '''Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray) ''' qw_list, qp_list = [], [] for ii in np.arange(len(self.J_list)): d, Q = np.linalg.eig(self.J_list[ii]) qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d) qw = (Q[0, qpi]**2).reshape([d.size, 1]) qw_list.append(qw) qp_list.append(qp) umesh = np.meshgrid(*qp_list) upoints = np.vstack([m.flatten() for m in umesh]).T wmesh = np.meshgrid(*qw_list) wpoints = np.vstack([m.flatten() for m in wmesh]).T return upoints, wpoints def getQuadraturePoints(self): '''Gets the quadrature points at which the output values must be found in order to train the polynomial expansion using gaussian quadrature. :return: upoints - a np.ndarray of size (num_polynomials, num_dimensions) :rtype: np.ndarray ''' upoints, _ = self.getQuadraturePointsAndWeights() return upoints
lwcook/horsetail-matching
horsetailmatching/surrogates.py
PolySurrogate.getQuadraturePointsAndWeights
python
def getQuadraturePointsAndWeights(self): '''Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray) ''' qw_list, qp_list = [], [] for ii in np.arange(len(self.J_list)): d, Q = np.linalg.eig(self.J_list[ii]) qp, qpi = d[np.argsort(d)].reshape([d.size, 1]), np.argsort(d) qw = (Q[0, qpi]**2).reshape([d.size, 1]) qw_list.append(qw) qp_list.append(qp) umesh = np.meshgrid(*qp_list) upoints = np.vstack([m.flatten() for m in umesh]).T wmesh = np.meshgrid(*qw_list) wpoints = np.vstack([m.flatten() for m in wmesh]).T return upoints, wpoints
Gets the quadrature points and weights for gaussian quadrature integration of inner products from the definition of the polynomials in each dimension. :return: (u_points, w_points) - np.ndarray of shape (num_polynomials, num_dimensions) and a np.ndarray of size (num_polynomials) :rtype: (np.ndarray, np.ndarray)
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/surrogates.py#L140-L169
null
class PolySurrogate(object): '''Class for creating surrogate models using non-intrusive polynomial chaos. :param int dimensions: number of dimensions of the polynomial expansion :param int order: order of the polynomial expansion [default 3] :param str/list poly_type: string of the type of polynomials to use in the expansion, or list of strings where each entry in the list is the type of polynomial to use in the corresponding dimension. Supported polynomial types are legendre and gaussian. [default legendre] *Example Declaration*:: >>> thePC = PolySurrogate(dimensions=3) >>> thePC = PolySurrogate(dimensions=3, order=3) >>> thePC = PolySurrogate(dimensions=3, order=3, poly_type='legendre') ''' def __init__(self, dimensions, order=3, poly_type='legendre'): self.dims = dimensions self.P = int(order) + 1 if isinstance(poly_type, basestring): self.poly_types = [poly_type for _ in np.arange(self.dims)] else: self.poly_types = _makeIter(poly_type) self.J_list = [_define_poly_J(p, self.P) for p in self.poly_types] imesh = np.meshgrid(*[np.arange(self.P) for d in np.arange(self.dims)]) self.index_polys = np.vstack([m.flatten() for m in imesh]).T self.N_poly = len(self.index_polys) self.coeffs = np.zeros([self.P for __ in np.arange(self.dims)]) def surrogate(self, u_sparse, q_sparse): '''Combines the train and predict methods to create a surrogate model function fitted to the input/output combinations given in u_sparse and q_sparse. :param numpy.ndarray u_sparse: input values at which the output values are obtained. Must be the same as the qaudrature points defined by the getQuadraturePoints method. :param numpy.ndarray q_sparse: output values corresponding to the input values given in u_sparse to which the surrogate is fitted :return: surrogate model fitted to u_sparse and q_sparse :rtype: function *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> surrogateFunc = thePC.surrogate(U, Q) ''' self.train(q_sparse) def model(u): return self.predict(u) return model def predict(self, u): '''Predicts the output value at u from the fitted polynomial expansion. Therefore the method train() must be called first. :param numpy.ndarray u: input value at which to predict the output. :return: q_approx - the predicted value of the output at u :rtype: float *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(U, Q) >>> thePC.predict([0, 1]) ''' y, ysub = 0, np.zeros(self.N_poly) for ip in range(self.N_poly): inds = tuple(self.index_polys[ip]) ysub[ip] = self.coeffs[inds]*eval_poly(u, inds, self.J_list) y += ysub[ip] self.response_components = ysub return y def train(self, ftrain): '''Trains the polynomial expansion. :param numpy.ndarray/function ftrain: output values corresponding to the quadrature points given by the getQuadraturePoints method to which the expansion should be trained. Or a function that should be evaluated at the quadrature points to give these output values. *Sample Usage*:: >>> thePC = PolySurrogate(dimensions=2) >>> thePC.train(myFunc) >>> predicted_q = thePC.predict([0, 1]) >>> thePC = PolySurrogate(dimensions=2) >>> U = thePC.getQuadraturePoints() >>> Q = [myFunc(u) for u in U] >>> thePC.train(Q) >>> predicted_q = thePC.predict([0, 1]) ''' self.coeffs = 0*self.coeffs upoints, wpoints = self.getQuadraturePointsAndWeights() try: fpoints = [ftrain(u) for u in upoints] except TypeError: fpoints = ftrain for ipoly in np.arange(self.N_poly): inds = tuple(self.index_polys[ipoly]) coeff = 0.0 for (u, q, w) in zip(upoints, fpoints, wpoints): coeff += eval_poly(u, inds, self.J_list)*q*np.prod(w) self.coeffs[inds] = coeff return None def getQuadraturePoints(self): '''Gets the quadrature points at which the output values must be found in order to train the polynomial expansion using gaussian quadrature. :return: upoints - a np.ndarray of size (num_polynomials, num_dimensions) :rtype: np.ndarray ''' upoints, _ = self.getQuadraturePointsAndWeights() return upoints
lwcook/horsetail-matching
horsetailmatching/demoproblems.py
TP0
python
def TP0(dv, u): '''Demo problem 0 for horsetail matching, takes two input vectors of any size and returns a single output''' return np.linalg.norm(np.array(dv)) + np.linalg.norm(np.array(u))
Demo problem 0 for horsetail matching, takes two input vectors of any size and returns a single output
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/demoproblems.py#L3-L6
null
import numpy as np def TP1(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' factor = 0.1*(u[0]**2 + 2*u[0]*u[1] + u[1]**2) q = 0 + factor*(x[0]**2 + 2*x[1]*x[0] + x[1]**2) if not jac: return q else: grad = [factor*(2*x[0] + 2*x[1]), factor*(2*x[0] + 2*x[1])] return q, grad def TP2(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) + 0.2*z*u[1]**3 + 7 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10. ) dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 return q, [dqdx1, dqdx2] def TP2b(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) +\ 0.2*z*u[1]**3 + 7 + u[0]*(y + z)*0.02 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10.) + 0.01*u[0] dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 + 0.01*u[0] return q, [dqdx1, dqdx2] def TP3(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input values of size 1''' q = 2 + 0.5*x + 1.5*(1-x)*u if not jac: return q else: grad = 0.5 -1.5*u return q, grad
lwcook/horsetail-matching
horsetailmatching/demoproblems.py
TP1
python
def TP1(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' factor = 0.1*(u[0]**2 + 2*u[0]*u[1] + u[1]**2) q = 0 + factor*(x[0]**2 + 2*x[1]*x[0] + x[1]**2) if not jac: return q else: grad = [factor*(2*x[0] + 2*x[1]), factor*(2*x[0] + 2*x[1])] return q, grad
Demo problem 1 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/demoproblems.py#L8-L18
null
import numpy as np def TP0(dv, u): '''Demo problem 0 for horsetail matching, takes two input vectors of any size and returns a single output''' return np.linalg.norm(np.array(dv)) + np.linalg.norm(np.array(u)) def TP2(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) + 0.2*z*u[1]**3 + 7 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10. ) dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 return q, [dqdx1, dqdx2] def TP2b(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) +\ 0.2*z*u[1]**3 + 7 + u[0]*(y + z)*0.02 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10.) + 0.01*u[0] dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 + 0.01*u[0] return q, [dqdx1, dqdx2] def TP3(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input values of size 1''' q = 2 + 0.5*x + 1.5*(1-x)*u if not jac: return q else: grad = 0.5 -1.5*u return q, grad
lwcook/horsetail-matching
horsetailmatching/demoproblems.py
TP2
python
def TP2(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) + 0.2*z*u[1]**3 + 7 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10. ) dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 return q, [dqdx1, dqdx2]
Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/demoproblems.py#L20-L34
null
import numpy as np def TP0(dv, u): '''Demo problem 0 for horsetail matching, takes two input vectors of any size and returns a single output''' return np.linalg.norm(np.array(dv)) + np.linalg.norm(np.array(u)) def TP1(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' factor = 0.1*(u[0]**2 + 2*u[0]*u[1] + u[1]**2) q = 0 + factor*(x[0]**2 + 2*x[1]*x[0] + x[1]**2) if not jac: return q else: grad = [factor*(2*x[0] + 2*x[1]), factor*(2*x[0] + 2*x[1])] return q, grad def TP2b(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) +\ 0.2*z*u[1]**3 + 7 + u[0]*(y + z)*0.02 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10.) + 0.01*u[0] dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 + 0.01*u[0] return q, [dqdx1, dqdx2] def TP3(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input values of size 1''' q = 2 + 0.5*x + 1.5*(1-x)*u if not jac: return q else: grad = 0.5 -1.5*u return q, grad
lwcook/horsetail-matching
horsetailmatching/demoproblems.py
TP3
python
def TP3(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input values of size 1''' q = 2 + 0.5*x + 1.5*(1-x)*u if not jac: return q else: grad = 0.5 -1.5*u return q, grad
