text stringlengths 454 608k | url stringlengths 17 896 | dump stringclasses 91 values | source stringclasses 1 value | word_count int64 101 114k | flesch_reading_ease float64 50 104 |
|---|---|---|---|---|---|
must be imported into your program before you can use it. The simplest use of the regular expression library is the search() function.
The following program demonstrates a trivial use of the search function.
import rehand = open('mbox-short.txt')for line in hand: line = line.rstrip() if re.search('From:', line) : print line to special characters to the search string that allow us to more precisely control which lines match the string. Adding these special characters to our regular expression allow us to do sophisticated matching and extraction while writing very little code.
For example, the caret character is uses in regular expressions to match “the beginning” of a line. We could change our application to only match lines where “From:” was at the beginning of the line as follows:
import rehand = open('mbox-short.txt')for line in hand: line = line.rstrip() if re.search('ˆFrom:', line) : print.
- 瀏覽次數：823 | http://www.opentextbooks.org.hk/zh-hant/ditatopic/6774 | CC-MAIN-2021-17 | refinedweb | 152 | 60.82 |
#include <wx/volume.h>
wxFSVolume represents a volume (also known as 'drive') in a file system under wxMSW.
Unix ports of wxWidgets do not have the concept of volumes and thus do not implement wxFSVolume.
Create the volume object with the given name (which should be one of those returned by GetVolumes()).
Stops execution of GetVolumes() called previously (should be called from another thread, of course).
Create the volume object with the given name (which should be one of those returned by GetVolumes()).
Returns the name of the volume meant to be shown to the user.
Returns the flags of this volume.
See wxFSVolumeFlags enumeration values.
This function is available only when
wxUSE_GUI is
1.
Returns the icon used by the native toolkit for the given file system type.
Returns the kind of this volume.
Returns the name of the volume; this is the internal name for the volume used by the operating system.
Returns an array containing the names of the volumes of this system.
Only the volumes with flags such that the expression
is true, are returned. By default, all mounted ones are returned. See wxFSVolumeFlags enumeration values for a list of valid flags.
This operation may take a while and, even if this function is synchronous, it can be stopped using CancelSearch().
Is this a valid volume?
Returns true if this volume is writable. | https://docs.wxwidgets.org/3.0/classwx_f_s_volume.html | CC-MAIN-2019-18 | refinedweb | 227 | 76.42 |
Firefox java applet fails when the firefox is launched by a java program
RESOLVED WONTFIX
Status
()
▸
Plug-ins
People
(Reporter: walter.garcia@upf.edu, Unassigned)
Tracking
Firefox Tracking Flags
(Not tracked)
Details
User Agent: Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:23.0) Gecko/20100101 Firefox/23.0 (Beta/Release) Build ID: 20130807180628 Steps to reproduce: R Run a simple Java program (a runnable jar file). The Java program code is as following import java.io.IOException; public class Main { public static void main(String args[]) throws IOException { //this website has a small Java applet inside Runtime.getRuntime.exec("firefox"); System.out.println("hello world"); } } My JRE (yes, JRE, not JDK) is Oracle JRE 1.6.0_38. What the program does is essentially opening in Firefox. This website has a small Java applet, so performance won't be a problem. Actual results: What happens then is that, Firefox is opened, is opened, BUT the applet inside the website FAILS to play. The same thing happens if my Java program opens any other website which has applet. However, if I run "firefox" from command line, everything works perfectly. Similarly, if I open Firefox, then I type at the URL field, everything also works perfectly. This problem happens in Firefox, but NOT in Chromium (which I installed from your standard repository). This problem doesn't happen in Oracle Linux distro. I haven't tried in other Linux distros, though. Expected results: Open in Firefox
Component: Untriaged → Plug-ins
Product: Firefox → Core
Priority: -- → P5
I'm marking this bug as WONTFIX per bug #1269807. For more information see -
Status: UNCONFIRMED → RESOLVED
Last Resolved: 11 months ago
Resolution: --- → WONTFIX | https://bugzilla.mozilla.org/show_bug.cgi?id=913857 | CC-MAIN-2018-09 | refinedweb | 276 | 58.18 |
Invalid duplicate class definition
"Invalid duplicate class definition...One of the classes is a explicit generated class using the class statement, the other is a class generated from the script body based on the file name. Solutions are to change the file name or to change the class name. "
When I was first learning Groovy I would get this error from time to time. It was puzzling to me because sometimes I'd get this error, and sometimes it would seem like the same situation and I wouldn't. It seemed quite random to me when the error would crop up. When it did, I'd usually just rename the class and go on, resolving to figure it out later. To save you the trouble, here's what is happening.
Groovy has two ways to treat a .groovy file: either as a script, or as a class definition file. If it is a script you can not have a class by the same name as the file. If it is a class definition file you can. It is very easy to tell whether a .groovy file is going to be treated as a script or as a class definition file. If there is any code outside a class statement in the file (other than imports), it is a script. What is happening is that if there is any code to be executed in the file then Groovy needs a containing class for that code. Groovy will implicitly create a containing class with the name of the file. So if you have a file called Grapher.groovy that has some code in it that isn't inside a class definition, Groovy will create an implicit containing class called Grapher. This means that the script file Grapher.groovy can not itself contain a class called Grapher because that would be a duplicate class definition, thus the error. If, on the other hand, all you do in the file Grapher.groovy is define the class Grapher (and any number of other classes), then Groovy will treat that file as simply a collection of class definitions, there will be no implicit containing class, and there is no problem having a class called Grapher inside the class definition file Grapher.groovy.
It's worth mentioning that the script version of Grapher.groovy will be compiled into a class called Grapher that extends groovy.lang.Script. In the other case, when Grapher.groovy merely defines classes, one of which is Grapher, that Grapher class will be compiled into a class that implements groovy.lang.GroovyObject.
I'm sure this is all explained somewhere in the Groovy documentation, but it didn't soak in to me until I read this Nabble post from which I extracted this explanation.
UPDATE: The text of this error message has changed (at least in some cases) to be a bit more informative. Now it reads:
One of the classes is an explicit generated class using the class statement, the other is a class generated from the script body based on the file name. Solutions are to change the file name or to change the class name.The underlying mechanics behind this error are the same.
Thank you. I've been scratching my head about this too....well no need for that anymore. (June 26, 2009 at 12:41 PM) top
Thanks a lot. I was really confused getting this problem. Your explanation helped. The problem was I occasionally added a line 'package com.xxx.yyy.zzz' outside the class. (May 21, 2010 at 2:34 AM) top
Annoying, by convention nice to have same Class/File name.
Groovy is a great lang, but the gotchas can be a hassle, particulary when error messages have apparently nothing to do with the root cause (not the case here, but in groovlet environment can be completely maddening) (December 20, 2010 at 12:23 PM) top
If you save your file as UTF-8 and accidentally save it with a Byte Order Mark, you will also get this error. Save as UTF-8 without BOM, and you're good. (February 5, 2013 at 10:50 AM) top
Nicely explain,Thanks a lot
This is my mail:
shoaibista@gmail.com
plz ping me,,,i would like to ask much more doubts regarding groovy (May 7, 2013 at 12:57 AM) top
Thank you! I just got this error message.
I had just changed this file from .java to .groovy, so it was very reasonable to believe that there actually was a duplicate class definition sitting somewhere in my state. I spent several minutes trying to track that down, to no avail.
That's what I get for taking a Groovy error message at face value -- I should have just Googled it immediately.
It turns out the culprit was an extraneous } halfway down the page, and Groovy interpreted everything after that as "code outside the class". Of course. (August 25, 2013 at 7:30 PM) top
Nice ! Worth reading !
It's clear and complete.
Thank you very much. (November 10, 2013 at 12:46 PM) top
Thank you for the explanation! It helped to solve my error. (May 8, 2014 at 7:07 AM) top
Was totally confused. Now, no longer. Thank you. (August 9, 2015 at 8:34 PM) top
I lost like an hour on this. One full hour of my life I can never get back.
In my case I had a simple syntax error in the 'imports' section of my Groovy source file. Instead of:
import com.example.myapp.MyObject
I forgot the "import" statement and had:
com.example.myapp.MyObject
Sarcastic "thank you" to the Groovy language/compiler. Sincere thank you to all of my fellow victims above that helped clue me that this was a syntax issue and nothing more. Mehhhh (October 22, 2015 at 9:44 AM) top
What is really confusing is injecting a service into a controller class outside the class name but renaming either the file or the class name does not work. The real solution is to move the def someService back below the class declaration. Thanks to this article for the understanding to guess the solution. (June 21, 2016 at 8:17 AM) top | http://bayesianconspiracy.blogspot.com/2009/03/invalid-duplicate-class-definition.html | CC-MAIN-2019-22 | refinedweb | 1,039 | 73.68 |
jGuru Forums
Posted By:
Andre_TheMunchkin
Posted On:
Tuesday, June 19, 2001 12:21 PM
Solution (well...):
use URLConnection and turn
caching off
Solution problem:
urlconection.getContent() returns an object
that needs be deciphered (it's a gif). i'd rather not write
the gif handler code so....???
What doesn't work:
the existing content handler provided
by sun does work in some browsers (linux netscape, ie5 win98),
but generates security exceptions in others (classloader
can't find the content/image/gif class, it seems).
import sun.net.;
...
// open urlconnection to image file
...
gif gi = new gif();
smImg = createImage((java.awt.image.ImageProducer)gi.getContent(urlconnection));
I've contemplated trying to find, decompile, and reverse
engineer some gif.class code, but that sounds like about
as much of a hassle as translating some C language gif
decoder routines.
anDY
Re: preventing caching with urlconnection and sun.net.
Posted By:
Finlay_McWalter
Posted On:
Wednesday, June 20, 2001 10:04 PM | http://www.jguru.com/forums/view.jsp?EID=441584 | CC-MAIN-2014-52 | refinedweb | 158 | 50.33 |
NAME
libstorage - InterNetNews Storage API library routines
SYNOPSIS
#include "inn/storage.h"); void SMshutdown(void); int SMerrno; char *SMerrorstr; #include "inn/ov.h", char **data, int *len, TOKEN *token); bool OVexpiregroup(char *group, int *lo);. ‘‘OV’’ is a common utility routines for accessing newsgroups and overview data independent of particular overview method. The ‘‘OV’’ function is to isolate the applications from the individual methods. All articles passed through the storage API are assumed to be in wire format. Wire format means ‘‘\CR\LF’’ at the end of lines, ‘‘.’’ at the beginning of lines, and ‘‘.\CR\LF’’ at the end of article on NNTP stream are not stripped. This has a performance win when transferring articles. For the ‘‘tradspool’’ method, wire format can be disabled. This is just for compatibility which is needed by some old tools written for traditional spool. IsToken checks to see if the text is formatted as a text token string. It returns true if formatted correctly or returns false if not. TokenToText converts token into text string for output. false if not. SMinit calls the setup function for all of the configured methods based on SMsetup. This function should be called prior to all other storage API functions which begin with ‘‘SM’’ except SMsetup. It returns true if initialization is successful or returns false if not. SMinit returns true, unless all storage methods fail initialization. SMstore stores an article specified with article. If arrived is specified, SMstore uses its value as article’s arrival time; otherwise SMstore uses the current time for it. SMstore returns token type as type, or returns TOKEN_EMPTY if article is not stored because some error occurs or simply does not match any uwildmat(3) in storage.conf. SMstore fails if SM_RDWR has not been. SMfreearticle frees all allocated memory used by SMretrieve and SMnext. If SMnext will be called with previously returned ARTHANDLE, SMfreearticle should not be called as SMnext frees allocated memory internally. SMcancel removes the article specified with token. It returns true if cancellation is successful or returns false if not. SMcancel fails if SM_RDWR has not beenprintfiles shows file name or token usable by fastrm(8). SMflushcacheddata flushes cached data on each storage method. Type is one of following. SM_HEAD flushes cached header SM_CANCELLEDART flushes articles which should be cancelled SM_ALL flushes all cached data SMshutdown ‘‘ovmethod’’ in inn.conf. Mode is constructed from following. OV_READ allow read open for overview method OV_WRITE allow write open for overview method This function should be called prior to all other OV functions which begin with ‘‘OV’’. OVctl probes or setsSTATICSEARCH if results of OVsearch are stored in a static buffer and must be copied before the next call to OVsearch OVgroupstats retrieves specified newsgroup information from overview method. OVgroupadd informs overview method that specified newsgroup is being added. OVgroupdel informs overview method that specified newsgroup is being removed. OVadd stores an overview data. OVcancel requests the overview method delete overview data specified with token. OVopensearch.) OVsearch retrieves information; article number, overview data, or arrival time. OVsearch ‘‘0’’ if there is no overview data for article. Note that the retrieved data is not neccessarily null-terminated; you should only rely on len octets of overview data being present. OVclosesearch frees all resources which have been allocated by OVopensearch. OVgetartinfo retrieves overview data and token specified with artnum. OVexpiregroup expires overview data for the newsgroup. OVexpiregroup checks the existense of the article and purges overview data if the article no longer exists. If ‘‘groupbaseexpiry’’ in inn.conf is true, OVexpiregroup also expires articles. OVclose frees all resources which are used by the overview method.
HISTORY
Written by Katsuhiro Kondou <kondou@nec.co.jp> for InterNetNews. This is revision 8451, dated 2009-05-07.
SEE ALSO
expire(8), inn.conf(5), storage.conf(5). LIBSTORAGE(3) | http://manpages.ubuntu.com/manpages/maverick/man3/libstorage.3.html | CC-MAIN-2013-48 | refinedweb | 633 | 59.8 |
Try creating a compound index instead of two indexes.
db.collection.ensureIndex( { 'loc':'2d','lastActiveTime':-1 } )
You can also suggest the query which index to use:
db.collection.find(...).hint('myIndexName')
You should use the background option.
db.collection.ensureIndex({ text: 'text', background: true })
From mongodb's documentation:
Builds the index in the background so that building an index does not
block other database activities.
More information here
We use CircleCi for our continuous integration. Circle makes it easy to do
deployment workflows based on the branch that is pushed. Plug for
CircleCi.
We had a Jenkins server, please just quit using it because it was a hassle
compared to a hosted service.
We do a similar process for one of our tool sets. We use the master branch
for development, and the release branch for production. The exception is a
successful test run builds our release branch, not a human clicking a
button. Do something like the following:
Develop your code in a master branch
Changes to your master branch are pushed to your development machines
You create a button that executes an action to merge into the release
branch, and push back to origin:
git fetch origin/release && git rebase origin/master &a
It's probably better to create a local .env file and use Foreman (installed
with the Heroku Toolbelt) to start your application.
You should put this in a local .env file:
DATABASE_URI=mongodb://localhost/minidatabase
and edit your source code to reference the database as:
mongoose.connect(process.env.DATABASE_URI)
Then start you app using:
foreman start
If you don't know how to install foreman, have a look at
Also, remember to run the following command to set the URI in Heroku:
heroku config:set DATABASE_URI=[the mongolab URI goes here]
If you mean how you should implement removal of a comment: you first
retrieve the article document using articleId, and then you can find and
remove the sub-document:
// find article
...
// find the comment by id, and remove it:
article.comments.id(commentId).remove();
// since it's a sub-document of article, you need
// to save the article back to the database to
// 'finalize' the removal of the comment:
article.save(function(err) { ... });
You could write a wrapper, a new module where you store the db instance,
something similar to this:
//db.js
var HOSTNAME = ...
var PORT = ...
var db = module.exports = {};
var instance;
db.connect = function (){
...
instance = <db_instance>;
};
db.disconnect = function (){
...
instance = null;
};
db.instance = function (){
return instance;
};
Now, every time you need the db instance retrieve it by doing:
var db = require ("./path/to/db");
db.instance ();
This is as close as I got without spending a ton of time on it. I don't
know all of the requirements or how much your data varies. For instance, is
recipient always a single value even though it's wrapped in an array?
Regardless, it should be close enough for you to take the rest of the way.
It's grouping by the recipient and the subject, and just sending the
message along. The reduce function creates the container, and for each
message passed in, pushes the value into trainingblock.
db.collection.mapReduce(function() {
var recipient = this.recipient[0];
emit(recipient + "#" + this.subject, { message: this.message });
}, function(key, values) {
var parts = key.split('#'),
recipient = parts[0],
subject = parts[1],
block = { recipient: recipient, subject: subject, t
This is how it works for me,
var connectionString =
"mongodb://username:password@localhost:27017/db_name";
var dbOptions = {
server:{
'auto_reconnect': true,
'poolSize': 20,
socketOptions: {keepAlive: 1}
}
}
// For long running applictions it is often prudent to enable keepAlive.
Without it,
// after some period of time you may start to see "connection closed"
errors for what
// seems like no reason.
MongoClient.connect(connectionString, dbOptions, function(err, db) {
if(err){
console.log(err);
}
app.use(express.session({
store:new mongoStore({db: db}),
secret: 'secret'
}));
})
This works perfectly for me and it will not give you not authorized issues
as well. Previously we don't need to give keepAlive option and it wor
Along the same lines of Dylan's comment, you should probably provide some
more information for an optimal response. In addition to his comments, one
that comes to mind is if you do full scans every time you look for
heartbeats. That is, you could potentially group some nodes in a doc as an
array (or create new collections based on access patterns) and manipulate
in the app layer.
Using xml.etree.ElementTree:
import xml.etree.ElementTree as ET
root = ET.fromstring('''<?xml version="1.0"?>
<BCPFORMAT
...
</BCPFORMAT>''')
# Accessing parent node:
parent_map = {c: p for p in root.getiterator() for c in p} child =
root.find('.//*[@ID="1"]')
print(list(parent_map[child]).index(child)) # => 0
Using lxml:
import lxml.etree as ET
root = ET.fromstring('''<?xml version="1.0"?>
<BCPFORMAT
...
</BCPFORMAT>''')
child = root.find('.//*[@ID="1"]')
print(child.getparent().index(child)) # => 0
Subfields in documents can be accessed by using dot notation (e.g.
outerfield1.innerField).
In Mongo shell you can do this sorting by using:
sort({'outerField1.innerField': -1})
If you are on Python, you may need to write instead:
sort([("outerField1.innerField": -1)])
The reason you need to do this is that Python's dict is an unordered data
structure. For details, see:
Since:
the user would come back from mongo as guestuser
... you should probably put some logging and/or error handling here:
passport.deserializeUser(function(id, done) {
User.findOne(id, function (err, user) {
done(err, user);
});
});
to help you track it down. Since you're just getting started, I'd suggest
you take a look at winston and loggly if you want to explore tooling up
your logging in node/heroku.
Your application looks very "bouncy", to coin a phrase -- lots of redirects
that seem sort of all over the place. Have you looked at fnakstad's
technique for node/angular authentication? (Note that the github page
references two blog posts that explain things). It might give you some
ideas on how to get things under control.
how can I find an item only by the unique_id?
the problem with this question is that the list contains Foo classes and
ordered by priority. this makes it problematic to search for items by
unique_id.
what i'm suggesting is to create a new std::map
std::map <int, foo> uniqueId_To_FooClass;
and when adding to bla a new item, add it to uniqueId_To_FooClass. that way
you can find a foo class by unique_id
I would more like to store the priority in the value (not as the key),
but I don't know how I can sort by value then. Is a std::map the right
class/template for that?
As far as I can remember, std::map will give you the iterator that will go
through the items sorted by the key. Only way to go through the sorted
items by the value, and still use the map, is to rewrite whole
A hack would be to change the order of the levels:
In [11]: g
Out[11]:
Sales
Manufacturer Product Name Product Launch Date
Apple iPad 2010-04-03 30
iPod 2001-10-23 34
Samsung Galaxy 2009-04-27 24
Galaxy Tab 2010-09-02 22
In [12]: g.index = g.index.swaplevel(1, 2)
Sortlevel, which (as you've found) sorts the MultiIndex levels in order:
In [13]: g = g.sortlevel()
And swap back:
In [14]: g.index = g.index.swaplevel(1, 2)
In [15]: g
Out[15]:
Sales
Manufacturer Product Name Product Launch Date
Apple iPod 2001-10-23 34
iPad 2010-04-03
Why not put some satellite data? Instead of sorting the numbers, just sort
pairs of numbers and their indices. Since the sorting is first done on the
first element of the pair, this shouldn't disrupt a stable sorting
algorithm.
For unstable algorithms, this will change it to a stable one.
But note that if you try sorting this way it generates the index while
sorting, not after.
Also, since knowing the permutation index would lead to a O(n) sorting
algorithm, so you cannot do it faster than O(nlogn).
Based on the answer
post.sort(function(a, b){
var keyA = a.date, keyB = b.date;
if(keyA < keyB) return 1;
if(keyA > keyB) return -1;
return 0;
});
console.log (post);
With XSLT 2.0 you could easily do
<xsl:stylesheet version="2.0"
xmlns:
<xsl:output
<xsl:strip-space
<xsl:template
<xsl:copy>
<xsl:apply-templates
</xsl:copy>
</xsl:template>
<xsl:template
<xsl:copy>
<xsl:apply-templates
<xsl:sort
</xsl:apply-templates>
</xsl:copy>
</xsl:template>
<xsl:template
<xsl:copy>
<xsl:apply-templates
<xsl:apply-templates
<xsl:sort
The below code should work
String[][] theArray = new String[][]{{"Enter values","Enter values"}, {more
values, more values}};
Comparator<String[]> comparator= new Comparator<String[]>() {
Arrays.sort(theArray, new Comparator<String[]>(){
@Override
public int compare(String[] o1, String[] o2) {
return o1[0].compareTo(02[0]);
}
});
Arrays.sort(theArray, comparator);
Casting should always try to be avoided. Although it should work to use
your approach (object and cast to string) so you probably have some other
error in your code. Anyway, it's better to explicit tell the compiler that
you are sorting Strings if you know that this is what you will be sorting.
You need to rethink your approach. Where to begin? This is a clear
example, basically of the limits, performance-wise, of the sort of
functional approach you are taking to SQL. Functions are largely planner
opaque, and you are forcing two different lookups on data_table for every
row retrieved because the stored procedure's plans cannot be folded
together.
Now, far worse, you are indexing one table based on data in another. This
might work for append-only workloads (inserts allowed but no updates) but
it will not work if data_table can ever have updates applied. If the data
in data_table ever changes, you will have the index return wrong results.
In these cases, you are almost always better off writing in the join as
explicit, and letting the planner figure out the best way to retri
A trivial way would be:
int firstBurst = jobs[0][2];
jobs[0][2] = Integer.MIN_VALUE;
Arrays.sort(jobs, new Comparator<int[]>(){
public int compare(int[] a, int[] b) {
// don't use subtraction, this can lead to underflows
return a[2] < b[2] ? -1 : (a[2] == b[2] ? 0 : 1);
}
});
jobs[0][2] = firstBurst;
Simply set the burst of the first item to Integer.MIN_VALUE (the integer
equivalent of minus infinity). That way, it guaranteed that the first item
is the smallest, so after sorting it will still be the first element. After
sorting, reset the burst of the first item to its original value.
EDIT
By checking the documentation to verify that Arrays.sort is stable, I
accidentally found the simplest version to solve this problem: use
Arrays.sort(T[] a,
I would suggest you to do two specific things:
Take in consideration the data-set you would need to sort, that usually
helps in sorting faster. (as mentioned in the comment, if its limited range
do a counting sort)
Start using multi-threading (actually called worker threads). YES
JAVASCRIPT DOEST SUPPORT IT NOW. So do a merge sort and start showing
results partially. For more details on how to use multi-threading, refer
worker threads. One good tutorial I can think of is());
}
The solution is to use MongoUpdateStorage:
Works like a charm
You can index some property of each user such as a userId to easily find
all user nodes ()
For the cars, do you simply wish to search for cars of a certain year and
color? Or do you want to use those to do more detailed querying?
If you just want a straight search, then you might consider the index there
as well. Otherwise the year nodes and maybe even color nodes would be the
way I'd do it.
Note that you can use both an index (index on year and color), as well as
the year/color nodes. The index might be helpful to find a starting set of
nodes before you do more involved querying. If color is important in these
queries, then having the car related to a color would be much better than
the color as a property on the car (because the propert
Unfortunately I was unable to solve the problem in Java. However, after
investigating the Python implementation of the API I was able to match the
functions well enough to get by with the PyDoc info.
The original goal was to export Gmail mailbox contents. The solution, in
Python, is as follows:
from gdata.apps.audit.service import AuditService
audit_service = AuditService(domain="example.com")
audit_service.ClientLogin(admin_user, passwd)
audit_service.createMailboxExportRequest(user="target_user")
#check the status
audit_service.getAllMailboxExportRequestsStatus()
As typical Google/gData caveats you need to make sure your user account has
appropriate permissions, you have enabled service, the API is authorized to
use the service, gdata is install, updated, and in your PYTHONPATH, an
You have two errors in your recursive method:
Before calling addAtRec(obj, k, current); you should decrease k by 1, so it
would be better to call k-- before this line.
Once you have reached the base case (when k == 0)and executed the logic to
add the new node, your recursive method must stop, probably with a simple
return; statement. In this case, you're not stopping it so you will call it
every time until get to the end of your list.
Based on these 2 advices, your code should look like:
public void addAtRec(Object obj, int k, ListNode current)
throws ListIndexOutOfBoundsException {
//always use braces even for single line block of code
if(current.nextNode == null && k != 0) {
throw new ListIndexOutOfBoundsException();
}
if(k == 0) {
ListNod
This solved my problems. Its adding another field to index and you can
create facet out of it.
function wtc_glossary_search_api_alter_callback_info() {
$callbacks['wtc_glossary_alter_add_first_letter_title'] = array(
'name' => t('First letter of listing title'),
'description' => t("This module provides first letter of title for
glossary view."),
'class' => 'WtcAlterAddFirstLetter',
);
return $callbacks;
}
/**
* Search API data alteration callback that adds the first letter of title
for glossary mode
*/
class WtcAlterAddFirstLetter extends SearchApiAbstractAlterCallback {
public function alterItems(array &$items) {
foreach ($items as $id => &$item) {
if (!isset($item->FIELD_YOU_NEED)) {
$item->search_api_title_first_lett
From the wiki:
A GestureDetector is an InputProcessor in disguise. To listen for
gestures, one has to implement the GestureListener interface and
pass it to the constructor of the GestureDetector. The detector is
then set as an InputProcessor, either on an InputMultiplexer or as
the main InputProcessor
I admit that is rather dense. But a bit farther down on the wiki you'll
see:
Gdx.input.setInputProcessor(new GestureDetector(new MyGestureListener()));
To rephrase the above in hopefully less dense English: Your GestureHandler
instance is passed to a Libgdx GestureDetector instance. That object will
accumulate "raw" inputs and convert them into higher-level "gestures". To
get the raw inputs, it needs to be installed where the raw inputs will be
delivered to it. The most
Well, I found the answer after speaking to an experienced colleague of
mine. Posting it here for any other beginners who might be getting aboard
my ship:
<table border="5" cellpadding="5" cellspacing="5">
<tr><td>EmpId</td><logic:iterate
<td width="8"><bean:write</td>
</logic:iterate></tr>
<tr><td>Name</td><logic:iterate
<td width="8"><bean:write</td>
</logic:iterate></tr>
</table>
The table tag is used for presentation. The key tag here is the
logic:iteration tag. It helps you iterate through the list(s) you pass to
the page. Two separate logic:iterate tags were created for each list.
In ord
Not possible. Once the app goes into background mode - the screen no longer
belongs to the app - but the home screen (Or whatever other app is on
screen) - also, due to the way apps on iOS devices are sandboxed - this
would be crossing that line..
Why do you need to register taps outside of the app?
Below is the code that uses the event of a point being clicked on a Shield
UI Chart:
events: {
pointSelect: function pointSelectHandler(args) {
var Information= "Point Value: " + args.point.y;
alert(Information);
},
Don't forget to enable the point selection:
enablePointSelection:true,
Is this the actual problem?
Yes. If you access a selected element via bracket notation, you get the raw
DOM element back. DOM elements don't have an animate method. By passing the
DOM element to jQuery again ($($('.myclass')[v])) you are creating a jQuery
object (again).
You can avoid this and use .eq to get a jQuery object for the element at
that index:
$('.myclass').eq(v);
It would be better to keep a reference to the selected elements outside the
loop though:
var indizes = [0, 3];
var $elements = $('.myclass');
$.each(indizes, function(i, v) {
$elements.eq(v).animate({
left: 100
}, 1000);
});
Alternatively, you can use .filter to filter out the elements you want to
animate, which at least looks a bit more concise:
$('.myclass').filter(function(i) {
retu
In Neo4j 1.9 you can't. In Neo4j 2.0, you can get the Labels and Indexes on
a node, e.g.
CREATE n:Person{name:'Jim'}
Make an index
CREATE INDEX on :Person(name)
List labels
MATCH n
RETURN LABELS(n) | http://www.w3hello.com/questions/Node-Mongo-Utilizing-an-index-in-sorting | CC-MAIN-2018-17 | refinedweb | 2,876 | 55.74 |
A vulnerability was recently discovered in the doorkeeper gem. It taught me the hard way how to deal with security issues in OSS, and I documented what I’ve learned in the process.
Keep it private at first
When you become aware of a vulnerability in a project you maintain, keep it private. A vulnerability shouldn’t be made public until it’s been fixed.
Rails developers may have experienced this: a new patch version of Rails is suddenly announced, and you should upgrade. What happens in such cases is:
- Someone discovers a vulnerability.
- Rails core team work with the discoverer to fix it.
- A new patch version is released for affected Rails versions.
- The vulnerability and how to upgrade is announced in Ruby on Rails Twitter account, their blog, security mailing lists, etc.
Let’s see in detail how all this happens.
Request a CVE identifier
Before you start fixing the bug (or while you are doing it) you should request a “CVE id”. An id can be requested from any of the “CVE Numbering Authorities”. I myself sent my request to RedHat, and Kurt Seifried provided me with an identifier in less than an hour. He hosts a wiki with more information.
CVE stands for “Common Vulnerabilities and Exposures”. This allows us to sanely talk about security issues (“issue CVE-2009-3555” instead of “the OpenSSL vulnerability, from like 2009, the DoS one… no, not that one”). CVE allows multiple vendors, products, and customers to properly track security vulnerabilities and make sure they are dealt with. CVE Identifiers are from an international information security effort that is publicly available and free to use.
Write the accompanying CVE report
The CVE report specifies:
- The project (name and related links)
- A description of the vulnerability
- Affected and fixed versions
- What’s the vulnerability’s impact (how many people are affected and how)
- What is the upgrade process
- What workarounds can users take, if any
- Credits
- Any other kind of relevant information you can provide
Here is my example:
Cross-site request forgery (CSRF) vulnerability in doorkeeper 1.4.0 and earlier allows remote attackers to hijack the user's OAuth autorization code. This vulnerability has been assigned the CVE identifier CVE-2014-8144. Versions Affected: 1.4.0 and below Fixed Versions: 1.4.1, 2.0.0 ## Impact Doorkeeper's endpoints didn't have CSRF protection. Any HTML document on the Internet can then read a user's authorization code with arbitrary scope from any Doorkeeper-compatible Rails app you are logged in. ## Releases The 1.4.1 and 2.0.0 releases are available at and. ## Upgrade Process Upgrade doorkeeper version at least to 1.4.1. ## Workarounds There are no feasible workarounds for this vulnerability. ## Credits Thanks to Sergey Belov of DigitalOcean for finding the vulnerability, Phill Baker of DigitalOcean for reporting and fixing it, and to Egor Homakov of Sakurity.com for raising awareness.
Fix it and publish releases
Work on the vulnerability in private. Only publish the fixes when you release new patched versions of your project. This keeps people from learning about the vulnerability before it’s been fixed, potentially taking advantage from affected deploys of your software. The goal is to reach most users of your project so they can upgrade as soon as possible.
If you take too long to release, the attacker might announce it before you have a fix ready. The person who reported the vulnerability is the “white hat”. There may already be “black hats” taking advantage of it. CVE reports typically go public after 2 weeks since an id was granted to address this issue.
Spread the word
After you get the CVE identifier and report, the fix and releases ready, publish this information to security lists and to users of your library, as widely as you’re able to, using any communication techniques available to you.
I was advised to post doorkeeper’s report to the following lists:
- oss-security@lists.openwall.com Mailing list.
- ruby-security-ann Google Group.
- ruby-advisory-db GitHub project.
After all is done you can relax. Until next time!
One thing end users can do
Adding the report to
ruby-advisory-db is particularly useful for end users.
Ruby developers can use
bundler-audit, which uses
ruby-advisory-db, to
automatically alert themselves of security issues. We made
bundle-audit a
dependency in all our Rails apps by adding it to
Suspenders in version 1.19.0.
Here’s how it is set up in our Rails apps:
# Gemfile group :development do gem "bundler-audit" end # Rakefile task default: "bundler:audit" # lib/tasks/bundler_audit.rake if Rails.env.development? || Rails.env.test? require "bundler/audit/cli" namespace :bundler do task :audit do %w(update check).each do |command| Bundler::Audit::CLI.start [command] end end end end
We hook the rake task into our default test suite so that we are sure it is run often. Running this task in an app using an insecure version of Doorkeeper would print out something like:
Name: doorkeeper Version: 1.3.0 Advisory: CVE-2014-8144 Criticality: Unknown URL: Title: Cross-site request forgery (CSRF) vulnerability in doorkeeper 1.4.0 and earlier. Solution: upgrade to ~> 1.4.1, >= 2.0.0
Things not to do
Don’t do what I did with Doorkeeper:
- Tell the person who reports the vulnerability to send a pull request, which is public.
- Wait until the next scheduled release to bump the patch version with the fix.
Instead:
- Keep it private, following the guidelines described above, making it public only after the fix is released.
- Release the fix as soon as possible.
A week after the Doorkeeper fix, Egor Homakov raised awareness, calling users to upgrade, and myself to finally release. Thank you Egor for your pat on the back that morning and for the ongoing help you are providing. | https://thoughtbot.com/blog/handling-security-issues-in-open-source-projects | CC-MAIN-2019-47 | refinedweb | 982 | 57.37 |
02 April 2009 22:54 [Source: ICIS news]
HOUSTON (ICIS news)--US polyethylene (PE) and polypropylene (PP) exports have gotten a strong boost over the past two weeks from surging demand in Asia, especially China, traders and sellers said on Thursday.
“The market is changing on an hourly basis,” a producer said.
The source said PP demand in particular has benefited from the Chinese government’s policy to promote the proliferation of electrical appliances in rural areas, which has caused appliance manufacturers in the region to work overtime.
US PP export prices rose 2-4 cents/lb ($44-88/tonne or €33-67/tonne) over the past 10 days, according to traders.
Bagged homopolymer resin was at 37-38 cents/lb FOB (free on board) US Gulf, up from the lowest deals heard at 34 cents/lb FOB in previous weeks, sources said.
Further increases on PP exports appeared likely as sellers raised the price tag on remaining inventory.
New offers of homopolymer PP have emerged at 38-39 cents/lb FOB ?xml:namespace>
“This is 100% due to
PE exports were also trending up, with high density PE (HDPE) for blow moulding quoted at around 38 cents/lb FOB US Gulf in bags, up 2-3 cents from March levels.
Further increases were likely on PE as well, but there was no firm consensus as to how long the current Asia-driven price spike would hold.
“I think this
Another trader agreed with that assessment, saying the export boom could last as little as one week or as long as four weeks.
“I don’t think it will last two months,” he said.
The effect on domestic prices was not yet clear either, but sellers said it was unlikely they would lower contract prices when export business was so strong.
Major
($1 = €0.76) | http://www.icis.com/Articles/2009/04/02/9205661/us-polyolefins-exports-building-on-asia-demand.html | CC-MAIN-2015-18 | refinedweb | 305 | 66.17 |
Hello,
I have a pololu maestro 18 channel servo controller . I have written a python code to control the maestro via pc using serial bluetooth HC 05
first of all I have made a library like this
code /
class Controller : self.PololuCmd = "0xaa" + "0xc" def setTarget(self, chan, target): lsb = target & 0x7f #7 bits for least significant byte msb = (target >> 7) & 0x7f #shift 7 and take next 7 bits for msb # Send Pololu intro, device number, command, channel, and target lsb/msb cmd = self.PololuCmd + "0x04" + hex(chan) + hex(lsb) + hex(msb) self.usb.write(cmd)
/ code
it is not full code but a logic
this seems to work
when i use this library
like this
code /
import maestro maestro = maestro.Controller() maestro.setTarget(0,9000) #moves to max angle
/ code
but when i want to set targets at multiple channels
like this
code/
import maestro maestro = maestro.Controller() maestro.setTarget(0,9000) #moves to max angle maestro.setTarget(1,9000)
/code
this will successfully set servo 1 to max angle
but servo 0 will not have any effect
like it haven’t received any commands for servo 0
and this also happens vice versa
maestro.setTarget(1,9000) #moves to max angle maestro.setTarget(0,9000)
in this case servo 0 moves to max angle
i think the problem is somewhere in sending some bytes
i mean it only understands ending commands
also there is a serial error of 0x0010 this lights up red led always
i have used set multiple target command and it also works for both servo
but set target method doesnt work for me
any help would be appreciated | https://forum.pololu.com/t/controlling-two-servos-at-same-time-problem/12282 | CC-MAIN-2018-13 | refinedweb | 274 | 54.97 |
Hello, I'm a new baby to coLinux. Thanks for your wonderful job! now, i have some trouble in creating disk image! here is my steps: first, I installed a xubuntu system in VirtualBox(my host OS is Windows7) second, I turn off the Virtual machine, using VBoxManage.exe covert the "vdi" disk file to RAW format third, I use dd coverting the RAW image with para "bs=512 skip=63" could you tell me whether these steps has something wrong? And How to make rootfs image? wish to hear from you ! Happy new year! Err info: VFS: Mounted root (ext2 filesystem) on device1:0. =========================================================================== # This process will install (if necessary) the coLinux modules for the # coLinux kernel. input: AT Translated Set 2 keyboard as /devices/serio0/input/input0 =========================================================================== Determining /, Found. Mounting / EXT3-fs (cobd0): error: can't find ext3 filesystem on dev cobd0. EXT2-fs (cobd0): error: can't find an ext2 filesystem on dev cobd0. EXT4-fs (cobd0): VFS: Can't fi nd ext4 filesystem ISOFS: Unable to identify CD-ROM format. mount: Mounting /dev/cobd0 on /mnt/linux failed: Invalid argument List of all partitions: 7500 4194272 cobd0 (driver?) No filesystem could mount root, tried: ext3 ext2 ext4 iso9660 Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(117,0)Pid: 1, comm: swapper Not tainted 2.6.33.5-co-0.7.8 #1 Call Trace: [<c122f6af>] ? printk+0x18/0x21 [<c122f681>] panic+0x4e/0x64 [<c12eea9c>] mount_block_root+0x242/0x254 [<c108dac7>] ? sys_mknod+0x27/0x30 [<c12ee0c7>] ? kernel_init+0x0/0xea [<c12eeb07>] mount_root+0x59/0x5f [<c12ef66f>] initrd_load+0x277/0x38c [<c12ee0c7>] ? kernel_init+0x0/0xea [<c12eebcb>] prepare_namespace+0xbe/0x183 [<c1080d40>] ? sys_access+0x20/0x30 thanks | https://sourceforge.net/p/colinux/mailman/attachment/4D505209.3070902@henry.ne.arcor.de/1/ | CC-MAIN-2016-30 | refinedweb | 275 | 60.01 |
The C++ function std::unordered_map::cend() returns a constant iterator which points to past-the-end element of the unordered_map.