Demo problem 1 for horsetail matching, takes two input values of size 1
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/demoproblems.py#L53-L62
null
import numpy as np def TP0(dv, u): '''Demo problem 0 for horsetail matching, takes two input vectors of any size and returns a single output''' return np.linalg.norm(np.array(dv)) + np.linalg.norm(np.array(u)) def TP1(x, u, jac=False): '''Demo problem 1 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' factor = 0.1*(u[0]**2 + 2*u[0]*u[1] + u[1]**2) q = 0 + factor*(x[0]**2 + 2*x[1]*x[0] + x[1]**2) if not jac: return q else: grad = [factor*(2*x[0] + 2*x[1]), factor*(2*x[0] + 2*x[1])] return q, grad def TP2(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) + 0.2*z*u[1]**3 + 7 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10. ) dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 return q, [dqdx1, dqdx2] def TP2b(dv, u, jac=False): '''Demo problem 2 for horsetail matching, takes two input vectors of size 2 and returns just the qoi if jac is False or the qoi and its gradient if jac is True''' y = dv[0]/2. z = dv[1]/2. + 12 q = 0.25*((y**2 + z**2)/10 + 5*u[0]*u[1] - z*u[1]**2) +\ 0.2*z*u[1]**3 + 7 + u[0]*(y + z)*0.02 if not jac: return q else: dqdx1 = (1./8.)*( 2*y/10.) + 0.01*u[0] dqdx2 = (1./8.)*( 2*z/10. - u[1]**2) + 0.1*u[1]**3 + 0.01*u[0] return q, [dqdx1, dqdx2]
lwcook/horsetail-matching
horsetailmatching/densitymatching.py
DensityMatching.evalMetric
python
def evalMetric(self, x, method=None): '''Evaluates the density matching metric at a given design point. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theDM = DensityMatching(myFunc, u) >>> x0 = [1, 2] >>> theDM.evalMetric(x0) ''' return super(DensityMatching, self).evalMetric(x, method)
Evaluates the density matching metric at a given design point. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theDM = DensityMatching(myFunc, u) >>> x0 = [1, 2] >>> theDM.evalMetric(x0)
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/densitymatching.py#L114-L134
[ " def evalMetric(self, x, method=None):\n '''Evaluates the horsetail matching metric at given values of the\n design variables.\n\n :param iterable x: values of the design variables, this is passed as\n the first argument to the function fqoi\n :param str method: method to use to evaluate the metric ('empirical' or\n 'kernel')\n\n :return: metric_value - value of the metric evaluated at the design\n point given by x\n\n :rtype: float\n\n *Example Usage*::\n\n >>> def myFunc(x, u): return x[0]*x[1] + u\n >>> u1 = UniformParameter()\n >>> theHM = HorsetailMatching(myFunc, u)\n >>> x0 = [1, 2]\n >>> theHM.evalMetric(x0)\n\n '''\n # Make sure dimensions are correct\n# u_sample_dimensions = self._processDimensions()\n\n if self.verbose:\n print('----------')\n print('At design: ' + str(x))\n\n q_samples, grad_samples = self.evalSamples(x)\n\n if self.verbose:\n print('Evaluating metric')\n\n return self.evalMetricFromSamples(q_samples, grad_samples, method)\n" ]
class DensityMatching(HorsetailMatching): '''Class for using density matching within an optimization. The main functionality is to evaluate the density matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a DensityMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Is a list of UncertainParameter objects, or a list of functions that return samples of the each uncertainty. :param function ftarget: function that returns the value of the target PDF function. :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param list integration_points: The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. ''' def __init__(self, fqoi, prob_uncertainties, ftarget=None, jac=False, samples_prob=1000, integration_points=None, kernel_bandwidth=None, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi self.prob_uncertainties = prob_uncertainties self.int_uncertainties = [] self.ftarget = ftarget self.jac = jac self.samples_prob = samples_prob self.samples_int = 1 self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose # Note that this class makes heavy use of the HorsetailMatching parent # class's methods ############################################################################## ## Public Methods ############################################################################## def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the density matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' return self._evalDensityMetric(q_samples, grad_samples) def getPDF(self): '''Function that gets vectors of the pdf and target at the last design evaluated. :return: tuple of q values, pdf values, target values ''' if hasattr(self, '_qplot'): return self._qplot, self._hplot, self._tplot else: raise ValueError('''The metric has not been evaluated at any design point so the PDF cannot get obtained''') ############################################################################## ## Private methods ## ############################################################################## def _evalDensityMetric(self, q_samples, grad_samples=None): if self.integration_points is None: q_min = np.amin(q_samples) q_max = np.amax(q_samples) q_range = q_max - q_min fis = np.linspace(q_min - q_range, q_max + q_range, 1000) self.integration_points = fis else: fis = self.integration_points # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = ((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth fjs = np.array(q_samples) N = len(fis) M = self.samples_prob t = np.array([float(self.ftarget(fi)) for fi in fis]).reshape([N, 1]) # column vector - row vector to give matrix delf = fis.reshape([N, 1]) - fjs.reshape([1, M]) const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) K = const_term * np.exp((-1./2.) * (delf/bw)**2) Ks = np.dot(K, np.ones([M, 1])).reshape([N, 1]) W = np.zeros([N, N]) # Trapezium rule weighting matrix for i in range(N): W[i, i] = (fis[min(i+1, N-1)] - fis[max(i-1, 0)])*0.5 l2norm = float((t - Ks).T.dot(W.dot((t - Ks)))) self._qplot = fis self._hplot = Ks self._tplot = t if grad_samples is None: return l2norm else: ndv = grad_samples.shape[2] gradjs = grad_samples[0, :, :] Kprime = const_term * np.exp((-1./2.) * (delf/bw)**2) *\ delf / bw**2 * -1. Fprime = np.zeros([M, ndv]) for kdv in range(ndv): Fprime[:, kdv] = gradjs[:, kdv] gradient = 2*(t - Ks).T.dot(W.dot(Kprime.dot(Fprime))).reshape(ndv) return l2norm, gradient
lwcook/horsetail-matching
horsetailmatching/densitymatching.py
DensityMatching.getPDF
python
def getPDF(self): '''Function that gets vectors of the pdf and target at the last design evaluated. :return: tuple of q values, pdf values, target values ''' if hasattr(self, '_qplot'): return self._qplot, self._hplot, self._tplot else: raise ValueError('''The metric has not been evaluated at any design point so the PDF cannot get obtained''')
Function that gets vectors of the pdf and target at the last design evaluated. :return: tuple of q values, pdf values, target values
train
https://github.com/lwcook/horsetail-matching/blob/f3d5f8d01249debbca978f412ce4eae017458119/horsetailmatching/densitymatching.py#L152-L165
null
class DensityMatching(HorsetailMatching): '''Class for using density matching within an optimization. The main functionality is to evaluate the density matching metric (and optionally its gradient) that can be used with external optimizers. The code is written such that all arguments that can be used at the initialization of a DensityMatching object can also be set as attributes after creation to achieve exactly the same effect. :param function fqoi: function that returns the quantity of interest, it must take two ordered arguments - the value of the design variable vector and the value of the uncertainty vector. :param list prob_uncertainties: list of probabilistic uncertainties. Is a list of UncertainParameter objects, or a list of functions that return samples of the each uncertainty. :param function ftarget: function that returns the value of the target PDF function. :param bool/function jac: Argument that specifies how to evaluate the gradient of the quantity of interest. If False no gradients are propagated, if True the fqoi should return a second argument g such that g_i = dq/dx_i. If a function, it should have the same signature as fqoi but return g. [default False] :param int samples_prob: number of samples to take from the probabilsitic uncertainties. [default 1000] :param list integration_points: The integration point values to use when evaluating the metric using kernels [by default 100 points spread over 3 times the range of the samples of q obtained the first time the metric is evaluated] :param number kernel_bandwidth: The bandwidth used in the kernel function [by default it is found the first time the metric is evaluated using Scott's rule] :param function surrogate: Surrogate that is created at every design point to be sampled instead of fqoi. It should be a function that takes two arguments - an array with values of the uncertainties at which to fit the surrogate of size (num_quadrature_points, num_uncertainties), and an array of quantity of interest values corresponding to these uncertainty values to which to fit the surrogate of size (num_quadrature_points). It should return a functio that predicts the qoi at an aribtrary value of the uncertainties. [default None] :param list surrogate_points: Only with a surrogate. List of points at which fqoi is evaluated to give values to fit the surrogates to. These are passed to the surrogate function along with the qoi evaluated at these points when the surrogate is fitted [by default tensor quadrature of 5 points in each uncertain dimension is used] :param bool/function surrogate_jac: Only with a surrogate. Specifies how to take surrogates of the gradient. It works similarly to the jac argument: if False, the same surrogate is fitted to fqoi and each component of its gradient, if True, the surrogate function is expected to take a third argument - an array that is the gradient at each of the quadrature points of size (num_quadrature_points, num_design_variables). If a function, then instead the array of uncertainty values and the array of gradient values are passed to this function and it should return a function for the surrogate model of the gradient. :param bool reuse_samples: If True will reuse the same set of samples of the uncertainties for evaluating the metric at any value of the design variables, if False wise will re-sample every time evalMetric is called [default True] :param bool verbose: If True will print out details [default False]. ''' def __init__(self, fqoi, prob_uncertainties, ftarget=None, jac=False, samples_prob=1000, integration_points=None, kernel_bandwidth=None, surrogate=None, surrogate_points=None, surrogate_jac=False, reuse_samples=True, verbose=False): self.fqoi = fqoi self.prob_uncertainties = prob_uncertainties self.int_uncertainties = [] self.ftarget = ftarget self.jac = jac self.samples_prob = samples_prob self.samples_int = 1 self.integration_points = integration_points self.kernel_bandwidth = kernel_bandwidth self.reuse_samples = reuse_samples self.u_samples = None self.surrogate = surrogate self.surrogate_points = surrogate_points self.surrogate_jac = surrogate_jac self.verbose = verbose # Note that this class makes heavy use of the HorsetailMatching parent # class's methods ############################################################################## ## Public Methods ############################################################################## def evalMetric(self, x, method=None): '''Evaluates the density matching metric at a given design point. :param iterable x: values of the design variables, this is passed as the first argument to the function fqoi :return: metric_value - value of the metric evaluated at the design point given by x :rtype: float *Example Usage*:: >>> def myFunc(x, u): return x[0]*x[1] + u >>> u1 = UniformParameter() >>> theDM = DensityMatching(myFunc, u) >>> x0 = [1, 2] >>> theDM.evalMetric(x0) ''' return super(DensityMatching, self).evalMetric(x, method) def evalMetricFromSamples(self, q_samples, grad_samples=None, method=None): '''Evaluates the density matching metric from given samples of the quantity of interest and gradient instead of evaluating them at a design. :param np.ndarray q_samples: samples of the quantity of interest, size (M_int, M_prob) :param np.ndarray grad_samples: samples of the gradien, size (M_int, M_prob, n_x) :return: metric_value - value of the metric :rtype: float ''' return self._evalDensityMetric(q_samples, grad_samples) ############################################################################## ## Private methods ## ############################################################################## def _evalDensityMetric(self, q_samples, grad_samples=None): if self.integration_points is None: q_min = np.amin(q_samples) q_max = np.amax(q_samples) q_range = q_max - q_min fis = np.linspace(q_min - q_range, q_max + q_range, 1000) self.integration_points = fis else: fis = self.integration_points # If kernel bandwidth not specified, find it using Scott's rule if self.kernel_bandwidth is None: if abs(np.max(q_samples) - np.min(q_samples)) < 1e-6: bw = 1e-6 else: bw = ((4/(3.*q_samples.shape[1]))**(1/5.) *np.std(q_samples[0,:])) self.kernel_bandwidth = bw else: bw = self.kernel_bandwidth fjs = np.array(q_samples) N = len(fis) M = self.samples_prob t = np.array([float(self.ftarget(fi)) for fi in fis]).reshape([N, 1]) # column vector - row vector to give matrix delf = fis.reshape([N, 1]) - fjs.reshape([1, M]) const_term = 1.0/(M * np.sqrt(2*np.pi*bw**2)) K = const_term * np.exp((-1./2.) * (delf/bw)**2) Ks = np.dot(K, np.ones([M, 1])).reshape([N, 1]) W = np.zeros([N, N]) # Trapezium rule weighting matrix for i in range(N): W[i, i] = (fis[min(i+1, N-1)] - fis[max(i-1, 0)])*0.5 l2norm = float((t - Ks).T.dot(W.dot((t - Ks)))) self._qplot = fis self._hplot = Ks self._tplot = t if grad_samples is None: return l2norm else: ndv = grad_samples.shape[2] gradjs = grad_samples[0, :, :] Kprime = const_term * np.exp((-1./2.) * (delf/bw)**2) *\ delf / bw**2 * -1. Fprime = np.zeros([M, ndv]) for kdv in range(ndv): Fprime[:, kdv] = gradjs[:, kdv] gradient = 2*(t - Ks).T.dot(W.dot(Kprime.dot(Fprime))).reshape(ndv) return l2norm, gradient
glue-viz/echo
echo/core.py
add_callback
python
def add_callback(instance, prop, callback, echo_old=False, priority=0): p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.add_callback(instance, callback, echo_old=echo_old, priority=priority)
Attach a callback function to a property in an instance Parameters ---------- instance The instance to add the callback to prop : str Name of callback property in `instance` callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). Examples -------- :: class Foo: bar = CallbackProperty(0) def callback(value): pass f = Foo() add_callback(f, 'bar', callback)
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L334-L372
[ "def add_callback(self, instance, func, echo_old=False, priority=0):\n \"\"\"\n Add a callback to a specific instance that manages this property\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n func : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n \"\"\"\n\n if echo_old:\n self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority)\n else:\n self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority)\n" ]
from __future__ import absolute_import, division, print_function import weakref from itertools import chain from weakref import WeakKeyDictionary from contextlib import contextmanager from .callback_container import CallbackContainer __all__ = ['CallbackProperty', 'callback_property', 'add_callback', 'remove_callback', 'delay_callback', 'ignore_callback', 'HasCallbackProperties', 'keep_in_sync'] class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return else: raise ValueError("Callback function not found: %s" % func) class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name) def remove_callback(instance, prop, callback): """ Remove a callback function from a property in an instance Parameters ---------- instance The instance to detach the callback from prop : str Name of callback property in `instance` callback : func The callback function to remove """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.remove_callback(instance, callback) def callback_property(getter): """ A decorator to build a CallbackProperty. This is used by wrapping a getter method, similar to the use of @property:: class Foo(object): @callback_property def x(self): return self._x @x.setter def x(self, value): self._x = value In simple cases with no getter or setter logic, it's easier to create a :class:`~echo.CallbackProperty` directly:: class Foo(object); x = CallbackProperty(initial_value) """ cb = CallbackProperty(getter=getter) cb.__doc__ = getter.__doc__ return cb class delay_callback(object): """ Delay any callback functions from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. Each callback will be called once on exit Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to delay Examples -------- :: with delay_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # callbacks triggered at this point, if needed """ # Class-level registry of properties and how many times the callbacks have # been delayed. The idea is that when nesting calls to delay_callback, the # delay count is increased, and every time __exit__ is called, the count is # decreased, and once the count reaches zero, the callback is triggered. delay_count = {} old_values = {} def __init__(self, instance, *props): self.instance = instance self.props = props def __enter__(self): delay_props = {} for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) if (self.instance, prop) not in self.delay_count: self.delay_count[self.instance, prop] = 1 self.old_values[self.instance, prop] = p._get_full_info(self.instance) delay_props[prop] = p._get_full_info(self.instance) else: self.delay_count[self.instance, prop] += 1 p.disable(self.instance) if isinstance(self.instance, HasCallbackProperties): self.instance._delay_global_callbacks(delay_props) def __exit__(self, *args): resume_props = {} notifications = [] for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): # pragma: no cover raise TypeError("%s is not a CallbackProperty" % prop) if self.delay_count[self.instance, prop] > 1: self.delay_count[self.instance, prop] -= 1 else: self.delay_count.pop((self.instance, prop)) old = self.old_values.pop((self.instance, prop)) p.enable(self.instance) new = p._get_full_info(self.instance) if old != new: notifications.append((p, (self.instance, old[0], new[0]))) resume_props[prop] = new if isinstance(self.instance, HasCallbackProperties): self.instance._process_delayed_global_callbacks(resume_props) for p, args in notifications: p.notify(*args) @contextmanager def ignore_callback(instance, *props): """ Temporarily ignore any callbacks from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. In contrast with :func:`~echo.delay_callback`, no callbakcs will be called on exiting the context manager Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to ignore Examples -------- :: with ignore_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # no callbacks called """ for prop in props: p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.disable(instance) if isinstance(instance, HasCallbackProperties): instance._ignore_global_callbacks(props) yield for prop in props: p = getattr(type(instance), prop) assert isinstance(p, CallbackProperty) p.enable(instance) if isinstance(instance, HasCallbackProperties): instance._unignore_global_callbacks(props) class keep_in_sync(object): def __init__(self, instance1, prop1, instance2, prop2): self.instance1 = weakref.ref(instance1, self.disable_syncing) self.prop1 = prop1 self.instance2 = weakref.ref(instance2, self.disable_syncing) self.prop2 = prop2 self._syncing = False self.enabled = False self.enable_syncing() def prop1_from_prop2(self, value): if not self._syncing: self._syncing = True setattr(self.instance1(), self.prop1, getattr(self.instance2(), self.prop2)) self._syncing = False def prop2_from_prop1(self, value): if not self._syncing: self._syncing = True setattr(self.instance2(), self.prop2, getattr(self.instance1(), self.prop1)) self._syncing = False def enable_syncing(self, *args): if self.enabled: return add_callback(self.instance1(), self.prop1, self.prop2_from_prop1) add_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = True def disable_syncing(self, *args): if not self.enabled: return if self.instance1() is not None: remove_callback(self.instance1(), self.prop1, self.prop2_from_prop1) if self.instance2() is not None: remove_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = False