Iterator obtained by this member function can be used to iterate container but cannot be used to modify the content of object to which it is pointing even if object itself is not constant.
Following is the declaration for std::unordered_map::cend() function form std::unordered_map header.
const_iterator cend() const noexcept;
None
Returns a constant iterator.
This member function never throws exception.
Constant i.e. O(1)
The following example shows the usage of std::unordered_map::cend() function.
#include <iostream> #include <unordered_map> using namespace std; int main(void) { unordered_map<char, int> um = { {'a', 1}, {'b', 2}, {'c', 3}, {'d', 4}, {'e', 5} }; cout << "Unordered map contains following elements: " << endl; for (auto it = um.cbegin(); it != um.cend(); ++it) cout << it->first << " = " << it->second << endl; return 0; }
Let us compile and run the above program, this will produce the following result −
Unordered map contains following elements: e = 5 d = 4 c = 3 b = 2 a = 1 | https://www.tutorialspoint.com/cpp_standard_library/cpp_unordered_map_cend_container.htm | CC-MAIN-2021-17 | refinedweb | 175 | 51.38 |
Neutron/VendorSplitPackaging
Contents
Neutron Vendor split: Packager perspective
Neutron has started on the route towards splitting vendor libraries from its tree into separate, vendor governed, repositories. Now vendors are in charge of releasing python (pypi) packages to the public to consume. Recent observations show that in some cases python packages released by vendors are not optimal for distribution consumption. Below you can see some random notes on how to help packagers to do their work by enhancing vendor library package contents.
Oslo modules
Neutron maintains some of oslo modules copied from oslo-incubator in neutron.openstack.common.* namespace. The tree is not supposed to be used by any code outside neutron core and *aas repositories. That said, multiple (if not most) vendor libraries still rely on the code in the tree.
The solution from vendor side would be one of the following:
- stop using neutron.openstack.common.* modules and instead maintain the needed modules inside the vendor library. More details in the corresponding oslo policy.
- switch to using oslo.* libraries that are graduating from oslo-incubator.
- pinning neutron revision in requirements.txt to avoid breakages at random time. Note that in this case vendors should also make sure they update the pin quite frequently to keep up with neutron development. Note that revision pinning has its own benefits beyond oslo modules' consumption.
PyPI package contents
It's observed that some vendor libraries already released at PyPI miss some files that are present in corresponding git repositories, and are useful for packaging. Specifically, the following files should be present in PyPI packages to help packagers:
- requirements.txt and test-requirements.txt: used to refer to dependencies that vendors expect to be used with their libraries.
- LICENSE file: most distributions have a requirement that license text must be included in their packages.
- all files needed to run tests without virtualenv: packages may run unit tests as part of their build process.
Dependencies
- Dependencies are communicated to packagers via requirements.txt and test-requirements.txt, so those files should contain proper and full dependencies that are needed in runtime and when running unit tests. It's adviced that vendor libraries do not rely on neutron dependency to fetch all the needed dependencies for them, and instead put packages that are explicitly used in their code into those files. The rationale behind it is that neutron may drop some of those dependencies later, and vendor libraries will become broken.
- Be sure that your dependencies fit with the dependencies listed in the requirements repository (see global-requirements.txt and test-requirements.txt) for the OpenStack version you want to support (i.e. for Kilo, use the dependencies from the Kilo branch in that repository). That is important to avoid conflicts between the overall OpenStack requirements and the requirements for your project and makes life for distributions much easier. Note that you may also add your stackforge project to projects.txt file to get automatic updates to your requirements.
Git repository
It's also advised that vendor libraries maintain a public git repository with the code in addition to PyPI releases. It's useful in cases where distributions run packaging CI against master (for reference, RDO for master called Delorean does it). | https://wiki.openstack.org/wiki/Neutron/VendorSplitPackaging | CC-MAIN-2022-40 | refinedweb | 536 | 57.06 |
Odoo Help
Odoo is the world's easiest all-in-one management software. It includes hundreds of business apps:
CRM | e-Commerce | Accounting | Inventory | PoS | Project management | MRP | etc.
Function field Save in database
In Openerp 7 Functional field added with store in Database
I used the below code:-
def _concat_attached_file(self, cr, uid, ids, name, arg, context=None): res = {} attachment_obj = self.pool.get('ir.attachment') for id in ids: procedure_attachment = attachment_obj.search(cr, uid, [('res_model', '=', 'model_name'), ('res_id', '=', id)], context=context) filelist = [] for val in attachment_obj.browse(cr, uid, procedure_attachment): filelist.append(val.name) filevalue = ', '.join(filelist) if not filevalue: filevalue = None res[id] = filelist return res 'attached_file' : fields.function(_concat_attached_file, method=True, store=True, string='Attached File Name', type='char'),
But the above functional fields generate output only with Store=False. Suppose used Store=True output not generated. I want to store the db also how to fix ?Thanks
'store'=True will store the value of the field in database. Once stored then the functional fields function will not be executed again. if 'store'=False, then every time(any change in the record) the functional field will be executed
But if the value of 'store' is a dictionary then (key of the dictionary will be a model name and value will a tuple with list of ids, list of field name and 10-i dont know ) any change/update in the model specified as the key of the dictionary and change/ update is in the ids specified in the tuple and change or update is in the field names specified in the list then the function of the functional field will be loaded and new data will be saved in database
Hi Omal Bastin, Thanks for reply I have another doubt In openerp 7 i want to shows attached file shows in list view and store in DB. After attached the file [using default options Wizard Attachments Add] method the function not called. So write method overridden and called the function. But how to shows the attachment document after attached to call function automatically [without using write method override] Thanks
Hi, goto res.company model and check how the image is saved and showed in the ERP. There you can find a better example for what your are looking for
About This Community
Odoo Training Center
Access to our E-learning platform and experience all Odoo Apps through learning videos, exercises and Quizz.Test it now | https://www.odoo.com/forum/help-1/question/function-field-save-in-database-35002 | CC-MAIN-2017-43 | refinedweb | 407 | 60.14 |
Trying to retrieve the page source from a website, I get a completely different (and shorter) text than when viewing the same page source through a web browser.
Python: Getting a wrong source code of the web page (asp.net)
This fellow has a related issue, but obtained the home page source instead of the requested one - I am getting something completely alien.
The code is:
from urllib import request
def get_page_source(n):
url = '' + str(n) + '/live'
response = request.urlopen(url)
return str(response.read())
n = 1006233
text = get_page_source(n)
b' src="/_Incapsula_Resource?CWUDNSAI=24&
xinfo=0-12919260-0 0NNY RT(1462118673272 111) q(0 -1 -1 -1) r(0 -1)
B12(4,315,0) U2&incident_id=276000100045095595-100029307305590944&edet=12&
cinfo=04000000" frameborder=0Request unsuccessful. Incapsula incident ID:
276000100045095595-100029307305590944</iframe></body></html>'
There is a couple of issues here. The root cause is that the website you are trying to scrape knows you're not a real person and is blocking you. Lots of websites do this simply by checking headers to see if a request is coming from a browser or not (robot). However, this site looks like they use Incapsula, which is designed to provide more sophisticated protection. You can try and setup your request differently to fool the security on the page by setting headers - but I doubt this will work.
import requests def get_page_source(n): url = '' + str(n) + '/live' headers = {'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36'} response = requests.get(url, headers=headers) return response.text n = 1006233 text = get_page_source(n) print text
Looks like the site also uses captchas - which are designed to prevent web scraping. If a site is trying this hard to prevent scraping - it's likely because the data they provide is proprietary. I would suggest finding another site that provides this data - or try and use an official API.
Check out this () answer from a while back. It looks like the whoscored.com uses the OPTA API to provide info. You may be able to skip the middleman and go straight to the source of the data. Good luck! | https://codedump.io/share/8nfPZTehKlLc/1/getting-wrong-page-source-when-calling-url-from-python | CC-MAIN-2017-13 | refinedweb | 368 | 66.54 |
I've seen similar questions to this one but none of them really address the trackback.
If I have a class like so
class Stop_if_no_then():
def __init__(self, value one, operator, value_two, then, line_or_label, line_number):
self._firstvalue = value_one
self._secondvalue = value_two
self._operator = operator
self._gohere = line_or_label
self._then = then
self._line_number = line_number
def execute(self, OtherClass):
"code comparing the first two values and making changes etc"
`Syntax Error (Line 3): No -THEN- present in the statement.`
You can use a
try: and then
except Exception as inst:
What that will do is give you your error message in a variable named inst and you can print out the arguments on the error with
inst.args. Try printing it out and seeing what happens, and is any item in
inst.args is the one you are looking for.
EDIT Here is an example I tried with pythons IDLE:
>>> try: open("epik.sjj") except Exception as inst: d = inst >>> d FileNotFoundError(2, 'No such file or directory') >>> d.args (2, 'No such file or directory') >>> d.args[1] 'No such file or directory' >>>
EDIT 2: as for closing the program you can always
raise and error or you can use
sys.exit() | https://codedump.io/share/EMnuYdQZ29zv/1/in-python-how-do-i-print-an-error-message-without-printing-a-traceback-and-close-the-program-when-a-condition-is-not-met | CC-MAIN-2017-34 | refinedweb | 201 | 67.55 |
Blur image effect in Actionscript 3
I previously created a Blur image effect tutorial that used the timeline to create the blurred effect. This time I will create the same effect using only Actionscript 3.0. You will need the tweenMax plug-in for this tutorial which can be downloaded from: blog.greensock.com/tweenmax/.
Blur image effect in Actionscript 3
Step 1
Open a new Flash AS3 file.
Import the image you wish to add the blur effect to by selecting File > Import > Import to Stage. You can alternatively create your own objects on the stage. I have used a free stock image that can be found at.
Step 2
Convert your image into a movie clip (F8) and give your movie clip an appropriate instance name. I have used the instance name: car_mc.
Step 3
Add a button component onto the stage by selecting Window > Component and drag the button component onto the stage. Then give the button the instance name: my_button. For more information on the button component take a look at this tutorial.
Step 4
On the timeline insert a new layer called Actions. Then open up the Actions panel and enter the following code.
//Imports the packages needed.For more information on the blur filter checkout the AS3 blur filter Component reference.
import flash.filters.BlurFilter;
import gs.*;
import gs.easing.*;
//Creates a new instance of the blur filter and adds the filter to the
//car movie clip.
var my_bf:BlurFilter=new BlurFilter(150,150,1);
car_mc.filters=[my_bf];
//Adds 'play' to the button label.
my_button.label="Play";
//Add an event listener with the mouse click event to the button.
my_button.addEventListener(MouseEvent.CLICK, blurEffect);
//Add the blur effect to the car and replay the effect.
function blurEffect(event:MouseEvent):void {
if (my_button.label=="Play") {
TweenMax.to(car_mc, 1, {blurFilter:{blurX:0, blurY:0, quality:1}});
my_button.label="Replay";
}
else{
car_mc.filters=[my_bf];
my_button.label="Play";
}
}
Step 5
Test your blur image effect Ctrl + Enter.
You should now have a blur image effect in Actionscript 3.0. | http://www.ilike2flash.com/2009/10/blur-image-effect-in-actionscript-3.html | CC-MAIN-2017-04 | refinedweb | 342 | 60.11 |
NEW: Learning electronics? Ask your questions on the new Electronics Questions & Answers site hosted by CircuitLab.
Project Help and Ideas » Serial comunication between Nerdkit and an MS Excel sheet
I have decided to post a more detailed example of a direct serial link between the Nerdkit and an excel sheet. I do not feel qualified to give an explanation of the possible advantages and disadvantages of this type of Nerdkit/PC link. I have used this type of two way serial link many of times to achieve precise PC control of up to 3 stepper motors at once and relay circuit control. The inherent power of excel makes data processing charting and other detailed analyses quick and relatively simple on data received from the Nerdkit. In fact this type of serial connection is just about the only system I use to set up a link between the PC and the Nerdkit, It has been fast enough for my needs.
First set up the VBA communication modules. Follow the detailed instructions at.
link to VBA serial communication setup
You should not be concerned about the VBA code. In most cases it will not be necessary to have an understanding of any of that code. Once you have the modules in place they work in the background you don’t even know that they are there. You will likely have to have a basic understanding of basic VBA coding to write modules to make efficient use of the connection.
I would recommend one simple addition to the existing modules to help you confirm that the serial connection is open and ready to send/receive data.
Place the one additional line of code in module that opens the com port.
Private Sub CommandButton1_Click()
Dim intPortID As Integer ' Ex. 1, 2, 3, 4 for COM1 - COM4
Dim lngStatus As Long
intPortID = 4
' Open COM port
lngStatus = CommOpen(intPortID, "COM" & CStr(intPortID), _
"baud=115200 parity=N data=8 stop=1")
Worksheets("Sheet1").Cells(4, 1).Value = lngStatus
End Sub
I have found that it is very useful to have an on the sheet indication of the state of the serial connection. Often after working in the VBA editor or using cmd.com to change coding on the micro it is necessity to re-establish the serial connection to your Excel sheet. In order to see what the current state of the connection is, click the open connection button on “sheet1” of the excel workbook. The number displayed in cell A4 will indicate if the com connection is actually open. You are looking for a 0 (zero) which should mean that excel is ready to read/write on the serial. A -1 can usually be corrected by clicking the close button then clicking the open button again. A 5 indicates that excel thinks the com port is in use by someone else and can be corrected by unplugging the Nerdkit USB cable from the PC for a second.
After you have established a viable serial connection between the PC and the nerd kit you may wish to try it out. I am posting some code to set up a simple two way data swap between the Nerdkit and the PC. I wanted to post an example that would clearly demonstrate that the connection is bidirectional and that the data from the Nerdkit is ending up in the excel cells that we intended. Using VBA to place incoming data in appropriate cells and the built in features of excel almost anything is possible. From this example I hope that it will be apparent how powerful and useful this type of connection could be.
Place this VBA code in a module that can be triggered from a control button on “sheet1’. Ether make a new control button or use your existing read button.
Private Sub CommandButton3_Click()
Dim intPortID As Integer ' Ex. 1, 2, 3, 4 for COM1 - COM4
Dim lngStatus As Long ' indicates if operation was successful
Dim strData As String ' data recieved from serial connection
Dim strData2 As String 'data held for sheet
Dim row_top As Integer
Dim col_left As Integer
Dim row_number As Integer
Dim col_number As Integer
Dim row_count As Integer
Dim col_count As Integer
Dim i As Integer
intPortID = 4 'set com port id
row_top = 12
col_left = 2
Worksheets("Sheet1").Cells(6, 2).Value = " Enter table size that you wish to fill with data from Nerdkit"
Worksheets("Sheet1").Cells(7, 3).Value = "Enter the number or rows in table into cell A7"
Worksheets("Sheet1").Cells(8, 3).Value = "Enter the number of columns in table into cell A8"
While ((Worksheets("Sheet1").Cells(7, 1).Value * Worksheets("Sheet1").Cells(8, 1).Value) < 2)
Worksheets("Sheet1").Cells(10, 4).Value = "Enter data table size - click start button"
Exit Sub
Wend
row_number = Worksheets("Sheet1").Cells(7, 1).Value
col_number = Worksheets("Sheet1").Cells(8, 1).Value
For i = 1 To 10
lngStatus = CommRead(intPortID, strData, 6) 'clear the buffer
Next i
Worksheets("Sheet1").Cells(10, 4).Value = "Waiting for data from NerdKit"
lngStatus = CommWrite(intPortID, "s") 'send a charitor "s" to micro to indicate ready to recieve serial data
While (Len(strData2) < 6) ' do we have 6 character to work with
lngStatus = CommRead(intPortID, strData, 6) 'read 6 character from com 4 store in strData
strData2 = strData2 & strData 'append serial read to previous read data
Wend 'end while- we do have at least 6 char
Worksheets("Sheet1").Cells(10, 4).Value = "Recieveing data from NerdKit"
For col_count = col_left To (col_left + col_number - 1) ' row selection on sheet
For row_count = row_top To (row_top + row_number - 1) 'col selection on sheet
If (Left(strData2, 6) = " stop") Then
Worksheets("Sheet1").Cells(10, 4).Value = "Data transmitions was stopped by the NerdKit- table size to large- resize table "
Exit Sub
End If
Worksheets("Sheet1").Cells(row_count, col_count).Value = Val(Left(strData2, 6)) 'val of the left 6 char to sheet
strData2 = Right(strData2, (Len(strData2) - 6)) ' remove left 6 char from string
While (Len(strData2) < 6) ' do we have 6 char to work with
lngStatus = CommRead(intPortID, strData, 6) 'read 6 character from com 4 store in strData
strData2 = strData2 & strData 'append serial read to previous read data
Wend
Next row_count 'next row
Next col_count 'next col
lngStatus = CommWrite(intPortID, "t")
Worksheets("Sheet1").Cells(10, 4).Value = "Data transmit ion completed table is full"
End Sub
Use this code on the Nerdkit.
#define F_CPU 14745600
#include <avr/io.h>
// #include <avr/pgmspace.h> // needed for PSTR with printf_P
#include "../libnerdkits/uart.h"
#include <avr/interrupt.h>
#include <avr/pgmspace.h>
#include <util/delay.h>
#include <inttypes.h>
#include <stdlib.h>
#include <string.h>
#include "../libnerdkits/uart.h"
#include "../libnerdkits/delay.h"
#include "../libnerdkits/lcd.h"
int main() {
char incoming;
uart_init();
FILE uart_stream = FDEV_SETUP_STREAM(uart_putchar, uart_getchar, _FDEV_SETUP_RW);
stdin = stdout = &uart_stream;
lcd_init();
FILE lcd_stream = FDEV_SETUP_STREAM(lcd_putchar, 0, _FDEV_SETUP_WRITE);
uint16_t i;
while(1) {
lcd_line_one();
lcd_write_string(PSTR(" NerdKit - Excel "));
lcd_line_two();
lcd_write_string(PSTR(" Waiting for start "));
lcd_line_three();
lcd_write_string(PSTR(" code from Excel "));
incoming= uart_read(); // read data from uart
while (incoming != 115) { // is the incoming data a "s"
incoming= uart_read(); // read data from uart
} // end while - when excel is ready for data
lcd_line_two();
lcd_write_string(PSTR(“Transmiting data "));
lcd_line_three();
lcd_write_string(PSTR(" "));
while(uart_char_is_waiting()){ // clear the uart
incoming = uart_read();
}
for (i=1;i<=25000;i++) { //count to 25000
while(uart_char_is_waiting()){ //check for new char
incoming = uart_read();
while(incoming == 116){ //is new char a "t"
lcd_line_two();
lcd_write_string(PSTR("Excel table is full"));
lcd_line_three();
fprintf_P(&lcd_stream, PSTR("last value %6d"),i);// print to LCD
lcd_line_four();
lcd_write_string(PSTR("power down to start"));
while(1){ //table is full do nothing just show message
}
} //while char is a "t"
} //while there is a char waiting
lcd_line_three();
fprintf_P(&lcd_stream, PSTR("sending %6d "),i);// print to LCD
printf_P(PSTR("%6d"), i); //send count number to serial
}//next i
lcd_line_two();
lcd_write_string(PSTR(" Excel table large "));
lcd_line_three();
fprintf_P(&lcd_stream, PSTR("Tx stopped at%6d"),i);// print to LCD
printf_P(PSTR(" stop")); //send count number to serial
lcd_line_four();
lcd_write_string(PSTR("power down to start"));
while (1){
} //do nothing just show mmessage
}
return 0;
}
Connect Nerdkit USB cable to the PC. Make sure the com port matches up with the VBA code(com4). Power up the Nerdkit. Click the open port button on the excel sheet. Check that you have a 0 in cell A4. Click on the command button that will run the VBA module listed here. Enter the desired table size for the first test (cells A7 and A8), don’t forget to press enter after change. Click on the command button to run the VBA code and hopefully receive the data from the Nerdkit.
This setup is working on my system. There is a small glitch that occurs when my VBA module ends I am looking into the exact nature of that odd result.
I would be very interested in knowing if anyone gets this working.
Darryl
I noticed the link I posted does not work, I will try again.
link to Serial port communication in excel
Darryl
Thanks for the link to SERIAL PORT COMM.
This may be the basis I am looking for to merge data from
my TRACTOR PULL SLED to a spread sheet for each puller.....
via a radio link from the sled..
Jim
I should also mention that you will have to change the baud rate in the excel VBA code to match up with the Nerdfkit baud rate. The baud rate is set in the open port sub listed above. You must simply change the baud=9600 to baud=115200.
Please log in to post a reply. | http://www.nerdkits.com/forum/thread/1880/ | CC-MAIN-2018-09 | refinedweb | 1,584 | 60.45 |
This illustrates that union members shares memory and that struct members does not share memory.
#include <stdio.h> #include <string.h> union My_Union { int variable_1; int variable_2; }; struct My_Struct { int variable_1; int variable_2; }; int main (void) { union My_Union u; struct My_Struct s; u.variable_1 = 1; u.variable_2 = 2; s.variable_1 = 1; s.variable_2 = 2; printf ("u.variable_1: %i\n", u.variable_1); printf ("u.variable_2: %i\n", u.variable_2); printf ("s.variable_1: %i\n", s.variable_1); printf ("s.variable_2: %i\n", s.variable_2); printf ("sizeof (union My_Union): %i\n", sizeof (union My_Union)); printf ("sizeof (struct My_Struct): %i\n", sizeof (struct My_Struct)); return 0; }
Some C implementations permit code to write to one member of a union type then read from another in order to perform a sort of reinterpreting cast (parsing the new type as the bit representation of the old one).
It is important to note however, this is not permitted by the C standard current or past and will result in undefined behavior, none the less is is a very common extension offered by compilers (so check your compiler docs if you plan to do this).
One real life example of this technique is the "Fast Inverse Square Root" algorithm which relies on implementation details of IEEE 754 floating point numbers to perform an inverse square root more quickly than using floating point operations, this algorithm can be performed either through pointer casting (which is very dangerous and breaks the strict aliasing rule) or through a union (which is still undefined behavior but works in many compilers):
union floatToInt { int32_t intMember; float floatMember; /* Float must be 32 bits IEEE 754 for this to work */ }; float inverseSquareRoot(float input) { union floatToInt x; int32_t i; float f; x.floatMember = input; /* Assign to the float member */ i = x.intMember; /* Read back from the integer member */ i = 0x5f3759df - (i >> 1); x.intMember = i; /* Assign to the integer member */ f = x.floatMember; /* Read back from the float member */ f = f * (1.5f - input * 0.5f * f * f); return f * (1.5f - input * 0.5f * f * f); }
This technique was widely used in computer graphics and games in the past due to its greater speed compared to using floating point operations, and is very much a compromise, losing some accuracy and being very non portable in exchange for speed. | https://sodocumentation.net/c/topic/7645/unions | CC-MAIN-2021-21 | refinedweb | 384 | 65.52 |
I can't seem to figure this out, though it should be pretty simple. Here are my instructions:
Compound interest
Compound.cpp
Write a program that calculates compound interest. The program should ask the user for the starting dollar amount and the daily increase (as a percentage), and the number of days. A loop should then be used to display the day, the amount of interest earned on that day and the account balance on that day. The program should also display the total interest earned. Output should be as follow
Initial amount in dollars? 100
Interest rate in percentage? 10
Number of days? 3
Day Earned interest Balance
-----------------------------------------------
1 $10 $110.00
2 $11 $121.00
3 $12.10 $133.10
Total Interest earned: $33.10
Validation:
Dollar amount should be between 10 and 10000
Interest rate should be between 1 and 22
Number of days should be between 2 and 30
The output should be formatted and aligned according to the above.
Here is my code:
Why is my interest 0?Why is my interest 0?Code:#include <cstdio> #include <math.h> #include <iostream> using namespace std; int main () { double dollars, amount, interest; int days, rate, decimalrate, a=1; bool notValid; do{ notValid=false; //Validation for amount cout << "\nPlease enter dollar starting amount?" ; cin >> dollars; if (dollars <10 || dollars >10000) { cout << "\nYou have entered invalid data"; notValid=true; } } while (notValid); do{ notValid=false; //Validation for rate cout << "\nWhat is the daily increase (as a percentage)?" ; cin >> rate; if (rate <1 || rate >22) { cout << "\nYou have entered invalid data"; notValid=true; } } while (notValid); do{ notValid=false; //Validation for days cout << "\nEnter the number of days:" ; cin >> days; if (days<2 || days >30) { cout << "\nYou have entered invalid data"; notValid=true; } } while (notValid); { cout << "\n\nDay Earned Interest Balance"; cout << "\n-----------------------------------\n"; for(a==1; a <= days; ) //Loop for display { decimalrate=rate/100; amount=dollars +(dollars*decimalrate); interest=amount-dollars; cout << a; cout << " "; printf ("%10.2lf", interest); cout << " "; printf ("%10.2lf", amount); cout << "\n"; a++; amount=amount+interest; //Increments a } system("pause"); } } | https://cboard.cprogramming.com/cplusplus-programming/93652-compound-interest-formula.html | CC-MAIN-2017-22 | refinedweb | 342 | 65.73 |
Jeremy Falcon wrote:Totally sounds like an environment I just came out of where the original devs used a lot of FoxPro.
Jeremy Falcon wrote:I'm just surprised to still see ads for it on CP, it's up to version 11. I suppose if Crystal Reports can last that long, anything:The problem is that they don't have a clue about good coding practices and their results are usually plagued with easily avoidable problems.
JimmyRopes wrote:I could be mistaken but by the testimony of the CR evangelists it sounds like one of those so simple anyone can use it products.
Jeremy Falcon wrote:JimmyRopes wrote:The problem is that they don't have a clue about good coding practices and their results are usually plagued with easily avoidable problems.
This is how I feel about MS Access
Dan Neely wrote:Because it seemed like a good idea 15 years ago and the bean counters won't pay to rewrite it in something sane today?
Joe Woodbury wrote:The company that truly keeps throwing sh*t against the wall, hoping one will stick
Mladen Janković wrote:The same joke
public class SanderRossel : Lazy<Person>
{
public void DoWork()
{
throw new NotSupportedException();
}
}
Sander Rossel wrote:Additionally it has cats
Sander Rossel wrote:it seems clearing your browser history will get it back.
General News Suggestion Question Bug Answer Joke Praise Rant Admin
Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages. | http://www.codeproject.com/Lounge.aspx?msg=4869945 | CC-MAIN-2017-17 | refinedweb | 253 | 59.47 |
A few weeks ago, I introduced you to functional programming in Python. Today, I'd like to go further into this topic and show you so more interesting features.
Lambda Functions
What do we call lambda functions? They are in essence anonymous functions. In order to create them, you must use the
lambda statement:
>>> lambda x: x <function <lambda> at 0x102e23620>
In Python, lambda functions are quite limited. They can take any number of arguments; however they can contain only one statement and be written on a single line.
They are mostly useful to be passed to high-order functions, such as
map():
>>> list(map(lambda x: x * 2, range(10))) [0, 2, 4, 6, 8, 10, 12, 14, 16, 18]
This will apply the anonymous function
lambda x: x * 2 to every item returned by
range(10).
functools.partial
Since lambda functions are limited to being one line long, it's often that they are used to specialize longer version of an existing function:
def between(number, min=0, max=1000): return max > number > min # Only returns number between 10 and 1000 filter(lambda x: between(x, min=10), range(10000))
Our lambda is finally just a wrapper of the
between function with one of the argument already set. What if we would have a better way, without the various lambda limitations, to write that? That's where
functools.partial comes handy.
import functools def between(number, min=0, max=1000): return max > number > min # Only returns number between 10 and 1000 atleast_10_and_upto = functools.partial(between, min=10) # Return number betweens 10 and 1000 filter(atleast_10_and_upto, range(10000)) # Return number betweens 10 and 20 filter(lambda x: atleast_10_and_upto(x, max=20), range(10000))
The
functools.partial function returns a specialized version of the
between function, where
min is already set. We can store them in a variable, use it, reuse it, as much as we want. We can pass it a
max argument, as shown in the second part — using a
lambda! You can mix and matches those two as you prefer and what seems clearer for you.
Common lambda
There is a type of lambda function that is pretty common: the attribute or item getter. They are typically used a
key function for sorting or filtering.
Here's a list of 200 tuples containing two integers
(i1, i2). If you want to use only
i2 as the sorting key, you would write:
mylist = list(zip(range(40, 240), range(-100, 100))) sorted(mylist, key=lambda i: i[1])
Which works fine, but make you use
lambda. You could rather use the
operator module:
import operator mylist = list(zip(range(40, 240), range(-100, 100))) sorted(mylist, key=operator.itemgetter(1))
This does the same thing, except it avoids using
lambda altogether. Cherry-on-the-cake: it is actually 10% faster on my laptop.
I hope that'll make you write more functional code! | https://julien.danjou.info/python-functional-programming-lambda/ | CC-MAIN-2019-51 | refinedweb | 485 | 62.27 |
Closed Bug 604381 Opened 12 years ago Closed 12 years ago
Panorama stops working, if set javascript
.options .methodjit .chrome to true
Categories
(Core :: JavaScript Engine, defect, P2)
Tracking
()
Future
People
(Reporter: alice0775, Assigned: dvander)
References
Details
(Keywords: regression, Whiteboard: fixed-in-tracemonkey)
Attachments
(1 file)
Build Identifier: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:2.0b8pre) Gecko/20101014 Firefox/4.0b8pre ID:20101014041748 Panorama stops working, if set javascript.options.methodjit.chrome to true Reproducible: Always Steps to Reproduce: 1. Start Minefield with new profile+ javascript.options.methodjit.chrome=true 2. Open Tabs 3. Open Panorama Ctrl + E Actual Results: no tab in panorama Regression window: Pushlog:
Confirmed.
I reported same bug for linux, and Ben for Mac, so OS -> All.
Status: UNCONFIRMED → NEW
Ever confirmed: true
OS: Windows 7 → All
Hardware: x86 → All
It seems to have appeared between the 10/12 nightly and the 10/13 nightly, which should be this window:
Is this a JavaScript bug or a Tab Candy bug? Who should be working on it?
Do we know if we will be turning this setting (javascript.options.methodjit.chrome to true) on for Firefox 4. My understanding is that we aren't, which means this bug probably doesn't need to be fixed as a P1 or P2 for shipping. CC'ing Dave for insight.
I'm not the Dave you're looking for. Luckily one of them (dmandelin) is CC'ed. CC'in another (dvander). Dave #24601
I haven't heard of any plans to turn on chrome for the method JIT, but if there's a correctness issue, it's worth diagnosing since it could affect content too.
(In reply to comment #7) > Do we know if we will be turning this setting > (javascript.options.methodjit.chrome to true) on for Firefox 4. My > understanding is that we aren't, which means this bug probably doesn't need to > be fixed as a P1 or P2 for shipping. We're still thinking about that, but leaning no per dvander in comment 9. As far as we can tell, turning on jits for chrome has no effect on subjective performance. The minus to turning it on is the extra bugs, like this one. The pluses are that bugs discovered that way probably affect content as well, and would improve the overall engine once fixed; and that some users will turn it on anyway, and we want things to work well for them too. My personal opinion at this point is that the new jit is in a fairly early stage, and we have enough bugs to keep us busy on the content side. Once that stabilizes, whenever that is, may be the time to turn it on for chrome as well.
Thanks for the info. In that case, it sounds like this is not something we will look at or focus on for the Firefox 4 timeline.
Priority: -- → P2
Target Milestone: --- → Future.
(In reply to comment #12) >. Interesting. Feel free to try it out any time and report how it does--I know some of our testers run with methodjit on for chrome, and it seems to work for the most part.
Yeah, at least keep testing -- best case, the crashes go away. Less likely but could be a lifesaver if content hits the same bug, we get some crucial diagnostic information. Please leave any such skidmarks here if you get them. /be
I use JM for chrome. I saw no improvement with Panorama compared to TM only. :(
blocking2.0: --- → ?
blocking2.0: ? → final+
Assignee: general → dvander
Status: NEW → ASSIGNED
This affects the web but it's pretty obscure - you need to have a constructor that, inside a catch block, returns an object other than |this|. In chrome it's easier because it can also trigger with "let". The bug is that constructors weren't checking fp->rval before assuming an implicit return (RETRVAL, STOP) are returning undefined.
Attachment #494529 - Flags: review?(dmandelin)
Whiteboard: fixed-in-tracemonkey
Status: ASSIGNED → RESOLVED
Closed: 12 years ago
Resolution: --- → FIXED | https://bugzilla.mozilla.org/show_bug.cgi?id=604381 | CC-MAIN-2022-27 | refinedweb | 674 | 74.29 |
Dave Massy
Hi Dave,
This is great to hear! It is constantly what I’m telling people. Missing tags, mismatched tags, the forgotten semicolon in javascript. Are these really reasons to dismiss an entire web page? No. Should companies and professionals create well formed HTML? Of course. But there are so many hobbists not to mention the millions of families and friends that create web pages to inform loved ones. Why should their little spot on the web be trashed because browser developers are too lazy to put in the extra effort. Anyone can write a parser that checks tags and whatnot dismissing any page with an error. It is the browser that can handle these small mistakes and still render something readable that is truly useful.
Then you have the small businesses that use WYSIWIG programs to create their pages. Valid tags are going to get deprecated, so all of their websites should be rendered useless in the next browser update…. That’s rediculous. Keep doing what you’re doing. IE has it right and that is shown by the huge market share it still commands despite the competition and doomsayer media doing their best to oust it. Ken.
> Missing tags, mismatched tags, the forgotten semicolon in javascript. Are these really reasons to dismiss an entire web page? No.
I sort of have to disagree with you there. Consider especially the case of a missing closing-comment tag.
By allowing sloppy code to go through, you’re not allowing developers to get an accurate world-view of their pages during the development process.
I’d like to see a browser that can be switched into "strict interprative mode" for developers, but defaults to "loose interprative mode" for browsing. Of necessity, any site that works in strict mode should still work in loose mode.
This will allow developers to catch the missing semicolons, tags, etc., while still grandfathering in all the existing bad code out there that worked in the loose rendering model of previous versions of IE.
WYSIWYG (I think you have a Freudian slip there… what you see is what *I* get?)
On the other hand, if you’re just adding functionality I don’t see how sites could break.
Is it telling that there is only one comment on this page? I think that’s an inidication that the two people who have written on this page are completely oblivious to the world of "web standards"..
The world would be a much nicer place to live in if IE finally got around to doing the same!
I dont see how fixing bugs like these:
and adding more CSS support would break existing applications. This just sounds like a bad excuse of being the browser with the worst standars support. You did improve the CSS and DOM support from version 5 to 6, why just not continue doing it?
I also want the C++ compiler to understand that when I write if( i = 0 ) then I actually meant i == 0 instead of just breaking my code. There’s tons of hobbyists who do not understand C++ well, so Microsoft, stop breaking our code
Maurits – thats exactly what the strict doctype is kinda doing, although once you leave it switched on…
."
I partially agree with this, in that in an ideal world all browsers should be designed to render pages to the w3c standards. However, if the IE rendering engine were to all of a sudden be fixed, all websites that actually *rely* on bugs in IE would appear broken to 90% of web users. Whether the site contains invalid markup is irrelevant to the vast majority of people… It would appear to them that the new IE breaks most websites. This is obviously not an ideal situation.
I think that being able to specify strict doctypes and have IE render html to the standards is a nice compromise. However it’s still not even close to working adequately in that fashion yet.
How can you possibly make a big thing of the dozen or so little things on positioniseverything when you have browsers out there that are so unbelievably non-conforming to standards that it isn’t funny anymore. Take mozilla for instance. Do a search on bugzilla for SSL. There are 283 open bugs, page through and you’ll see about 3 dozen SSL security standards that mozilla fails to comply with. *Security Standards* Ouch! Other searches (security, hang, freeze, crash) read through and discover that close to 1 out of every 15 of these thousands are due to standards non-compliance. Maybe security isn’t important, fine. Let’s talk straight semantics. One of the rising stars in web development is being able to call web services from the client. Very cool stuff. But…mozilla can’t seem to get a handle on those standards either. Try using a W3C compliant WSDL that uses complex types…your out of luck.
Now for the punchline go to this link in IE:
Is that hilarious or what? They can’t even get an SSL certificate to work properly on their own bug tracking site. If you want compliance with standards use IE.
Sure the HTML/CSS bugs on the website above are annoying. But I’d rather use the easy to use hasDisplay property or the Holly Hack any day than find work arounds for security failures. Ken.