glue-viz/echo
echo/core.py
remove_callback
python
def remove_callback(instance, prop, callback): p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.remove_callback(instance, callback)
Remove a callback function from a property in an instance Parameters ---------- instance The instance to detach the callback from prop : str Name of callback property in `instance` callback : func The callback function to remove
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L375-L391
null
from __future__ import absolute_import, division, print_function import weakref from itertools import chain from weakref import WeakKeyDictionary from contextlib import contextmanager from .callback_container import CallbackContainer __all__ = ['CallbackProperty', 'callback_property', 'add_callback', 'remove_callback', 'delay_callback', 'ignore_callback', 'HasCallbackProperties', 'keep_in_sync'] class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return else: raise ValueError("Callback function not found: %s" % func) class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name) def add_callback(instance, prop, callback, echo_old=False, priority=0): """ Attach a callback function to a property in an instance Parameters ---------- instance The instance to add the callback to prop : str Name of callback property in `instance` callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). Examples -------- :: class Foo: bar = CallbackProperty(0) def callback(value): pass f = Foo() add_callback(f, 'bar', callback) """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.add_callback(instance, callback, echo_old=echo_old, priority=priority) def callback_property(getter): """ A decorator to build a CallbackProperty. This is used by wrapping a getter method, similar to the use of @property:: class Foo(object): @callback_property def x(self): return self._x @x.setter def x(self, value): self._x = value In simple cases with no getter or setter logic, it's easier to create a :class:`~echo.CallbackProperty` directly:: class Foo(object); x = CallbackProperty(initial_value) """ cb = CallbackProperty(getter=getter) cb.__doc__ = getter.__doc__ return cb class delay_callback(object): """ Delay any callback functions from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. Each callback will be called once on exit Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to delay Examples -------- :: with delay_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # callbacks triggered at this point, if needed """ # Class-level registry of properties and how many times the callbacks have # been delayed. The idea is that when nesting calls to delay_callback, the # delay count is increased, and every time __exit__ is called, the count is # decreased, and once the count reaches zero, the callback is triggered. delay_count = {} old_values = {} def __init__(self, instance, *props): self.instance = instance self.props = props def __enter__(self): delay_props = {} for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) if (self.instance, prop) not in self.delay_count: self.delay_count[self.instance, prop] = 1 self.old_values[self.instance, prop] = p._get_full_info(self.instance) delay_props[prop] = p._get_full_info(self.instance) else: self.delay_count[self.instance, prop] += 1 p.disable(self.instance) if isinstance(self.instance, HasCallbackProperties): self.instance._delay_global_callbacks(delay_props) def __exit__(self, *args): resume_props = {} notifications = [] for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): # pragma: no cover raise TypeError("%s is not a CallbackProperty" % prop) if self.delay_count[self.instance, prop] > 1: self.delay_count[self.instance, prop] -= 1 else: self.delay_count.pop((self.instance, prop)) old = self.old_values.pop((self.instance, prop)) p.enable(self.instance) new = p._get_full_info(self.instance) if old != new: notifications.append((p, (self.instance, old[0], new[0]))) resume_props[prop] = new if isinstance(self.instance, HasCallbackProperties): self.instance._process_delayed_global_callbacks(resume_props) for p, args in notifications: p.notify(*args) @contextmanager def ignore_callback(instance, *props): """ Temporarily ignore any callbacks from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. In contrast with :func:`~echo.delay_callback`, no callbakcs will be called on exiting the context manager Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to ignore Examples -------- :: with ignore_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # no callbacks called """ for prop in props: p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.disable(instance) if isinstance(instance, HasCallbackProperties): instance._ignore_global_callbacks(props) yield for prop in props: p = getattr(type(instance), prop) assert isinstance(p, CallbackProperty) p.enable(instance) if isinstance(instance, HasCallbackProperties): instance._unignore_global_callbacks(props) class keep_in_sync(object): def __init__(self, instance1, prop1, instance2, prop2): self.instance1 = weakref.ref(instance1, self.disable_syncing) self.prop1 = prop1 self.instance2 = weakref.ref(instance2, self.disable_syncing) self.prop2 = prop2 self._syncing = False self.enabled = False self.enable_syncing() def prop1_from_prop2(self, value): if not self._syncing: self._syncing = True setattr(self.instance1(), self.prop1, getattr(self.instance2(), self.prop2)) self._syncing = False def prop2_from_prop1(self, value): if not self._syncing: self._syncing = True setattr(self.instance2(), self.prop2, getattr(self.instance1(), self.prop1)) self._syncing = False def enable_syncing(self, *args): if self.enabled: return add_callback(self.instance1(), self.prop1, self.prop2_from_prop1) add_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = True def disable_syncing(self, *args): if not self.enabled: return if self.instance1() is not None: remove_callback(self.instance1(), self.prop1, self.prop2_from_prop1) if self.instance2() is not None: remove_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = False
glue-viz/echo
echo/core.py
callback_property
python
def callback_property(getter): cb = CallbackProperty(getter=getter) cb.__doc__ = getter.__doc__ return cb
A decorator to build a CallbackProperty. This is used by wrapping a getter method, similar to the use of @property:: class Foo(object): @callback_property def x(self): return self._x @x.setter def x(self, value): self._x = value In simple cases with no getter or setter logic, it's easier to create a :class:`~echo.CallbackProperty` directly:: class Foo(object); x = CallbackProperty(initial_value)
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L394-L418
null
from __future__ import absolute_import, division, print_function import weakref from itertools import chain from weakref import WeakKeyDictionary from contextlib import contextmanager from .callback_container import CallbackContainer __all__ = ['CallbackProperty', 'callback_property', 'add_callback', 'remove_callback', 'delay_callback', 'ignore_callback', 'HasCallbackProperties', 'keep_in_sync'] class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return else: raise ValueError("Callback function not found: %s" % func) class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name) def add_callback(instance, prop, callback, echo_old=False, priority=0): """ Attach a callback function to a property in an instance Parameters ---------- instance The instance to add the callback to prop : str Name of callback property in `instance` callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). Examples -------- :: class Foo: bar = CallbackProperty(0) def callback(value): pass f = Foo() add_callback(f, 'bar', callback) """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.add_callback(instance, callback, echo_old=echo_old, priority=priority) def remove_callback(instance, prop, callback): """ Remove a callback function from a property in an instance Parameters ---------- instance The instance to detach the callback from prop : str Name of callback property in `instance` callback : func The callback function to remove """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.remove_callback(instance, callback) class delay_callback(object): """ Delay any callback functions from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. Each callback will be called once on exit Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to delay Examples -------- :: with delay_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # callbacks triggered at this point, if needed """ # Class-level registry of properties and how many times the callbacks have # been delayed. The idea is that when nesting calls to delay_callback, the # delay count is increased, and every time __exit__ is called, the count is # decreased, and once the count reaches zero, the callback is triggered. delay_count = {} old_values = {} def __init__(self, instance, *props): self.instance = instance self.props = props def __enter__(self): delay_props = {} for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) if (self.instance, prop) not in self.delay_count: self.delay_count[self.instance, prop] = 1 self.old_values[self.instance, prop] = p._get_full_info(self.instance) delay_props[prop] = p._get_full_info(self.instance) else: self.delay_count[self.instance, prop] += 1 p.disable(self.instance) if isinstance(self.instance, HasCallbackProperties): self.instance._delay_global_callbacks(delay_props) def __exit__(self, *args): resume_props = {} notifications = [] for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): # pragma: no cover raise TypeError("%s is not a CallbackProperty" % prop) if self.delay_count[self.instance, prop] > 1: self.delay_count[self.instance, prop] -= 1 else: self.delay_count.pop((self.instance, prop)) old = self.old_values.pop((self.instance, prop)) p.enable(self.instance) new = p._get_full_info(self.instance) if old != new: notifications.append((p, (self.instance, old[0], new[0]))) resume_props[prop] = new if isinstance(self.instance, HasCallbackProperties): self.instance._process_delayed_global_callbacks(resume_props) for p, args in notifications: p.notify(*args) @contextmanager def ignore_callback(instance, *props): """ Temporarily ignore any callbacks from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. In contrast with :func:`~echo.delay_callback`, no callbakcs will be called on exiting the context manager Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to ignore Examples -------- :: with ignore_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # no callbacks called """ for prop in props: p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.disable(instance) if isinstance(instance, HasCallbackProperties): instance._ignore_global_callbacks(props) yield for prop in props: p = getattr(type(instance), prop) assert isinstance(p, CallbackProperty) p.enable(instance) if isinstance(instance, HasCallbackProperties): instance._unignore_global_callbacks(props) class keep_in_sync(object): def __init__(self, instance1, prop1, instance2, prop2): self.instance1 = weakref.ref(instance1, self.disable_syncing) self.prop1 = prop1 self.instance2 = weakref.ref(instance2, self.disable_syncing) self.prop2 = prop2 self._syncing = False self.enabled = False self.enable_syncing() def prop1_from_prop2(self, value): if not self._syncing: self._syncing = True setattr(self.instance1(), self.prop1, getattr(self.instance2(), self.prop2)) self._syncing = False def prop2_from_prop1(self, value): if not self._syncing: self._syncing = True setattr(self.instance2(), self.prop2, getattr(self.instance1(), self.prop1)) self._syncing = False def enable_syncing(self, *args): if self.enabled: return add_callback(self.instance1(), self.prop1, self.prop2_from_prop1) add_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = True def disable_syncing(self, *args): if not self.enabled: return if self.instance1() is not None: remove_callback(self.instance1(), self.prop1, self.prop2_from_prop1) if self.instance2() is not None: remove_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = False
glue-viz/echo
echo/core.py
ignore_callback
python
def ignore_callback(instance, *props): for prop in props: p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.disable(instance) if isinstance(instance, HasCallbackProperties): instance._ignore_global_callbacks(props) yield for prop in props: p = getattr(type(instance), prop) assert isinstance(p, CallbackProperty) p.enable(instance) if isinstance(instance, HasCallbackProperties): instance._unignore_global_callbacks(props)
Temporarily ignore any callbacks from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. In contrast with :func:`~echo.delay_callback`, no callbakcs will be called on exiting the context manager Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to ignore Examples -------- :: with ignore_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # no callbacks called
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L511-L555
null
from __future__ import absolute_import, division, print_function import weakref from itertools import chain from weakref import WeakKeyDictionary from contextlib import contextmanager from .callback_container import CallbackContainer __all__ = ['CallbackProperty', 'callback_property', 'add_callback', 'remove_callback', 'delay_callback', 'ignore_callback', 'HasCallbackProperties', 'keep_in_sync'] class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return else: raise ValueError("Callback function not found: %s" % func) class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name) def add_callback(instance, prop, callback, echo_old=False, priority=0): """ Attach a callback function to a property in an instance Parameters ---------- instance The instance to add the callback to prop : str Name of callback property in `instance` callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). Examples -------- :: class Foo: bar = CallbackProperty(0) def callback(value): pass f = Foo() add_callback(f, 'bar', callback) """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.add_callback(instance, callback, echo_old=echo_old, priority=priority) def remove_callback(instance, prop, callback): """ Remove a callback function from a property in an instance Parameters ---------- instance The instance to detach the callback from prop : str Name of callback property in `instance` callback : func The callback function to remove """ p = getattr(type(instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) p.remove_callback(instance, callback) def callback_property(getter): """ A decorator to build a CallbackProperty. This is used by wrapping a getter method, similar to the use of @property:: class Foo(object): @callback_property def x(self): return self._x @x.setter def x(self, value): self._x = value In simple cases with no getter or setter logic, it's easier to create a :class:`~echo.CallbackProperty` directly:: class Foo(object); x = CallbackProperty(initial_value) """ cb = CallbackProperty(getter=getter) cb.__doc__ = getter.__doc__ return cb class delay_callback(object): """ Delay any callback functions from one or more callback properties This is a context manager. Within the context block, no callbacks will be issued. Each callback will be called once on exit Parameters ---------- instance An instance object with callback properties *props : str One or more properties within instance to delay Examples -------- :: with delay_callback(foo, 'bar', 'baz'): f.bar = 20 f.baz = 30 f.bar = 10 print('done') # callbacks triggered at this point, if needed """ # Class-level registry of properties and how many times the callbacks have # been delayed. The idea is that when nesting calls to delay_callback, the # delay count is increased, and every time __exit__ is called, the count is # decreased, and once the count reaches zero, the callback is triggered. delay_count = {} old_values = {} def __init__(self, instance, *props): self.instance = instance self.props = props def __enter__(self): delay_props = {} for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): raise TypeError("%s is not a CallbackProperty" % prop) if (self.instance, prop) not in self.delay_count: self.delay_count[self.instance, prop] = 1 self.old_values[self.instance, prop] = p._get_full_info(self.instance) delay_props[prop] = p._get_full_info(self.instance) else: self.delay_count[self.instance, prop] += 1 p.disable(self.instance) if isinstance(self.instance, HasCallbackProperties): self.instance._delay_global_callbacks(delay_props) def __exit__(self, *args): resume_props = {} notifications = [] for prop in self.props: p = getattr(type(self.instance), prop) if not isinstance(p, CallbackProperty): # pragma: no cover raise TypeError("%s is not a CallbackProperty" % prop) if self.delay_count[self.instance, prop] > 1: self.delay_count[self.instance, prop] -= 1 else: self.delay_count.pop((self.instance, prop)) old = self.old_values.pop((self.instance, prop)) p.enable(self.instance) new = p._get_full_info(self.instance) if old != new: notifications.append((p, (self.instance, old[0], new[0]))) resume_props[prop] = new if isinstance(self.instance, HasCallbackProperties): self.instance._process_delayed_global_callbacks(resume_props) for p, args in notifications: p.notify(*args) @contextmanager class keep_in_sync(object): def __init__(self, instance1, prop1, instance2, prop2): self.instance1 = weakref.ref(instance1, self.disable_syncing) self.prop1 = prop1 self.instance2 = weakref.ref(instance2, self.disable_syncing) self.prop2 = prop2 self._syncing = False self.enabled = False self.enable_syncing() def prop1_from_prop2(self, value): if not self._syncing: self._syncing = True setattr(self.instance1(), self.prop1, getattr(self.instance2(), self.prop2)) self._syncing = False def prop2_from_prop1(self, value): if not self._syncing: self._syncing = True setattr(self.instance2(), self.prop2, getattr(self.instance1(), self.prop1)) self._syncing = False def enable_syncing(self, *args): if self.enabled: return add_callback(self.instance1(), self.prop1, self.prop2_from_prop1) add_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = True def disable_syncing(self, *args): if not self.enabled: return if self.instance1() is not None: remove_callback(self.instance1(), self.prop1, self.prop2_from_prop1) if self.instance2() is not None: remove_callback(self.instance2(), self.prop2, self.prop1_from_prop2) self.enabled = False
glue-viz/echo
echo/core.py
CallbackProperty.notify
python
def notify(self, instance, old, new): if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new)
Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L97-L120
null
class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return
glue-viz/echo
echo/core.py
CallbackProperty.add_callback
python
def add_callback(self, instance, func, echo_old=False, priority=0): if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority)
Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority).