More and more web developers are developing in Mozilla these days. More and more are developing sites that look fine in Mozilla, Opera and Safari only to discover they look completely wrong in IE. This is a growing trend Microsoft can’t afford to ignore.
As for IE’s compatibility, it leaves the rendering engine in a barmy and unpredictible state. I spend ages this week hunting down a bug whereby our sidebar was not appearing at all. I found that it was because padding-left was set on an h1 inside the banner, which is completely outside of the context of the sidebar. Why should the left (and only left) padding on an h1 element inside a banner div affect a totally separate sidebar that’s absolutely positioned?
I sure hope that eventually MS bites the bullet. As mentioned previously in these comments, C++ has "millions of hobbyists" but it would just be stupid to care for their mistakes in every release of VS… don’t ruin our code!
I think, if you were to poll the average serious web developer they would feel pretty strongly *against* the way the IE team sees backward compatibility. Surely these people are worth listening to? Pleeeeease?
I wish people didn’t feel the need to promote other browsers and bash stuff here. Compare all you want, but unrelated bugs and "Firefox is great" just don’t do any good. You can get that stuff anywhere but where else will you find information straight from the IE team?
The thing I wonder about with IE’s compatibility is, what do you do with rendering bugs in strict mode? Do you break pages that rely on those bugs, or do you end up with multiple strict modes?
Non-strict parsing was a bad idea. Now you have all that old broken markup, much of which no one will ever touch again. IMO the best approach would be to continue allowing it, but put one of those warning bars across the top of the page to discourage new broken markup.
[quote]However, if the IE rendering engine were to all of a sudden be fixed, all websites that actually *rely* on bugs in IE would appear broken to 90% of web users.[/quote]
The way for IE to handle this is to fix those rendering bugs completely. That way, to won’t render the hacks, and it should render fine without the hacks.
I wrote something about this:
Using the XHTML MIME-type to trigger full standard compliant mode IE could continue not breaking the web, while supporting standards for those we need it.
lowercase josh! That’s exactly what i was thinking! Like in FireFox when it blocks a popup, IE could have the same kind of bar, but for particular html discrepancies! nice… are you listening IE team?
Anne, I totally agree with you! IE should have standard support in application/xhtml+xml, while still support the crappy old bug/features in text/html.
> I’d like to see a browser that can be switched into "strict interprative mode" for developers, but defaults to "loose interprative mode" for browsing. Of necessity, any site that works in strict mode should still work in loose mode.
I would fully agree. As a developer, it would be extremely useful to be able to switch one’s browser into a (strictly) development mode where ONLY valid pages display.
That’s a nice story.
Fact of the matter though is that my PC really didn’t like XPSP2, if it was human it would get a fever and start sneezing.
I took the SP off and I’m not gonna install it again before it’s a little bit less likely to screw up my computer.
Good luck
It’s good to see that you care so much for backwards compatibility. Is that why you removed the support for the Netscape plugin API sometime around version 5.5???. A decision which is to blame for some of your biggest security holes.
So please cut the crap about you caring about backwards compatibility, you only care when it suite Microsoft.
The only benefit I see for IE keeping the sloppy backwards compatibility is to give a disadvantage to the rival browsers. Some hobbyist creates a site it looks fine in IE, they then hear about Firefox (for example) see the page looks bad and then posts F1R3F0X 1Z TH3 SUX0RZ all over the forums.
The important thing is to move the web forward, not hold it back. If a new standard conflicts with some of IE’s backward compatibility code then the backwards compatibility MUST go. Please stop holding back the web and stop the crap that maintaining backwards compatibility is in our interests. It’s not! The whole Netscape plugin issue proves that Microsoft does what suits Microsoft, users be damned.
Internet Explorer should stop compensating for the bad code of the past, and work in the official code of today, and more importantly.
I use Firefox, and it is VERY RARE that a web page doesn’t display perfectly.
In Internet Explorer (which I stopped using for security reasons) valid pages should be displayed correctly, and if the user sees a page broken they can force IE to use the old engine.
Regards,
Stephen O’Brien
StopIE.com
> forgotten semicolon in javascript
Javascript doesn’t require semicolons. The parser is required to add an "implicit semicolon" wherever necessary
>.
Internet Explorer’s support for Netscape plugins comes from an ActiveX plugin, plugin.ocx, that loaded the appropriate Netscape plugin for the MIME type. I have IE 6 SP 1 and plugin.ocx is still there. Of course, the reverse is impossible, since the COM-based interface of ActiveX plugins is so much better than the obsolescent and ad-hoc interface of the Netscape plugin API. So you’re wrong on both accounts
> A decision which is to blame for some
> of your biggest security holes.
No. Netscape plugins not only are loaded through an ActiveX control, but nothing in their API protects them. They are still scriptable, run as native code and are initialized with untrusted input. ActiveX is a powerful and solid technology, there’s nothing intrinsically broken about it
The real issues are IE allowing scripts to instantiate any ActiveX control, and lots of UI issues (warning the user when failing the operation period would have been better, and other such things). And badly written controls, but the IE team can’t do a lot about it
This whole web standard stuff isn’t about backward compatibility. We – developers – have to work twice more because we have to make a standard compliant webpage and an ie_bugfixed version. All the old and not-so-well-made pages will have their problems if IE would use W3 standards? Perhaps. But decide which is the best: having some buggy old hobby-pages or have these bugs conserved forever.
I would chose to respect web standards and if the IE team is only thinking of ‘backward compatibility’, pray for the success of Firefox. (or any other real browser)
I think what this post is trying to say is
"A lot of our user base is corporate intranets with web apps that abuse bugs in IE. If we fix the bugs, the web apps will break. Then they’ll be updated to be standards compliant.
"Once they’re standards compliant, the companies can switch over to Firefox, Safari, or whatever other browser they want.
"Therefore, fixing the bugs would lead to a (possibly dramatic) drop in our market share. Since we have to do what’s best for the shareholder, we can’t fix our bugs, or we have to at least be very slow and cautious about doing it."
P.S.: bugzilla.mozilla.org’s security certificate is issued to (bugzilla|bonsai|tinderbox|despot|mecha).mozilla.org–apparently IE doesn’t understand that syntax.
P.P.S.: Why is there no preview button?
The real problems are not caused by attempting to maintain backwards compatibility. The real problems are caused by not bothering to conform to spec.
For instance, missing units for non-zero lengths in CSS. The specification clearly states that the rule should be ignored. Internet Explorer treats them as pixels.
This means that when somebody is checking their work in just Internet Explorer (the vast majority of the ‘hobbyists’ Ken refers to when defending Microsoft), they aren’t going to be aware of their mistake. It means that they are going to be destroying the layout for anybody not using Internet Explorer. It does web authors and non-Internet Explorer users a disservice. But as long as the web authors don’t realise this, and the non-Internet Explorer users don’t realise what causes it, I guess Microsoft can get away with it, right?
If Microsoft were to fix this issue, the only people who would notice are the ones whose sites were already broken in other browsers. And it might break temporarily, but it would be quickly fixed, and it would _raise_ the reliability of the website, as it would be fixed for _all_ browsers. But the status quo is that there are tremendous amounts of clueless developers out there generating what _appears_ to be bugs in other browsers – so it’s in Microsoft’s interests for these websites to remain broken and to deviate from spec.
I’m sorry, but this post doesn’t really say we’re not going to be fixing bugs or moving towards better functionality.
The first three paragraphs really capture the entire point of this post.
> I’m sorry, but this post doesn’t really say we’re not going to be fixing bugs or moving towards better functionality.
It indicates that you are probably not going to fix _certain_ bugs:
> We feel it is vitally important for web sites and applications that worked with yesterday’s IE work with today’s IE, and continue to work with tomorrow’s IE.
There are bugs in Internet Explorer that, if fixed, would break some websites. The deviation from spec. I noted earlier is an example of one of them.
Hey, that’s sweet – they’ve started deleting comments! And all I did was point to the fact that the US law requires the owners of websites to make them perfectly accessible to people with various disabilities and that MS is liable for new lawsuits because their pathetic excuse for a browser doesn’t really like pages that are made according to standards because it’s so full of bugs it’s hard to make it interpret valid HTML+CSS.
Go, Microsoft! Censorship über alles!
I deleted your earlier post because I felt it broke our posting guidelines. The tone and language were just a bit too abusive and insulting.
Feel free to repost it if you like, if you can leave the insults and abuse out.
Bruce
"… moving towards better functionality"
If by this you mean adding (say) tabbed browsing or RSS/Atom feed reading support into the browser app, can we hope that this will come *after* you have at least matched the standards support already achieved by Gecko and Opera?
If you really do have to manage with limited development and test man-hours, then it is core capability that has to come before the consumer-facing bells and whistles, however necessary it may appear to play ‘catch-up’.
"we have an incredible number of different users and developers using IE in many different ways …"
You’ve had a couple of really good suggestions in these comments on ways to keep the browser product both backwards compatible, and standards-compliant and secure going forward.
IE (the browser product) can never function as a ‘universal canvas’, so don’t go there. The features required by those delivering convincing web applications are properly the province of custom hosts for WebBrowser and MSHTML; for example, customers of our Zeepe rich client framework are wiring together and deploying some really complex and powerful systems of the sort that cannot (and should never be allowed to) run in the browser.
Just my 2c.
Gee, what an unexpected post. Never in my 2 years of reading blogs focused on web development have I heard that Microsoft believes they are doing the right thing by maintaining bug compatibility.
Right.
Actually, I’m tired of hearing it. Stop saying it. No one thinks its even the real motive at this point (even if it is).
And for those of you who look to Mozilla’s bug tracking site for ammo, I have two things to say. First, I hope you’re not MVPs. MVP is a good program. I have been fortunate to get an MVP award once or twice (Halloween always brings fond memories), and maybe someday I’ll do something to deserve it again. But sometimes the awardees in that program lose perspective and I hate to see that program lose credibility. Second, let’s see Microsoft’s bugzilla.
In other words, you deleted my post because I said "stop the crap". Ask any decent web developer what he thinks about IE – i’m pretty sure that the word "crap" will appear in the first three words of his answer.
I honestly don’t understand why the IE dev team isn’t doing anything to improve the browser. You’re holding the web back.
MVP program is just BS. "Dont ever talk bad about Microsoft, ok?"
IE maintaining backward compatibility with obviously INCORRECT behavior is just plain dumb. Or "its just business… nothing personal"
I can relate to that last saying… in a sense, think about how difficult it has to be to be an IE team dev and have to test blantantly incorrect HTML to make sure it shows up properly?
IE for me is a love-hate relationship… the only reason its so popular is because it’s bundled with the OS… [we cant possibly separate IE from the OS… yeah sure…]
Well.
It would be nice if Microsoft would actually stick to it. For instance, patch MS03-015 has broken IE by adding the apostrophe character to the set of "unsafe" URL characters. It escapes me how a character in a URL can be "unsafe" (except for the different meaning used for that term in RFC2396), and why server programmers all over the world should change their URL-generating code to workaround a bug in IE.
To make things worse, the problem only occurs when the URL triggers content to be passed to an external plugin (such as Acrobat).
BTW: the problem could be trivially solved by IE escaping the apostrophe as "x27" before passing it on to whatever considers it "unsafe".
And yes, there is an open support case about this problem. For almost a year now.
Best regards, Julian
This biggest sin of the internet was to put the name of the browser in the HTTP protocol. This invites browser oriented code instead of standard code.
I call here all the browser manufacturers: stop identifying, just tell the version of each protocol you implemented.
Site builders must use a Validator to make their site standard.
If C++ compilers behaved like browsers, one couln’t port any C++ code!
Isnt this what doctype switching is for? You have standard compliance mode (correctly implemented css, html) and quirks mode (backwards compability for html, css).
Fixing the bugs in standard compliance mode will not affect old websites, since they probably dont have a standard compliance doctype.
> I call here all the browser manufacturers: stop identifying, just tell the version of each protocol you implemented.
This isn’t realistic. The User-Agent header is there so people can work around bugs in particular implementations. Nobody is perfect, and it’s very short-sighted to not allow for a small margin of error.
The _real_ problem is that once these bugs were identified, the browser developers didn’t fix them.
> Isnt this what doctype switching is for? You have standard compliance mode (correctly implemented css, html) and quirks mode (backwards compability for html, css).
Unfortunately, "standards compliance mode" is far from perfect, and even people who trigger this will often rely on bugs in Microsoft’s implementation.
In the past I have suggested that Microsoft pay attention to a HTTP header that turns on a spec. conforming mode – which is similar to the doctype switching, only better.
Jim: what about the millions of people whose sites are hosted by an ISP, or a webspace provider? We *cannot* monkey with HTTP headers because we simply don’t have access. We can only upload static pages. Any conformance control *has* to be in the source document.
Not to mention what happens if you obtain a document from some other protocol. HTML, while most often used with HTTP, should not be tied to it, nor vice-versa.
> what about the millions of people whose sites are hosted by an ISP, or a webspace provider? We *cannot* monkey with HTTP headers because we simply don’t have access.
That’s what meta http-equiv is for.
The early IE rendering engines were developed at a time when the web was developing too fast for its own good. Commerce tends to dictate these things. If it weren’t so then North America wouldn’t be using AC power and AM radio.
I’m much less concerned about how much extra work it takes me to do CSS builds on sites and much more concerned with whether the end users of the web site are getting the best possible experience. Every site I build is about compromise – getting the best possible standards support with the best possible end user experience. How can you use an accessibility argument to criticize Microsoft for not breaking web user experience?
That said, I spend about 30% of every project making sites work properly in IE(s) and breaking good semantic structure to force a page to render properly in IE. But that’s the job of a web developer. Do criticize Microsoft creatively. But a competent web developer shouldn’t ever suggest that the IE team break the web for people. But you SHOULD perhaps criticize the company as a whole for underallocating resources to the team…
I’m in the office all day today – Sunday – finishing the build for Capital Health’s new website (local regional health authority). We’re doing the CSS/XHTML/WCAG build. I’m nearly there and it’s looking pretty good. There are a few methodologies that Dave and I have fought over, and it’s been very productive. I’ve had some good discussions with folks from all over the world on my test case for using JavaScript and DOM to balance columns. My conclusion right now is that although Javascript for layout is frought with issues at this point, there are limited cases where it can be applied as long as it is exclusively for visual layout and where the lack of it in no way impedes accessibility. If the real world of the web was closer to the theoretical world of the Semantic Web, then one would regularily use ECMA scripting to modify the Document Object Model rather than using redundant, non-semantic clearing divs, wrapper divs and extraneous intra-document CSS hacks to force layout issues. But the real world dictates alternate strategies. This is one of the reasons that I’m particularly enthralled with one of the recent documents to come out of the W3C this week. "Authoring Techniques for XHTML & HTML Internationalization: Specifying the language of content 1.0" is an excellent resource, but is also an excellent guide. Generally the W3C presents excellent academic recommendations that often are impossible to support in the real world of development. This document actually includes implementation guidelines that recognize implicitely the current browser environment with specific implementation notes for certain particularly obstinate rendering engines. Although I like a group like the W3C to remain somewhat academic and aloof from the real world of implementation (let’s design based on principles moreso than compromises) it’s still important to recognize the fact that the best design actually has to be applied somewhere. To see the converse position – where compromise beats principles hands down every time – check out the latest posting to the IE Blog. Not that I criticize their stance. I want people to trust the web as a publishing medium. Some of the standards enthusiasts who encourage the IE Team to break people’s web experience in the name of Standards support are forgetting that the web is still pretty new and unwieldy. I bet they never used rainbow <hr>s and <blink>, or stayed up all night to wait for their copy of Netscape Navigator 1.0 to download. Suck it up, junior, making web is hard work with problems to solve. Let’s make the web a friendly place. The W3C has also published a first kick at developing a query protocol for the Resource Description Framework. I think RDF is cool….
I think there’s a lot of unfairness in the rhetoric posted on this subject, here and other places like slashdot. The issue of standards and compatibility involves a lot of people with very different, competing, and conflicting interests. As a producer of content I sure do hate writing duplicate code, but I also hate the idea of being allowed to do nothing more with my content than what the W3C thinks appropriate. From my point of view they sometimes call my baby, bathwater, and I don’t like the idea of being obligated to live under the rule of their definitions and imagination Whether you call the extra abilities of Internet explorer bugs or features I supposed strictly depends on what you find useful, and what you have no use for. Obviously looking at this toolbox I’ve found a lot of Internet explorers "bugs" to be quite useful and friendly– without them my Internet experience would be diminished. Sadly You Internet explorer guys have had to kill some of my favorite bugs (control over the clipboard, and url bookmarklet script size etc) I imagine your pop-up blocker is also not healthy for some of my bugs. I profoundly lament that the rotten liars and cheaters are forcing us to produce content in a much narrower imagination universe.
Those of you who quickly dismiss the innovation of Internet explorer (quick and easy div placement and movement, cut and paste, and the incredibly powerful filters that I only wish I had another lifetime to explore the possibilities of) are throwing a lot of baby out with the bathwater IMHO. The ever increasing speed of client side processors should be encouraging enhancements to client side JavaScript– in theory very powerful applications can be efficiently coded this way, right inside the browser– the only real obstacle is security, and I think that can be most easily and efficiently fixed by allowing content producers to register (with browser companies) and take accountability for, the content they provide. I also believe this would be a solution to search engine Spam… but that’s a whole other subject..
To the I E guys, thanks for not putting my imagination out of business….yet.
PS I’m not a professional (or even schooled) programmer, so please no critiques regarding my incredibly sloppy code, or any other mistakes in precise literary "syntax"
> As a producer of content I sure do hate writing duplicate code, but I also hate the idea of being allowed to do nothing more with my content than what the W3C thinks appropriate.
The W3C generally include methods for extending their work. Can you give an example of how the W3C is holding you back?
> Those of you who quickly dismiss the innovation of Internet explorer… are throwing a lot of baby out with the bathwater IMHO.
Hey, when Microsoft actually offers useful stuff, I don’t complain. It’s when they do it in a non-standard way that causes problems that I start to criticise. For example, proprietary CSS properties that do new things I would not criticise *if they used a prefix instead of polluting the global namespace*. Mozilla uses a -moz- prefix. KHTML uses a -html- profix. Opera uses an -opera- prefix. Internet Explorer doesn’t use a prefix, forcing future specifications to either copy Microsoft no matter how badly designed it was to begin with, or break things, or use a less intuitive property name for the standard way of doing things.
>.
Microsoft employees were part of the CSS working group that published the CSS specifications. You can’t criticise the W3C for their user-oriented stance without criticising Microsoft for the exact same thing.
> PS I’m not a professional (or even schooled) programmer, so please no critiques regarding my incredibly sloppy code, or any other mistakes in precise literary "syntax"
I’ll limit myself to one comment: I would take your comments regarding the specifications more seriously if you had not used incorrect syntax in your web pages. All that says to me is "I haven’t read or understood the specifications I am criticising".
We need to move forward. The sites that don’t work in standards compliant browsers are in the minority (otherwise no one would use Opera, Safari or even Mozilla) so therefore it’s your duty to help move the web forward by supporting new standards and allowing people to push the limits of the web. IE is a sad call from the late nineties when it was a pioneer, it’s become a laughing stock and that’s a real shame.
Most of Microsoft’s CSS extensions were created before the -whatever- convention came about, so I wouldn’t blame them too much for that. I remember early builds of Mozilla supported the "opacity" property long before it was finalised.
Frankly, I don’t care about malformed sites. These are the people that actually build the web, not the web savvy (no matter how much we wish it). IE can render malformed sites for all I care.
The thing I want is for IE to render well formed sites as they should be without using CSS hacks. If IE properly supported CSS and PNG (and continued to do so), I would be happy.
You mentioned this as one mitigation of compatibility issues. This one needs work. IE6 has some real problems with this feature. Pages that validate, and that look and act as expected in IE5.x and Mozilla, will break utterly and bizarrely in IE6 when a "strict switch" doctype is used. We have seen this so much since IE6 came out that our development rule is not to use any doctype statement which could put IE6 into this very quirky mode.
Recently a technical writer came to me with such a page. Looked as desired in IE5.x, the code validated and I confirmed the layout in Mozilla. IE6 displayed the layout in a way hard to describe without symbolist poetry. As it turned out, Dreamweaver had helpfully added a strict switch doctype.
Once we deleted that, all was fine in all test environments.
Brett
Dave Massy explains IE’s compatibility choices…
*sigh* it’s simple, if you want the web to move forward you have to sacrifice some compatibility for the sake of correctness. However, there’s no point in getting rid of backwards compatibility if it doesn’t interfere with published standards.
I’m sure that a lot of people would be very happy if we just removed the backwards compatibility that conflicts with the standards and left the rest in – that seems to be the way of the other browsers, I can’t see any other mainstream browser that forces standards they all have some backward degree of compatibility.
So please, move the web forward, support the standards, you only have to drop compatibility if the standards conflict with the current situation.
Ken — the next time you post up here, I’d like to see a link to your site. If you’re so expeienced, I’d like some proof. From what I read, it sounds like you don’t have that much talent, or that much knowledge. Who paid you to play devil’s advocate?
Backwards compatibility does not have to be sacrificed for the advancement of standards technology, nor vice versa. That type of mindset is what makes the difference between an industry leader, and an industry follower. All I see Microsoft is doing is forcing the industry to further stagnate – why else, if other smaller companies have not seen such fear in this issue?
You can keep your bugs for older, or doctype less pages — and build new MS WYSIWYG editors and Doctype pages to conform to standards. That leaves open a whole new market, but I don’t see it happening.
And remember, with all the hacks I have in my Websites to make them work for IE, I’ll have to go back and fix it too if you update. But believe me, I want you to break it, I’d be happy with that — as long as you break it by using standards.
It’s sad, because there’s a part of me that started with Microsoft, and I’d really like to love this company — but I don’t. I don’t think Microsoft cares for it’s developers at all, I think it cares for it’s own goals and wishes to shift us in the path they choose. How can you say that what a developer wants is not reflective of the industry? We build the sites that are engaging to users?
I agree with "Dave".
The backwards compatibility ain’t the problem per se. But when such compatibility conflicts with current standards, it has to go.
There might be a few intranet-applications that rely on the IE-bugs, but catering to those is throwing good money after bad. Every serious web-project right now spend up to 30% of their time applying fixes to various MSIE bugs – time is money – ergo: MSFT cost a lot of businesses a lot of development money because of their lax support for standards.
MSIE is turning my hair grey, is it our fate then to always have an old dinosaur that make our lives complicated? It used to be NS4 now it is MSIE6?
How many of the websites out there, with DOCTYPEs that trigger standards mode, don’t actually work with Opera, KHTML, Gecko, and a hypothetical better IE6? 1%? 0.1%?
I don’t think anyone seriously expects IE to break support for DOCTYPE-less sites, but a lot of us would like you to finish what IE6 started.
>""<
In my opinion, there is one WWW and there should be one standard. Imagine if car manufacturers hadn’t agreed on the mechanisms for operating a vehicle. You’d have to learn how to drive a car all over again everytime you wanted to drive something made by a different company. This would be frustrating to consumers, and I think the same principle will inevitably apply to IE. Whether anyone likes it or not, the computer proficiency of the average individual is steadily rising. As more people grow wise about options other than Microsoft, they’ll see benefits with the other options and MS will be forced to compete on a level playing field. People will want standards-compliance, and if MS wants to keep customers they’ll have to offer it. Of course, this requires not being forced into a browser by your operating system, but it looks like MS’s stranglehold on the OS market is quickly loosening. I think the day will come that Microsoft is forced to innovate on the same playing field as the rest of the industry. That’s when we’ll see if they really can make the best software out there anymore.
I think IE would be going along the right track to maintain it’s current behaviour for pages that don’t specify a doctype, but why are we STILL waiting for standards compliance in strict mode? If a page says it’s html4 or xhtml then treat it as such – if it breaks, the author can either fix it, or remove the doctype declaration.
Fix the css implementation, and I’ll no longer have one of the key reasons I have to convert people to Firefox. Although I prefer non-Microsoft solutions, I really don’t care what my users have, so long as it doesn’t make life more difficult for me.
My tip of the day – drop the IE rendering engine, and build on top of Gecko; open source isn’t something to fear 😉
I work for local government. Some applications that we use internally require Internet Explorer and, if the IE team ceased their policy of maintaining backwards compatibility, these applications would need to be updated. This may involve internal development work or procuring new systems. Inevitably there would be costs associated with this and these costs would ultimately fall on the taxpayer. That is, *you* would have to pay for it.
Dear Richard, this may surprise you, but – the world isn’t the US.
This post does bring up some good points. When you consider the licensing of Internet Explorer (a commercial product) vs. Mozilla (an open-source free product), several differences arrive.
Now, Internet Explorer’s rendering engine was developed and used before W3C standards were such a big deal. There are MANY commercial products that rely on IE’s current rendering engine to create the correct output. Removing backwards compatibility with that would be suicide for Microsoft, much larger than the problem with Firefox’s increasing market share.
Sure, if it was just "hobby" websites that would be broken, I wouldn’t have a problem with a complete standards compliant IE. But the fact of the matter is IE is used both as a client’s browser and an integrated portion of applications.
I too work as a software engineer for the government. There have been several internal projects I’ve seen that break with a standards compliant web browser.
I know the frustration of seeing a cool new CSS feature you’d love to do on your website, only to find IE doesn’t support it. The hacks that most users go through right now are ridiculous. My support goes for the idea that all standards that do NOT interfere with current rendering should be met, and any other proprietary standards should be slowly fazed out (.NET 1.1 brought breaking changes).
I guess the real blame for holding the web back lies partly on Microsoft for not having the insight to comply with standards when they were first published, and partly on developers who program specifically for IE.
My original suggestion for a browser that operates in "lenient mode" by default, but which can be forced into "strict mode" while developing, is a generalization of the DOCTYPE thing.
A DOCTYPE is a switch on a page. What I’m suggesting is a switch in the Preferences panel that can say:
* Give me a big ugly error message if there’s a problem with the page I’m looking at
* Just figure it out as best you can
The first option is useful for developers when looking at sites you develop.
The second option is useful when looking at other sites.
The general rule I’m trying to convey here is:
BE LENIENT IN WHAT YOU CONSUME
BE STRICT IN WHAT YOU PRODUCE
This has various analogs in other areas, but is particularly pertinent for web development.
please, i’ve been waiting for your ie port to texas instruments-99 for years.. when is it coming?
It seems to me an update to IE that contained some fashion of a meta tag that can force IE to render in truly w3c compliant mode would be the simplist answer (as someone else has mentioned).
Sites developed without such a meta tag in place would be unaffected, and new sites developed with that meta tag in place could be coded with the minimal amount of hacks to work as intended.
Wouldn’t something that simple for the ‘end developer’ be a route to consider?
Calvin, that would be mega-neat, but I doubt it will happen any time soon; in fact, 2010. sounds like a nice year for Microsoft to release IE7 which finally supports transparent PNG images.
um, unknOwn, did I suggest that the world was the US? I didn’t mean to and I didn’t realise that I did.
I don’t live in the US so I’ve no idea why I would suggest that.
"realise"… I’m guessing UK
Why does this website not validate?
Well then, the world is not the UK 😉
It’s great that when slashdot writes something that might give IE the slightest positive comment then you’re pasting it on your front page, but when it’s saying things against Microsoft it’s just a useless site full of zealots.
If you don’t like slashdot, never cite one of its stories, don’t be selective.
Erm that was meant for the other thread!
Ken writes:
<em>
Is that hilarious or what? They can’t even get an SSL certificate to work properly on their own bug tracking site. If you want compliance with standards use IE.
</em>
Uh, what are you talking about? Did you even read the cert’s details? The cert’s completely valid, it’s just that IE doesn’t support certs across multiple subdomains.
<em>
It seems to me an update to IE that contained some fashion of a meta tag that can force IE to render in truly w3c compliant mode would be the simplist answer (as someone else has mentioned).
</em>
That’s done already: it’s called strict mode, and it’s triggered by the presence of the appropriate !DOCTYPE in the page.
What I think would be best is if MSHTML rendering engine was forked into a strict version and a quirks version. The quirks version would be frozen at the place where IE6 is now. The strict version would have development continued upon it. This would have the carrot of supporting existing sites with the stick of not supporting newer features, forcing them to use standards compliant markup to take advantage of new features.
Bugger! I’m just after realising that I’m after entering my email address rather than my site URL in those last two comments by accident. If somebody’d correct that, I’d appreciate it. No big deal though.
> That’s done already: it’s called strict mode, and it’s triggered by the presence of the appropriate !DOCTYPE in the page.
Yes, but that mode is also buggy and doesn’t follow spec., so there will be sites that rely on those bugs. If Microsoft fix those bugs and just roll them into strict mode, they’ll be breaking those sites, which they don’t want to do.
This is why I suggested authors actually ask for the new rendering engine instead of having it used by default in common situations. Or, to avoid this situation in future, allow authors to ask for a specific rendering version, e.g.:
X-IE-Render: 2
Where 0 is quirks mode, 1 is strict mode, and 2 is the next version that Microsoft attempt to get things right with. Five years down the line, they can drop support for 0 and introduce 3 if they didn’t manage to get things right with 2.
It must be a horrific mess to maintain multiple rendering engines, but they are doing it already, and trying to kludge compliance with specifications into an already broken rendering engine while keeping compatibility with buggy websites must be far harder and more risky.
I really like Jim’s idea. The only way to have complete compliance is one step at a time. For example, remember Microsoft’s proprietary MARQUEE tag? That’s pretty much been phased out.
If W3C could implement a header tag that would specify what version of their recommendations the page follows, then the rendering engine could work the way we want. As time goes on (e.g. 2+ years), we would have complete compliance with all W3C standards.
Its not "backwards compatibility" its "backwards incompatibility". Its the freak extensions that professionals distain. Step 1: quit creating freakish extensions. Step 2: phase out the past garbage that pollutes the web. Then there will (eventually) be no issue with the "backwards incompatibility" that IE currently strives for and achieves.
This post is just a silly excuse. MS ran so fast to get to the no. 1 spot in the browserworld, that they forgot to do a proper job. Now looking back, MS sees the mess they’ve created with IE (and Frontpage) and come up with the user as an excuse. It _should_ be obvious that holding back the web is _not_ good for the users, however, it’s been said that for IE to change course, it takes time for the top to realise the new course is better.
Which brings up the question: why post this article here? You know it’s not read by your users, it’s read by developers. Look at the reactions. I can only say: "a penny for your thoughts" ’cause I’m wondering if you guys are reading the replies and thinking: "oops, we could we be wrong" or thinking: "they’re just developers, who cares about them or their extra time spent tweaking for IE" or thinking: "… "
when someone creates a web page he/she is NOT a IE user, even if he/she wants to. The code must be valid for all browsers!
Now finally the tides are turning… When firefox has say 20% or 30% market share nobody will create IE only pages anymore.
.. is that Microsoft tolerates hobbyists who have no clue how to write correct HTML and punishes those of the pro developers who actually try to do the right thing – get the site working in Safari, Opera, FFox/Mozilla and IE…
And suprise – getting it to work in IE is always the hardest part..
I have to disagree with those who say that supporting the bugs/nonstandard stuff is ok as long as IE gets better standards support.
By continuing to support bugs/nonstandard code, Microsoft is encouraging apathy on the part of sloppy website designers. Without any pressure to fix their sites, this tolerance allows a large part of the web to persist being inaccessible to any browser that doesn’t go to great lengths to duplicate the numerous bugs and quirks of past IE versions. Thereby making the web inaccessible to the myriad of new devices that do not run some form of Windows+IE.
This is something that’s not in the interest of a company that truly cares about the health and well-being of the web. This IS in the interest of a company who has less-than-admirable ulterior motives. Microsoft’s action (not words in some PR pseudo-blog) here will determine which they are.
Since when is supporting old bugs a good policy? Bugs are bugs and should be fixed, not left to fester and encouraging sites to actually DEPEND on them. That is utter nonsense. The only reason that so many sites depend on them now is because MS has (intentonally) neglected for so long their responsibility of coding to the accepted standards of the internet. Why "intentionally"? Browser lock-in. IE-dependence on one hand, and needing a "fixed" version of IE involving paying for XP so you can get SP2? All too convenient for raking in extra profits.
The time to eliminate the bugs is NOW. The longer you wait, the worse the problem gets. Microsoft needs to take the hardline and just stop supporting them, if they are to remain true to this image they are trying to spin about themselves. Until IE just drops/fixes the bugs, websites won’t wake-up and finally get fixed so that they use standard code, making the job of web designers easier like they’ve been longing for so many years now.
Your ‘backwards compatability’ arguement just wont hold up forever guys. Allowing these ‘hobbists’ to continue wih their sloppy code is like telling them its ok to do a half-assed job. It’s not ok. When Macromedia release AS 2.0 half of my scripts were broken and needed to be fixed, so what did I do? Well, I didn’t complain about fixing them. I am glad they took the time to update the maturing actionscipt language to provide more reliable results, even if that means I have to change my code a little. By the time you release the new version of IE with Longhorn, the damage will be done and your dominance in the market will have withered away, making the significance of this new, hopefully standard complaint browser, no more important than any other Microsoft update. Developers have already made the switch to Mozilla. Coding their websites to W3 standards, and providing alternate fixes for the IE ‘bugs’. By 2006 there will be nothing left for you.
I’m curious about something, Rick: what was wrong with your scripts before Macromedia broke them with AS 2.0?
Did you have regular, functioning scripts, or was there something unusual about them that made the AS 2.0 changes incompatible?
So basically… after reading through the post’s comments, I’d give everyone the main points it in a bulleted form combined with my own retorts… save yourself the time:
* IE doesn’t conform to web standards
– (well duh)
* Microsoft uses the excuse of maintaining backwards compatibility to allow them to continue to support broken (X)HTML
– (well.. thats fair enough actually)
* The IE dev team is more concerned over security than adding new features, which will come with the advent of Longhorn
– (considering Microsoft’s record profits, yet high employee turnover… why not move more employees to the IE project? Or for a simple solution to all the security holes… release IE as Open Source
…Then we’ll have protests about IE being propriety, well here comes the clue-train, and the last stop is you: No-one with the right mind would use IE because there are already better open-source alternatives, such as the Gecko renderer, the Firefox browser, and loads more besides.
* Microsoft is more concerned about shareholders and the mass-market than the people who CREATED THE MARKET IN THE FIRST PLACE
– (the internet would not be where it is today if it weren’t for the efforts of web developers, much like myself and the millions of others who feel strongly of IE’s depreciation amongst Microsoft’s HQ, by turning their backs on the real innovators for the purpose of making more money, they’re losing out the opportunity of making more money in the future…. I can see parellels between Microsoft and IE with EA and Westwood Studios (long live Westwood, Kane lives in death))
* Regarding "hobbyists"
Hobbyists are mainly what we affectionatly call "newbs", they troll around on WebMonkey.com and TutorialForums.com, there, most (if not all) of the members there are fully aware of web-standards and encourage beginners to adopt said standards from the get-go and thus, not be corrupted with depreciated code (death to <marquee>)
Secondly, Hobbyists do NOT create websites with any real information on them, usually they set up an account on Angelfire or Tripod and get all hyped up about it for the first few weeks, then the site becomes a cobweb full of link-rot and stagnant content ("omfg! t’was my 13th birthday last weekend in July 2002!")
…The so-called "Mom and Pop" market… when they create websites, much in the same fashion as the people who use Microsoft’s "PictureIt!" or "PhotoStory", are more likely to use something like FrontPage-Express, FrontPage, or some "Happy-Family-Website-Page-Creator 1999"
I’m aware it wouldn’t be the Microsoft’s fault if the latter were to be employed, but FrontPage (and its little brother) are both Microsoft products, and thus… it is Microsofts fault for:
a) Creating editors that produce broken code
b) Creating a browser than works with broken code
…Then have the nerve to tell us they need to support broken code
WHEN ITS THEIR OWN FAULT!
Okay… so I can give them a little slack…
FrontPage Express (98 and IE5 editions) were both made before the HTML4.0 specification was finalised, and HTML3.02 was the language of the time… and as we all know… HTML3.02 = teh l0se!
But then we hit back again
Microsoft ASP.Net 1.0, 1.1, and the 2.0 Beta
The built-in web-controls serve HTML3.02 content up, when HTML4.01 code could easily have sufficed, if not XHTML1.0
And ASP.Net is a recent innovation, released AFTER the XHTML1.0 spec
Result being, I have to re-author all the controls to work the way they NEED to work
Again, Microsoft is breaking the internet and using their own "product features" as an excuse to keep their browser in the past.
So finally, I pitch this one question:
If Microsoft is so commited to web-standards, then why do *NONE* of the sites, mini-sites, or pages under Microsoft.com or MSN.com validate as compliant (X)HTML? (Above 3.02)
…As a recent StopDesign investigation showed that if Microsoft switched to XHTML, their bandwidth costs could be reduced by up to **62%**:
URI:
-W3b
quote:
As a recent StopDesign investigation showed that if Microsoft switched to XHTML, their bandwidth costs could be reduced by up to **62%**:
URI:
That is such a wonderful and great argument! But I get the feeling that although this is a weblog where people can comment, the editors are not reading the comments, or not caring, because how could one argue with these kind of examples? They obviously can’t, or only by using the enduser and their feedback as a bulletproof shield.
My question: Helllloooooooo? Anyone of the editors listening here? Care to comment on this? Probably not. Then why don’t you shut this weblog off, or at least the comments, because this is ridiculous!!
We, the IE team, have absolutely no day-to-day responsibility for how Microsoft.com, MSN, MSDN, Hotmail, or any other Microsoft website decides to code their pages. Even this blog is being hosted for us, and frankly we have better things to do than fiddle with the blog code.
If you click the "Contact Us" link at the bottom of Microsoft.com, you can get to a form where you send feedback to the people who actually work on that site.
>>But there are so many hobbists not to mention the millions of families and friends that create web pages to inform loved ones.