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L134-L156
null
class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return
glue-viz/echo
echo/core.py
CallbackProperty.remove_callback
python
def remove_callback(self, instance, func): for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return else: raise ValueError("Callback function not found: %s" % func)
Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L158-L176
null
class CallbackProperty(object): """ A property that callback functions can be added to. When a callback property changes value, each callback function is called with information about the state change. Otherwise, callback properties behave just like normal instance variables. CallbackProperties must be defined at the class level. Use the helper function :func:`~echo.add_callback` to attach a callback to a specific instance of a class with CallbackProperties Parameters ---------- default The initial value for the property docstring : str The docstring for the property getter, setter : func Custom getter and setter functions (advanced) """ def __init__(self, default=None, docstring=None, getter=None, setter=None): """ :param default: The initial value for the property """ self._default = default self._callbacks = WeakKeyDictionary() self._2arg_callbacks = WeakKeyDictionary() self._disabled = WeakKeyDictionary() self._values = WeakKeyDictionary() if getter is None: getter = self._default_getter if setter is None: setter = self._default_setter self._getter = getter self._setter = setter if docstring is not None: self.__doc__ = docstring def _default_getter(self, instance, owner=None): return self._values.get(instance, self._default) def _default_setter(self, instance, value): self._values.__setitem__(instance, value) def __get__(self, instance, owner=None): if instance is None: return self return self._getter(instance) def __set__(self, instance, value): try: old = self.__get__(instance) except AttributeError: # pragma: no cover old = None self._setter(instance, value) new = self.__get__(instance) if old != new: self.notify(instance, old, new) def setter(self, func): """ Method to use as a decorator, to mimic @property.setter """ self._setter = func return self def _get_full_info(self, instance): # Some callback subclasses may contain additional info in addition # to the main value, and we need to use this full information when # comparing old and new 'values', so this method is used in that # case. The result should be a tuple where the first item is the # actual primary value of the property and the second item is any # additional data to use in the comparison. return self.__get__(instance), None def notify(self, instance, old, new): """ Call all callback functions with the current value Each callback will either be called using callback(new) or callback(old, new) depending on whether ``echo_old`` was set to `True` when calling :func:`~echo.add_callback` Parameters ---------- instance The instance to consider old The old value of the property new The new value of the property """ if self._disabled.get(instance, False): return for cback in self._callbacks.get(instance, []): cback(new) for cback in self._2arg_callbacks.get(instance, []): cback(old, new) def disable(self, instance): """ Disable callbacks for a specific instance """ self._disabled[instance] = True def enable(self, instance): """ Enable previously-disabled callbacks for a specific instance """ self._disabled[instance] = False def add_callback(self, instance, func, echo_old=False, priority=0): """ Add a callback to a specific instance that manages this property Parameters ---------- instance The instance to add the callback to func : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``func(old, new)``. If `False` (the default), will be invoked as ``func(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if echo_old: self._2arg_callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) else: self._callbacks.setdefault(instance, CallbackContainer()).append(func, priority=priority) def remove_callback(self, instance, func): """ Remove a previously-added callback Parameters ---------- instance The instance to detach the callback from func : func The callback function to remove """ for cb in [self._callbacks, self._2arg_callbacks]: if instance not in cb: continue if func in cb[instance]: cb[instance].remove(func) return
glue-viz/echo
echo/core.py
HasCallbackProperties.add_callback
python
def add_callback(self, name, callback, echo_old=False, priority=0): if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name))
Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority).
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L245-L268
[ "def is_callback_property(self, name):\n \"\"\"\n Whether a property (identified by name) is a callback property.\n\n Parameters\n ----------\n name : str\n The name of the property to check\n \"\"\"\n return isinstance(getattr(type(self), name, None), CallbackProperty)\n" ]
class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name)
glue-viz/echo
echo/core.py
HasCallbackProperties.remove_callback
python
def remove_callback(self, name, callback): if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name))
Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L270-L289
[ "def is_callback_property(self, name):\n \"\"\"\n Whether a property (identified by name) is a callback property.\n\n Parameters\n ----------\n name : str\n The name of the property to check\n \"\"\"\n return isinstance(getattr(type(self), name, None), CallbackProperty)\n" ]
class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty) def iter_callback_properties(self): """ Iterator to loop over all callback properties. """ for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name)
glue-viz/echo
echo/core.py
HasCallbackProperties.iter_callback_properties
python
def iter_callback_properties(self): for name in dir(self): if self.is_callback_property(name): yield name, getattr(type(self), name)
Iterator to loop over all callback properties.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/core.py#L325-L331
[ "def is_callback_property(self, name):\n \"\"\"\n Whether a property (identified by name) is a callback property.\n\n Parameters\n ----------\n name : str\n The name of the property to check\n \"\"\"\n return isinstance(getattr(type(self), name, None), CallbackProperty)\n" ]
class HasCallbackProperties(object): """ A class that adds functionality to subclasses that use callback properties. """ def __init__(self): from .list import ListCallbackProperty self._global_callbacks = CallbackContainer() self._ignored_properties = set() self._delayed_properties = {} self._delay_global_calls = {} self._callback_wrappers = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): prop.add_callback(self, self._notify_global_lists) def _ignore_global_callbacks(self, properties): # This is to allow ignore_callbacks to work for global callbacks self._ignored_properties.update(properties) def _unignore_global_callbacks(self, properties): # Once this is called, we simply remove properties from _ignored_properties # and don't call the callbacks. This is used by ignore_callback self._ignored_properties -= set(properties) def _delay_global_callbacks(self, properties): # This is to allow delay_callback to still have an effect in delaying # global callbacks. We set _delayed_properties to a dictionary of the # values at the point at which the callbacks are delayed. self._delayed_properties.update(properties) def _process_delayed_global_callbacks(self, properties): # Once this is called, the global callbacks are called once each with # a dictionary of the current values of properties that have been # resumed. kwargs = {} for prop, new_value in properties.items(): old_value = self._delayed_properties.pop(prop) if old_value != new_value: kwargs[prop] = new_value[0] self._notify_global(**kwargs) def _notify_global_lists(self, *args): from .list import ListCallbackProperty properties = {} for prop_name, prop in self.iter_callback_properties(): if isinstance(prop, ListCallbackProperty): callback_list = getattr(self, prop_name) if callback_list is args[0]: properties[prop_name] = callback_list break self._notify_global(**properties) def _notify_global(self, **kwargs): for prop in set(self._delayed_properties) | set(self._ignored_properties): if prop in kwargs: kwargs.pop(prop) if len(kwargs) > 0: for callback in self._global_callbacks: callback(**kwargs) def __setattr__(self, attribute, value): super(HasCallbackProperties, self).__setattr__(attribute, value) if self.is_callback_property(attribute): self._notify_global(**{attribute: value}) def add_callback(self, name, callback, echo_old=False, priority=0): """ Add a callback that gets triggered when a callback property of the class changes. Parameters ---------- name : str The instance to add the callback to. callback : func The callback function to add echo_old : bool, optional If `True`, the callback function will be invoked with both the old and new values of the property, as ``callback(old, new)``. If `False` (the default), will be invoked as ``callback(new)`` priority : int, optional This can optionally be used to force a certain order of execution of callbacks (larger values indicate a higher priority). """ if self.is_callback_property(name): prop = getattr(type(self), name) prop.add_callback(self, callback, echo_old=echo_old, priority=priority) else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def remove_callback(self, name, callback): """ Remove a previously-added callback Parameters ---------- name : str The instance to remove the callback from. func : func The callback function to remove """ if self.is_callback_property(name): prop = getattr(type(self), name) try: prop.remove_callback(self, callback) except ValueError: # pragma: nocover pass # Be forgiving if callback was already removed before else: raise TypeError("attribute '{0}' is not a callback property".format(name)) def add_global_callback(self, callback): """ Add a global callback function, which is a callback that gets triggered when any callback properties on the class change. Parameters ---------- callback : func The callback function to add """ self._global_callbacks.append(callback) def remove_global_callback(self, callback): """ Remove a global callback function. Parameters ---------- callback : func The callback function to remove """ self._global_callbacks.remove(callback) def is_callback_property(self, name): """ Whether a property (identified by name) is a callback property. Parameters ---------- name : str The name of the property to check """ return isinstance(getattr(type(self), name, None), CallbackProperty)
glue-viz/echo
echo/qt/autoconnect.py
autoconnect_callbacks_to_qt
python
def autoconnect_callbacks_to_qt(instance, widget, connect_kwargs={}): if not hasattr(widget, 'children'): return for child in widget.findChildren(QtWidgets.QWidget): full_name = child.objectName() if '_' in full_name: wtype, wname = full_name.split('_', 1) if full_name in connect_kwargs: kwargs = connect_kwargs[full_name] elif wname in connect_kwargs: kwargs = connect_kwargs[wname] else: kwargs = {} if hasattr(instance, wname): if wtype in HANDLERS: HANDLERS[wtype](instance, wname, child, **kwargs)
Given a class instance with callback properties and a Qt widget/window, connect callback properties to Qt widgets automatically. The matching is done based on the objectName of the Qt widgets. Qt widgets that need to be connected should be named using the syntax ``type_name`` where ``type`` describes the kind of matching to be done, and ``name`` matches the name of a callback property. By default, the types can be: * ``value``: the callback property is linked to a Qt widget that has ``value`` and ``setValue`` methods. Note that for this type, two additional keyword arguments can be specified using ``connect_kwargs`` (see below): these are ``value_range``, which is used for cases where the Qt widget is e.g. a slider which has a range of values, and you want to map this range of values onto a different range for the callback property, and the second is ``log``, which can be set to `True` if this mapping should be done in log space. * ``valuetext``: the callback property is linked to a Qt widget that has ``text`` and ``setText`` methods, and the text is set to a string representation of the value. Note that for this type, an additional argument ``fmt`` can be provided, which gives either the format to use using the ``{}`` syntax, or should be a function that takes a value and returns a string. Optionally, if the Qt widget supports the ``editingFinished`` signal, this signal is connected to the callback property too. * ``bool``: the callback property is linked to a Qt widget that has ``isChecked`` and ``setChecked`` methods, such as a checkable button. * ``text``: the callback property is linked to a Qt widget that has ``text`` and ``setText`` methods. Optionally, if the Qt widget supports the ``editingFinished`` signal, this signal is connected to the callback property too. * ``combodata``: the callback property is linked to a QComboBox based on the ``userData`` of the entries in the combo box. * ``combotext``: the callback property is linked to a QComboBox based on the label of the entries in the combo box. Applications can also define additional mappings between type and auto-linking. To do this, simply add a new entry to the ``HANDLERS`` object:: >>> echo.qt.autoconnect import HANDLERS >>> HANDLERS['color'] = connect_color The handler function (``connect_color`` in the example above) should take the following arguments: the instance the callback property is attached to, the name of the callback property, the Qt widget, and optionally some keyword arguments. When calling ``autoconnect_callbacks_to_qt``, you can specify ``connect_kwargs``, where each key should be a valid callback property name, and which gives any additional keyword arguments that can be taken by the connect functions, as described above. These include for example ``value_range``, ``log``, and ``fmt``. This function is especially useful when defining ui files, since widget objectNames can be easily set during the editing process.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/autoconnect.py#L27-L105
null
from __future__ import absolute_import, division, print_function from qtpy import QtWidgets from .connect import (connect_checkable_button, connect_value, connect_combo_data, connect_combo_text, connect_float_text, connect_text, connect_button, connect_combo_selection) __all__ = ['autoconnect_callbacks_to_qt'] HANDLERS = {} HANDLERS['value'] = connect_value HANDLERS['valuetext'] = connect_float_text HANDLERS['bool'] = connect_checkable_button HANDLERS['text'] = connect_text HANDLERS['combodata'] = connect_combo_data HANDLERS['combotext'] = connect_combo_text HANDLERS['button'] = connect_button HANDLERS['combodatasel'] = connect_combo_selection
glue-viz/echo
echo/qt/connect.py
connect_checkable_button
python
def connect_checkable_button(instance, prop, widget): add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False)
Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L20-L36
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_text
python
def connect_text(instance, prop, widget): def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop))
Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L39-L74
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n", "def update_widget(val):\n if hasattr(widget, 'editingFinished'):\n widget.blockSignals(True)\n widget.setText(val)\n widget.blockSignals(False)\n widget.editingFinished.emit()\n else:\n widget.setText(val)\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_combo_data
python
def connect_combo_data(instance, prop, widget): def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L77-L114
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n", "def update_widget(value):\n try:\n idx = _find_combo_data(widget, value)\n except ValueError:\n if value is None:\n idx = -1\n else:\n raise\n widget.setCurrentIndex(idx)\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_combo_text
python
def connect_combo_text(instance, prop, widget): def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L117-L154
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n", "def update_widget(value):\n try:\n idx = _find_combo_text(widget, value)\n except ValueError:\n if value is None:\n idx = -1\n else:\n raise\n widget.setCurrentIndex(idx)\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_float_text
python
def connect_float_text(instance, prop, widget, fmt="{:g}"): if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop))
Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L157-L200
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n", "def update_widget(val):\n if val is None:\n val = 0.\n widget.setText(format_func(val))\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_value
python
def connect_value(instance, prop, widget, value_range=None, log=False): if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop))
Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L203-L253
[ "def add_callback(instance, prop, callback, echo_old=False, priority=0):\n \"\"\"\n Attach a callback function to a property in an instance\n\n Parameters\n ----------\n instance\n The instance to add the callback to\n prop : str\n Name of callback property in `instance`\n callback : func\n The callback function to add\n echo_old : bool, optional\n If `True`, the callback function will be invoked with both the old\n and new values of the property, as ``func(old, new)``. If `False`\n (the default), will be invoked as ``func(new)``\n priority : int, optional\n This can optionally be used to force a certain order of execution of\n callbacks (larger values indicate a higher priority).\n\n Examples\n --------\n\n ::\n\n class Foo:\n bar = CallbackProperty(0)\n\n def callback(value):\n pass\n\n f = Foo()\n add_callback(f, 'bar', callback)\n\n \"\"\"\n p = getattr(type(instance), prop)\n if not isinstance(p, CallbackProperty):\n raise TypeError(\"%s is not a CallbackProperty\" % prop)\n p.add_callback(instance, callback, echo_old=echo_old, priority=priority)\n", "def update_widget(val):\n if val is None:\n widget.setValue(0)\n return\n if log:\n val = math.log10(val)\n if value_range is not None:\n imin, imax = widget.minimum(), widget.maximum()\n val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin\n widget.setValue(val)\n" ]
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
connect_button
python
def connect_button(instance, prop, widget): widget.clicked.connect(getattr(instance, prop))
Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L256-L269
null
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
_find_combo_data
python
def _find_combo_data(widget, value): # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,))
Returns the index in a combo box where itemData == value Raises a ValueError if data is not found
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L272-L284
null
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_text(widget, value): """ Returns the index in a combo box where text == value Raises a ValueError if data is not found """ i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/qt/connect.py
_find_combo_text
python
def _find_combo_text(widget, value): i = widget.findText(value) if i == -1: raise ValueError("%s not found in combo box" % value) else: return i