–an increasing number of people are using msn groups, photo sites and pre-packaged blog templates to accomplish this far more effectively than doing it all on their own. There aren’t that many html ‘hobbists’ and of the ones I do know, they are all quite capable of running their pages through a quick validation and 5 minutes of fixing missing tags before they upload.
>>the small businesses that use WYSIWIG programs to create their pages.
— if someone shells out money to buy a WYSIWYG program I’d expect the professional company to make a professional program (even if its made for hobbists) that creates valid html.
quote -> Should companies and professionals create well formed HTML? Of course.
Após ler artigo "How Microsoft can support CSS2 without breaking the Web", referenciado no The Web Standards Project, nem precisa ser muito anti-Microsoft para perceber o jogo cruel de Redmond. Segundo uma entrevista citada de Gary Schare, a Microsoft se…
That post from before about dar…
PingBack from
PingBack from
PingBack from
PingBack from
PingBack from
PingBack from
PingBack from | https://blogs.msdn.microsoft.com/ie/2004/10/15/compatibility-or-just-dont-break-my-site/ | CC-MAIN-2016-30 | refinedweb | 9,625 | 61.36 |
eyeBuild
In order to publish an .SWF file which can be served without additional setup by eyereturn, you must implement eyereturn's eyeBuild class in your code. This will enable the file to communicate and interact with eyereturn's ad server. eyeBuild contains the basic functionality required for most ads, with methods to be called on basic user interactions such as click, close, and tracking events.
More complex ads which need to communicate and/or synchronize may require the eyeBuildPlus component, which shares all the functionality of eyeBuild and adds the necessary features to support more advanced ad units. See the eyeBuildPlus page for more information.
IMPORTANT! The eyeBuild component will add 2.5k to the final size of your published .SWF, so plan your file accordingly in order to meet file size specifications (found on the specifications page)
eyeBuild cannot be used for files that eyereturn is not serving (ie – most CPC placements)
Importing, Creating and Initializing eyeBuild
Before integrating any of eyereturn's eyeBuild functionality into the unit, you must first add the eyeBuild component to the FLA's library. Open your Components panel (Command-F7 on Mac, Ctrl-F7 on PC), locate the eyeBuild component in the eyereturn subfolder of your Components panel, and drag it into the Library panel. This will allow you to create and initialize the eyeBuild class in your document, following these steps:
for ActionScript 2: Add the following code to the first frame on the root timeline of the ad unit:
import com.eyeReturn.eyeBuild; eyeBuild.init();
for ActionScript 3: Add the following code to the first frame on the root timeline of the ad unit:
import com.eyeReturn.eyeBuild; eyeBuild; // import the eyeBuild initializing eyeBuild } private function init(evt:Event):void { removeEventListener(Event.ADDED_TO_STAGE, init); eyeBuild.init(this); // when the document class has been added to the stage, the listener is removed and the eyeBuild class is initialized } } }
Clickthroughs
To create the clickthrough for your ad unit, you need to call the doClick method of the eyeBuild class.
eyeBuild.instance.doClick();
MULTIPLE CLICKTHROUGHS - If your ad has multiple clickthrough destinations, pass the ID number of the click as a parameter to the doClick call. The index numbers of the clickthroughs are 0-based. This means that the first clickthrough's ID is not 1, but 0. The above single-click examples call doClick with no ID number argument, which is equivalent to "eyeBuild.instance.doClick(0)", as 0 is the first (and default) clickthrough ID.
For example, the calls in an ad unit with 3 unique clickthroughs would look like:
eyeBuild.instance.doClick(); // Main clickthrough // OR eyeBuild.instance.doClick(0); eyeBuild.instance.doClick(1); // Second clickthrough eyeBuild.instance.doClick(2); // Third clickthrough
Be sure to make your contact at eyereturn aware of which URLs correspond to each clickthrough IDs, as they will be responsible for ensuring the URLs are called by the appropriate click events.
DYNAMIC CLICKTHROUGHS - Some clickthroughs need to be dynamically contructed, including variables from the ad unit as part the destination. In these cases, you may pass up to three strings as optional arguments to the doClick method.
For example, imagine that your ad unit features ten cars, each a different colour, and a mix of car models. All the cars are clickable to the same destination, but you need the colour of the car clicked included in the query string in order to affect which version appears on landing page.
To accomplish this, all of the clicked cars should make slightly different doClick calls. Specifically, the Black Sedan indicated by the above URL would be called as follows:
eyeBuild.instance.doClick(0, "Black", "Sedan", "None");
Provide your contact at eyereturn with instructions on how to construct the URL from the variables passed. In this case, they would need to know that the clickthrough needs to be constructed as follows:[VALUE_1]&carModel=[VALUE_2]&carOptions=[VALUE_3]
Expanding Units
eyereturn hosts expanding units in two pieces; unexpanded and expanded. In order to communicate the opening and closing between the files, the doOpen() and doClose() calls are used. These calls require no arguments, simply apply them to the appropriate buttons in each file; the clickable area in the unexpanded file, and the close button in the expanded file.
eyeBuild.instance.doOpen(); eyeBuild.instance.doClose();
Only click-to-expand banners require a doOpen call; the more common rollover-to-expand configuration does not require a doOpen call, or any code to detect the user's rollover All expanded files require a close button with the doClose call.
Event Tracking
If your ad unit requires tracking of content specific interactions (ie: a replay button, or the tracking of transport controls in a video unit), you must implement a doIAT() call when the event occurs. Each call requires a corresponding ID to be passed as an argument to the call.
The following table contains the eyeBuild calls for the most common tracking events. Implement these calls wherever applicable in your ad unit. For events not covered by this table, use the doCustomIAT() method as described in the next section.
Custom Event Tracking
For unique events not covered in the previous section, the doCustomIAT() method allows you to create up to 100 your own tracking events, which will appear in reporting with the names of your choosing. First, create a list of the events you wish to track with the event name to be displayed in reporting, and a corresponding ID number, starting from 0 and incrementing for each new event. You will need to provide this list to your eyereturn contact upon delivery of the ad unit.
To implement these calls, you must call do doCustomIAT() method, which accepts two arguments; the ID number of the event, and a string to be returned as a trace in Flash' output window, for testing and debugging purposes. Please note that this string is only for testing/debugging purposes, and does not have any effect upon serving. The names and IDs must be provided to your eyereturn contact in order for eyereturn to enable the events in reporting. Use the traced messages to make sure that the the correct events are being called throughout your ad unit.
eyeBuild.instance.doCustomIAT(0, "Event A"); // Custom Event A eyeBuild.instance.doCustomIAT(1, "Event B"); // Custom Event B eyeBuild.instance.doCustomIAT(2, "Event C"); // Custom Event C
Feel free to use both standard doIAT() and custom doCustomIAT() tracking events in the same unit as required, but for continuity in reporting we prefer that the pre-defined doIAT() calls are used wherever applicable. For instance, we would advise against creating a custom event for "Send" where the "Submit" event doIAT(18) would suffice.
Importing External Resources
Many polite and rich media ad units need to load external resources into the main swf, such as videos, mp3s, XML documents, or additional images. In order to maintain a relative path to these files during development and testing and still publish a finished unit, eyeBuild provides a method called getPolite() which should be implemented every time you load external files into your unit.
During development, your unit will behave normally, loading your resources from the relative path you specify (preferably a single subdirectory consolidating all the external resources necessary for the unit). However, during trafficking, eyereturn is able to append the beginning of the absolute path to our resource server, where the resources will be hosted prior to trafficking. This eliminates the need to manually change file paths and republish the unit. The following code illustrates how this method should be implemented.
var holderClip:MovieClip = this.createEmptyMovieClip("holderClip", this.getNextHighestDepth()); // The above creates an empty movieClip to use as a holder, but you may target any clip on your timeline as needed
holderClip.loadMovie(eyeBuild.instance.getPolite("politeFile.swf"));
This differs from providing the path directly, ie: holderClip.loadMovie("politeResources/myResource.swf"). The getPolite() method will behave the same way locally, but provides the ability to modify the absolute location of resources during eyereturn's setup. | http://specs.eyereturnmarketing.com/eyebuild.html | CC-MAIN-2017-30 | refinedweb | 1,332 | 50.67 |
Activity
From 09/10/2009 to 10/09/2009
-
09/23/2009
09/21/2009
- 07:19 pm Bug #625 (Fixed): Flaky Encoders on Bot 7
- Both Encoders on Bot 7 return 1025, unless they are manually pulled away from the center of the bot. Rich suggests ad...
- 07:10 pm Bug #624 (Fixed): Flaky Left Encoder on Bot 14
- Bot 14 keeps retuning 1025 on Rich's encoder unit test. Rich says this is because the magnet is getting to close to t...
09/20/2009
- 06:47 pm Bug #622 (Fixed): Update dragonfly_init & dragonfly_lib.h
- Applied in changeset r1426.
- 05:32 am Bug #622 (Fixed): Update dragonfly_init & dragonfly_lib.h
- To use the encoders library, you currently must
#include <encoders.h>
and manually call
encoders_init();
- 06:47 pm Revision 1426: Fixes #622.
- encoders.h is not included in dragonfly_lib.h and encoders_init() is called within dragonfly_init().
- 12:17 pm Revision 1425: updated wireless basic library code and docs
-
- 10:37 pm Revision 1424: Includes working naive version of Target Practice
-
09/19/2009
- 06:53 pm Revision 1423: Renamed code folder for bot being tested to testBot.
-
- 06:51 pm Revision 1422: Added BOM test code for general use.
-
- 06:45 pm Revision 1421: Lowered the initial wait time for the target bot.
-
- 06:27 pm Revision 1420: Updated code for testing BOMs. Added rudimentary diagram for analysis.
-
- 04:11 pm Revision 1419: Added a folder for developing Target Practice Demo.
-
- 08:38 pm Revision 1418: Added encoder_get_x and encoder_get_v.
- use get_v at your own risk
If encoder_read returns -1, this usually means battery is low.
If encoder_read returns a v...
09/18/2009
09/17/2009
09/16/2009
- 03:56 pm Revision 1415: Removed while(1) to allow looping through different tests.
-
- 03:35 pm Revision 1414: Changed rangefinder unit test to use all orbs and not take forever.
-
09/11/2009
- 04:36 pm Revision 1413: uses left and right motor in motor test instead of motor1 and motor2
-
- 03:59 pm Revision 1412: motor 1 and motor 2 changed to motor L and motor R
-
- 12:56 am Revision 1411: Hunter-prey works! its a decent proof of concept but could use improvement
- using robots Edgar (3), 7, 5
- 12:03 am Revision 1410: Hunter-prey sort of works!
-
- 11:20 pm Enhancement #579 (Wontfix): better naming for BOM numbers
- Use the compass convention (N, NE, NNE), etc
- 10:43 pm Revision 1409: Changed BOM threshold to 120, seems to be helping for some robot and causing probl...
- hunter-prey is done except the hunting part
- 09:40 pm Enhancement #578 (Wontfix): usb_put* function with fixed width
- like %3d so we can have nicely outputted columns of data
- 09:22 pm Enhancement #577 (Assigned): Improve library error codes
- Each library function should check a global value to see if the library is properly init'd, if not it should return a...
- 09:19 pm Enhancement #576 (Wontfix): BOM calibration
- Consider a mode which (perhaps when button 2 is held during dragonfly_init) calibrates the bom assuming it is in nois...
- 08:20 pm Bug #575 (Assigned): Battery Level Indicator
- Make an battery level indicator behavior using the orbs
09/10/2009
- 07:39 pm Revision 1408: behavior without BOM seems to be working
-
- 04:54 pm Task #490 (Fixed): Spec out new wheels for the robot
- this was bought and installed a while ago.
- 04:53 pm Bug #563 (Fixed): Buy a Xbee programmer
-
- 04:53 pm Bug #557 (Fixed): Batteries
-
- 04:52 pm Bug #558 (Fixed): USB Cables
- | http://roboticsclub.org/redmine/projects/colony/activity?from=2009-10-09 | CC-MAIN-2014-35 | refinedweb | 593 | 61.36 |
perlsvc - Convert Perl program into a Windows service
perlsvc [options] perlscript
perlsvc [options] project
perlsvc
perlsvc --help
perlsvc --version
The PerlSvc utility converts a Perl program into a Windows service. This utility combines a Perl program, all of the required Perl modules and a modified Perl interpreter into one binary unit. When the resulting service is run, it searches for modules within itself before searching the filesystem.
Most commonly, PerlSvc is invoked with the name of the Perl program that you want converted as an argument. This produces a working service. Some of the options described below make it possible to control which modules are included and how the generated service behaves.
If PerlSvc PerlSvc replaces each one of these with the arguments parsed from the corresponding file.
perlsvc myservice.pl --add IO::Socket --add XML::Parser::Expat
...would include IO::Socket and XML::Parser in your service..
PerlSvc::get_bound_file()and
PerlSvc: service start. It is deleted when the service terminates. The extraction directory is added to the
PATHenvironment variable. It is also added to the front of
@INC.
extractoption or the
PerlSvc::extract_bound_file()function. File permissions must be specified as an octal number (0555 by default);
PerlSvc
PerlSvc: service either via the
PERL5LIB environment variable (for dependent
services) or via the
--lib PerlSvc command-line option (for freestanding
services). For example:
perlsvc - PerlSvc generated service..
perlsvc - PerlSvc explain all files it includes.
--dependentoption to built a non-freestanding service.
perlsvc --help FUNCTIONS perlsvc -.
PerlSvc
$PerlSvc: PerlSvc is assumed to
be the input script filename. Thus
perlsvc myservice.pl
...is equivalent to:
perlsvc --script myservice.
Note: PerlSvc does not automatically create this directory; it must exist before the service service created by
PerlSvc. They are available via the
PerlApp:: namespace in addition to
PerlSvc::, to simplify sharing modules between PerlApp applications and
PerlSvc services.
my $datafile = "data.txt"; my $filename = PerlSvc::extract_bound_file($datafile); die "$datafile not bound to service (PerlSvc::get_bound_file("data.txt")) { # ... process $line ... }
If the file is not bound,
get_bound_file() returns
undef in scalar
context or the empty list in list context.
The following predefined variables are available to the service created by PerlSvc.
All
PerlSvc:: variables documented here are also available via the
PerlApp:: namespace
$PerlSvc::BUILDvariable contains the PerlSvc build number.
$PerlSvc::PERL5LIBvariable contains the value of the PERL5LIB environment variable. If that does not exist, it contains the value of the PERLLIB environment variable. If that one does not exists either,
$PerlSvc::PERL5LIBis
undef.
$PerlSvc::RUNLIBvariable contains the fully qualified path name to the runtime library directory specified by the
--runliboption. If the
--norunliboption is used, this variable is
undef.
$PerlSvc::TOOLvariable contains the string: "PerlSvc", indicating that the currently running executable has been produced by the PerlSvc tool.
$PerlSvc::VERSIONvariable contains the PerlSvc version number: "major.minor.release", but not including the build number.
When the service built with PerlSvc runs, it extracts its dynamic object
files in the pdk-username subdirectory of the temporary directory. The
temporary directory is located using the
TEMP environment variable. It is
also possible to hardcode the location with the
--tmpdir command-line
option.
If the service was built using the
--clean option, PerlSvc also appends the
process id to the username when creating the temporary directory (e.g.,
pdk-username-1234). This avoids race conditions during cleanup. Unless the
--clean option is used, extracted files are left behind when the service
terminates. They are reused by later incarnations of the same service (or by
other PDK-created executables).
PerlSvc uses the
PERLSVC_OPT environment variable to set default
command-line options. PerlSvc treats these options as if they were specified
at the beginning of every PerlSvc command line. Note: Perl must be in your
PATH if you want to use
PERLSVC_OPT.
All directories specified in the
PERL5LIB environment variable are treated
as if they had been specified with the
--lib command-line option. Therefore
modules located in
PERL5LIB directories will be included even in dependent
services. If
PERL5LIB is not set, PerlSvc will use the value of
PERLLIB instead (just like regular Perl).
PerlSvc will pipe the output of
perlsvc --help through the program
specified in the
PAGER environment variable if
STDOUT is a terminal.
The following environment variables are not visible to the service built with
PerlSvc:
PERL5LIB,
PERLLIB,
PERL5OPT,
PERL5DB and
PERL5SHELL.
The temporary extraction directory is automatically added to the
PATH
environment variable when a file is bound using the
[extract] option.
When PerlSvc can't locate a module that seems to be used or required by the service, it produces an error message:
VMS\Stdio.pm: warn: Can't locate VMS\Stdio.pm refby: C:\perl\lib\File\Temp.pm
In general, PerlSvc. PerlSvc includes a number of platform-specific rules telling it that certain dependencies are likely not required. In those cases, the error messages are downgraded to a warning. In all other cases it is the responsibility of the user to verify if the module is needed or not. PerlSvc. PerlSvc PerlSvc internally uses a case-sensitive file name lookup and otherwise does not load the file at runtime.
PerlSvc PerlSvc.
The first thing PerlSvc needs to do is to determine which modules and
external files the converted script depends upon. The PerlSvc program starts
out by scanning the source code of the script. When it finds occurrences of
use,
do or
require, it tries to locate the corresponding module and
then parse the source of that module. This continues as long as PerlSvc
finds new modules to examine.
PerlSvc does not try to run the script. It will not automatically determine which modules might be loaded by a statement such as:
require $module;
In cases like this, try listing additional modules to traverse with the
--add option.
The PerlSvc program has some built-in heuristics for major Perl modules that
determine additional modules at runtime, like
DBI,
LWP,
Tk. PerlSvc
anticipates which additional modules are required so that they are available in
freestanding executables.
PerlSvc then decides which modules to include in the generated service. service is built with all the modules compressed (unless the
--nocompress option is used) and included. When the service runs it arranges
for any
use,
do and
require statements to look for and extract the
corresponding modules in itself.
It can check for the
$PerlSvc::VERSION variable. It will be set to the
version number of PerlSvc that was used to build the executable.
The Windows Service control Manager uses $ENV{SystemRoot} (e.g. C:\WINDOWS) as the current working directory, not the directory where the PerlSvc executable is stored.
It will always have the value:
perl. The
$^X is a special variable that
normally contains the filename of the Perl interpreter that is executing the
script. It is sometimes used in calls to system or exec to invoke perl from
within the script.
No. PerlSvc service built with PerlSvc running with an evaluation license expires
when the evaluation license times out. Use the
--version option to view the
time limit of your current license.
perl(1)
PerlSvc is part of the Perl Dev Kit. More information available at
This manpage documents PerlSvc version 9.5.0 (build 300008) | http://docs.activestate.com/pdk/9.5/PerlSvc.html | CC-MAIN-2017-39 | refinedweb | 1,193 | 57.77 |
i did, and i got it working now. but do you know how to add music files to a java applet? i have the code and everything working now, thanks to your assistance and some critical thinking, but i need...
Type: Posts; User: LovellHoliday
i did, and i got it working now. but do you know how to add music files to a java applet? i have the code and everything working now, thanks to your assistance and some critical thinking, but i need...
the timer doesn't start until the button is clicked, right? and yes i dont get the null error or any other exception now, the only problems im having is:
a) trying to get the images to start back at...
i moved the code that i had up there to the TimerHandler:
//this is the code for my timer in the init:
timer1 = new Timer(50, new TimerHandler());
//this code is for my timer to work...
would the:
String text1 = one.getText();
int Car1 = Integer.parseInt(text1);
be assigning it a value? i was giving it the text of the JTextField. the problem i guess is that until the app...
i was trying to set the value of text1 up with the text of the JTextField, and was gonna use the text in an Integer.parseInt for the Car1. i never declared text1 as null or any of that.
that's what im wondering. the line 72 in my code is:
[CODE]
Car1 = Integer.parseInt(text1);
[\CODE]
it seems to not want to do anything with turning the text into an int.
this is the new error im getting:
java.lang.NumberFormatException: null
at java.lang.Integer.parseInt(Integer.java:417)
at java.lang.Integer.parseInt(Integer.java:499)
...
Is there a way i could use my textfield text(for example: 25 is entered) to change the speeds for my pictures instead of making them all go the same speed? also a way to change the text into an int,...
the variable was basically all the JTextFields, one, two, three, and four. something about turning the text inside them into the int Car1 made my program quite angry.
This is the code for my...
i took out some program:
[CODE]
Car1 = Integer.parseInt(one.getText());
Car2 = Integer.parseInt(two.getText());
Car3 = Integer.parseInt(three.getText());
Car4 =...
i got problem handled now. the only error im running into is when i run my program, i get this error whenever i press the button linked to the start timer:
[CODE]
Exception in thread...
Ok i did it like you posted, and got this now
big=javax.swing.JPanel[,0,0,0x0,invalid,layout=java.awt.FlowLayout,alignmentX=0....
big.add(one);
java.lang.NullPointerException
at java.awt.Container.addImpl(Container.java:1045)
at java.awt.Container.add(Container.java:365)
at...
How do i add a println? i did a System.out.println() over the whole line but then it skipped to line 92 and so on and so on lol. it's really frustrating as i've never seen this error before.
I've recently run into a new problem, am on break from school but have to have this done by next Tuesday. I've gotten a lot done, but a new problem has arrisen. this is my full code:
import...
g.drawImage(iron, a, 100 100, 100, this);
g.drawImage(exp, 100, 100, 500, 500, this);
g.drawImage(hulk, b, 100, 100, 100 this);
g.drawImage(thor, c, 100, 100, 100, this);
g.drawImage(cap, d, 100,...
click.addActionListener(new ActionListener() {
public void actionPerformed(ActionEvent e)
{
timer1.start();
}
});
I have an action listener connected to my button, so when it's clicked...
superclass.java:84: addActionListener(java.awt.event.ActionListener) in java.swing.AbstractButton cannot be applied to (superclass)
click.addActionListener(this);
Ok I got that settled and for...
I worked out some of the kinks and this is what i have working for right here:
import javax.swing.JApplet;
import java.awt.Graphics;
import javax.swing.JTextField;
import javax.swing.Timer;...
Im making a racecar app for my java class for a final exam grade, and i am rushing to get this done, but want it to work well. i consulted my teacher in class and she said that i would need to change...
I am creating a racecar/moving pictures app for my Java class project, and am on a good track, have recently came into some very helpful information from some classmates finished with their projects...
thank you! i feel like i go on and finish this soon, perhaps by 12 tonight for at least the textfield and the labels. much appreciated.
OK, i really appreciate that because i must've been getting the wrong info. Did you see my code? i was wondering if i could cut out the frame and the panel and paint the objects i have there on it,...
so could i do all my coding and painting without a JFrame? i was under the impression that they had to be together
I've run into some sort of a conundrum in my code. I am a 1st year college student in Java, so im not really experience, but i am doing a applet for a test grade and am having trouble trying to get... | http://www.javaprogrammingforums.com/search.php?s=23dd6102fc3fdca83799411cca4f7937&searchid=1583594 | CC-MAIN-2015-22 | refinedweb | 887 | 76.52 |
Form::Processor - validate and process form data
In an application you might want a controller to handle creating and updating a "User" record. And not want to write much code. Here's using Catalyst as an example:
package MyApplication::Controller::User; use strict; use MyApplication::Form::User; sub edit : Local { my ( $self, $c, $id ) = @_; # Create the form object my $form = MyApplication::Form::User->new( $id ); # Update or create the user record if form posted and form validates $form->update_from_from( $c->request->parameters ) if $c->form_posted; $c->stash->{form} = $form; }
The above form class might then look like this:
package MyApplication::Form::User; use strict; use base 'Form::Processor::Model::CDBI'; sub object_class { 'DB::User' } sub profile { my $self = shift; return { required => { name => 'Text', age => 'PosInteger', sex => 'Select', birthdate => 'DateTimeDMYHM', }, optional => { hobbies => 'Multiple', address => 'Text', city => 'Text', state => 'Select', email => 'Email', }, dependency => [ [qw/ address city state /], ], }; } sub options_sex { return ( m => 'Male', f => 'Female', ); } sub validate_age { my ( $self, $field ) = @_; $field->add_error('Sorry, you must be 18') if $field->value < 18; }
Or when you need a quick, small form do this in a controller:
my @fields = qw/ first_name last_name email /; $c->stash->{form} = Form::Processor->new( profile => { required => { map { $_ => 'Text' } qw/ first_name last_name email /, }, }, );
[Docs under construction. The docs are probably, well, less concise then they could be. Editors are welcome..]
Note: Please see HTML::FormHandler for a well-supported, Moose-based derivation of Form::Processor.
This is a class for working with forms. A form acts as a layer between your internal data representation (such as a database) and the outside world (such as a web form). Moving data between these areas often requires validation and encoding or expanding of the data. For example, a date might be a timestamp internally but externally is a collection of year, month, day, hour, minute input fields.
A form is made up of a collection of fields of possibly different types (e.g. Text, Email, Integer, Date), where the fields require validation before being accepted into their internal format. The validation process is really made up of a number of steps, where each step can be overridden to customize the process. See Form::Processor::Field for methods specific to fields.
Forms are (typically) defined by creating a separate Perl module that includes methods for defining the fields that make up the form, plus any special and additional validation checks on the fields.
Form::Processor does not generate any HTML. HTML should be generated in a "view" (and often using templates). And besides, HTML forms are trivial to create and in real life almost always needs customization. The use of a good template system makes this nearly painless.
Likewise, there is also no method to spit out an entire web form with a single method. Having a single method to generate a complete HTML form is often only useful for the most simple web forms.
This module is not restricted to use in a web environment, although that is the typical application. It was designed for use with Catalyst, Class::DBI, Template-Toolkit, and HTML::FillInForm. But, those are not required.
The design of this class is based a lot on the design of Rose::HTML::Objects, but, as mentioned, HTML widget generation is not part of the class. This class focuses more on moving data between the data store to the form that from the form to html. It's recommended that you look over Rose::HTML::Objects if not already done so.
As shown above in the synopsis, a "form" class is where a collection of "fields" are defined via a profile (that looks a lot like a Data::FormValidator profile). In general, the fields know how to validate input data, but the form class can also include additional validation methods for each field and can also cross-validate fields. The form class is what is used in your application code.
A form's "fields" are really small individual classes and they are often sub-classed to make more specific classes with additional constraints. For example, an Integer field might be a subclass of the basic Text field that limits input values to digits. And a year field might be a subclass of an Integer field that limits the range of integer values.
It's recommended that you create new field classes for each specific type of data you have. That is, create a "DeptNumber" field that knows what a department number will look like instead of using a generic "Text" field and then validating that in your form. Save field validation in the form for validation that can't be done in a generic way (like validating that the department number actually exists by doing a database lookup).
Unlike Rose::HTML::Objects, this class does not generate (x)html. I prefer to leave that up to the view (templates). But there is a plan to add that ability via a plug-in system for those that want it. I just find anything to do with HTML is better in the templates where it can be easily tweaked.
A method is provided to generate a hash of current values. This makes populating forms via HTML::FillInForm very easy. HTML::FillInForm is one of those modules that people either love or hate. I love it because HTML forms can be written in a very clean and generic way (i.e. no extra code needed to populate the form widgets). It also makes it easy to populate forms in a number of different ways in your application, which an be handy.
Rose::HTML::Objects is really nice (you should take a look), and one of its features is it handles compound fields -- fields that are made up of other fields such as a collection of fields that are used to specify a date and time. This class doesn't have compound fields, but there's nothing stopping you from defining a field that is made up of a form that includes multiple fields. See Form::Processor::Field::DateTimeDMYHM for an example of this. After all, a field's job is to take input from something and create an internal value. So, its input can be another form made up of multiple fields.
To help with this there's a "name_prefix" form setting that can be used to help with nested forms.
The base class for your forms is Form::Processor, and Form::Processor can be used on its own. But, the fun is when used with a "form model class" -- a class that knows how to work with your data objects.
For example, the SYNOPSIS uses Form::Processor::Model::CDBI for working with CDBI objects. When Form::Processor::Model::CDBI is used then valid options for a field are automatically pulled from the database by looking at the relationships set up in the CDBI classes. When working with an field that "has_a" relationship with another table, then possible options can be fetched from the other table. These options can then be displayed in a HTML select list. And when validating input, the field can check that the input matches one of the available options.
As shown in the SYNOPSIS above, when using a form model class complete controllers can be written in two lines of code. Here's the first line:
my $form = MyApplication::Form::User->new( $id );
That creates a form object. If $id is defined then the $id is fetched from the database for pre-populating the form. The fetched data object is stored in $form->item. A hash suitable for HTML::FillInForm is available in $form->fif (which can be used in a WRAPPER in Template-Toolkit or in the end() sub in Catalyst).
Then, the next line:
$form->update_from_from( $c->request->parameters ) if $c->form_posted;
If a form was posted then call $form->update_from_form. That method validates the parameters and then updates (or creates) the object. Link tables are also updated (e.g. a user "has_many" roles using a mapping/link table).
In the template the fields can be fetched with form.field('name'). Fields have an error method to return the error(s) found during validation. Methods on the form object can be used to tell if validation has run or if an object was updated or created. See Methods below.
Each form field is associated with a general type. The type name is used to load a module by that name:
my $profile = { required => { title => 'Text', age => 'Integer', }, };
Type "Text" loads the Form::Processor::Field::Text module and likewise, type 'Integer' loads Form::Processor::Field::Integer.
The most basic type is "Text" which takes a single scalar value. A "Select" class is similar, but its value must be a valid choice from a list of options. A "Multiple" type is like "Select" but it allows selecting more than one value at a time.
Each field has a "value" method, which is the field's internal value. This is the value your database object would have (e.g. scalar, boolean 0 or 1, DateTime object). A field's internal value is converted to the external value by use of the field's
format_value() method. This method returns a hash which allows a single internal value to be made up of multiple fields externally. For example, a DateTime object internally might be formatted as a day, month, and year externally.
There's a form method called fif, that generates a hash of all the field's external values. This is quite useful for populating a form using HTML::FillInForm.
When data is passed in to validate the form, it is trimmed of leading and trailing whitespace by default and placed in the field's "input" attribute. Each field has a validate method that validates the input data and then moves it to the internal representation in the "value" attribute. Depending on the model, it's this internal value that is stored or used by your application.
By default, the validation is simply to copy the data from the "input" to the "value" field attribute, but you might have a field that must be converted from a text representation to an object (e.g. month, day, year to DateTime).
These are the methods that can be called on a form object. See Form::Processor::Field for methods called on individual fields within a form.
Gets or set the form's name. This can be used to set the form's name when using multiple forms on the same page.
It's also prefixed to fields when asked for the field's id.
The default is form + a one to three digit random number.
sub name { 'useform' }
Returns the profile as a hashref as shown in the SYNOPSIS. This is the one method that you *must* override in your form class. This is what describes your form's fields, after all.
The profile provides a concise and easy way to define the fields in your form. Fields can also be added indiviually to a form, but using a profile is the recommended and common approach.
Fields typically fall into two major categories: required or optional. Therefore, the profile definition is grouped by those categories:
my $profile => { required => { # required fields }, optional => { # optional fields }, };
The individual field names are the hash keys, and the field type is the value:
my $profile = { required => { title => 'Text', age => 'Integer', }, };
The field type maps directly to a field module, as described above. The values may optionally be a hash:
my $profile = { required => { age => { type => 'Integer', }, }, };
The only required key is "type". Any other keys are considered method names and will be called on the field once created:
my $profile = { required => { favorite_color => { type => 'Select', label_column => 'color_name', active_column => 'is_active', }, }, };
Is basically:
require Form::Processor::Field::Select; my $field = Form::Processor::Field::Select->new; $field->name( 'favorite_color' ); $field->type( 'Select' ); $field->form( $form ); $field->required( 1 ); $field->label_column( 'color_name' ); $field->active_column( 'is_active' ); $form->add_field( $field );
This points to a hash reference of field names as the keys and field types as the values. The field types are suffixes of the name space Form::Processor::Field:: and will be require()ed automatically. For example:
sub profile { return { required => { first_name => 'Text', roles => 'Multiple', }, }; }
causes Form::Processor::Field::Text and Form::Processor::Field::Multiple to be loaded and calls their new() method. See Form::Processor::Field for more information on the field types.
As mentioned above, the value can optionally be a hash reference instead of a scalar. In this case the hash must contain a "type" key.
Each of these fields have their "required" attribute set true.
Like above, but listed fields are not set as required.
sub profile { return { required => { first_name => 'Text', roles => 'Multiple', }, optional => { age => 'Integer', } }; }
This just make the above a bit easier.
This list an array of field names. The field types will try and be determined by various means (by calling $form->guess_field_type). For example, with Form::Processor::Model::CDBI it will look at the meta_info() to guess the field type.
auto_required => [qw/ name age sex birthdate /], auto_optional => [qw/ hobbies address city state /],
With CDBI, if it has a has_many relationship with another CDBI object it will be a Select (pick one from a set of options), where a has_many relationship would be a Multiple select.
Other methods might be used such as asking the DBI layer for the column type information, or maybe via a method in your object classes that returns the type for each column.
The hope is that the Form::Processor::Model:: classes can get smart about determining the field type.
*this method is not implemented*
If this is set then the value represents the method used to fetch all the field names from the object class.
auto_all => 'columns', # for cdbi objects
This is not implemented yet, but something like:
map { $_ => 'Auto' } $form->object_class->columns;
This is an array of arrays of field names. During validation if any of the fields in a given group are found to contain the pattern /\S/ then they are considered non-blank and then *all* of the fields in group are set to required. This should work like DFV's dependency_groups profile entry.
sub profile { my @address_group = qw/ address city state zip /; my @credit_card_group = qw/ cc_no cc_expires /; return { required => { name => 'Text', age => 'Integer', date => 'DateTimeDMYHM', }, optional => { comment => 'Text', ... }, dependency => [ \@address_group, \@credit_card_group, ], }; }
This class doesn't have DFV's "dependencies" option at this time.
This is an array ref field names that should be unique in the data base. This feature depends on the model class being used.
New creates a new form object. The constructor takes name/value pairs:
MyForm->new( item => $item, item_id => $item->id, verbose => 1 );
Or, as is commonly done, only an item or item_id needs to be passed to the constructor. In this case a single parameter may be supplied:
MyForm->new( $id );
or
MyForm->new( $item );
If the value passed in is a reference then it is assumed to have and "id" method. So:
MyForm->new( $item_object );
is the same as:
MyForm->new( item => $item_object, item_id => $item_object->id, );
The constructor can accept the following parameters:
The id (primary key) of the item (object) that the form is updating or has just created. The form's model class (e.g. Form::Processor::Model::CDBI) should have an init_item method that can fetch the object from the object_class for this id.
An existing object (i.e. the object that id points to). This can be passed in to the new constructor, but typically it's loaded by the form's model class by its init_item method.
Name of the form. See the name method above.
Prefix used for all field names listed in profile when creating each field. This is useful for creating compound form fields where a single field is made up of a collection of fields. The collection of fields can be a complete form. An example might be a field that represents a DateTime object, but is made up of separate day, month, and year fields.
This defines the object class of the item (used by the form's model class to load, create, and update the object.
Typically, this would be defined as the "object_class" method in your form class, but can be specified in the constructor, for example, for small forms that do not use a form class (and specify the profile directly in the constructor -- see profile below).
This is useful for very short forms where you do not wish to define a subclass for your form.
my $form = = Form::Processor::Model::CDBI->new( item => $item, item_id => $id, object_class => $class, profile => { required => { name => 'Text', active => 'Boolean', }, }, );
If init_object is supplied then it will be used instead of item to pre-populate the values in the form when init_from_object is called.
This can be useful when populating a form from default values stored in a similar but different object than the one the form is creating.
See init_from_object below.
The new() method will return false if the init() method returns false. Typically this would happen if passed in an invalid $id. You may override the init() method in your form class, but make sure you call
return unless $self->SUPER::init(@_);
from your method.
This option is used to load form fields into memory. This can be called in a persistent environment such as mod_perl or FastCGI to pre-load modules.
This method is not called during normal use of the form.
This simply creates a dummy object and calls the method to load form fields via the profile. Any fields your form dynamically creates outside of of the form's
profile method are not loaded. Options are not loaded as this may require reading from a data store which may not be available.
Clears out state information on the form. Normally this does not need to be called by external code. An exception might be if the form stays in memory between uses -- but that's not the idea quite yet
This is called when the form object is first created. Parameters are passed unchanged from the new() call.
Returning false causes new() to return false.
As mentioned in new() above, if a single option is passed then it's considered as a "item" parameter if it's a reference, otherwise it's considered an "item_id".
If an "item_id" is passed in (either as a single parameter or as a named parameter) the init method will return false if the init_item method returns false. (Calling the item method when item is undefined automatically calls the init_item method.) The init_item method is typically defined in the form's model class and should know how to translate an item_id into an item object. See "init_item" below.
So, the idea is you can pass in an item_id into the constructor and have the init_item method validate the item_id and avoid validating the $id in the calling code (e.g. in a controller method).
MyApp::Form->new( $id ) or return 'Invalid id supplied';
Note that if $id is undefined then new() will still return true. This allows the same code to be used for both create and update forms.
The init method calls the build_form method which reads the profile and creates the field objects. See that method for its magic.
The method init_from_object is called. This is typically specific to the type of form model used (e.g. CDBI) and is used to load each field's internal value from the object (which can then be used to populate the HTML form with $form->fif). init_from_object does nothing if no item_id (or item) is available. This would be the case when filling in a new blank form.
If an "init_object" is passed into the constructor then init_from_object will use this obeject (instead of "item") to load the initial field values. This is useful when initializing a form with values from another object.
Finally, the load_options method is called to load options for each field value used on multiple-choice fields. Typically, the form's model class will know how to load the options for each field by looking at the form's class relationships. See load_options below.
Again, this method will return false if an item_id is supplied and an item cannot be loaded from that id.
This parses the form profile and creates the individual field objects. It calls the make_field() method for each field. See the profile() method above for details on the profile format.
For "Auto" field types it calls the guess_field_type() method with the field name as a parameter. Form model classes will override guess_field_type(), or you can override in your own form class. You might do that if your field names are labeled with their type -- "event_time" "age_int", etc. Although that would be an odd thing to do.
The above can also return an array reference.
For all fields, if the field is a "Select" or "Multiple" (i.e. has an "options" method) then will call "options_$field_nane" if that method exists, otherwise will call the "lookup_options" method.
This should be called after $self->item is loaded because existing values may be needed in setting the valid options.
In general, "options_$field_name" would be defined in your form class, where "lookup_options" would be defined in the model form class and handle the more general case of looking up the available options in the database.
Here's an example of a method defined in your form's class to populate the "fruit" field with possible options:
sub options_fruit { return ( 1 => 'Apple', 2 => 'Grape', 3 => 'Cherry', ); }
Dumps the fields of the form. For debugging.
$field = $form->make_field( $name, $type );
Maps the field type to a field class, and returns the field by calling new() on that field class.
The "$name" parameter is the field's name (e.g. first_name, age).
If the second parameter is a scalar it's taken as the field's type (e.g. Text, Integer, Multiple).
If the second parameter is a hash reference then the field type is determined from the required "type" value (i.e.
$type-{type}> ).
The fields are assumed to be in the Form::Processor::Field name space. If you want to explicitly list the field's package prefix it with a plus sign:
required => { name => 'Text', # Form::Processor::Field::Text foo => '+My::Field::Foo', },
This method populates each field's value ($field->value) with either a scalar or an array ref from the object stored in $form->item. It does this by calling init_value() passing in the field object and $form->item. init_value() must return the value(s).
init_value() should be overridden in the form model subclass. For example, in ::Model::CDBI objects are expanded to primary keys for object methods that return a list of objects (e.g. has_many relationships).
If a method "init_value_$name" is found then that method is called instead. This allows overriding specific fields in your form class.
Returns a hash of parameters. The parameters are initialized from the item (see init_params() below), or are the last set of parameters passed to the validate() function.
See also the fif() method.
Calls each field's format_value() method to populate a parameters hash from each field's internal value. This is used to build a hash of all values in the form for use in populating the HTML form (using HTML::FillInForm). That has is returned by this method.
This method is called automatically when $form->params and params are not defiend. You may need to call this method directly if $form->item changes while the form object is in memory to force a refresh of params.
Clears the internal and external values of the form
Returns a hash of values suitable for use with HTML::FillInForm. It's a copy of $self->params with any passowrd fields removed.
Calls fields and returns them in sorted order by their "order" value.
Searches for field named "NAME". dies on not found. Useful for entering the wrong field.
my $field = $form->field('first_name');
Pass a second true value to not die on errors.
Returns true (the field) if the field exists
Set or get the Locale::Maketext language handle. If not set will look for a language handle in the environment variable $ENV{LANGUAGE_HANDLE} and otherwise will Create a default language handler using the name space:
Form::Processor::I18N
You can add your own language classes to this name space, but a more common use might be to provide an application-wide language handler.
The language handler can be passed in when creating your form instance or set after the object is created.
Validates the form from the CGI parameters passed in. The parameters must be a hash ref with multiple values as array refs.
Returns false if validation fails.
Note that this returns the cached validated result if $form->ran_validation is true. So to force a re-validation call $form->clear. This should only happen if the $form object stays in memory between requests.
For each field:
1) hash parameters are trimmed (override in field class) and saved to each field's "input" attribute. 2) dependency fields are set by setting fields to required if needed 3) validate_field is called for each field. This tests that required fields are not blank, and that only fields marked as multiple can include multiple values. For Selects and Multiple type fields the values must match existing options. If the above tests pass then the fields "validate" method is called. The validate method tests the input value (or values) and sets the field's input value based on the input data. The default validate method simply copies the input attribute to the value attribute: $field->value( $field->input ); 4) The form's validate_$fieldname is called, if the method exists AND if there's a value in the field. Use cross_validate if you need to validate fields that may be blank (such as setting defaults). 5) The models validation method is called, if exists. For example, this is used to check that a value is unique in the database.
Finally, after all fields have been processed:
6) The form's cross_validate is called. This allows access to all inflated values. This is called even if not all fields validated. This just makes it easier to do bulk validation where fields may be in common.
If you override validate() make sure you set the flag fields like the validate here does.
This item can be overridden in the base class for the form. It's useful for cross checking *values* after they have been saved as their final validated value.
This method is called even if some fields did not validate.
Returns true if validate has been called and the form did not validate.
Returns list of field with errors.
Returns the names of the fields with errors.
Returns either "required" or "optional" for the specified field.
Something like:
<div class="[% field.required_text %]">
Short cut for:
$form->field($name)->value;
Can pass a second true value to avoid die on not found.
Returns true if the value in the item has changed from what is currently in the field's value.
This only does a string compare (arrays are sorted and joined). And note that:
'foo' != ['foo']
which is probably incorrect.
Default: false
By default all required fields MUST be supplied to the validate method. This works great for forms submitted by browsers because browsers submit all fields (except un-selected checkboxes and radio buttons).
When used with an API or AJAX it may be useful to allow a subset of fields to be submitted. Fields not submitted retain their existing value.
When this flag is true if the field is not supplied in the hash supplied to the validate method then will default to the field's current formatted value, if any.
Do not enable this on web forms, though. Otherwise, any checkboxs will always remain checked once checked.
If this feature is enabled it can be selectively disabled on a field by setting the field's "allow_existing" attribute. This simply means that a key of the field's name must be supplied. It does not inspect the key's value.
This is implemented by set_existing_values.
This method will load existing values into params for any not supplied.
This does not work on compound fields.
Generates a hidden html field with a unique ID which the model class can use to check for duplicate form postings.
This value can be used to link a sub-form to the parent field.
One way to create a compound field -- a field that is composed of other fields -- is by having the field include a form that is made up of fields. For example, a date field might be made up of a form that includes fields for the day, month, and year.
If a form has a parent_field associated with it then any errors will be pushed onto the parent_field instead of the current field. In the date example, an error in the year field will cause the error to be assigned to the date field, not directly on the year field.
This stores a weakened value.
Form model classes are used to moved form data between a database and the form, typically via an object relational mapping tool (ORM).
See Form::Processor::Model for details.
The CGI parameters passed in are stored in Form::Processor instead of in each field object.
When a field is entered and then changed to a different format, what format should be displayed? That is, a form with a date is updated. The text "tomorrow" is entered. If the form doesn't validate what should display? The actual formatted date for tomorrow, or still the text "tomorrow"?
Currently, if the form doesn't validate "tomorrow" is displayed. But if the form validates (and is updated by the model class) then the form will display the formatted date for tomorrow. That still may be different from what the date might look like next time it's fetched from the database (due to timezone settings). Another way to go would be to re-load from the database object to make the date look like it will next time it's fetched on a fresh form.
Init from object happens in Form::Processor, too. It would be nice to have each field know how to initalize from the source object. But, that doesn't work well with overriding Form::Processor with the Model class.
Bill Moseley - with *much* help from John Siracusa
Form::Processor is Copyright (c) 2006-2007 Bill Moseley. All rightes reserved.
This library is free software, you can redistribute it and/or modify it under the same terms as Perl itself.
Form::Processor is free software and is provided WITHOUT WARRANTY OF ANY KIND. Users are expected to review software for fitness and usability. | http://search.cpan.org/~hank/Form-Processor-0.31/lib/Form/Processor.pm | CC-MAIN-2018-17 | refinedweb | 5,139 | 63.59 |
I ran into this post, which was pretty interesting to read. It compares a bunch of ways to index inside CouchDB, so I decided to see how RavenDB 4.0 compares.
I wrote the following code to generate the data inside RavenDB:
public class User { public int Score; public string Name; public DateTime CreatedAt; } private static char[] _buffer = new char[6]; private static string RandomName(Random rand) { _buffer[0] = (char)rand.Next(65,91); for (int i = 1; i < 6; i++) { _buffer[i] = (char) rand.Next(97, 123); } return new string(_buffer); } static void Main(string[] args) { using (var store = new DocumentStore { Url = "", DefaultDatabase = "bench" }.Initialize()) { var sp = Stopwatch.StartNew(); using (var bulk = store.BulkInsert()) { var rand = new Random(); for (int i = 0; i < 100*1000; i++) { bulk.Store(new User { CreatedAt = DateTime.Today.AddDays(rand.Next(356)), Score = rand.Next(0, 5000), Name = RandomName(rand) }); } } Console.WriteLine(sp.Elapsed); } }
In the CouchDB post, this took… a while. With RavenDB, this took 7.7 seconds and the database size at the end was 48.06 MB. This is my laptop, a a roughly Erlang native views.
This took 1.281 seconds to index 100,000 documents, giving us a total of 78,064 indexed documents per second.
The working set grew to 312MB during the indexing process..
var entity = new User { CreatedAt = DateTime.Today.AddDays(rand.Next(356)), Score = rand.Next(0, 5000), Name = RandomName(rand), }; for (int j = 0; j < rand.Next(150,1500); j++) { entity.CustomProperties[RandomName(rand)] = Random600CharString(rand); }
This generates 100,000 documents in the 90–900KB consumed four times faster than the faster CouchDB option.
{{ parent.title || parent.header.title}}
{{ parent.tldr }}
{{ parent.linkDescription }}{{ parent.urlSource.name }} | https://dzone.com/articles/benchmarking-ravendb-40-vs-couchdb | CC-MAIN-2017-34 | refinedweb | 282 | 53.17 |
ASP.NET and .NET from a new perspective>
<ItemTemplate>
<tr>
<td><%# Path.GetFileName((string)Container.DataItem) %></td>
<td>
<asp:LinkButton
</td>
</tr>
</ItemTemplate>
<FooterTemplate>
</table>
</FooterTemplate>
</asp:Repeater>() {
this.rptFiles.DataSource = files;
this.rptFiles.DataBind();
This is what Databound controls are good at. Let them do their job. Doing things dynamically when you don't really need to only complicates things. This design is so much better in so many ways. For one, notice we really have no UI related code in the code-behind (except for the 'status' message). ASP.NET's code-behind model is meant to separate code and UI. Also, notice that there's considerably less code! Why? Because the repeater takes care of the following things for us: (1) it implements the foreach loop. We give it something to enumerate over and call DataBind, it does the rest. (2) it implements the creation of the controls for us. Controls are still being created dynamically at runtime, but we've handed that responsibility to the repeater by describing for it what an item should look like. (3) it implements INamingContainer. Calling DataBind on a databound control throws away its contents and rebuilds it from scratch, just like we had to do. All we need to worry about is maintaining the data and letting the repeater know when the data has changed.
Example: You don't know "what" controls should be rendered at design time, or you want to avoid loading controls you don't need because of performance or because there are too many possibilities.
If you don't know what controls will be rendered in the first place, you have to use dynamic controls, right? Well, even then, it depends. Maybe you don't know what control you will need, but the possibilities are limited. In that scenario, you can avoid dynamic control complexities by simply declaring every possible control with Visible=false, then switch the one you want on by making it visible. Like the repeater example above, this takes the burden of being responsible for the control tree out of your hands. It will also work well if it's possible for the control that is loaded to change during a postback due to a change in state. Since you're loading them all anyway, it doesn't matter.
What about performance?
I've suggested this solution a number of times to readers who sent me their problems. A lot of the time, they were doing it dynamically instead of using the visibility trick because they considered it better for performance. True, loading two controls is more work than loading one, especially when you can know one isn't needed early on. But you're splitting hairs, my friend! Controls, whether they be built in controls, custom server controls, or user controls, are efficient. It doesn't take many resources to instantiate one. It doesn't take many resources to add it to the control tree.
The only time I'd be worried about the performance of a control on the page that doesn't need to exist is if the code in the control is going to do something it doesn't need to do, or if it has a lot of ViewState associated with it. Most of the time, that's not the case. If all the control does is render some html and maybe process some postback data, it isn't worth the effort. If the problem is the amount of viewstate it contains... well, then turn it off, and just enable it for the one control that you do need. If the problem is the operations the control is going to do, like query a database, then that control isn't coded correctly. Control's typically shouldn't go off and do things on their own. They should be told when to do things by the page it lives on. Controls should almost never call databind on themselves, unless they are designed for a really specific purpose where you just want the control to do its own thing. Refactor the control so it has to be told when to do that expensive operation, such as by putting the logic in the DataBind method.
Ok. Let's move on. Say you can't just load them all ahead of time. Maybe the control to load depends on a setting in your database. Maybe there are hundreds of possibilities based on input. The next question to ask is whether the control to load is dependent on state information on the page.
For example, if your database holds the path to a user control which acts as the footer to your layout, that's probably pretty static. It probably doesn't depend on any state information. So all you have to do is load the control and add it to the tree every request. No issues. Just make sure you do it as soon as you can in the page life cycle. OnInit preferably.
If the control(s) to load depend on state data, then that's another story. In this example, we have two radio buttons and a place holder. The dynamic control is loaded into the placeholder, but which control we load depends on which radio button is selected. If you wish, you can also imagine that the control we load is dependent on a setting in web.config. But for this example we'll just use two hard controls.
First we'll define two user controls, UserControl1.ascx and UserControl2.ascx.
<%@ Control Language="C#" %>
<script runat="server">
private void ButtonClick(object sender, EventArgs args) {
this.lbl.Text = "Clicked UserControl1";
</script>
UserControl1<hr />
<asp:Label<br />
<asp:Button
UserControl2 looks the same except has "UserControl2" instead of "UserControl1". Each has a label and a button. When the button in control 1 is pressed, it updates the label in control 1. When the button in control 2 is pressed, it updates the label in control 2. Keep in mind that each control has its very own button and label. There is no label on the form:
<%@ Page Language="C#" %>
if (this.opt1.Checked) {
ph.Controls.Add(LoadControl("~/UserControl1.ascx"));
else if (this.opt2.Checked) {
ph.Controls.Add(LoadControl("~/UserControl2.ascx"));
<html xmlns="" >
<head runat="server">
<title>Pick-a-control</title>
</head>
<body>
<form id="form1" runat="server">
<div>
<asp:RadioButton
<asp:RadioButton
<hr />
<asp:PlaceHolder
</div>
</form>
</body>
</html>
The devil is in the details. Let's load it up. UserControl1 will load by default, so lets just go ahead and make sure it's working by clicking on the button:
So far so good. It says "Clicked UserControl!" like we expected. It loaded, we posted back, it reloaded, and it successfully processed the click as if it were a declared control. Beautiful. Let's swap over to UserControl2 now... but we won't click on its button yet...
What? That's weird. According to this, I've loaded UserControl2, but it's label has the data I put into UserControl1's label! Voodoo!
That's nothing. Now let's have some fun. Change the label control in UserControl2 to a TextBox. Don't even give it the same ID.
this.txt1.Text = "Clicked UserControl2";
UserControl2<hr />
<asp:TextBox<br />
Now what happens when we switch from control 1 to control 2?
Ohhh yah... we're hacking now. We've successfully loaded the ViewState for the Label in UserControl1 into the TextBox in UserControl2. They both have a "Text" property, which both happen to use "Text" as the ViewState key to remember the value in. *High five*
I mean this to drive home an important point. The control tree into which viewstate is loaded must basically match the control tree which was used to save that viewstate. Coming from more procedural web frameworks (like ASP), you might tend to think of the life of a control to be over once it renders itself. But really, I like to think of it as if a control's lifecycle straddles the request/response boundary. It does indeed get re-instantiated upon every request, but because ASP.NET manages state data for us, the control created on a postback is intimately connected with its "predecessor", for lack of a better word. This more "logical" lifecycle ends after the control has loaded its ViewState and its postback data from its previous life, if any. After that, but before it begins its next life in PreRender, you can make persistent changes to the control tree with no worries.
So here is the solution to the above problem. Every scenario is different, so this isn't necessarily a real general pattern you should follow to the tee, but it shows how we do things the right way for this scenario. Hopefully you can adapt the solution to your specific needs.
if (!Page.IsPostBack) {
// no viewstate on initial request, load the default control
ViewState["state"] = opt1.Checked ? 1 : 2;
LoadUserControl();
protected override void LoadViewState(object savedState) {
base.LoadViewState(savedState);
// viewstate loaded, now we know which control to show.
LoadUserControl();
private void CheckChanged(object sender, EventArgs args) {
// state has changed. Remove the loaded control and load the new one.
ViewState["state"] = opt1.Checked ? 1 : 2;
ph.Controls.Clear();
private void LoadUserControl() {
Control c;
int state = (int)ViewState["state"];
if (state == 1) {
c = LoadControl("~/UserControl1.ascx");
else {
c = LoadControl("~/UserControl2.ascx");
// id assigned to avoid shifting IDs when control changed on a postback
c.ID = "foo";
ph.Controls.Add(c);
The idea is to load the control that existed on the previous request by utilizing a ViewState field to remember which control was active. We override LoadViewState, then immediately after calling base.LoadViewState, we can look for the ViewState value to tell us which control existed previously. On the initial request, there is no ViewState and therefore no call to LoadViewState, so we detect this in OnLoad and make sure the default control is loaded at first. Then, we listen for the CheckChanged event on the radio buttons to tell us when the control to be loaded has changed (note: I actually really dislike the CheckChanged event, but that's a different discussion. It works well for this simple scenario).
At the time the CheckChanged event fires, we will have already loaded a user control -- whichever one was active previously. So we have to remove it before adding the new one. That is why we call Controls.Clear() on the placeholder. And finally, we assign the control a specific ID so there's no way we can run into the ID problem mentioned earlier, which would happen when removing the old control and adding the new one in response to the CheckChanged event. There's no way we could accidentally post data from one control into another as you switch from one control to another, because we're loading the correct control prior to the loading of post data. Post data is loaded right after LoadViewState. That is why we don't do the logic from OnLoad -- the user control would miss that phase. Actually, ASP.NET loads post data in two phases, one before OnLoad / after LoadPostData, and one right after OnLoad. The purpose of the 2nd pass after OnLoad is to load postdata for any controls that may have been dynamically added (or created through databinding) during OnLoad. That would actually work just fine for our scenario, but there are consequences to this late loading that are best avoided if possible. Normally, you can rely on post data to have been completely loaded from the OnLoad phase. But for late-comers, that isn't the case. It's best to provide consistent behavior if you can, so LoadViewState is where it's at!
Actually, this very closely emulates what data bound controls do. If you were to examine the code for the Repeater, for example, you would find that from its CreateChildControls method it examines a ViewState field. If it exists, it calls its CreateControlHiearchy method, which rebuilds its control tree based on ViewState. When it is DataBound, it clears the control collection and rebuilds it again, calling the same CreateControlHiearchy method.
THAT'S ALL FOLKS. In the next part we'll cover custom server controls and some of the things to watch out for there.
One more thing...
Many of you apparently were so anxious to read part 4 of the series, you cleverly deduced that the url to the article must be the same as Part 3, only with a "4". You URL HAXXOR, you!!!!111... You see, I've been working on a draft of this article for a long time now (which as it turns out has been completely redone), and I unknowingly had the article saved in a state that would allow you to access it if you happened to know the address to it!
In all, by the time I realized what was happening, this article received 121 hits to it even before it was published! All you hackers got to read my embarrassingly terrible draft. I thwarted you by renaming the article temporarily. I added "abc" to the end. I was waiting for someone to guess that, too. If you did, there would have been a surprise in it for you. But no winners. Oh well....
Until next time.... part 5 will come much sooner than part 4 did, I promise.
ASP.NET Dynamic controls
I like the fact that you often come back to alternatives to dynamic controls.
I've answered a lot of questiong on aspmessageboard.com about dynamic controls and most of the time the best answer is to use one of the alternatives.
In a recent project, I needed tables where you could add and remove rows so I extended the gridview so I could extract its data, remove a row from the data, then rebind.
well good article - It is a great pity that you did not publish it earlier (like a month ago). Due to the problem which you described under example 2 I had to disable the ViewState - and I have to go with it now. If I had known that solution before life could have been easier.
Btw. to all very abiscious ASP coders:
forget about using dynamic controls unless there is NO OTHER WAY (and I mean it). ASP dynamic controls show a *lot* of problems if you do not know very well the ASP (inner structure, and very good understaing of databinding process, event flows etc).
e.g. - even if you have to switch between 5 GridViews with their SQLDataSources etc - forget about it. Just have them all statically and change a Visible property accordingly to your needs. I did not do it that way, AND I REGRET IT BADLY now.
I am recently having a similar problem with a datalist. No events from any control inside the EditItemTemplate getting fired. I have EmableViewState=false for page. Any solution? Please email info@webcosmo.com
I'm really having hard time trying to read all these dark-blue, grey, dark-green etc. words on the black background. You should consider switching at least the background color.
Hi
I've read a lot of your articles and they have been very helpful!
So, here is my issue that i have not been able to resolve. I have custom controls (.ascx) being added dynamically to statically placed placeholders on the page. I need to add the controls dynamically beause their location, which placeholder they belong in, is driven by xml files configured by the some administrator.
I've followed the implemention you have with loading the dynamic controls using the Init() and LoadViewstate. But, for some reason, when a postback occurs, i lose all entered data for Textbox, state of Checkbox, state of Combobox and basically every control who's state can change. If i make the Textbox Readonly=True, then the values remain after each postback. I see you had to update ViewState for the checkboxes in your example above to maintain the state the user selected. I can't do that since my controls are outside the scope of my page. i've tried doing FindControl but with no success. So, if you have any suggestions for preserving the state of these controls (textbox, checkbox, combo) i would really appreciate it!
Thanks again for your articles!
Dani -- I would have to see some code, or at least get some more details. Exactly when and where are you loading the control? And are you giving it an ID?
Hi,
Thanks for the quick response.
I have a LoadCustomControls() function tha is being called by before base.OnInit(e) and if it's not a postback and also after base.LoadViewState(savedState). Every control has an ID and i am also clearing call controls from each placeholder by calling the PlaceHolder.Controls.Clear().
Also, inside each control i am loading the data on the condition that is not a postback. So if(!IsPostback){LoadControlData();}
I can email you some code examples of exactly what i'm doing if that helps.
Thanks again.
Why do you only load in OnInit when it is not a postback? Does the control that you are loading depend on any state data? You said it depends on a configuration file, so unless there's more to it, the answer is no. And if the answer is no, you should just load it from OnInit every request and forget about LoadViewState.
hi, i have a similar problem as in example 2.
however, i have textbox in the ascx for the user to enter. there is a 'save' button in the aspx page. the save button is suppose to capture the textbox entry and save it to the database. how do i capture the entry by the user? i know that you need to cast it and access the expose the property set in the ascx page. but, i want to know how would u implement it.
cheers and thanks
linus -- so if I understand correctly, you just need to get the value in the textbox from the page that is hosting the user control? You could use FindControl to get to the textbox if you know it's ID. But I wouldn't do it that way. I would do as you say -- create a property on the user control that exposes the textbox value. Then on the aspx, cast the user control to the type of the code behind (depending on the project model you are using, you may need to add a @reference directive) and access the property. That method allows the communication you need without hard coding the control ID, or even hard coding the fact that it's a textbox. You are free to change those details in the ascx without breaking the page.
ok thanks for the guidance... my next question would be...
if my dynamic ascx is only to be loaded at runtime or should i say on postback, do i have to do the creation of the dynamic content in the init event as adviced in this article, aspnet.4guysfromrolla.com/.../092904-1.aspx. ?
i know that i can't do the creation on the button event as it would have reset the textbox entries in the ascx when i click the 'save' button. hmm what should i do?
what if you had Labels and label update Buttons in UserControl1 & 2. would the events get fired and correctly update their associated label?
CurlyFro -- The controls will behave just as they normally would.
Great articles, keep up the good work :)
I just started building a solution with dynamic controls, but stopped immediately after reading your part 1-4 above.
Though, after reading, I am all confused about what to do - I am new to .NET so that might be why I cannot figure it out myself.
I have to create a - dynamic!? - form builder with anywhere from 0 to N groups, each group with 1 to N checkboxes or 2 to N radiobuttons inside. And each grouped list of either checkboxes or radiobuttons can be displayed in 1-5 columns. The number of groups, checkboxes, radiobuttons, and columns vary from form to form. Though the form is not dynamically in the sense, that user input can change it. Whenever the form has been build (by web editors), that's what it looks like.
If not creating dynamic controls to build the groups of checkboxes/radiobuttons, what could I do instead?
charlotte --
Nice scenario. I think you should be able to do it without anything dynamic. Even though your data is dynamic on a couple different dimensions, you always know WHAT controls would be rendered. Here's a rough outline that would get you close...
<asp:Repeater DataSource="<%# GetGroups() %>" ...>
<ItemTemplate>
<%# Eval("GroupName") %>
<asp:CheckBoxList
<asp:RadioButtonList
</ItemTemplate>
The GetGroups method returns a list of group objects, each which contains info about the group as well as a 'Items' property or method that returns a list representing the choices within that group. They also contain a property IsCheckBoxList which returns true if it should be a checkbox list or false if it should be a radio button list (a more UI-agnostic name might be better, like AllowMultiple).
Then you simply build both a CheckBoxList and a RadioButtonList, but their visibility is mutually exclusive based on the boolean value of IsCheckBoxList/AllowMultiple.
That control itself allows you to specify the number of columns the options should be rendered in, which you databind with another reference like RepeatColumns="<%# Eval('Columns') %>". Note that even the datasource of these controls is set declaratively. All you have to do is call DataBind() on the root repeater and everything is set into motion automatically.
Note that it may be that RadioButtonList/CheckBoxList do not render in the way you would require it, but the concept would be the same even if you had to create your own custom control that renders them in TDs or something.
wow - thanks !!! for your quick reply - I'll go ahead right away - you are a darling :))
I need to put two controls with the same id in two different panels.
I understood that it would be possible since I gave to the two panels two different id so that, say, the uniqueids of my two controls would be something like:
Panel1$Textbox1 and
Panel2$Textbox1
VS2005 doesn't allow me to do this.
Isn't panel implementing INamingContainer ? Maybe I misunderstood the meaning of INamingContainer ....
Nicola -- Panel does not implement INamingContainer. That would be way too much for the main scenario panel is used for.
INamingContainer is really meant for situations where you know for sure there's going to be duplicate IDs, or its a possibilty and you don't know.
For example, repeater implements INamingContainer, because it is going to be repeating the declared item template, which likely has a control within it with an ID.
UserControls implement INamingContainer, because they might contain a control with a specific ID and someone might put multiple instances of the user control on the same page.
You don't have either of those risks with a panel.
Even if it did implement it, ASP.NET assigns controls to the fields named after the ID. So you can't expect "this.foo" to refer to one of the panels, because how does asp.net know which one it should refer to? For that reason you generally can't have multiple controls with the same ID declared within the same markup on the same page or user control.
Why do you require this? Perhaps there's a better way.
I don't understand this:
"ASP.NET assigns controls to the fields named after the ID"
But, if I understand the rest, I can create a user control with a panel inside and use it! I will try.
I need to do this because I am involved in this fool project: we have a win32 form designer which stores form definitions in a database.
A client application (again win32) can read this form definitions (along with lot of other data dictionary informations: user profiles, data catalogues and so on) and renders the UI dynamically, actually running a complete application. It is a kind of framework, something similar to Access (we called this app of ours "Ouverture"). Now I am trying to port this logic to asp.net to have a web based client. The form definitions are hierarchical, one form could be placed inside another, and so I can have components with same name, provided that they are in *different containers* (if it wouldn't be so, it would be a problem even for the win32 client).
I sometimes feel frustrated but I keep fighting :-) (and your articles are a great support for me indeed..)
Thanks
Bye
Nicola
With a user control containing just a panel it works. But perhaps are there any risks / dangers in this approach ? You seem quite careful about user controls.. or not ?
Nicola -- UserControls implement INamingContainer, that's why it works. If all you need is to be able to contain the item with a naming container, then you should implement a simple Container control that implements INamingContainer and use that instead. It would be much simpler. And easy too.. here's the entire class
public class NamingContainer : Control, INamingContainer
{
done :)
Hi.
I'm trying to build a custom server control for creating a datagrid.
And I can't solve the following problem:
I created a custom template where I build these LinkButton controls: Edit, Delete/Update,Cancel to represent all those actions but when I click on a button the ItemCommand Event isn't raised. I only occurs when I press the button a second time.
Any help would be apreciated.
Joao
Joao -- sounds like exactly what would happen if your IDs were different on postbacks. The first time its ignored because the id changes on the postback, then the second time it works because it was a postback before and after. You'll have to send me some code to tell you exactly why that might be happening. Turn on tracing and watch the control tree as you click the button -- if its id changes, find out why by thinking about how and when it is created.
I'm trying to create a custom server control and I can't make it work.
It's a dynamic datagrid, I created several ITemplate classes and one of them has the buttons: Edit,Delete/Update,Cancel. When I click in one of the buttons for the first time it doesn't fire the ItemCommand event, the event only occurs when I click a second time in the same button.
Any idea why?
Joao -- see above comment
ok ok, I already know that is not a very good idea use OnLoad, I don't want to commit a ViewState crime!!
How can I dynamically populate a control (DropDown, CheckBoxList)??
julianmj -- databinding?
cbl.DataSource = GetMyData();
cbl.DataBind();
I think you can add to the items collection manually if you'd rather. Whether you do that from OnLoad or elsewhere depends on your scenario. Earliest you can do it, is when you should, and disable viewstate on the thing if you can do it every request.
Sorry but my first comment does not appears.
I wrote:
I have made some custom controls(Dropdownlist, Checkboxgroup) with new properties, I use this properties to create a query and populate the control dynamically. I do this sending to a method this collection: Form.Controls, then search for every custom control and do the databind for each one.
Now, after I read your articles I modify the custom control class and add this method:
public class GrupoRadio : RadioButtonList,IControlBind
{
...
protected override void OnInit(EventArgs e)
{
//IdOrigenDatos is a new propertie
this.DataSource = GetData(this.IdOrigenDatos);
this.DataBind();
base.OnInit(e);
}
Is this correct??
Again sorry and thanks !!! for your quick reply.
julianmj -- Is it correct... well is it working or are you having any problems? You might want to disable viewstate on that control since you're binding it every request anyway.
Thanks again. It is working and I already disable ViewState. I'm leaving that way, it worked for me.
I will read again all your articles to find a way to improve my controls...
(sorry for all the typos, I'm still learning English) :)
Hi Dave!
Several month ago in one of the topis on gotdotnet.ru forum we got similar question about 2 dynamically loaded user controls. But my solution was a bit different (adapted to your example):
private bool optChanged;
base.OnLoad(e);
if (IsPostBack) {
((IPostBackDataHandler)opt1).RaisePostDataChangedEvent();
LoadControls(opt1.Checked ^ optChanged);
ph.Controls.Clear();
}
LoadControls(opt1.Checked);
private void LoadControls(bool loadFirst) {
Control ctl;
if (loadFirst) {
ctl = LoadControl("~/UserControl1.ascx");
} else {
ctl = LoadControl("~/UserControl2.ascx");
ctl.ID = "foo";
ph.Controls.Add(ctl);
protected void CheckChanged(object sender, EventArgs e) {
optChanged = true;
What can you say about it?
Oops, sorry! My last piece of code will not work ;) Correct variant might look like this:
protected override void OnLoadComplete(EventArgs e) {
base.OnLoadComplete(e);
if (IsPostBack) {
LoadControls(opt1.Checked ^ optChanged);
ph.Controls.Clear();
}
LoadControls(opt1.Checked);
Control ctl;
if (loadFirst) {
ctl = LoadControl("~/UserControl1.ascx");
} else {
ctl = LoadControl("~/UserControl2.ascx");
}
ctl.ID = "foo";
ph.Controls.Add(ctl);
optChanged = true;
Alexander -- events like CheckChanged occur after OnLoad so I'm not following what the code is doing... it seems like optChanged would always be false. It also seems that on postbacks you're always going to load the control(s) twice even if the active one has not changed.
If you use LoadViewState instead you are able to get the value of opt1 before it loads its postdata, which means it will always be whatever it was on the last request. You load the appropriate control there. The the changed event, if it has indeed changed, is raised. You clear the existing control then load the new control.
> events like CheckChanged occur after OnLoad
Yes, that's why in second example I've used OnLoadComplete instead. So it works.
But I don't understand about LoadViewState. As I know, page LoadViewState will be called before opt1.LoadViewState. Surely, I can create new control, inherited from RadioButton, override LoadViewState method and create something like WasChecked property or add event that will occur after LoadViewState and before Page.ProcessPostData. But can I do it without creating new controls?
>> As I know, page LoadViewState will be called before opt1.LoadViewState
Yes, thats why you call base.LoadViewState from your override first. It's recursive, so by then the checkbox/radiobutton will have loaded its ViewState.
> It's recursive, so by then the checkbox/radiobutton will have loaded its ViewState.
Well, please correct me if I'm wrong. LoadViewState is called recursively for all controls that have something in viewstate, but it doesn't produce this recursion. As I understood, recursion is produced by three internal methods - LoadViewStateRecursive, LoadChildViewStateByID and LoadChildViewStateByIndex. And LoadViewState is called before LoadChildViewStateByID or LoadChildViewStateByIndex. That's why I think opt1 viewstate will be loaded after Page.LoadViewState is completed, and LoadViewState override will not help, especially when page viewstate is empty.
You see, I'm trying to reduce ViewState size and avoid adding data that already in there (radiobuttons already store their old values). May be itis better to look on the problem from another angle - disable opt1 and opt2 viewstates and manage changes manually?
No you're absolutely right about LoadViewState -- child state is loaded after it is completed. My bad. But my lapse of reason had a purpose because I was intending for it to be used as part of the "manual" tracking (when you'd always have a key to load, and it would always be loaded after calling base). Doing that and disabling the opt's ViewState would probably result in smaller state, slightly. Kind of splitting hairs at that point though.
I still think you're doing too much with calling LoadControls() twice even when the state didn't change. Why not add an IF around it...
if (optChanged) {
LoadControls(opt1.Checked);
I understand that, using dynamic controls, I have to rebuild the control tree at each postback. Would it be possible to save the control tree (once generated the first time) somewhere in some format and restore it on next postbacks ? Doing so may save me some processing..
Thanks bye Nicola
i'm working on a UserControl that has a customizable number of rows of data. There can be 0 or there can be infinity or anything in between. I've got add and remove link buttons.
The code is in the URL . I originally did all the stuff without a repeater and dyamically created all the controls on Page_Load, which as you have noted, was a logistical nightmare. So, as per your suggestion I've recoded with the repeater, trying to bind it to a dataset in the viewstate. I can add items, and remove items with the add and remove button. The data in the repeater items persist through postbacks... but any time I click on the add or remove linkbuttons all my data in the text fields of all the items disappear. I figure it has to do with the fact that my underlying datasource (The dataset in the viewstate) doesn't contain the modified data. but if it is data bound, why not? and what is the best way to fix this. The repeater is a much simpler solution than I had before. but I'm not able to get this to work.
Nicola -- Say you did somehow save the control tree. Where would you put the data? On the client for postback? How would you process that data? Parse it, and rebuild the tree? So, you're going to be processing something anyway. Why not cut out the middle man and just process it the way you normally do? One way or another something has to rebuild the tree -- even if asp.net did it for you, it would be using resources that you can eliminate.
I got it figured out. The dataset and the repeater was not quite in sync. So I caught it in the Page_Load and updated the information in the dataset with the information in the Repeater, foreach repeateritem item in Repeater.Controls, then a FindControl("txtboxes") on each of the controls and saved them to the datatable using the index of item.ItemIndex. If anyone is lost on what I'm saying, a link to my source is in the post above. I had just got done converting a mess of dynamic controls to use the repeater control and was losing my information between postbacks. The actual code I added is below. Not beautiful, but it works. Your article series was GREAT and allowed me to refactor a large unwieldy code base to something a lot more comfortable to deal with.
// Fill the ViewState Datasource with information.
foreach (Control item in Redux1.Controls)
if (item is RepeaterItem)
{
if (((RepeaterItem)item).ItemIndex > -1)
{
// Find if the controls are in this Repeater Item
TextBox txtName = (TextBox)(item.FindControl("fldName"));
DropDownList ddlMonth = (DropDownList)(item.FindControl("iMonth"));
TextBox txtDay = (TextBox)(item.FindControl("fldDay"));
TextBox txtYear = (TextBox)(item.FindControl("fldYear"));
TextBox txtPerc = (TextBox)(item.FindControl("fldPercent"));
// if the controls are found save them to the dataset.
if (txtName != null){
((DataSet)ViewState["dsetOwners"]).Tables[0].Rows[((RepeaterItem)item).ItemIndex]["fldName"] = txtName.Text;}
if (ddlMonth != null)
{
((DataSet)ViewState["dsetOwners"]).Tables[0].Rows[((RepeaterItem)item).ItemIndex]["iMonth"] = ddlMonth.SelectedValue;}
if (txtDay != null){ ((DataSet)ViewState["dsetOwners"]).Tables[0].Rows[((RepeaterItem)item).ItemIndex]["fldDay"] = txtDay.Text; }
if (txtYear != null)
{ ((DataSet)ViewState["dsetOwners"]).Tables[0].Rows[((RepeaterItem)item).ItemIndex]["fldYear"] = txtYear.Text;}
if (txtPerc != null)
{ ((DataSet)ViewState["dsetOwners"]).Tables[0].Rows[((RepeaterItem)item).ItemIndex]["fldPercent"] = txtPerc.Text; }
}
}
Stephen --
I love that you converted to repeater and it's working well. It's definitely the way to go...
As for your problem... well, when you are rebinding the repeater you have to think of it as if you are starting over. Anything not in the data you are binding is going to be thrown away.
But -- you probably have a way to get the data out of the repeater, correct? Like a save button that goes into each item and updates the corresponding row? You can use that exact same logic to aid you here. It's the reverse of setting the DataSource on the repeater -- you just need a method that gives you the DataSource back again from the state of the repeater. You do that, then you add or remove the row and rebind that. When it finally comes time to save, you use the same method to get back the dataset from the repeater and do what you want with it.
Stephen -- nevermind you read my mind!