Returns the index in a combo box where text == value Raises a ValueError if data is not found
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/qt/connect.py#L287-L297
null
# The functions in this module are used to connect callback properties to Qt # widgets. from __future__ import absolute_import, division, print_function import math from functools import partial from qtpy import QtGui from qtpy.QtCore import Qt from ..core import add_callback from ..selection import SelectionCallbackProperty, ChoiceSeparator __all__ = ['connect_checkable_button', 'connect_text', 'connect_combo_data', 'connect_combo_text', 'connect_float_text', 'connect_value', 'connect_combo_selection'] def connect_checkable_button(instance, prop, widget): """ Connect a boolean callback property with a Qt button widget. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setChecked`` method and the ``toggled`` signal. """ add_callback(instance, prop, widget.setChecked) widget.toggled.connect(partial(setattr, instance, prop)) widget.setChecked(getattr(instance, prop) or False) def connect_text(instance, prop, widget): """ Connect a string callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. """ def update_prop(): val = widget.text() setattr(instance, prop, val) def update_widget(val): if hasattr(widget, 'editingFinished'): widget.blockSignals(True) widget.setText(val) widget.blockSignals(False) widget.editingFinished.emit() else: widget.setText(val) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_combo_data(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the userData. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_text: connect a callback property with a QComboBox widget based on the text. """ def update_widget(value): try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_combo_text(instance, prop, widget): """ Connect a callback property with a QComboBox widget based on the text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QComboBox The combo box to connect. See Also -------- connect_combo_data: connect a callback property with a QComboBox widget based on the userData. """ def update_widget(value): try: idx = _find_combo_text(widget, value) except ValueError: if value is None: idx = -1 else: raise widget.setCurrentIndex(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemText(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_float_text(instance, prop, widget, fmt="{:g}"): """ Connect a numerical callback property with a Qt widget containing text. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. fmt : str or func This should be either a format string (in the ``{}`` notation), or a function that takes a number and returns a string. """ if callable(fmt): format_func = fmt else: def format_func(x): return fmt.format(x) def update_prop(): val = widget.text() try: setattr(instance, prop, float(val)) except ValueError: setattr(instance, prop, 0) def update_widget(val): if val is None: val = 0. widget.setText(format_func(val)) add_callback(instance, prop, update_widget) try: widget.editingFinished.connect(update_prop) except AttributeError: pass update_widget(getattr(instance, prop)) def connect_value(instance, prop, widget, value_range=None, log=False): """ Connect a numerical callback property with a Qt widget representing a value. Parameters ---------- instance : object The class instance that the callback property is attached to prop : str The name of the callback property widget : QtWidget The Qt widget to connect. This should implement the ``setText`` and ``text`` methods as well optionally the ``editingFinished`` signal. value_range : iterable, optional A pair of two values representing the true range of values (since Qt widgets such as sliders can only have values in certain ranges). log : bool, optional Whether the Qt widget value should be mapped to the log of the callback property. """ if log: if value_range is None: raise ValueError("log option can only be set if value_range is given") else: value_range = math.log10(value_range[0]), math.log10(value_range[1]) def update_prop(): val = widget.value() if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - imin) / (imax - imin) * (value_range[1] - value_range[0]) + value_range[0] if log: val = 10 ** val setattr(instance, prop, val) def update_widget(val): if val is None: widget.setValue(0) return if log: val = math.log10(val) if value_range is not None: imin, imax = widget.minimum(), widget.maximum() val = (val - value_range[0]) / (value_range[1] - value_range[0]) * (imax - imin) + imin widget.setValue(val) add_callback(instance, prop, update_widget) widget.valueChanged.connect(update_prop) update_widget(getattr(instance, prop)) def connect_button(instance, prop, widget): """ Connect a button with a callback method Parameters ---------- instance : object The class instance that the callback method is attached to prop : str The name of the callback method widget : QtWidget The Qt widget to connect. This should implement the ``clicked`` method """ widget.clicked.connect(getattr(instance, prop)) def _find_combo_data(widget, value): """ Returns the index in a combo box where itemData == value Raises a ValueError if data is not found """ # Here we check that the result is True, because some classes may overload # == and return other kinds of objects whether true or false. for idx in range(widget.count()): if widget.itemData(idx) is value or (widget.itemData(idx) == value) is True: return idx else: raise ValueError("%s not found in combo box" % (value,)) def connect_combo_selection(instance, prop, widget, display=str): if not isinstance(getattr(type(instance), prop), SelectionCallbackProperty): raise TypeError('connect_combo_selection requires a SelectionCallbackProperty') def update_widget(value): # Update choices in the combo box combo_data = [widget.itemData(idx) for idx in range(widget.count())] choices = getattr(type(instance), prop).get_choices(instance) choice_labels = getattr(type(instance), prop).get_choice_labels(instance) if combo_data == choices: choices_updated = False else: widget.blockSignals(True) widget.clear() if len(choices) == 0: return combo_model = widget.model() for index, (label, choice) in enumerate(zip(choice_labels, choices)): widget.addItem(label, userData=choice) # We interpret None data as being disabled rows (used for headers) if isinstance(choice, ChoiceSeparator): item = combo_model.item(index) palette = widget.palette() item.setFlags(item.flags() & ~(Qt.ItemIsSelectable | Qt.ItemIsEnabled)) item.setData(palette.color(QtGui.QPalette.Disabled, QtGui.QPalette.Text)) choices_updated = True # Update current selection try: idx = _find_combo_data(widget, value) except ValueError: if value is None: idx = -1 else: raise if idx == widget.currentIndex() and not choices_updated: return widget.setCurrentIndex(idx) widget.blockSignals(False) widget.currentIndexChanged.emit(idx) def update_prop(idx): if idx == -1: setattr(instance, prop, None) else: setattr(instance, prop, widget.itemData(idx)) add_callback(instance, prop, update_widget) widget.currentIndexChanged.connect(update_prop) update_widget(getattr(instance, prop))
glue-viz/echo
echo/callback_container.py
CallbackContainer._wrap
python
def _wrap(self, value, priority=0): if not callable(value): raise TypeError("Only callable values can be stored in CallbackContainer") elif self.is_bound_method(value): # We are dealing with a bound method. Method references aren't # persistent, so instead we store a reference to the function # and instance. value = (weakref.ref(value.__func__), weakref.ref(value.__self__, self._auto_remove), priority) else: value = (value, priority) return value
Given a function/method, this will automatically wrap a method using weakref to avoid circular references.
train
https://github.com/glue-viz/echo/blob/6ad54cc5e869de27c34e8716f2619ddc640f08fe/echo/callback_container.py#L20-L43
null
class CallbackContainer(object): """ A list-like container for callback functions. We need to be careful with storing references to methods, because if a callback method is on a class which contains both the callback and a callback property, a circular reference is created which results in a memory leak. Instead, we need to use a weak reference which results in the callback being removed if the instance is destroyed. This container class takes care of this automatically. """ def __init__(self): self.callbacks = [] def _auto_remove(self, method_instance): # Called when weakref detects that the instance on which a method was # defined has been garbage collected. for value in self.callbacks[:]: if isinstance(value, tuple) and value[1] is method_instance: self.callbacks.remove(value) def __contains__(self, value): if self.is_bound_method(value): for callback in self.callbacks[:]: if len(callback) == 3 and value.__func__ is callback[0]() and value.__self__ is callback[1](): return True else: return False else: for callback in self.callbacks[:]: if len(callback) == 2 and value is callback[0]: return True else: return False def __iter__(self): for callback in sorted(self.callbacks, key=lambda x: x[-1], reverse=True): if len(callback) == 3: func = callback[0]() inst = callback[1]() yield partial(func, inst) else: yield callback[0] def __len__(self): return len(self.callbacks) @staticmethod def is_bound_method(func): return hasattr(func, '__func__') and getattr(func, '__self__', None) is not None def append(self, value, priority=0): self.callbacks.append(self._wrap(value, priority=priority)) def remove(self, value): if self.is_bound_method(value): for callback in self.callbacks[:]: if len(callback) == 3 and value.__func__ is callback[0]() and value.__self__ is callback[1](): self.callbacks.remove(callback) else: for callback in self.callbacks[:]: if len(callback) == 2 and value is callback[0]: self.callbacks.remove(callback)