One thing that concerns me with the code is that you're saving the DataSet in ViewState. You shouldn't need to do that -- you can always rebuild the DataSet from scratch using the data in the repeater. By storing the DataSet you're really storing the data twice -- once in the dataset, once in the repeater.
> I still think you're doing too much with calling LoadControls() twice even when the state didn't change. Why not add an IF around it...
Yes, you absolutely right and I'm really ashamed of this mistake.
Well I was thinking that, at the end, the result of all this process is "simply" some html/javascript. My idea was: why don't you save the html result somewhere (in a database ?) and then re-send it "as it is" from the second request on ? (just trying to understand better..)
Nicola -- the result of rendering controls is just html/javascript, but thats not all they are good for. They react to postbacks, and change their state. They may render different data each time.
If you're rendering the same thing all the time and want to optimize for that, take a look at OutputCaching.
How would you solve the issue of dynamic Controls being loaded inside different containers than the last time the page posted back. For example if you look at Web 2.0 startpages like iGoogle or PageFlakes you will notice that you can move around widgets entirerly using clientscript (without PostBack) & a callback to the server saves their new location. However, with the default ASP.Net ViewState management, the ViewState will never be able to get applied to the control because now the control was dynamically added to a different container than in the previous postback. Sure, you can re-add the control back to the same container first - so that ViewState can get applied - and then move it but that seems like such a hack and awfully slow. Would you recommend building custom ViewState in the Load/SaveViewState methods of the page, similar to the DynamicPlaceHolder?
Arif -- the structure of the control tree does not have to dictate the structure of the rendered html. In this case I'd recommend a container control that knows how to selectively render its child 'modules' into the right locations, without physically moving them in the control tree. See my post on "Rendering ASP.NET Controls out of place" for an example of doing something like that. This particular solution can be more specific though without the use of that specialized Renderer control.
Another solution is to just position the 'modules' with css. It wouldnt matter what order they were in the html, the style would just put them in the right location. You'd have to have some client side positioning logic anyway if you're going to support moving them client side.
hi, thanks for your article. i'm currently chewing on some of the things you said about viewstate.
my situation: i make a webservice call, from which i asynchronously receive a rendered dropdown, which gets placed on the page that makes the call. the dropdown is a customized control, which has overriden events (eg. prerender) that conditionally add rows to the dropdown. i'm instantiating the custom control in the webservice to bind it and render the html for the return. what i'm noticing is that the prerender event of the control does not fire. why would this be?
the thing that's different here from your explanations is that the control is not being added to the page control tree, as it's not exactly participating in the page life cycle.
i've been looking for some resource to speak to this area but haven't found any yet. your thoughts would be appreciated.
g --
Events like Init and PreRender are driven by asp.net itself, not each individual control. That's why PreRender doesn't occur... no one told it to. Controls weren't designed to be 'hosted' independently so I don't think there's a great workaround for you there. You certainly wouldn't be able to participate in postback events with that control (which you may not care about anyway).
But... if what I feel about the control is right, then you dont need to do what you're doing from PreRender anyway. If you dont care about viewstate and postbacks then it wouldnt make any difference if you just did your dynamic item additions from Render instead of PreRender. Just keep in mind that since the control isnt in a control tree, it isn't going to have a fully unique ID either. Render this thing twice on the same page, and you've got two dropdowns with the same "name".
dave,
thanks for the prompt response! your comments help fit some of the pieces together. do you think the fact that a control doesn't go through its events is a design oversight?
in the interim i've started looking at script callbacks as well, which may be a good middle ground to settle in (vis-a-vis webservice calls). we'll see.
g -- no problem. No I don't think its a design oversight. It's a limitation of the design I suppose. Controls going through lifecycle events without being in a page just doesn't make sense in asp.net because of how viewstate and postback data are processed, and because of INamingContainer.
Do you know about UpdatePanel? Allows you to make partial updates to the page without opting out of the page lifecycle.
a prompt response again - thanks :). re: your comments on controls going through events - i'll chew on that. what are the implications of inamingcontainer?
i do know about updatepanel. i'm currently exploring my options for asynchronous page renderings - i'd like to know how the plumbing works. using the updatepanel means subscribing to the ajax.asp framework, which means a couple of things that make me pause: 1) going through the whole page lifecycle on asynch postbacks; 2) leaving too much of the functionality to the black box (though maybe wonderful) of ajax.asp and resting on my laurels.
Thank you for such a great article. It has taught me a lot around dynamic controls. I have fought most of these issue on a project where we used dynamic user controls. I am playing around using templates to replace some of our very complex dynamic control code and I have a question. I am using the following code to add my controls to the page:
<asp:Repeater
<ItemTemplate>
<gv:Fees ID="fee" runat="server" Amount=<%# DataBinder.Eval(Container.DataItem, "Fee_Amount") %>
FeeID=<%# DataBinder.Eval(Container.DataItem, "Fee_ID") %>
PaidBy=<%# DataBinder.Eval(Container.DataItem, "Fee_PaidBy") %>
Type=<%# DataBinder.Eval(Container.DataItem, "Fee_Type") %> />
</ItemTemplate>
</asp:Repeater>
This works great and the right number of gv:Fee user controls get added. The problem comes in with post back events. This user control has a button that when clicked needs to call a function back on the server.
When I run the page, the controls get added but when I click the button, the server side code is never called. Is there something extra you have to do to get the repeater to allow for this?
Thanks,
Josh
Is it possible to "reverse-engineering" a control collection to a .aspx page ? I have a function that dynamically builds at runtime a page. I am thinking about changing it in an offline process to statically build pages to be compiled, instead of build them on the fly. Would it be possible ? I looked at HtmlWriter class but it seems suited for pure html, what if you need to write an .aspx file ?
Thanks bye nicola
Nicola -- I'm afraid nothing built-in is gonna help you here. But you can use an xml writer or xml document. Determining all the attributes and converting the values to strings won't be a trivial task though :) Don't forget about dynamically writing the @register directives.
Hi, I read (skimmed) all the 4 parts. Lots of good knowledge, thanks. I have a problem where I'm using dynamic controls, and I don't know whether I can use a repeater like you showed in this part.
I have a page which allows the user to look up products. To begin with, that page has 10 rows, which have a textbox for entering item num and a textbox for entering qty.
When the user runs out of the rows, they can click the "More" button to give them more rows.
My Draw() method adds rows and cells dynamically, to a statically declared table, and adds a textbox to each one of the cells dynamically. For my textboxes, I do declare IDs, based on the numrow, and I keep count of current number of rows, so when I need to read in the user input values, I can do that.
Finally when the user is done, they click "Validate" which validates the items and quantities against a database by first reading in all the values and adding them to an object and calling object.validate(). The object then contains the validation info, basically text description for an item if it was valid, and a valid status, and error message if it was invalid, and an invalid status.
Based on this info, I get an enumerator into that object's list and while drawing my textboxes dynamically, I input the values from the list into the textboxes, and highlight the rows appropriately (red for invalid, green for valid).
I have to call my Draw() method from Pre_render event handler, which is a problem when I want to do validation using validators. I'd love to call it from Pre_Init, but that doesn't work because firstly, I don't know how many rows I have till I get to the button More's event handlers, and secondly I don't have the validated item list from the database till I get to the Validate's event handler. So as far as I know it, it's impossible for me to get the validators to work.
Any ideas on how to solve this?
I understand, thanks. Another doubt: what about thread synchronization ? If I am dynamically building controls into a web page do I need to stick around some "lock(this)" statements ?
Or does the framework handle this process automatically ?
Thanks again! Bye Nicola
Nicola -- there's only one thread that processes the page. No worries.
Really ? it is quite different from what I was used to with isapi development. What about static class or Application / Session properties ?
I guess my question is a little off topic... Perhaps may you point me to a good resource about this topic ? thanks again again. Bye Nicola
Nicola -- yes really. There's only one thread processing the request. The page, the instances of the controls, the entire control tree and all its state -- these are all unique to each request which is served by only one thread, so you're safe to go willy nilly. There are many requests occuring for different users, of course, so accessing static members or shared instance members still needs to worry about thread sync. Session doesn't, because unless you put it into read only mode, only one request from the user can access session at a time (it is internally syncrhonized already).
Great article! I've been doing .Net for a few years but this helps me understand some of the behaviors I've seen in past projects.
I like your idea of using a Repeater in cases where you know "what" you're going to put on a page, but not "how many" and I'm trying to use this in a test project. I have a user control (RxWebObject.ascx) which contains an image button and a gridview for now. It's associated with a business object which can have "child objects". You can think of it as a hierarchy of objects. When the user clicks on the user control, I want to have it display its children as user controls (so they can be clicked on too).
For performance reasons (and because I don't know how the user might want to navigate through the children) I don't want to read all the data and display the entire structure at page load time.
Of course, VisualStudio.Net doesn't allow me to use circular references, so I can't put a reference to RxWebObject.ascx into the ItemTemplate of my Repeater in RxWebObject.ascx.
I AM able to use LoadControl in the codebehind of RxWebObject to create my child RxWebObjects and then dynamically add them to a placeholder inside the user control, but then I run into plenty of problems that you've discussed in your article...they disappear, aren't clickable, etc. Also, the business object associated with the control becomes "nothing" again on postback.
Do you have any suggestions for how I might handle this situation?
Thank you!
I have a survey system which creates questions of one of six different types within a databound repeater. Each question type is a user control, which is created dynamically using the ItemDataBound event. All of this looks great. I have the system navigating back and forth through the sets of questions and they display well. The one problem which I cannot seem to solve is how to save the user's answer for each question, which could be a selection in a radiobuttonlist, a checkboxlist or even simply free text in a textbox. When I iterate through the control collection in the repeater I can't see the user control that was created dynamically, even though I've made sure it has a unqiue ID. Do you have any suggestions?
Stev -- the fact you are creating the controls from ItemDataBound worries me a bit :) That's because unless you are databinding the repeater every request, the control isn't going to exist on a postback. ItemDataBound only occurs when you DataBind the repeater.
What you'd need to do is store in ViewState information about which controls were created during databinding, then from the ItemCreated event you lookup that value and add the appropriate control. That way they can continue to exist even if no databinding occurs on a postback.
One test you must always put your controls through is this -- put a button that does a postback on the page. Don't have it do anything, it just posts back. You should be able to click that button at any stage in your page progression and everything should remain the same. If things disappear, or revert their state, you have a problem.
Jenny -- Just create a control that does nothing except load the user control dynamically.
<abc:MyPlaceHolder
Where MyPlaceHolder overrides OnInit and loads the user control into its control collection, nothing more. It's just a level of indirection so you can still approach the case where you have a user control nested within itself.
Thanks for the articles and especially for sticking around for all the comments so far (please don't stop now :)
I have a question if i may.
i think i fall in the category of your 2nd example:
"Example: You don't know "what" controls should be rendered at design time, or you want to avoid loading controls you don't need because of performance or because there are too many possibilities."
But now i don't know if i need to use dynamic controls or templates or ??.
Here's my dilemma: I have about 10 tables in my sql 2005 db. they all are related to one master table which is the employee master table. And there's a possibility more tables will be added later.
what i'm trying to do is be proactive and build a general routine kinda like the formview (and maybe i need to use the formview programmatically i don't know yet.)
thanks,
rodchar
rodchar -- whether you use the FormView or not depends on whether you want some of the features it provides. Either way, you'd be building controls into a hiearchy that meets the needs of the table in question. For each, you'd be hooking into its DataBinding event, and from the handler you'd be assigning to the control the data from the bound data. You just need to ensure that you create these controls every request, such as by saving the selected table (its name or id, NOT the table itself!) into viewstate then from OnLoad or OnLoadViewState you look for that entry. It wouldn't be that different from the example in this article.
Hi InfinitiesLoop,
Firstly thank you for such a great article, i have read lot of article in these issue but this was the best. You have given very thorow explanation of the topics.
But still i am confused while creating one of the logic for my project, could you please tell me the reason, i am sending you a small demo code, where i am providing a simple functionality to provide the language known option for user according to their choice, for which i have provided a add more button, when that clicks and new textbox appers in the form, i managed to restore its view state but when i again click the add more button, the text box is added two times, i know i could add the invisible text boxes, but my problem is little bigger where i have to add a usercontrol with lot of control in it, it is just a simple demo, which if you could given me explanation i will be thankful.
public partial class DemoCommonControl : System.Web.UI.Page
private bool IsButtonClicked = false;
protected void Page_Load(object sender, EventArgs e)
protected override void OnInit(EventArgs e)
base.OnInit(e);
protected override object SaveViewState()
return base.SaveViewState();
protected override void LoadViewState(object savedState)
base.LoadViewState(savedState);
CreateTextBox();
protected void Button1_Click(object sender, EventArgs e)
if (ViewState["counter"] == null)
ViewState["counter"] = 0;
else
ViewState["counter"] = Convert.ToInt32(ViewState["counter"]) + 1;
IsButtonClicked = true;
private void CreateTextBox()
int x = Convert.ToInt32(ViewState["counter"]);
TextBox txtName;
if (x == 0)
for (int ctr = 0; ctr <= x; ctr++)
txtName = new TextBox();
PlaceHolder1.Controls.Add(txtName);
for (int ctr = 0; ctr <= x - 1; ctr++)
Digamber
Digamber -- your CreateTextBox method is going to create textboxes based on your ViewState counter each time it is called. But you are calling it twice -- once from LoadViewState and once from Button Clicked. Say the counter is 2. You call CreateTextBox. Then the button click event is raised and you change the counter to 3. Then you call CreateTextBox again, which creates the same number of textboxes as before plus 1 more.
Hi, and thanks for a great series of articles.
What if the control you load in LoadUserControl depends on ControlState (maybe in addition to ViewState)?
LoadControlState occurs before LoadViewState - doesn't that mean that your controls need to have been created before LoadControlState? And adding controls from within LoadControlState is not possible - you get a "Collection was modified after the enumerator was instantiated".
Thanks!
/Fredrik
Fredrik -- I'd try to stick to thinking of Control State as ViewState. Its really the same thing, just segregated out. The fact the event occurs whenever doesn't matter, controls added later still go through the essential event sequence.
But if I add controls after LoadControlState, will these controls have their control state loaded? I thought that if the control is not in the hierarchy at the time when LoadControlState occurs, the control state would be lost.
If I add controls in LoadViewState (occuring after LoadControlState), how can these controls ever receive their control state?
Or am I missing something here?
Fredrik -- when a control is added to the tree, it plays 'catch up' with the event sequence. If you add a control dynamically from Load, for example, which is after LoadViewState, it will still load its viewstate.
Hello InfinitiesLoop (Newman, if your a Seinfeld fan),
Regarding the following quote:
"At the time the CheckChanged event fires, we will have already loaded a user control -- whichever one was active previously. So we have to remove it before adding the new one. That is why we call Controls.Clear() on the placeholder."
I have adapted your CheckChanged example and depending on which radio button is selected i retrieve a DataTable from the database and manually display the fields in an html table.
my question is when CheckChanged is fired, say it was opt1 and now it's opt2, will opt1's data routine run as well, then clear the controls, then run opt2's data routine?
Helly Jerry -- err, Rodney,
Yes.
Does that answer your question? :)
Do you think this is ok to do? to go out to database on each postback twice?
It's only twice on postbacks in which the radio button selection has changed.
I think that's better than storing data in viewstate and rebuilding the controls based on that. If you're really concerned about perf/db contention then you can cache the data in asp.net cache pretty easily, and even use a SqlCacheDependency to make it automatically invalidate when the database changes. It doesn't get much better than that.
Great!! Thank you for your article and follow-ups. This has been educational.
Rod.
Thanks for the articles! I thought after much goggling I might find the answer here. I'm much wiser now, but I am still searching for a solution to a problem.
I can send/post code if needed, but first let me just say that there appears to be a few people like me out there with this problem (and without a resolution -- and the code involved is rather lengthy): I am creating dynamic user controls, each containing one particular type of input field (dropdownlist, radio list, text field, checkboxes, etc), all driven by field meta data stored in a database.
I can successfully process all field types I'm using, except for a dropdownlist -- I cannot fetch the SelectedValue from the dropdownlist (within a user control) after the user clicks save.
I have worked hard at understanding the points at which I should be creating and repopulating the control. This article and another led me to start overriding LoadViewState and recreating and repopulating the list after a call to base.LoadViewState. It simply doesn't seem to work... The frameworks via Viewstate just doesn't seem to be able to set the selectvalue of the dropdownlist; it's empty, even though via a trace I can see the the value was posted back correctly.
I am fairly certain that I am creating my controls in the same order, etc., etc., etc. Any thoughts are appreciated!
Never mind the above; I believe I have it... I just listened to what I was saying and decided that I might not be recreating the option list in time, and sure enough...
Looking forward to part 5!
Hi again InfinitiesLoop and all,
The following is a snippet from your Part4SampleCode:
protected override void OnLoad(EventArgs e) {
if (!Page.IsPostBack) {
// no viewstate on initial request, load the default control
ViewState["state"] = opt1.Checked ? 1 : 2;
LoadUserControl();
base.OnLoad(e);
protected override void LoadViewState(object savedState) {
LoadUserControl();
private void CheckChanged(object sender, EventArgs args) {
ViewState["state"] = opt1.Checked ? 1 : 2;
LoadUserControl();
public property DataTable dt;
public void LoadUserControl() {
int state = (int)ViewState["state"];
if (state == 1) {
dt = LoadDataTable1();
else {
dt = LoadDataTable2(); }
I'm trying to encapsulate this snippet inside another WebUserControl if possible. I ran into a wall with my implementation. I'm using a DataTable in the LoadUserControl (which i made public) and i would like to expose the datatable so that this new UserControl will accept a DataTable and use it from there. However, on the postback the datatable becomes null.
Any ideas? do you have a better way or suggestions? do you think this would be an unusual implementation?
Rod -- I'm not exactly sure what is going on based on your code snippet. The DataTable is assigned in LoadUserControl. When is it null and how?
I'm sorry for the confusion. What I'd like to do is extract the LoadDataTable methods from the new UserControl (from your LoadUserControl procedure) and let the code-behind do the LoadDataTable methods and just pass in the datatable.
so, when i try this any postback will cause the datatable to be null.
hope this is a little clearer.
rod.
Rod -- so I think what you are saying is that you'd just have a DataTable property, which is assigned by the page using this user control. And your problem is that it is null on a postback.
It's going to be null unless you assign a value to it. Remember posts represent an entirely new life for the page. Everything is reinstantiated from scratch. So the property won't have a value unless you assign it.
In the override LoadViewState the DataTable property is null on postback. now, i'm assigning a value to the DataTable everytime inside the code-behind page_load and i notice when i debug it and cause a postback from the usercontrol the code-behind page_load doesn't even run. am i doing the assigning of the datatable in the wrong place?
Rod -- LoadViewState comes before Load. You'll have to either do it sooner like Init.
So instead of using viewstate["state"] i'm going to have to use session? because when i checked in OnInit viewstate wasn't available, which i guess makes sense since LoadViewState hasn't occurred yet. am i thinking right? and is this ok to do this way?
Rod -- why would you use session? Apparently I really don't understand what you're trying to do. Why don't you send me some actual code privately?
Thanks once again for taking time out to follow-up on all our comments. This has TRULY been helpful and I appreciate it very much. Also, thank you for the long and short ways to handle my specific issues, I now feel I have better direction.
Hey all,
So if i'm using dynamic controls and i'm hitting a database on every postback could i turn off viewstate for the individual controls that are being dynamically generated? could i disable it for the entire page except for the viewstate mentioned in the article's context?
Rodney -- in general, yes. Sometimes even if you are providing data on every request you need viewstate enabled because of other features of the control, but you have to take that on a case by case basis, and even then there's usually a way to work around it.
Rodney, thanks for taking the time to help people. It is greatly appreciated by many of us!
I'm trying to do something simliar but a little different and can't seem to get it working right.
I've got a gridview that has a template field with a place holder in it. In the rowdatabound event I'm finding the placeholder control and adding a dynamic control to it based on a switch conditional. Generates the control fine.
In the updating row event, I'm trying to store the user inputed data to the database. When I use findcontrol for the placeholder, it finds it. However, the dynamically added controls don't exist. How can I fix it? I'm fine with using a statically creating it (if it will work) and setting visible = false. However, i'm trying to save the results in a database, so I'm wondering if i'm going to have problems binding multiple controls to a single field? Here's some code in how i'm doing it now... perhaps you could give me some guidance in how to approach this better. Thank you for your help... it is much appreciated!
protected void EditableGrid_RowDataBound(object sender, GridViewRowEventArgs e)
if( e.Row.RowType == DataControlRowType.DataRow ) { // we are binding a "data" row, as opposed to a header or footer or empty row.
GridView gv = sender as GridView;
// parse the data column, using a regex
string answer_type = DataBinder.Eval( e.Row.DataItem, "answer_type" ).ToString();
PlaceHolder p = e.Row.FindControl( "answer_placeholder" ) as PlaceHolder;
if( p != null ) { // ...and we found the panel to hold auction links
switch(answer_type) {
case "r":
break;
default:
//build radio buttons and mark them according to answer in db.
RadioButtonList rbl = new RadioButtonList() ;
rbl.ID = "rbl";
rbl.RepeatDirection = RepeatDirection.Horizontal;
string answer_int = DataBinder.Eval(e.Row.DataItem, "answer_int").ToString();
rbl.Items.Add(new ListItem("Yes", "1"));
rbl.Items.Add(new ListItem("No", "0"));
if (answer_int != "")
{
rbl.Items.FindByValue(answer_int.Trim()).Selected = true;
}
p.Controls.Add(rbl);
}
}
}
protected void EditableGrid_RowUpdating(object sender, GridViewUpdateEventArgs e)
{
GridView gv = sender as GridView;
if (gv.Rows[e.RowIndex].RowType == DataControlRowType.DataRow)
PlaceHolder p = gv.Rows[e.RowIndex].FindControl("answer_placeholder") as PlaceHolder;
RadioButtonList rbl = p.FindControl("rbl") as RadioButtonList;
if (rbl != null) //HERE - RETURNS NULL!!!
if (rbl.SelectedIndex > -1)
e.NewValues["answer_int"] = rbl.Items[rbl.SelectedIndex].Value.ToString();
Hi there,
Excellent article - very informative.
I've been waiting for part 5 to ask this, but I'm out of time. I'm doing everything possible to make things difficult and complicated in an application (not without reason, of course, but still...)
I am loading several custom controls into user control that is dynamically loaded into an update panel. I got it working just fine like that, then I dropped the whole kit 'n kaboodle into another user control that calls DataBind() from the PreRender for it's own reasons. Suddenly ONE of the controls (which renders 3 dropdown lists for selecting hour/minute/second) decides that the SelectedValue is illegal 'cause it's not in the (hour) list. Other dropdowns built on the fly have been processed already, and more are still to follow.
The debugger catches here:
public override void DataBind() {
this.EnsureChildControls();// the dropdowns are added by a subclass in an overidden CreateChildControls method
base.DataBind(); // error is thrown here - the base class is Panel at this point.
setInputFromProperty();// takes the generic Text property and sets the rendered conrol accordingly -- e.g. a DateTime.ToString() gets converted into SelectedValues for Hour/Minute/Second dropdowns
the top of the stack trace is:
at System.Web.UI.WebControls.ListControl.PerformDataBinding(IEnumerable dataSource)
at System.Web.UI.WebControls.ListControl.OnDataBinding(EventArgs e)
at System.Web.UI.WebControls.ListControl.PerformSelect()
The debugger shows good values for all the dropdown lists and all the SelectedValue properties even after it has halted for the exception. These were set by the DataBind() method in the user control container that's part of this subsystem. The entire subsystem worked in two user controls that didn't DataBind() in the PreRender.
I'm completely mystified. Before I swamp you with massive code, or put a lot of work into recreating the problem in a simplified version I could post here, is there something you could tell me that might help me find a solution? I have considered simply not supporting DataBind() in the PreRender - i.e. ignoring the problem and altering the several controls that need this subsystem so they don't do that anymore - but that seems extremely sloppy, and like I might be sweeping under the rug a real problem I should be addressing.
Is DataBind() in the PreRender a dumb thing to do in the first place?
If you can't get back to me before I need to have this fixed, I'll post to tell you what happened.
BriaN
Hmm...
I fixed it, but I have no clue how. Maybe I just had a bad binary someplace... If I find out what went wrong I'll post, but it goes into the unsolved mystery pile for now.
Thanks for this blog, anyway, it rocks!
BriaN -- glad you fixed it, no help from me! I don't know what was wrong without really getting into it, but perhaps these tidbits of info will help you.
1. Databinding from PreRender is fine -- in fact thats when binding occurs for controls hooked up declaratively to a datasource control.
2. DataBind is recursive -- it databinds the control it is called on and all of its child controls.
3. DropDownList will throw that error if you set the SelectedValue before binding, and then bind to it a list that doesn't contain that value. Binding without setting the datasource first would fall under this too.
4. PreRender is top-down... so when you bind from PreRender from parent, it is before PreRender in child.
Thanks for replying so fast, and thanks again for the great article.
I haven't fixed it, I had commented out the DataBind() statement in the parent - but just as was going to test it an unrelated database problem happened and I simply forgot I had made the change...
If I don't call base.DataBind() in the base class of my custom control (which extends panel), I don't get the error, but then I can't set the SelectedValue of the dropdown properly because it's passed in a databinding expression in the markup (as the Text property), which is kind of the whole point of the custom control - hiding the internal differences between different kinds of input controls, heavy with client-side code, behind a uniform markup syntax.
I can't figure this out, and I need to find the answer in under 48 hours or punt and just disallow databinding in the prerender. Is there a chance of you being able to help me in that time frame?
Thanks again,
PHEW!!!
I found it, it had nothing directly to do with any of this, but was instead caused by expecting a Items.Add(String.Format("{0:00}",i) to add an item with text and value both of just a regular 2 character string with 2 digit numbers - instead, in the PreRender, the formatting stuff was still in the values, so the SelectedValue I set before of "00" was invalid.
The funky thing is, if I only ran DataBind once, I could treat them as "00", "01", etc. to no ill effect. It was only going back to DataBind() without a round-trip to the browser that broke it... very, very odd.
I now add the items this way instead:
Items.Add(new ListItem(String.Format("{0:00}", i),i.ToString()))
and it seems to be working everywhere.
Can't decide whether to feel smart for finding it or dumb for doing it in the first place now...
I really appreciate having had this public forum to force me to be more rigorous in defining my problem, it helped a lot.
Thanks, and I can't wait for part 5,
Few days ago I've found a custom control, where child contols were created and added to Controls collection in constructor. Certainly, that isn't good because Controls property and AddedControl method are virtual. But - is that the only reason why we must not add child controls in constructor?
Alexander -- I'd at least create them from CreateChildControls and call EnsureChildControls from the constructor. There's no technical reason why you should avoid doing it that way though, other than the virtual wierdness you are referring to.
Just stumbled upon your blog yesterday, looking for information on dynamically creating controls. Some very excellent posts!
I realised that what I was doing was better done with a repeater and a template. So far so good.
I'm creating DropDownLists dynamically based on the user's choice in a first DDL. This DDL is static and the selected index determines the DataSource which determines the number of dynamically created DDLs.
The problem occurs because DropDownList1_SelectedIndexChanged runs after OnInit/OnLoad. If I understand it correctly, the new selected index of the first DDL is not set until DropDownList1_SelectedIndexChanged is invoked, which means my DataSource for the repeater is not setup until after OnInit has run.
Is the only solution to check in OnInit whether the DataSource is ready, skip setting up the DDLs and then call the method to set up the DDLs from DropDownList1_SelectedIndexChanged?
Cheers and thanks for a great article series.
Neil
Neil -- the selected index should be available in OnLoad (before the event).
Try this -- always create the child DDLs from OnLoad, based on the SelectedIndex of the first DDL (forget about SelectedIndexChanged, you don't need it). The child DDLs are within a repeater, which has ViewState disabled.
This means binding the repeater every request and therefore getting your data every request. As long as your data is relatively cheap to retrieve that should be better than using ViewState.
But if you'd rather ViewState then just binding the repeater from SelectedIndexChanged should do the trick. You didn't really say whether it was working, just that you seem not to like the order of things?
Thanks for your reply. I see your points and it's working now. I realize I was doing a bit of unnecessary extra work.
My only concern now is that the dynamically created DDLs will be too difficult to setup when they are created in the Repeater. I have to run through each DDL and setup their DataSources. The DataSources also depend on the Selected Index of the first DDL. I'm guessing this will have to be setup the hard way on every postback anyways. So I don't know "how many" controls and I only sort of know "what" controls. I know it's DDLs, but not what to DataBind them to. I'm in doubt whether I should still use the Repeater...
I still think you should use a repeater. You know you can set the DataSources of the child DDLs declaratively?
<asp:Repeater>
<ItemTemplate>
<asp:DropDownList
GetDataSource would be a method you define in the code-behind (of at least 'protected' protection). The parameter will be the data item for the repeater at the time. If you can manage to squeeze it all in the <%# %> expression then you wouldn't even need the extra function.
"Until next time.... part 5 will come much sooner than part 4 did, I promise." (c)
So where it is? ;)
It's coming it's coming! I still have time to live up to that promise, I think. I've been working on something else lately,
Excellent articles! I have been significanty struggling trying to get a RadioButtonList to work within a GridView. I downloaded your example and unfortunately that does not help me much. Here is my situation:
I am building a online survey and looking to populate a GridView with a list of questions. In the ItemTemplate, I am trying to pull in the list of Choices/Responses into a RadioButtonList. I need to obviously dynamically create these because all this is driven by a DB backend and the total list of questions as well as each set of Choices/Responses can vary. So questions typically have 2 to x number of Choices/Responses. I have paging setup on the GridView (set to 10 for now) to pull in the questions. I want to scroll through the GridView and extract the set of Choices/Responses from a separate DB table and display to the user. Obviously viewstate is an issue here as well. I have tried to follow your article(s) on how to accomplish and have been unsuccessful.
Can you please provide an example (hopefully in VB) that addresses a RadioButtonList embedded in a GridView please?
Much appreciate that!!
The problem (wrong file is deleted) can also be avoided by simply deleting the file in event handler and reloading the page:
Response.Redirect(Request.Path);
Of course no message for the user can be displayed (at least without effort).
I just wanted to say thank you. Not only is the article excellent, but your responses to the questions throughout the series clear up the issues even more.
Again. thanks!
Hey Dave,
I've tried to use your Part4SampleCode as a template for my page and I'm having a small problem with the way I'm trying to do things:
Posted Code Here:
I'm probably missing some concepts here so please bear with me. I am trying to create a dynamic gridview. Try to imagine 2 tabs on a page. I click the second tab the correct GridView loads fine. However, when I click the Select button for row nothing happens but when I select it again the row is then selected.
any ideas?
Rodney -- try giving the GridView a specific ID, preventing it from assuming an automatically generated ID a 2nd time when the tab changes.
Also -- you don't have to do everything so dynamic. Much of what you have in that dynamic code is always the same. Consider declaring the grid and everything about it that is always the same. Then create everything else dynamicaly, like your fields, etc. Might side-step the whole issue.... :)
Thanks Dave as usual.
Dave,
Could you see this semi-Dynamic GridView working as a UserControl like in the Part4SampleCode?
In the UserControl Page_Load I would:
1. Access any db table based on a condition
2. Build the template columns from db fields
3. Databind
could this be a possible workflow?
Rodney -- sure, in general creating a user control isn't any different than creating a page, if you're making it a self-contained system that doesn't require any data or method calls from the page its hosted on. If you do need to give the control data from the page, such as your tabindex or something, then make it a property on the control and allow the page to call databind when it wants the data reconstructed.
uc1.SomeInformation = info;
uc1.DataBind();
Ok, I declared the gridview and the Select is working for the GridView. When i select either RadioButtons it loads a GridView based on a different datasource. i can select a row press edit and it goes into edit mode fine.
my problem is what if i wanted to changed the selected index value of the GridView back to -1. How can i do that from the host? should i make my gridview object a public proberty inside my user control1 ?
Could you please take a look at my UserControl1.ascx and see if it looks ok so far. Again this host is exactly the same as Default4.aspx in your Part4SampleCode:
Yeah you could make it a property, or you could have a method like SetSelectedIndex or ClearSelection, either way.
Keep in mind since you are rebinding the grid each time you may be wasting viewstate. Normally I'd just say you should disable viewstate on the gridview, but I don't think its selected and edit indexes work without it... not sure... try it. If not then you should only perform the databinding and template construction when the view changes, so you are at least utilizing the viewstate you are storing.
Surprise...I have a question...
I posted my current UserControl1 and was wondering how difficult would it be to add an inline Insert to my dynamic gridview?
would the new line have to occupy the footer row? how do you dynamically add a template column to a footer row?
Hi Dave,
Scenario: Multi-tab UI with different info/modules presented on each tab. Tabs are dynamically generated depending on user action (e.g., selecting menu command which exists outside of any tab or clicking on an edit link in a list from a generated tab). My initial approach was to create a user control to encapsulate each info/module. On user action, I create a tab, then Page.LoadControl the associated user control to embed it within the associated pageview of the tab. The problem is that the loaded control disappears upon postback. I can re-load the user controls on postback but I have to maintain the state of the user controls. For example, think of a data entry module that is loaded onto one of the tabs. User enters information but keeps the tab "open". User then selects a menu command which posts back and launches a new module in a new tab (which is set as the currently active tab). User switches back to the previous data entry tab. This tab should maintain any info that was previously entered by the user. Any thoughts?
watson -- as long as you are loading the 'module' and putting it in the control tree, it doesnt matter whether it is visible or not, it will maintain its state perfectly. So as long as a tab is active in the sense that the user may not be done with what is on it, you should be loading the control that belongs to it. Its just that the 'active' tab is the only one whose module is actually visible.
I Truly apologize for the bombardment of questions several days ago. I do appreciate your replies and value them.
The last question i figured out that i'm just going to add a blank record to the db and the call it back up in edit. good enuff for now. i'll come up with something.
Could you please spot check something for me. When i go into edit mode on the gridview there's a custom control i built as a date picker. i've greatly simplified the control for debugging purposes. The problem i'm having is i can't toggle the calendar on and off with my button. i'm not sure why it's not persisting as it should. i understand if i've reached my limit of posts.
Given:
Default4.aspx of Part4SampleCode (untouched)
UserControl1.ascx
CustomControl
p.s. sorry :)
Rodney -- I assume ViewState isn't turned off? When you say it won't persist what do you mean exactly... does it do anything or is clicking the button just causing a postback with no observable effect?
Don't apologize I do enjoy helping people... I only regret that I'm not more responsive. I will eventually answer most inquiries, even if its weeks later.
Thanks Dave for your patience and efforts...
When I click on the button to show the calendar that part will work. However, it's when i click the button again to hide the calendar is what's wrong. The calendar doesn't disappear from that point on.
Great post(s). Long time ago passed from Part 4...when we can expect Part 5 which covers server controls?
Thanks for your excellent posts on the Viewstate. Keep up the good work. I would love to see more indepth articles like this from you and your team. Thanks again.
I have a few basic things I am still not clear on. What has been most confusing to me about asp.net in general is the order of events in the lifecycle. I understand from reading your articles that when a control gets added to the control tree, it automatically plays catchup to the point right before the current event in the lifecycle (pretty cool stuff IMO). I am still confused on overriding methods in general.
1) There is the virtual method Render(HtmlTextWriter writer) which can be overriden by classes that inherit the Control class. When we do that, are we required to call the base render method (is there stuff that the parent method is doing that is required in the render process)? If so where do we place our custom code? This question does not only apply to only the Render method but to any virtual method.
protected override void Render(HtmlTextWriter writer)
//Custom code here?
base.Render(writer);//Is this required?
//or here?
2) We have the Init event and the OnInit method. Should we be handling the Init event or overriding the Onint method? Why?
public _Default() {
this.Init += new EventHandler(_Default_Init);
void _Default_Init(object sender, EventArgs e)
//Cusom Code
I guess the answer is overriding the Oninit method. If so, should we include our custom code before or after calling the base OnInit?
protected override void OnInit(EventArgs e)
//Custom code here?
base.OnInit(e); //Required for base to inform other subscribers.
//or here?
3) Controls have properties. It is recommended that anytime you get/set a property, you should call EnsureChildControls to guarantee that the control tree is created. EnsureChildControls internally calls CreateChildControls. Since properties can be set/get anytime during the lifecycle, does this mean I may or may not have ViewState available in CreateChildControls? Example: I set a property in the Init stage.
4) When overriding CreateChildControls(), am I always required to call Controls.Clear() first? Is this because a post back could change something causing the control tree to be rebuilt? Should I always call base.CreateChildControls()?
5) Controls are initialized during the Init stage. This loads the default properties of the controls. How does it do this (I failed to find a method where this is done using Reflector)?
I would really really appreciate it if you could answer these questions.
Thanks in advance!
Andy
Thank you for your excellent post about viewstate. I have questions below.
I have textbox inside the repeater and how can i maintain the data key in inside the textbox by users after postback?
Thank you again.
Thanks for writing these articles and all the answers you provide us with, still I
got don't get my example to work. ;)
I have a repeater containing a User Control. To keep it simple this user controls only contains a textbox. I populate the repeater using a Arraylist containing 5 items (Text 1 to Text 5). The user control exposes a public property called myTest. In the ItemCreated event of the repeater I create an instance of the UC and set the public property. This all works fine. But, and I think you already know what I'm going to ask, when I click a button on the page to post-back. The textboxes are all rendered again, but empty. To me it seems the viewstate of the UC is not maintained. So the repeater keeps track of the number of controls rendered, but the controls itself don't have the data anymore. That is one question. I also want to add a this UC dynamically when I add one of the buttons on the page. But it is possible that data already in the repeater, has changed. How to accomplice adding a new item and keep the changed values of the other items. In the PDF document I read that you suggest first retrieve the changed values put them in a new datasource then rebind again. (Item 4585429 by Rob).
I hope you understand what I want to achieve, any help is appreciate.
Thx
Stephan --
ItemCreated is called on every postback, even when you aren't databinding again. So on a postback you'll end up creating the UC and then setting its text to something not from the data, since the data isn't available anymore.
What you want to do is either just put the UC in the ItemTemplate statically, or break it up and use the ItemDataBound event. You create the UC from ItemCreated, but you only assign its public property from the ItemDataBound event (alternatively you could hook into the DataBinding event from ItemCreated, and set it from the handler).
As for your second question... yes I recommend you retrieve all the data and rebind!
Andy --
These are all excellent questions. So excellent in fact I plan on a blog entry to answer some of them, as soon as I can...
To give you a quick answer to #5... you can't find it, because that's not where it happens. The framework builds methods on the fly when it compiles the markup, and they are what assign the property values, which is even before OnInit.
I am using a repeater to allow people to enter their names etc. for reservations.
The repeater currently uses a simple array to control the number of rows built by the repeater and the size of the array is supplied by the User when specifying the number of people requiring ticket reservations.
The entered data will be stored in the database when a User clicks on a button to finalise the reservation process.
Now, if someone wants to change the number of people or delete or add a row before clicking on the final button, how do I save previously entered data and bind it when the repeater is rebuilt during postback?
Should I base the Array on a Structure and store the Array in a SessionState variable? Can a Data Table or Dataset object be created in memory and stored in a SessionState variable and bound to a Repeater?
What's the most efficient way to achieve this?
Michael Holberton
Hospedaje Los Jardines & Sacred Valley Mountain Bike Tours
Cusco Database Development and Cycling Services E.I.R.L.
databaseservices.blogspot.com
Michael -- I suggest that when one of these operations are needed (adding, deleting, etc) and you want to save data already entered by the user, you first 'save' the data -- but not the database. You probably have a method already that saves the data to the database by basically copying it directly from the form to the database. Rather than that, refactor it into two parts. The 1st part creates an array of data items representing all the data (using a custom type or whatever), the 2nd part saves it to the database FROM that array of custom types. This provides a clean separation of UI and business logic. And it also gives you for free a way to save the existing data TEMPORARILY. So when the user wants to add/delete, etc, you first save it into the array, using the 1st part of the two methods I described, then you manipulate the array, then re-bind it to the repeater. No need to save anything in session.
Thanks for the reply Dave!
I'll try using an array based on a Structure as I am thinking that should help to self document the code instead of having to review the code each time to determine which field is associated with which array item.
Michael
I sent a message to you just the other day, but I'll make this one public.
I really only want to do a simple thing:)
I have and update panel with a placeholder inside. I Have a button outside the UdPanel that acts as an AsyncPBack Trigger.
I have designed a UserControl (a <tr>) which contains a few DDLs and Text boxes. It also contains a button called "Remove".
The idea is that clicking "add" adds rows dynamically and obviously, Remove, removes the row from the PlaceHolder.
Adding works to an extent but seems to muck up the Control hierarcy. Removing also does this. Data appears in the wrong controls etc. A control exists on the page when it shouldn't etc.
Added to this, I wanted to have OnSelectedIndexChanged events working on the DDLs in the UserControl.
Its quite annoying. I have done this with Javascript but I wanted to use server controls to simplify the posting of data plus I wanted partial postbacks to make it quicker.
Anyone have any ideas?
ezrider -- they key is to make sure the controls are added each request. For example if you click add, and then do an async postback via some other button, the control should remain there. It won't be if you aren't adding it, despite the add button not being clicked the 2nd time. But that's not the only thing you would need to worry about. You are probably getting 'messed up' data/hiearchy because you're getting a shifting-id problem, as explained in this article. If you have 3 rows and you remove the 2nd one, the 3rd row is going to have its IDs shift. You can solve that using the techniques I outlined. The simplest way for you may be to just make removed rows invisible rather than actually remove them.
Since you are adding the same control each time, it would be much cleaner if you just used a repeater:
<asp:repeater
<itemtemplate>
<abc:usercontrol
</itemtemplate>
</asp:repeater>
You would databind to this repeater a collection of data items that could contain data required by the user controls (or just a blank array if you only care about the number of them). The tricky part is that you must rebind the repeater whenever you add or remove an item, so you must be able to re-create the collection from the repeater's items, in order to preserve the data the user has already put into it so far.
Hey Thanks mate!
My "add row" functions were working fine. It was really only the shifting Ids problem with the Remove that was giving me a head ache.
I'm not sure how I would solve that problem despite the solutions in your article( I had problems with NamingContainer and fixed id names).
I had written a custom object that has the same structure as the control, allowing the values in each row(filter clauses) to be stored in the DB and then recreated when a query was loaded for editing.
Since I already have a Collection of objects maintaining the row values, the repeater approach would be best. It certainly looks cleaner!
I really appreciate your helping me. Great Blag too.
Many of the comments I've received in the various dynamic controls entries I've written have been questions
hi.
I was just reading through ur articles. It all good, but I think I need the #5 for my problem.
I'm trying to create dynamic "multi row edit" gridview with template. Why do I want it ? cause I would use the gridview as main source for lookup data for user.
well, of course most of it I knew what columns that I would need, but I would need the same grid view to look up different data (which of course had lots of different columns on size and format).
And i was thinking to even take it further to make some of the columns editable (and it is multirow in a sense).
I have successfully created it to be loadable, and editable.
what I need now, I wanted to get the data after the user edit it, while the controls would always gone after the postback.
what should I do to retrieve the data from the dynamically made columns on gridview ?
Fire -- columns are a StateManagedCollection. That means if you dynamically add or remove a column, it will persist that way on its own without you having to redo that operation on a postback. So I'm not really following what your problem is. You say you want to get the data the user entered, but the controls aren't recreated? They should be created by the grid view automatically based on the columns and their templates. Perhaps you are rebinding it or something?
I have solved it already.
So, it seems it was just my misunderstanding of asp.net state.
After reading ur viewstate article, and discussing with my friend, I have solved the problem.
So, dynamically created column template should always be recreated on gridview_init event (and the cool thing is their value persisted cause of the magic viewstate from the .net framework).
after, that, then I can access the data following the postback event because the controls had been recreated from the init event.
so, things are good for now.. Still perfect-ing my custom gridview though..
yeah, I know that u and some people here said (DONT USE CUSTOM CONTROL)...... :P
but this custom gridview would helped me alot on my projects cause I would definitely need it.
thx anyway for the great article :)
hi again infinities.
I have another problems that I wanted to consult with u.
so, I tried to make a custom textbox to be used for my custom dynamic template gridview.
here's some code from the textbox :
Public Enum TextBoxTypeEnum
Text
Money
End Enum
Public Property TextBoxType() As TextBoxTypeEnum
Get
Return ViewState("TextBoxType")
End Get
Set(ByVal Value As TextBoxTypeEnum)
ViewState("TextBoxType") = Value
GetDefaultFormatText()
End Set
End Property
Public Property FormatText() As String
Return ViewState("FormatText")
Set(ByVal Value As String)
ViewState("FormatText") = Value
Protected Overrides Sub OnLoad(ByVal e As EventArgs)
Page.ClientScript.RegisterClientScriptResource(Me.GetType(), "AVPControls.AvpTextBox.js")
If ViewState("TextBoxType") = TextBoxTypeEnum.Money Then
Me.Text = String.Format("{0:" & FormatText & "}", CDbl(Me.Text))
End If
MyBase.OnLoad(e)
End Sub
Private Sub GetDefaultFormatText()
If FormatText = "" Then FormatText = "#,##0.00"
ElseIf ViewState("TextBoxType") = TextBoxTypeEnum.Text Then
FormatText = ""
The idea is just that if I choose type money from the textbox then it would convert my text into numeric format (#,##0.00) in this example
now the problems :
1. the getdefaultformattext made my textbox error because it seems that it cannot format my text on that function. So, I decided to play with it a bit and changed it to :
Me.Text = "123"
now, my textbox rendered successfully, and when I tried to change the type into money, it then change the formattext property and my textbox text property. The problem is that the text property DID CHANGE but only on properties window, and it DIDN'T CHANGE the text on my textbox text in design view (on the design visible form).
2. I gave up trying the above problem (if I just deleted that text changer on getdefaultformattext function, the textbox did run successfully and change the text on runtime, so it's not really big deal, though I really want to know how to change the text on design time too). Now, the second problem is when I tried to use my custom textbox on my dynamic template gridview. Here, my numeric format wouldn't work. So, I thought maybe the onload didn't get called. So, I tried to add msgbox on onload method just to know the truth :
msgbox (me.text)
the msgbox return the RIGHT TEXT (that is the numeric formatted text) successfully. But the text on my textbox didn't change at all. So it seems te problem from number 1 occured again here. Could u suggest me how to fix it ?
thx
regards,
Fire.
hi infinities,
no answer for the 1st question yet ?
as for the 2nd question, sorry, again, it's part of my fault of my (still) misunderstanding of viewstate ordering process.
Fire -- the control is kind of confusing! The format text is set in so many different places. If you format the text once, then submit again, you're going to end up trying to format it twice. I'm not sure if the Parsing will work or not. Ideally you'd just output the formatted text from Render or AddAttributesToRender instead of actually CHANGING the text value. I wish I could help more than that at the moment but I gotta go! We might be having a baby today...
Your articles are so in-depth and have really pulled through for me, great work.
Could I get some pointers on using OnLoad and OnInit, I've read your onload-vs-page-load-vs-load-event article.
My Server Control consists of a GridView which is built in the OnInit() method. However the databinding for the gridview is called from the OnLoad() method which in turn calls a method RunQuery();
RunQuery() returns a DataTable which is used by a method named BindData() to bind the gridview but also create a Table using the DataTable column names for the headings and a TextBox control for each gridview column.
As RunQuery() will need to be called multiple times, it will return different sets of data, I'm clearing the Table and recreating the Table Header and Rows. However this occurs on every postback due to the RunQuery method being called from OnLoad()
Sorry if this seems a bit incoherent, I've probably confused my self ;-)
Just read your comment about having a baby, all the best and take care.
Regards,
I have a rather odd problem
I am loading an external (skin template) and applying it to my server control when I dynamicly load it onto my page.
( it's a plugin based site editor, I can't use static controls )
The wierd thing that is happening is that all of the controls will not persist thier state unless they are visible on the page, meaning if I have them in a multiview in my template file, only the visible controls persist thier values, from post back to post back event.
protected override void OnInit(EventArgs e)
this.EnsureChildControls();
protected override void CreateChildControls()
if (!_skinLoaded)
_skinLoaded = LoadDefaultThemedControl();
if (_skinLoaded)
AttachChildControls();
}
protected virtual bool LoadDefaultThemedControl()
if (this.Page != null && this.DefaultSkinFileExists)
Control defaultSkin = this.LoadControl(this.DefaultSkinPath);
defaultSkin.ID = "_";
this.Controls.Add(defaultSkin);
return true;
return false;
In my server control class I am finding the controls and wireing them up prior to the host controls onInit so the order should not be screwing me up, additionally the controls do work when visible ???
I am not really sure, but I am sure it is something stupid I am doing :)
Any help would be greatly appreciated.
You know after 5 hours of looking 5 minutes after I submited my question I figured out that I failed to specify the ID of my host control. :-(
Well the good news is it works just like it should and I thought it would, the bad news is I wasted 5 hours :)
Oh well I am just glad I found it.
I've read these articles and you seem like the best person to ask for this.
I've got a web page that holds a user control, which is really a factory that might load up one of several different user controls inside it. The web page doesn't really know what kind of control is getting loaded, and it doesn't care. It just passes the data in to the factory/container control and trusts that guy to load up the right control for the content provided.
This all works great, and I've used it in multiple projects. But I've always just had the child controls hold literals. They never held any interactive controls like a textbox. And now, I need them to do that, and I've discovered something that doesn't work the way I expected.
I'll post the code below, but basically, here's what's happening:
The default page has a textbox, a button, and a factory control. You enter text into the textbox and hit the button.
What should happen is that the page reloads. Then the page takes the text from the textbox and hands it to the factory control. The factory control loads the child control, and passes that text to the child control to display. The child control then displays that text twice - once in a textbox and once in a literal.
If you set a breakpoint in the child control, you can even see that the textbox.text shows the correct value. BUT, when the page loads, only the literal will actually display the value. The textbox itself is blank. Why is it doing that?
Here's the code:
------------------------------------------------------------------------------------------------------------
<%@ Page Language="C#" AutoEventWireup="true" CodeBehind="Default.aspx.cs" Inherits="Dirtbox._Default" %>
<%@ Register src="~/FactoryControl.ascx" TagName="edit" TagPrefix="admin" %>
<html>
<form id="form1" runat="server">
<asp:TextBox
<asp:Button
<admin:edit<br />
</form>
using System;
namespace Dirtbox
public partial class _Default : System.Web.UI.Page
protected void Page_Load(object sender, EventArgs e)
containerControl.PutStuffIntoUserControl(sourceText.Text);
<%@ Control Language="C#" AutoEventWireup="true" CodeBehind="FactoryControl.ascx.cs" Inherits="Dirtbox.FactoryControl" %>
<asp:Panel
public partial class FactoryControl : System.Web.UI.UserControl
public void PutStuffIntoUserControl(string useThis)
MyUserControl subControl = (MyUserControl)LoadControl("~/MyUserControl.ascx");
subControl.myText = useThis;
subControl.PopulateInteriorControl();
myPanel.Controls.Add(subControl);
<%@ Control Language="C#" AutoEventWireup="true" CodeBehind="MyUserControl.ascx.cs" Inherits="Dirtbox.MyUserControl" EnableViewState="false" %>
<asp:TextBox<br />
<asp:Literal
public partial class MyUserControl : System.Web.UI.UserControl
public string myText;
public void PopulateInteriorControl()
textWithinControl.Text = myText;
literalWithinControl.Text = myText;
Thank you for all the time you put into this, helping us out with things that we can't really find in any book. You should write one, by the way:). Looking forward to your next articles.
And congratulations for the baby!
SeanB --
Here is why I think. When the page loads for the first time, you are loading the child control with empty text. On the postback, you load it with 'newtext' or whatever you typed in. You assign this value to the textbox.text property. The textbox did not exist in the control tree before page load, since that is when it is being dynamically loaded. Normally textboxes load their text before page load. It missed out. But the framework knows that many pages will be loading controls dynamically from page load and that they might have postback data to load -- so, there is a 2nd phase in which postback controls load postback data, AFTER load. The textbox then loads the postback data, which is what the textbox had after the 1st time the page loaded -- BLANK! So the value is then reset back to blank.
To avoid this you must either give each distinct possible user control a different id (which is stable and remains the same if the text in the textbox didnt change, but is different for each possibility), which results in each textbox getting a different unique id, or you must do this loading in a two step process: first, you always load the control you had last time, by utilizing a viewstate field to remember what the textbox value was; second, you respond to new data by handling the buttons click event, at which point you remove the old child control and insert the new one, this ensures the first one gets the last request's postback data, not the new one.
I suggest you take a look at my dynamic controls by example post. It doesnt directly solve your particular problem, but has some very similar things in it.
So the 'real' solution uses a technique that relies on viewstate, where you are always sure to loa
Great series! It's answered alot of my questions about controls.
One thing I've discovered in my recent testing is that if you add a control dynamiclly to a part of the control tree (in OnInit) that is further down the line, it doesn't play catch-up. Which, if you think about it, makes sense since as InitRecursive() works through the tree it will eventually get to that newly added control.
The example I was working on had two content place holders on a master page. The first content place holder loaded several user controls onto the second content place holder control. I was expecting the dynamically loaded controls to init right away but they didn't.
Sorry if this is a restatement of something in this series but I don't think it was there.
Jon -- yes, if you manage to add a control to one that hasn't init'd yet, the added control won't suddenly init either, but will eventually when the normal init phase reaches it. The 'catch up' feature only comes into play if the parent you are adding the control to has already been through a phase.
This is a great series. Helped me understand.
I have a requirement where i create controls on the fly based on database values. The controls are not a set type, there can be textboxes, checkboxlists etc. and i don't know how many there will be either.
I want to load the controls based on a db condition. Then, when i click on a button, i want to save the control values and then load a next set of controls depending on a db condition, kind of like a wizard. Right now i was loading them using a user control in the Page_Init of the control.
Seems to work but it's not working very well. When i click the button, to submit the control values, it re-renders the previous controls, saves the values and then loads the next set using the Controls.Add. However, if i go to view source for the page, it still shows the previous controls in the html view.
Do you have some good suggestions on how to make it work the best way. I'm not sure how to implement a repeater solution, though i'd like to if possible with this kind of scenario.
Thanks for your help and suggestions.
Cristian
Cristian -- my article on understanding dynamic controls by example might help. It sounds like you have the right idea. You just need to remove the old control when moving on to the next one, with Controls.Remove or Controls.Clear. You also then should give the control you are loading an ID based on something short and concise in the db, such that each possible control will have a different ID but one that is always the same. This avoids the 'shifting id' problem I've talked about.
Hi, thanks for the quick reply. There seems to be an issue when using this type of functionality with Ajax. Turning off ajax makes everything work as expected but with ajax, my labels(Literals thanks to another one of your article saved me loads of time on css) don't render upon an async postback. The other controls that have id's such as textboxes render properly....well except validator controls which also don't seem to work. I tried giving them unique id's such as lblMyTitle + new Random().Next(100) but didn't help. Any ideas.
In my page, i am removing the user control where the questions are created using the Controls.Clear -- controls is mapped back to a placeholder control. After Clear i'm calling a LoadControl to load back that user control which should now have the new set of controls.
Cristian -- I'd have to see code or a minimized version of the code. Simply turning on ajax shouldnt cause problems like that if everything is done correctly. By 'turn off' I assume you mean setting EnablePartialRendering="false"? could you provide more details?
Hey man, these articles are very nice, however due to my very limited experience with .NET and developing webbased applications I cannot directly find a solution to the problem I'm having. Maybe you can help me a bit?
I posted the issue over here (code included):
forums.microsoft.com/.../ShowPost.aspx
It's kind of a story but it's not too difficult... You seem like someone who could push me in the right direction! It would be great if you could just take a look and share your first thoughts on the matter...
Man, you are still getting hits and I'm not surprised. Hey, I just wanted to run something by you and get your take on it.
I'm beginning to understand when you said that a lot of the times when people want dynamic controls the solution doesn't really require the degree of dynamic control one thinks (if that makes any sense).
anyways, i'm taking advantage of the ITemplate interface by creating my own class and implementing the interface. and my idea is that i load an xml file to determine the presentation of the fields. what do you think about that idea?
for example, as the dynamic template is being created, there's this xml file pre-loaded keyed by fieldname. And when it's that fields turn to go and look up the instructions in the xml file like whether it's a textbox in edit mode, if it needs a format string {0:d} for dates, if it needs javascript appended? stuff like that.
rodchar -- sure, could be useful as a generic form generation control. But be careful that you aren't actually reinventing asp.net markup with your xml. In other words, why create custom XML, and a custom parser, when the output is simply a control tree -- why not let the xml be actual asp.net markup, and your 'parser' is simply LoadControl()? Remember you can load an ascx file and use it as an ITemplate with LoadTemplate.
dude! that was good read.
i've been struggling with loading a custom web control through an XML config file.
i've tried a ton of stuff, but just can't stuff to work right. you clued me in on a few important points that i will attempt.
i'll let you know how it goes...if you're interested that is...i'm sure you get slammed with more questions than you care to talk about.
at any rate....
THANKS!!!!
-
cT
thanks a lot
Can I pick your brain for a minute?
I have a number of the same usercontrol(which consists of a checkbox and an image) that I need to load at runtime. I declare a default control(of the same usercontrol type) in the .aspx page initially because the default always needs to be there.
By default, the default control is checked and once all the other controls are loaded, if the user clicks one or more of the loaded user controls, the default control should uncheck, etc.....on the same note, if none of the dynamically loaded controls are checked, the default should check, etc...
I almost have this working but I can't quite get the functionality I've described above. I suspect I have the event handlers wired up wrong or not implemented the correct way.
The reason I feel I have to do it like this, is that if we declare all of the control types in the .aspx page, we have to republish the site code...yuck! So, I have an XML file that describes the control and I load them dynamically....that way we can add or remove from the XML file as we please.
At the moment, I load all of the controls(including the default control) in a panel/placeholder.
Can you possibly shed some light for me and/or is there a better way to approach this?
Thanks a lot...I appreciate it!
CrazyTasty --- hilarious name, by the way. Its making me hungry every time I see one of your comments!
Anyway -- sounds like you're on the right track. I dont know exactly what is wrong without more detail. The way I imagine doing it is having a helper method like SetCheckedImage(Control) where the passed in control is the one that should be checked. The method would first enumerate all the loaded controls and set their checkboxes to false, and then finally set the passed-in control to true. If the passed in control was null then it sets the default control to checked.
Then, in the user control, whatever it is that causes a postback (checkbox autopostback?), calls that method passing itself when the checkbox changes to true, and null when it changes to false.
Theres probably a faster way to do it -- this will mean enumerating all the controls whenever anything happens, but if you only have a dozen or so of them I doubt its worth trying to optimize too much.
Thanks man. I will try a few things tomorrow when I go in and let you know.
p.s. you remember the commercial product that had the tagline, Crazy Tasty!!! ...don't you? :-)
Bloody vikings!
Thanks for this article.Its an excellent article on Dynamic Controls.
But i have some trouble in working with dynamic controls.Here is the situation I have one dynamic Web user control named WebUserControl1.ascx.Its type name is same (i.e WebUserControl1) in the code behind file and its public , but when i want to access this type in default.aspx page then compiler gives error : "Type or namespace WebUserControl1 could not be found.are you missing an assembly reference?"
Why i am not getting this public type in default.aspx while both user control and default page are on the same level in web site and both have no namespace(means no namespace is specified for both).
Once i tried to access this type using ASP.webusercontrol_ascx , at one time this worked but at other time it does not work.What can be the problem with this .Remember i am working using VS.NET 2008
He Zee,
I had this problem, but when I put the
<%@ Register Src="~/WebUserControl3.ascx" TagName="Item" TagPrefix="uc1" %>
at the top of my page, everything works fine. If you put it at the top you can use it in the code behind as a typed control and access it's properties. Well maybe you should cast it first.
Thanks again for the update. Sorry i've been away for a while and i just saw your message.
I have attached some my code so you can see what i'm talking about.
By taking ajax off, i meant literally removing the update panels and all that.
Here's a smaller version of the code i'm using.
////////////////////////////////////////////////////////////////////////////
protected void Page_Load(object sender, EventArgs e)
((HtmlGenericControl)Master.FindControl("SiteMasterBodyTag")).Attributes.Add("onunload", "javascript:__doPostBack('PostBackLeavingPage', '');");
BindRepeater();
//set validation group of the currently loaded section
lnkContinue.ValidationGroup = "Section" + loadSectionID + "Validation";
LoadQSControl(loadSectionID);
//this is not working that great
protected void Page_SaveStateComplete(object sender, EventArgs e)
if (IsPostBack)
object eventTarget = Request["__EVENTTARGET"];
object eventArgument = Request["__EVENTARGUMENT"];
if (eventTarget != null && eventTarget.ToString().Trim().Equals("PostBackLeavingPage"))
SectionQuestions sq = (SectionQuestions)QuestionsPlaceHolder.Controls[0];
//sq.saveAnswers(false);
private void LoadQSControl(byte sectionID)
lnkContinue.ValidationGroup = "Section" + sectionID + "Validation";
SectionQuestions control = (SectionQuestions)LoadControl("/profile/controls/SectionQuestions.ascx");
control.ID = "SQ" + sectionID;
control.sectionID = sectionID;
QuestionsPlaceHolder.Controls.Clear();
QuestionsPlaceHolder.Controls.Add(control);
protected void listSections_ItemCommand(object source, DataListCommandEventArgs e)
if (e.CommandName.Equals("GoToSection"))
byte sectionID = Convert.ToByte(e.CommandArgument);
lnkContinue.ValidationGroup = "Section" + sectionID + "Validation";
LoadQSControl(sectionID);
////////////////////////////////IN THE CONTROL ITSELF///////////////////////////////////
protected void Page_Init(object sender, EventArgs e)
if (sectionID.Equals(0)) return;
RenderControls();
//RenderControls simply does a Controls.Add of the various controls such as checkbox, textbox etc.
//cutoff version of the saveMethod.
public void saveAnswers(bool continueToNextSection)
{
ArrayList relatedQuestions = new ArrayList();
if (Controls.Count != 0)
for (int i = 0; i < Controls.Count; i++)
if (Controls[i].ID != null)
if (Controls[i].ID.Contains("AnsForQID"))
short questionID = Convert.ToInt16(Controls[i].ID.Replace("AnsForQID", ""));
UserAnswersCollection uaColl = new UserAnswersCollection();
if (uac.Count != 0)
{
UserAnswersCollection.RemovePreviousAnswers(accountID, questionID);
}
string typeName = Controls[i].GetType().Name;
switch (typeName)
case ("RadioButtonList"):
RadioButtonList radList = Controls[i] as RadioButtonList;
AddUAEntityToColl(uaColl, radList.SelectedValue, accountID, questionID,false);
break;
case ("DropDownList"):
DropDownList ddList = Controls[i] as DropDownList;
AddUAEntityToColl(uaColl, ddList.SelectedValue, accountID, questionID,false);
}
}
}
}
//////////////////////////////////////////////////////////////////
I am not very experienced with .net.
I want to build a form where the meta data in the datbase dictates whether to display a checkbox, radio button, textfield, text area or select box? The exact labels/questions and the number of each type of elements is determined based on the specific survey or question set.
Does this have to be done using dynamic controls? Can this be done using repeaters/templates as described above?
Mike
Great article! In the case of sharepoint, specifically an editor for a web part, is it possible to load a .ascx control within that control? There is no page file; only .cs files.
Consider the following: I have a web part that has a variable number of properties that depend on a couple of other properties that are alway present.
Very similar to the property grid in visual studio. In visual studio when you focus on an item, the property grid binds to that object and dynamically populates the grid based on the items properties. Some of these properties are complex and require the use of type converters.
In sharepoint when you develop a web part and the properties of the web part are not simple types (string, int, etc.) you often times need to create an editor. The editor has a method named CreateChildControls. This is where you typically and dynamically would add your controls using Control.Add(myControl). I would like to convert my dynamic model to a templated model that a .ascx control would provide. By seperating the GUI and backend code it would provide me with a model that is easier to maintain, the ability to use databinding, and better unit testing. What is your recommeded approach?
~Ted
Great article very helpful.
I'm looking to for a way to solve the following problem without using dynamic controls but just can not figure it out.
I have a form for editing multiple XML node groups at once, each node group will be edited using a customised user control. My problem is that the frequency and type of node groups can vary, there can be 1 to n and of any type of node group.
Hi...HOw to create multiple rows and columns of textboxes and enter the values in the database....
Good stuff!! Eagerly awaiting part 5...
Wonderful post!!!
Had had this challenge for a while and had read thru the 4 parts of ur blessed article. When I saw the dates I was full of prayers that you would still be online that was before I got to the end of the list... :)
My challenge is simple. I have an e-testing app that automatically loads any specified number of questions from the database (this could range from 50 to 300 questions). I constructed a usercontrol containing all the required fields. this control is then "hosted" on the view controls of the multiview control.
So if the quiz master selects 68 questions to be loaded the app should load 68 questions on 68 dynamically created controls on the panel control added to the view control which is in turn added to the multiview control. This has been a nightmare. Any suggestions for alternatives are welcome.
Waiting eargerly for your reply
chris
jcc_nnannah@yahoo.co.uk | http://weblogs.asp.net/infinitiesloop/archive/2006/10/16/TRULY-Understanding-Dynamic-Controls-_2800_Part-4_2900_.aspx | crawl-002 | refinedweb | 20,537 | 65.22 |
Back
According.
and attempts to store custom type in a Dataset lead to following error like:
Unable to find encoder for type stored in a Dataset. Primitive types (Int, String, etc) and Product types (case classes) are supported by importing sqlContext.implicits._ Support for serializing other types will be added in future releases
Or:
Java.lang.UnsupportedOperationException: No Encoder found for ....
Are there any existing workarounds?
Just use kryo
The best suggestion that you will get for this problem very tedious. Especially if your code is manipulating all sorts of datasets, joining, grouping etc. in such case you end up collecting a bunch of extra implicits. So, as a better approach we can simply make an implicit that does this all automatically
import scala.reflect.ClassTagimplicit def kryoEncoder[A](implicit ct: ClassTag[A]) = org.apache.spark.sql.Encoders.kryo[A](ct)val d4 = d2.joinWith(d3, $"d2._1" === $"d3.)
d2.printSchema
// root
// |-- value: binary (nullable = true)
If you want more information regarding the same, refer the following. | https://intellipaat.com/community/5146/how-to-store-custom-objects-in-dataset?show=5219 | CC-MAIN-2021-43 | refinedweb | 168 | 50.53 |
I followed the Watson Conversation documentation instructions to create an action to call my IBM Cloud Function action. When I test the conversation (in the "Try it out" panel) I get the an error after my intent and entities are recognized. Clicking on the triangle error symbol shows the following message:
Direct Cloud Functions call was not successful. Http response code is [400]. (and there is 1 more error in the log)
How do I see the log? There was no error logged at the Cloud Function end.
What am I doing wrong? I have specified my credentials correctly (when I specify them incorrectly, I get a 401). One concern: The namespace I entered "/Natural%20Language%20Analytics_dev/" seems different from the namespaces in the documentation examples.
Answer by MitchMason (5336) | Feb 11, 2018 at 07:46 PM
We do have plans to improve the error messaging within the tool, there are just so many possibilities it has taken us some time. To solve your problem specifically, 400 usually refers to some sort of invalid request, typically this is with credentials, but could be a number of things. Im not sure if you can have % in your namespace, for example, mine was using an email address, so I just replaced the % with the @ sign like it was supposed to be, and it worked after that. Your namespace probably shouldn't have spaces in it though. To see the more descriptive error you can inspect the webpage, specifically the network tab and see the full response from CloudFunctions.
Id also recommend recreating the samples from the docs to make sure you have everything else set up right, then customizing one at a time so you can spot the error in your command above.
Answer by NatMills (1) | Mar 19 at 09:30 AM
@HH1T_Sugato_Bagchi I had trouble with this as well. It turns out I was supplying too much of the URI. The directions state you should begin with the /namespace but I had included /api/v1/namespaces. Once I cleaned up the name to begin with the /namespace/actions/package/function that solved the 400 error. I'd also had a dollar sign for the response_variable name. Removing the dollar sign allowed the response variable name I'd entered to be added to the context in the reply. Also, ensure the type is "cloud_function" and that you have provided your credentials in the context. Here is an example of what I'd used with credentials altered...:
{ "context": { "my_creds": { "user": "83f2455v-blah-blah-blah-ffb6dea876b2", "password": "1vWfiRj_typed_some_garbage_here_9GaM" } }, "output": { "text": { "values": [ "Called MyCloudFunction" ], "selection_policy": "sequential" } }, "actions": [ { "name": "/mynamespace/actions/mypackage/myCloudFunction", "type": "cloud_function", "parameters": { "input": "<?input?>", "entities": "<?entities?>" }, "credentials": "$my_creds", "result_variable": "entities" } ] }
The result was to add a new entities variable to the context (my Cloud Function filtered entities to clean up overlapping locations, and conflicts with sys-number/sys-date).
135 people are following this question.
Watson Assistant action to call IBM Cloud Function fails with Direct CloudFunctions calls are not supported on this platform 6 Answers
Is it possible to upload a video file to IBM OpenWhisk function and encode it? 0 Answers
File size limitation for Python action (Cloud Functions) 2 Answers
Dialog node error 3 Answers
IBM Cloud Functions - serverless deploy fails when multiple functions in yml 0 Answers | https://developer.ibm.com/answers/questions/430533/watson-conversation-action-to-call-ibm-cloud-funct.html?sort=votes | CC-MAIN-2019-39 | refinedweb | 550 | 52.39 |
Author: pedronis Date: Thu Feb 15 15:46:34 2007 New Revision: 38900 Modified: pypy/dist/pypy/doc/coding-guide.txt pypy/dist/pypy/doc/faq.txt Log: review and finish faq about the Developmet... kill some hard, no good answers, not really faq questions. Modified: pypy/dist/pypy/doc/coding-guide.txt ============================================================================== --- pypy/dist/pypy/doc/coding-guide.txt (original) +++ pypy/dist/pypy/doc/coding-guide.txt Thu Feb 15 15:46:34 2007 @@ -686,9 +686,12 @@ Only specified names will be exported to a Mixed Module's applevel namespace. -Sometimes it is neccessary to really write some functions in C (or whatever -target language). See the `external functions documentation`_ for details. +Sometimes it is necessary to really write some functions in C (or +whatever target language). See `rctypes`_ and `external functions +documentation`_ for details. The latter approach is cumbersome and +being phased out and former has currently quite a few rough edges. +.. _`rctypes`: rctypes.html .. _`external functions documentation`: translation.html#extfunccalls application level definitions Modified: pypy/dist/pypy/doc/faq.txt ============================================================================== --- pypy/dist/pypy/doc/faq.txt (original) +++ pypy/dist/pypy/doc/faq.txt Thu Feb 15 15:46:34 2007 @@ -12,31 +12,14 @@ What is PyPy? ------------- -XXX +PyPy is both a Python reimplemenation and a framework to implement +interpreters and virtual machines for programming languages, +especially dynamic ones. PyPy tries to find new answers about ease of +creation, flexibility, maintainability and speed trade-offs for +language implementations. For further details see our `goal and +architecture document`_ . ------------------------------------------------------- -Why a new implementation of Python? What does it add? ------------------------------------------------------- - -XXX - ------------------------------------ -What is the status of the project? ------------------------------------ - -XXX status - - --------------------------------- - Can it be used in practice yet? --------------------------------- - ! +.. _`goal and architecture document`: architecture.html .. _`drop in replacement`: @@ -55,12 +38,12 @@ On what platforms does it run? ------------------------------ . At the moment you need CPython 2.4 for the translation +process, 2.5 is not fully supported. Currently (due to time restrictions) we are not trying hard to make PyPy support 64 bit platforms. While this seems to still mostly work out, a few modules won't @@ -125,8 +108,19 @@ How do I write extension modules for PyPy? ------------------------------------------ -XXX +PyPy extension modules are in the form of so called `mixed modules`_, +at the moment they all need to be translated together with the rest of PyPy. +We have a proof concept in what we call the `extension compiler`_ and +our support for a static variant of ctypes interface (`rctypes`_) to +help with their development. At the moment both have quite some rough +edges, also cross compilation to CPython extensions which is possible +doesn't deliver completely satisfying results. This area is going to +improve over time. + +.. _`mixed modules`: coding-guide.html#mixed-module-mechanism +.. _`extension compiler`: extcompiler.html +.. _`rctypes`: rctypes.html .. _`slower than CPython`: @@ -188,12 +182,6 @@ .. _`project suggestions`: project-ideas.html .. _`contact us`: contact.html ----------------------------------- -Why so many levels of abstraction? ----------------------------------- - -XXX see pypy-vm-construction - ---------------------------------------------------------------------- I am getting strange errors while playing with PyPy, what should I do? ---------------------------------------------------------------------- | https://mail.python.org/pipermail/pypy-commit/2007-February/018927.html | CC-MAIN-2014-15 | refinedweb | 517 | 52.15 |
As suggested by Jochen Theodorou in this
<>
post, I have created this topic for `Text / String processing`
I have always felt that groovy platform itself can have rich set of DSLs (list given below) so that groovy will be in hands of many non-developers.
Also, the command line String/text processing capabilities can be greatly enhanced with java/groovy power.
Use Case 1 : command line usage Simple input processing can be done in the command line itself
$ groovy -e --complex 'def s = 0; EACHLINE {s+=line.numval}; END {println s}'Use Case 2 : DSL Script for more complex line processing functionalities (similar to 'awk') with java / groovy power
def s = .... ...... ...... //for choosing lines with regex matches (regex) { work for those lines } //for choosing lines with some expression matches { condition {expression for line selection} process { work for those lines } } END { //end closure after processing all matching lines of the input files } | http://mail-archives.apache.org/mod_mbox/groovy-dev/201810.mbox/raw/%3C1539371212935-0.post@n5.nabble.com%3E/2 | CC-MAIN-2019-22 | refinedweb | 151 | 50.5 |
Introduction to Tkinter Bind
In Tkinter, bind is defined as a Tkinter function for binding events which may occur by initiating the code written in the program and to handle such events occurring in the program are handled by the binding function where Python provides a binding function known as bind() where it can bind any Python methods and functions to the events. Therefore, in general, the Tkinter bind is defined as a function for dealing with the functions and methods of Python that are bind to the events that occur during the program execution such as moving the cursor using a mouse, clicking of mouse buttons, clicking of buttons on the keyboard, etc are events that are handled using bind function in Tkinter.
Working of Tkinter Bind in Python
In this article, we will discuss the Tkinter binding function that is usually used to bind any Python functions and methods to the events or actions that occur by initiating the out of box program or scope of the program and such actions can be handled by a small piece of code and this code is bind function in Tkinter. In Tkinter which is usually used for designing the web applications or desktop applications where there are different events created such as navigating from one page to another, clicking on one link and going to another page, etc are all events and such events can be handled by using a binding function known as bind() which we can bind any functions of Python to such events to handle the events properly and then these functions, in turn, can be bind with any widgets provided by the Tkinter module. Therefore events handling is not only done with the bind() function we have already used or seen when using the button widget, where on clicking the button widget we would get some message which is already written in the command option of the button widget. In the same way, the bind function is used to deal with some other events that are occurred through the system by the users some of the common events are like clicking the mouse button, pressing of any button on the keyboard, etc.
In this article, let us see a few examples of using the bind function to deal with a few system events like mouse clicks, keyboard button typing, or pressing. So first we will see a syntax where each widget provided we can bind it with Python methods which are possible using the bind() function.
Syntax:
widget.bind(event, handler)
Parameters:
- event: this parameter is used to define the events so that it can be handled using the handler function. E.g FocusIn, Enter, KeyPress, etc
- handler: this parameter is used to define the handler function so that it can describe the event that occurred using the event objects which are called along with the handler function such as having a mouse position using the x and y-axis in pixels, mouse button numbers, etc.
So using the above syntax we can bind an event to a function with a very simple process which is just if any event such as clicking the mouse or keyboard buttons an event occurs which will automatically trigger the handler function defined in the syntax and suppose if the event argument is not given or hidden in the function then there is a chance of you getting an error known as TypeError. There are other defined events provided by Tkinter such as the <Destroy> event which when defined or specified or bind with any widget then the widget is destroyed, therefore there are many such events that can be seen in the Tkinter module itself.
Now let us see an example of how to use the bind function by attaching the event to the widgets. In the below example we will see a keyboard button typing along with the key which is pressed is displayed.
Example
Code:
from Tkinter import *
import tkMessageBox
master = Tk()
master.geometry('500x200')
def func():
tkMessageBox.showinfo( "Hello Educba", "Press any key on the keyboard")
b1 = Button( master, text='Click me for next step', background = 'Cyan', fg = '#000000', command = func)
b1.pack()
def Keyboardpress( key):
key_char = key.char
print( key_char, 'key button is pressed on the keyboard')
master.bind( '<Key>', lambda i : Keyboardpress(i))
master.mainloop()
In the above program, we have seen a simple code for seeing the key pressed on the keyboard and this event <key> is bind with the Python function which we have defined in this program is “Keyboardpress()” where when this event is occurred then automatically when we type or press any key on the keyboard it prints the key name which you have typed on the output screen as shown in the above screenshot. In this above code, we have also seen the button widget calling an event by specifying the action to do in the function defined as a value to the command option in the button widget. Therefore we can see in the above first, when we run the program it will display a button having the name as “click me for the next step” so when we click this button an event is generated, and the function defined in the command option is triggered by handling this event which shows a message saying “Press any key on the Keyboard”. Then this pressing of any key is again an event <key> which we have bind this event with the function defined in the program to display the characters pressed on the keyboard which is done by using the bind() function.
Therefore there are many such events like <destroy> for destroying the widgets, <configure> for setting the window display according to the parent window, <key> for pressing the keys on the keyboard, <Enter> for pressing on the enter key on the keyboard, etc. These all events we can bind with the function which we define in the program and handle such events.
Conclusion
In this article, we conclude that the Tkinter bind is a binding function for binding any event to the Python function or method that is defined within the program to handle such events. In this article, we saw how the bind() function can be define using the syntax that holds both the event name and the handler which would be the function name. In this article, we also saw a simple example of pressing any key on the keyboard is an event and the pressed key names are displayed on the output screen.
Recommended Articles
This is a guide to Tkinter Bind. Here we discuss the Introduction and working of Tkinter Bind along with an example and code implementation. You may also have a look at the following articles to learn more – | https://www.educba.com/tkinter-bind/ | CC-MAIN-2021-49 | refinedweb | 1,127 | 53.07 |
I've written about NExpect a few times before:
If you'd like to learn more first, open those up in other tabs and have a look. I'll wait...
I've had a preference for using NExpect myself, but I'm obviously biased: I had a developer-experience with NUnit assertions which I found lacking, so I made a whole new extensible assertions library.
But there's always been something that I haven't been able to qualify about why I prefer NExpect over NUnit assertions. I've even gone so far as to tell people to just use either, because they're both good, and I don't want to be that guy who tells people to use his stuff.
Though the deep-equality testing is really convenient and NUnit doesn't do that...
Today, that changes. I have a good reason to promote NExpect over NUnit now, apart from all of the obvious benefits of NExpect:
- fluid expression
- extensibility
- deep-equality testing
- better collection matching
Today I found that NExpect can tell you earlier when you've broken something than NUnit can.
Explain how?
Consider this innocuous code:
using NUnit.Framework; [TestFixture] public class Tests { [Test] public void ShouldPassTheTest() { var result = FetchTheResult(); Assert.That(result, Is.EqualTo(1)); // or, the olde way: Assert.AreEqual(result, 1); } private int FetchTheResult() { return 1; } }
Of course, that passes.
The interesting bit here is the usage of
var, which, in C#, means "figure out the type of the result at compile-time and just fill it in". Long ago, that line would have had to have been:
int result = FetchTheResult();
var has some distinct advantages over the prior system. It's:
- shorter to write
- you only have to remember one "muscle-memory" to store a result (always
var ___ = ___)
- it means that if you do change return types, things are updated for you.
In theory (and practice), it makes you quicker on the first run and when you refactor.
The problem comes in when those strong types are discarded by code which compiles perfectly. The compiler can't save you from yourself every time!
Enter the refactor
When we update the above so that
FetchTheResult now returns a complex object, the code will still compile:
using NUnit.Framework; [TestFixture] public class Tests { [Test] public void ShouldPassTheTest() { var result = FetchTheResult(); Assert.That(result, Is.EqualTo(1)); // or, the olde way: Assert.AreEqual(result, 1); } public class Result { public int Value { get; set; } public DateTime Created { get; set; } } private int FetchTheResult() { return new Result() { Value = 1, Created = DateTime.Now }; } }
for the intent of the flow of logic, we're still returning the value
1, but we've also attached a
DateTime property to that to indicate when that result was created. Rebuilding, we find that everything builds just fine, and perhaps we forget to re-run tests (or perhaps this function is very far away from where the result is being used, so we don't realise that we just broke a test somewhere else).
This is because the NUnit assertions fall back on
object for types:
Assert.Thatis genericised for the first parameter, so it has a fixed type there, but takes a Constraint for the second parameter -- and a Constraint can house anything, because it casts down to
object
Assert.AreEqualhas an overload that expects two
objects, so it will fall back on that, and also compile.
The test will fail -- if you remember to run it (or when CI runs it).
So how does NExpect help?
If we'd written the first code like so:
using NUnit.Framework; using NExpect; using static NExpect.Expectations; [TestFixture] public class Tests { [Test] public void ShouldPassTheTest() { var result = FetchTheResult(); Expect(result).To.Equal(1); } private int FetchTheResult() { return 1; } }
then the refactor would have caused a compilation failure at second line of the test, since NExpect carries the type of
result through to the
.Equal method.
actually, it does a bit of up-casting trickery so that you
can, for example,
Expect((byte)1).To.Equal(1);, but that's
beside the point for this particular post...
So the second the refactor had gone through, the test wouldn't compile, which means I could find the failure even before running the tests and update them accordingly, instead of waiting for tests to fail at a later date.
Conclusion
Strongly-typed languages have a certain amount of popularity because the types can help us to avoid common errors. This is why TypeScript is so popular. C# is strongly typed, but there are ways that the strength of that typing can be diluted, and one of those ways is found in NUnit assertions.
NExpect protects you here and alerts you about potentially breaking changes to your code before even running tests. Neat, huh?
I think I'll pat myself on the back for that 🤣
Discussion | https://dev.to/fluffynuts/nexpect-not-just-pretty-syntax-ja5 | CC-MAIN-2020-50 | refinedweb | 810 | 60.85 |
The marriage license of Hannah Peatt (as was indexed) and Isaac Armfield is abstracted from Book 1, pg 42 Hamilton County, Indiana as follows:
Hannah Piatt resident of Hamilton County, Indiana, of lawful age and Isaac Armfield of lawful age. Proved by affidavit of Peter Case with Daniel R Brown clerk of the Circuit Court, Noblesville and the minister's return (24 Dec 1851) by William W Boyden 22 Dec 1851. She was his second wife.
Isaac and Hannah had the following known children:
Florence Armfield was born May 1869 in Montford, Grant Co, WI and died January 4, 1913 Minneapolis, Minnesota. She married 29 Jun 1887 Janesville, WI, Frank Pierce Williams who was born 1870 Janesville, WI and died 22 Feb 1942 Omaha, NE. Frank was the son of Robert Williams and Harriet Parker Travis.
Frank and Florence had the following children:
Bessie Maude Williams was born June 18, 1888, in Janesville, Wisconsin, and died April 1973 in Aberdeen, Washington, and is buried in Tacoma, Washington, in a family plot. She married Frank Cecil Strickland who was born April 6, 1883, in Chicago, Cook County, Illinois, and who died December 17, 1951 in San Bernardino, California. Frank was the son of Frank Cantelo Strickland and Margaret Martin. Frank's Jr's father, Frank, was born June 15, 1853, in England to Richard Strickland, a saddler, born in England, and married January 11, 1853, at St. Mary's Church, Alverstoke, Hampshire, England, to Harriet Cantelo who was born 1827 on the Isle of Wight. Frank's Jr's mother, Margaret, was born 1853 Chicago, Cook County, Illinois, and died May 6, 1933, in West Chicago, Illinois. She was the daughter of Moses Dewitt Martin and Katherine Beckman. Moses was born April 19, 1835, Prescott, Canada, and died March 25, 1911, Rockford, Illinois. Katherine was born November 21, 1832, in Germany, and died in 1915.
Bessie and Frank Strickland's children (not in birth order):
Cantello and Lillian's children:
Lillian and Chester Ream children:
Sources:
Back to Home | http://www.angelfire.com/ar/pyeatt/Armfield.html | CC-MAIN-2017-17 | refinedweb | 338 | 67.49 |
Textile makes it easy to run IPFS peers in your mobile apps
Does Textile Photos run an IPFS peer on every user’s mobile device? We get the question often, so we know the answer isn’t obvious. The answer is, yes. At Textile, we use IPFS because it is helping us get closer to our goal of helping people truly own their digital data. We believe that the stack of protocols available in IPFS can help all developers rethink the way they treat data on the web, and in particular, can help us use encryption, permissions, and decentralized sharing to put data ownership into the hands of people.
Big ideas aside, running IPFS on mobile introduces challenges to any app developer, such as deciding when and how it should run, how to efficiently transfer data in a mobile environment, and how to link the behaviors of the IPFS node to the native device APIs (e.g. background and foreground events). Over the past months of working with IPFS in Textile Photos, we’ve been tuning and improving our approach to launching, managing, and using IPFS in our mobile app. Now, it’s ready for any developer to build on that work.
To make IPFS run easily in our mobile apps, we’ve combined some scaffolding with a bunch of custom APIs into the Textile developer framework. Our first release of the framework, the React Native SDK, lets an app developer install and manage an IPFS node directly in their app with just a couple lines of code. Soon, our Objective-C (iOS) and Java (Android) libraries will make it accessible to even more developers. Here, I just want to walk through some of the simple steps to get IPFS into your mobile app using Textile.
Getting started
Assuming you have a React Native app setup, you should be able to simply add the following to the top of your app setup file. You can see examples of where this is run in one of the existing demo apps (e.g. Single Screen Boilerplate, Notes, AirSecure).
import Textile from '@textile/react-native-sdk';// Setup will prepare the environment
Textile.setup();
You may be familiar with the IPFS setup steps (e.g.
ipfs init) that result in a new peer-id, private key, etc. With Textile installed in your app, initiating Textile will handle setting up a new IPFS peer on the device, all in one simple line of code:
// A single line to create the Textile node and embedded IPFS peer
Textile.nodeCreateAndStart();
Simple controls
From there, you can let Textile manage the starting and stopping of your node based on phone behaviors, or you can take direct control. These methods enable a developer to quickly launch IPFS in a mobile app and restart it whenever new media needs to be stored or retrieved. While simple to use, there is actually a lot of complexity going on under the hood here. From bootstrapping peer connections, to fetching content updates, to processing network queries in the background. All of this is abstracted away into a simple start/stop API that is easy to use:
import { API } from '@textile/react-native-sdk';// Stop an already started node
API.stop();// Start it back up later
API.start();
Request data on IPFS
Decentralized data can benefit mobile apps and their users already (e.g. censorship resistance). So we’ve made it easy to request any IPFS hash (and path) using the IPFS protocol.
import { API } from '@textile/react-native-sdk';// Grab an image off of IPFS
const hashPath = 'QmTgtbb4LckHaXh1YhpNcBu48cFY8zgT1Lh49q7q7ksf3M/raster-generated/ipfs-logo-text-512-ice.png';
const imageData = API.dataAtPath(hashPath);// Display that image using React Native Image
<Image
style={{width: 150, height: 150, resizeMode: 'cover'}}
source={{uri: 'bbd1" class="lm ln du ap lo b lp lq lr ls lt lu lv lw lx ly lz">Boilerplate apps
Textile offers a few simple boilerplate repos for developers to clone, build, and tear apart. The first one, react-native-boilerplate, provides a single-screen React Native app that packs in quite a few Textile & IPFS commands (including those above) to simply and easily show what is possible.
Already have a more complete app idea in mind? Grab the advanced boilerplate, for multi-screen support, state management, Textile event management, and more. To give you an idea of what can be done quickly and easily with these boilerplate apps, check out the AirSecure app or Textile Notes app.
Why add IPFS or Textile to your mobile app
There are a number of reasons to add IPFS or Textile to your mobile app, both technical and not. Our opinions are biased obviously, but we believe IPFS is forging the way towards better systems of media and data exchange on the Internet. But the architecture of the IPFS network and how Textile uses it have some important side-effects. Here are some of our top reasons for being so optimistic.
Content Addressing
A content addressed approach to information retrieval on the Internet opens up a new way to think about, not only data availability, but data ownership as well. We can build new kinds of apps and services that don’t need to think about building silos of their own data when they can collaborate across an interlinked network of information created by many apps. That’s an exciting system to build applications on.
Decentralized Networks
At the root of IPFS are a handful of mechanisms that allow nodes on the network to retrieve information from other peers, even when not directly connected to the original provider. This gives rise to new benefits, like being able to create apps that are fully functional in small isolated subnets, cut off from the greater Internet.
Encrypted & P2P
Related to the decentralized nature of IPFS, Textile helps developers use encryption for all data storage and transfer. Encryption and permissions systems mean that apps can build secure chat and sharing features into their apps easily. In Textile Photos, you can see these two primitives used many times across the app.
Censorship Resistant
Components like those above have made the IPFS network resilient to censorship attempts. Examples include the Uncensorable Wikipedia in Turkey and the vote organizing in Catalonia. Censorship resistance and privacy through encryption may allow developers to build the future of journalism/reporting applications, knowledge platforms, voting applications, and more.
Personal Data Sovereignty
We’ve previously written about why we think this is important. Data is becoming an extension of our physical existence. Voting, dating, shopping , health are just a few examples of where humans and the important (and often private) information they create is being captured in digital data. We need to build technologies that let individuals own that data, forever. We are building our developer tools to help make that future a reality. Using Textile and IPFS, developers can start building apps that take the first steps.
There are many others, including new ways to think about GDPR, building secure patient or student services, connecting to blockchains, etc, but we’ll cover each of those in more depth later.
Next steps
If you’d like to try it out, grab one of the boilerplate apps and play around with your first mobile app with embedded IPFS peer. Or, grab the code for one of the existing apps released on Textile, including Textile Photos, AirSecure, and Textile Notes.
We have a great community of builders on our developer slack channel, so jump in if you have any questions about using IPFS in your mobile apps. | https://medium.com/textileio/textile-makes-it-easy-to-run-ipfs-peers-in-your-mobile-apps-f797af98311b | CC-MAIN-2019-35 | refinedweb | 1,257 | 59.53 |
[Lab] determine black or white color to draw text onto a CSS color
I made the most basic of functions to try to decide based on a CSS color background, what color to use for text on the background so it's not lost ( only black or white). The one below, seems to work for quick and dirty uses. It's just fun. But I know a lot more can be done mathematically with the r,g,b components. I tried a few silly things with sliders to filter the CSS colours also. Anyway have a gist here.
The function that I did is below. Yes, it's crap, but it sort of works if you have nothing else. Maybe someone with time on their hands can come up with something more exacting.
def safe_color(css_color): rgba = ui.parse_color(css_color) r = rgba[0] g = rgba[1] b = rgba[2] if (((r + g + b) / 3 ) > .55) : return 'black' else: return 'white'
Output from gist
Maybe
return 'black' if sum(rgba[:3]) / 3 > .55 else 'white';-)
Or
return 'black' if sum(rgba[:3]) > 1.65 else 'white';-)
return int(sum(rgba[:3]) <= 1.65)
- Webmaster4o
One tip for making text more readable is something the Material Design guidelines recommend: While white text should be
rgba(255, 255, 255, 1), but black text should be
rgba(0, 0, 0, 0.87). The slight opacity in the black makes it much more readable on colored backgrounds.
return int(sum(rgba[:3]) <= 1.65) or (0, 0, 0, .87)
Thanks guys. I have made a snippet for myself using
return int(sum(rgba[:3]) <= 1.65) or (0, 0, 0, .87)
I know it's not a big deal, but it's nice to have these things handy and to be a no brainer even if you ultimately tweak the colors. Will take a closer look at what @Webmaster4o did later , as I think he did some complimentary color work when he was working on themes.
@ccc , I did try briefly to simplify the expression. I was trying to use unpacking, slicing didn't occur to me 😱I didn't search stackflow, because I was playing with components also. But it seems like the average checked against a threshold seems to work ok. But I am glad you call us out on it. It does make me think before I post stuff. When I posted it, I was almost certain that I would be hearing from you. I was grinning for 5 mins when I saw your reply 😬😬😬
Don't think toooo hard before you post to the forum... we love to see your crazy ideas in their raw form. But do consider running
Check Styleand
Analyze (PyFlakes)as they find simple issues (like your unused variable) and push you to make your code more readable. I have become big fan of autopep8 on my Mac. It would be cool to see that level of automatic reformatting as an option in Pythonista. | https://forum.omz-software.com/topic/3527/lab-determine-black-or-white-color-to-draw-text-onto-a-css-color | CC-MAIN-2017-26 | refinedweb | 497 | 83.15 |
dirkbaechle repo owner 2014-03-05T16:59:13+00:00
I'm stuck using SCons v2.3.0. Should that be able to build the HTML manual out of the box? SCons tells me there's "No tool named DocBook".
After installing it manually, I get the following error messages:
What do you mean when you say "you're stuck"? Can't even download the "local" version of SCons (doesn't require you to install it systemwide) and start it?
SCons v.2.3.0 doesn't have the "docbook" Tool as core component, it was only added for v2.3.1. So the other option would be to download it separately from its repo (as given in the ToolsIndex at the Wiki).
If you don't have the Docbook stylesheets installed in your system, you'll also want to remove the "DOCBOOK_XSL=..." parameter in the SConstruct.
I switched over to v.2.3.1 temporarily, but I'm still getting the same error messages (both with and without the DOCBOOK_XSL parameter).
I'm pretty new to SCons, is there some way to get more verbose logs to help find the cause?
Please make sure that you have one of the libxml2 or lxmxl Python bindings installed. You should be able to open a Python interpreter, and type either
import lxml
or
import libxml2
without getting any errors. If you do get an ImportError, please install one of these Python bindings and their dependencies via your package manager.
Then try again.
If you should still have trouble getting the DocBook Tool to work, I've added a README.rst to the repo a few seconds ago. So you can simply go to the frontpage of scons_qt5 and read the manual there now.
Hope this helps.
No ImportError's if I try to import libxml2.
Apparently the message I'm getting is usually caused by SCons not being able to find gcc/g++, but if I try to manually set the paths to them in the Environment, I get even more cryptic messages:
Thanks for the README though, very helpful :)
The error message says that in l. 35 of your SConstruct "conf.CheckCXX()" gets called. Did you add it? I'm asking because I searched the full source tree for the Qt5 tool and SCons v2.3.1 itself, and found no match.
Sorry for asking this, but have you already successfully run some other build with SCons, or are these your very first steps? In the latter case you might want to write to our mailing list at scons-users@scons.org, where you can get more and faster help because more people listen.
Sorry, I added that to try and confirm that SCons was correctly detecting gcc/g++.
I've built a couple of projects with SCons already, with no problems.
Thanks for your help, I think I'll leave it for now as you've added the Readme, and the problem seems to be specific to my environment.
If it's okay with you like that, sure. I didn't want to cut off this conversation, but simply wondered what your level of knowledge with SCons would be.
From what I could see so far, I'd still suspect that (for some unknown reason) the DocBook tool is not able to find the libxml2/lxml bindings, and also (this is the last fallback usually) can't detect a "xsltproc" executable on your current PATH. That's where the funny name "o" for the executable comes from, that SCons is trying to call.
Anyway, we'll leave it at that for now. If you should experience further problems and have more questions, please feel invited to open a new issue report at the offending tool, or come over to the SCons user mailing list.
Thanks a lot for your patience.
Best regards,
Dirk
Closing this, since no follow-up questions arose for some time. | https://bitbucket.org/dirkbaechle/scons_qt5/issues/1/precompiled-manual | CC-MAIN-2017-13 | refinedweb | 655 | 81.83 |
To clarify my merge-tracking proposal more in detail:
It seems to me that the important aspect is that mergeinfo
should contain only entries from *direct* merges.
In this way, the mergeinfo build a tree.
(This merge history tree information is missing in the
current mergeinfo scheme, which contains both direct
*and* indirect merges.)
def smart_merge(source, -r X:Y, target):
def is_1_contained_in_tree_2(path1@r1, path2@r2):
if path1@r1 == path2@r2:
return True
for each subpath@subr which is in mergeinfo of path2@r2,
but not in mergeinfo of path2@(r2-1):
if is_1_contained_in_tree_2(path1@r1, subpath@subr):
return True
return False
def undo_existing_changes_1_in_2(path1@r1, path2@r2):
if is_1_contained_in_tree_2(path1@r1, path2@r2):
merge path1 -r r1:(r1-1) from target
else:
for each subpath@subr which is in mergeinfo of path1@r1,
but not in mergeinfo of path1@(r1-1)
in reverse(!) revision order:
undo_existing_changes_1_in_2(subpath@subr, path2@r2)
for i2 = LATEST_REV downto 1:
for i1 = Y downto X+1:
if mergeinfo of target does not contain source@i1:
undo_existing_changes_1_in_2(source@i1, target@i2)
for i1 = X+1 to Y:
if mergeinfo of target does not contain source@i1:
merge source -r i1-1:i1 into target
add source@i1 to mergeinfo of target.
This should handle both
and
correctly.
This is a brute-force implementation, which of course could be optmized
to work on revision ranges instead of individual revisions.
But this is "only" an implementation detail.
The above pseudo-code is just quickly written down
and may be buggy, but I think you get the idea.
Cheers,
Folker
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Dec 4 22:55:02 2007
This is an archived mail posted to the Subversion Dev
mailing list. | http://svn.haxx.se/dev/archive-2007-12/0137.shtml | CC-MAIN-2014-15 | refinedweb | 303 | 50.36 |
« Return to documentation listing
MPI_Init_thread - Initializes the MPI execution environment
#include <mpi.h>
int MPI_Init_thread(int *argc, char ***argv,
int required, int *provided)
INCLUDE 'mpif.h'
MPI_INIT(REQUIRED, PROVIDED, IERROR)
INTEGER REQUIRED, PROVIDED, IERROR
#include <mpi.h>
int MPI::Init_thread(int& argc, char**& argv, int required)
int MPI::Init_thread(int required)
argc C/C++ only: Pointer to the number of arguments.
argv C/C++ only: Argument vector.
required Desired level of thread support (integer).
provided Available level of thread support (integer).
IERROR Fortran only: Error status (integer).
This routine, or MPI_Init, must be called before any other MPI routine
(apart from MPI_Initialized) is called. MPI can be initialized at most
once; subsequent calls to MPI_Init or MPI_Init_thread are erroneous.
MPI_Init_thread, as compared to MPI_Init, has a provision to request a
certain level of thread support in required:
MPI_THREAD_SINGLE Only one thread will execute.
MPI_THREAD_FUNNELED If the process is multithreaded, only the
thread that called MPI_Init_thread will make
MPI calls.
MPI_THREAD_SERIALIZED If the process is multithreaded, only one
thread will make MPI library calls at one time.
MPI_THREAD_MULTIPLE If the process is multithreaded, multiple
threads may call MPI at once with no restric-
tions.
modifies, interprets, nor distributes them:
{
/* declare variables */
MPI_Init_thread(&argc, &argv, req, &prov);
/* parse arguments */
/* main program */
MPI_Finalize();
}
The Fortran version does not have provisions for argc and argv and
takes only IERROR. implementa-
tion, it should do as little as possible. In particular, avoid anything
that changes the external state of the program, such as opening files,
reading standard input, or writing to standard
Open MPI 1.2 September 2006 MPI_Init_thread(3OpenMPI) | http://icl.cs.utk.edu/open-mpi/doc/v1.2/man3/MPI_Init_thread.3.php | CC-MAIN-2015-32 | refinedweb | 265 | 57.37 |
Closed Bug 1021751 Opened 7 years ago Closed 7 years ago
Homepage contextual hint
Categories
(Firefox for Android Graveyard :: General, defect)
Tracking
(fennec34+)
Firefox 34
People
(Reporter: ibarlow, Assigned: liuche)
References
Details
Attachments
(5 files, 7 obsolete files)
A common piece of user feedback is that people can't find their history or bookmarks. So one of the first "tips" we want to test is something on the homepage that encourages people to swipe left or right to access them. (see) This rough mockup was already in our meta bug 998036 but I noticed we didn't have any bug to track this specific work. Including Anthony here for visual design polish, Chenxia for implementation.
OS: Mac OS X → Android
Hardware: x86 → ARM
Version: Firefox 31 → Trunk
tracking-fennec: --- → ?
(In reply to Ian Barlow (:ibarlow) from comment #0) > This rough mockup was already in our meta bug 998036 but I noticed we didn't > have any bug to track this specific work. Have overlays (e.g,) that other apps tend to use on first-run been ruled out?
I'm trying to push us away from that. There is a growing body of evidence that suggests people don't read this stuff, and handing out information in smaller, bite sized pieces at the right time is much more valuable.
Anthony / Ian: How final are the mockups?
Assignee: nobody → liuche
Flags: needinfo?(alam)
tracking-fennec: ? → 33+
(In reply to Mark Finkle (:mfinkle) from comment #3) > Anthony / Ian: How final are the mockups? They can always be more final... I think they are structurally accurate, but depending on what Anthony does in bug 1011712 and bug 1014293 it will likley need a unification pass from him. -- A note to the engineer -- consider transitions when building this. It's likely we will want this tip to animate in and animate out in some way.
As discussed with Ian, this may be an issue of not associating a "New Tab" with "Bookmarks" or "History". The short version of this reasoning is that users are most likely browsing a web page when they encounter a desire and therefore frustration accessing their "bookmarks". At this point, they would just be presented with a URL bar, Tabs button, or 3 dot menu along the top (neither of which are associated with "bookmarks" or "history"). But opening a "new tab" is actually the first thing they must do to access their "bookmarks" and so we can see how there needs to be some sort of mental connection to be made there. We want to focus more on developing a mental model of "I have shortcuts in my new tab" in the user so they will look to "+" when they want "bookmarks", "history", or maybe even other panels they've then baked into their daily routine. Mock ups to follow.
Adding to my comment above ^... To reinforce this mental association in the user and show off the side-swipe capability of the Homepage Panels, we're going to try a couple of things: Mainly, I would like to try and ease in (with a little bit of a bounce and delay) the labels that exist to the right and left of "Top Sites". I also want to extend the height of the navigational header where the labels currently reside ("Top sites", "Bookmarks", etc) and utilize the snippet to display some additional information.
Flags: needinfo?(alam)
Here's a quick animation clip to show what I mean. Timing here would be key, so we have to detail and fine tune this. If it helps, it's set at 2.25 seconds delay right now.
Looks like we could tie into the banner system to display the mockup. The animation clip isn't really showing any animation though. Might be a .mov issue? I was talking to Chenxia about this bug and proposed that we could use something like Ian's mockup in bug 1011712 () here if we wanted. The original banner (aligned to top) is causing a little grief in implementation given that it sits in "content" and would even be in conflict with promo banners at times. The popup "PRO TIP" approach gives more of a "tip" feel and can be made to be low-obtrusive, but offering the user a "this thing means something different than other UI in the app" feel. It seems like we are striving for a consistency in tips as well, and I like having a solid base fo code being reused in other situations. It makes adding new tip faster in the future (at least in places where the tip popup is appropriate).
In order to animate the titles in the home pager tab strip, we'll need to either 1) overlay a fake tab strip on top of the Android PagerTabStrip and hide it after the animation, or 2) copy the code for PagerTabStrip and PagerTitleStrip and make those elements protected (instead of package-private). I'm going to experiment with (2) because if we want this animation all the time, taking the approach of (1) and overlaying a FrameLayout will probably make for some negative perf, and we'll also have to handle the case where the user swipes during the animation.
Both of these approaches are less than good. I'm not keen on copying Android source code, and dealing with the maintenance, for this feature.
Friendly ping on status here?
I have a proof of concept with ripping ~100k of code from the Android source, but I started trying to finish some of the "close all tabs"/recent pages work. I'll finish up my rough mock and post something for UX to try in the next day or two.
Were you going to look into using reflection too?
Yep, that's what I'm looking at now.
This is obviously not going to make 33, so dropping that flag.
tracking-fennec: 33+ → ---
tracking-fennec: --- → ?
antlam, here's an apk for you to try: Let me know about the timing.
Attachment #8460555 - Flags: review?(lucasr.at.mozilla)
Flags: needinfo?(alam)
Hey Chenxia! Awesome work! Seeing the animation is nice. In regards to timing, I'd like to try a .25 second delay if possible. Could we also get it to slide in faster and bounce with a little less tension? It's definitely close, I think we just need to fine tune it. :)
Flags: needinfo?(alam)
To bring this back to the discover-ability issue - have we thought about also exposing these two as menu options tucked into the "3 dot" overflow icon?
Anthony and I have been going back and forth on IRC, but the various tweaked apks are here:
Currently going back and forth with antlam about timing and the feel of the animation. This is the current iteration, and has a somewhat complex chain of animations to get the correct bounce.
Attachment #8460555 - Attachment is obsolete: true
Attachment #8460555 - Flags: review?(lucasr.at.mozilla)
Attachment #8460666 - Flags: feedback?(lucasr.at.mozilla)
Thanks for all the builds Chenxia! It's coming along nicely! I'm testing out v.4 right now and I think it overshoots too much on the first time around. It also seems like it doesn't bounce "back" enough (I think it stops right now as soon as it comes back around?) You mentioned that this was essentially setting multiple "destination" points for the labels to hit, might I ask how many we have right now?
Flags: needinfo?(liuche)
Comment on attachment 8460666 [details] [diff] [review] Patch: WIP Homescreen contextual hint v2 Review of attachment 8460666 [details] [diff] [review]: ----------------------------------------------------------------- Let's not rely on reflection. Animation looks good, I'd like to see it refactored out into its own class. ::: mobile/android/base/home/HomePagerTabStrip.java @@ +59,5 @@ > + try { > + // Use reflection to make the Tab Strip titles accessible for animation. > + final Field prevText = PagerTitleStrip.class.getDeclaredField("mPrevText"); > + prevText.setAccessible(true); > + final View prevTextView = (View) prevText.get(this); Can't you simply do a getChildAt(0) here? As you know, this is likely to break in the future. @@ +63,5 @@ > + final View prevTextView = (View) prevText.get(this); > + > + final Field nextText = PagerTitleStrip.class.getDeclaredField("mNextText"); > + nextText.setAccessible(true); > + final View nextTextView = (View) nextText.get(this); Same here, maybe just do getChildAt(getChildCount() - 1)? Just bail (instead of crashing) if these views cannot be found. @@ +81,5 @@ > + // Two-part animator for a softer bounce. > + final PropertyAnimator softBounceAnimator = new PropertyAnimator(ANIMATION2_DURATION_MS, new AccelerateInterpolator()); > + softBounceAnimator.attach(prevTextView, Property.TRANSLATION_X, -bounceDistance + BOUNCE_OFFSET); > + softBounceAnimator.attach(nextTextView, Property.TRANSLATION_X, bounceDistance - BOUNCE_OFFSET); > + softBounceAnimator.addPropertyAnimationListener(new PropertyAnimationListener() { Maybe factor out this bounce animation code into a separate class or something? @@ +92,5 @@ > + // Bounce animation that leaves the text in its original position. > + final PropertyAnimator bounceAnimator = new PropertyAnimator(20, new BounceInterpolator()); > + bounceAnimator.attach(prevTextView, Property.TRANSLATION_X, 0); > + bounceAnimator.attach(nextTextView, Property.TRANSLATION_X, 0); > + bounceAnimator.start(); This reminds me we should revisit that idea of switching to NineOldAndroids... @@ +115,5 @@ > + @Override > + public void run() { > + animator.start(); > + } > + }, ANIMATION_DELAY_MS); Why the delay? ::: mobile/android/base/resources/layout/home_pager.xml @@ +13,5 @@ > android: android: > > <org.mozilla.gecko.home.HomePagerTabStrip android: + android:layout_height="40dip" Maybe this should go in a separate patch?
Attachment #8460666 - Flags: feedback?(lucasr.at.mozilla) → feedback+
tracking-fennec: ? → 34+
Anthony, a few more apks for you to try: See 5, 6A, and 6B.
Flags: needinfo?(liuche)
Anthony, I added another apk that has a smaller bounce and also does the stairstep bounce *mimes some hand motions* See 7-small-bounce.apk at I can tweak a lot of things there easily now - duration of each leg of the bounce, distance of each leg, fading speed, etc. Let me know what you think.
Flags: needinfo?(alam)
This is a custom BounceAnimator that chains AccelerateInterpolators so that in the future, we can have more control over making bounce animations the way we want instead of using Android's BounceInterpolator (which is hard-coded to have 4 bounces).
This will eventually be merged with the Homescreen contextual hint patch, but is just a WIP demonstrating the use of the BounceAnimator.
Comment on attachment 8465964 [details] [diff] [review] Part 2-ish: BounceAnimator On second thought, I'll ask for feedback if this turns out to be something that antlam likes.
(oops, uploaded a completely empty patch)
Attachment #8465965 - Attachment is obsolete: true
Animation looks great since the 8th iteration. I think next step will just be to wrap up the snippet and then we can wrap this up and let it simmer for a bit. I'm really interested to see how all the elements come together to help build more context around about:home in the users mind.
Flags: needinfo?(alam)
Bounce animator - need to fix gravity in the titlebar :/
Attachment #8460666 - Attachment is obsolete: true
Attachment #8465964 - Attachment is obsolete: true
Attachment #8466382 - Attachment is obsolete: true
Comment on attachment 8476396 [details] [diff] [review] Part 2: Home page snippet Review of attachment 8476396 [details] [diff] [review]: ----------------------------------------------------------------- Looking pretty good overall. Just not entirely sure about the telemetry aspect and copy. Up to you to decide. ::: mobile/android/components/Snippets.js @@ +384,5 @@ > + let id = Home.banner.add({ > + text: text, > + icon: "drawable://homepage_banner", > + onclick: function() { > + // Remove the message and never show it again. I'd keep the original comment here ("Remove the message, so that it won't show again for the rest of the app lifetime.") as it's more accurate. @@ +386,5 @@ > + icon: "drawable://homepage_banner", > + onclick: function() { > + // Remove the message and never show it again. > + Home.banner.remove(id); > + Services.prefs.setBoolPref("browser.snippets.homepage.enabled", false); Add a comment here stating that this will ensure we'll only show this banner once. @@ +388,5 @@ > + //. ::: mobile/android/locales/en-US/chrome/aboutHome.properties @@ +1,5 @@ > +# This Source Code Form is subject to the terms of the Mozilla Public > +# License, v. 2.0. If a copy of the MPL was not distributed with this > +# file, You can obtain one at. > + > .
Attachment #8476396 - Flags: review?(lucasr.at.mozilla) → review+
> > + //. The Extra ("homepage") is how we have been differentiating snippets. In this case I think "homepage" might be too generic. Maybe "firstrun-homepage" might be better. Maybe change the preference too: "browser.snippets.firstrunHomepage.enabled" since we use camelCase for the existing "browser.snippets.syncPromo.enabled"
> > . Anthony, any thoughts on this?
"Get back here" was actually intended to feel like an action because I have a sneaky suspicion that a part of the issue here is establishing a mental connection in the user's mind that "New tab" can show them so much more than just a URL bar to type in. That being said, I think it's still too much hope to pin on a single string of copy so I'd be open to changing it :). Let's try that for now and we can revisit the copy later cause I can't think of a much better one ATM and I don't want this to block it. :)
Addressed other comments from earlier feedback? request. > > @@ +115,5 @@ > > + @Override > > + public void run() { > > + animator.start(); > > + } > > + }, ANIMATION_DELAY_MS); > > Why the delay? > I think the delay is so that users will notice the animation more because it won't be happening with all the other startup motion (views appearing, etc).
Attachment #8476395 - Attachment is obsolete: true
Attachment #8477020 - Flags: review?(lucasr.at.mozilla)
Addressed comments, carrying over r+.
Attachment #8476396 - Attachment is obsolete: true
Attachment #8477021 - Flags: review+
(In reply to Chenxia Liu [:liuche] from comment #36) > I think the delay is so that users will notice the animation more because it > won't be happening with all the other startup motion (views appearing, etc). Yep! :D
Comment on attachment 8477020 [details] [diff] [review] Part 1: Bounce animator v2 Review of attachment 8477020 [details] [diff] [review]: ----------------------------------------------------------------- This looks good but I think you can simplify this patch by implementing a custom interpolator instead of the more complex bound animator (if OvershootInterpolator is not good enough). I don't feel strongly about it though. Up to you if you want to go ahead with this approach or try implementing with an interpolator. ::: mobile/android/base/animation/BounceAnimator.java @@ +17,5 @@ > + * > + * After constructing an instance, animations can be queued up sequentially with the > + * {@link #queue(Attributes) queue} method. > + */ > +public class BounceAnimator extends ValueAnimator { It seems to me that what you need is a custom Interpolator instead. ::: mobile/android/base/home/HomePagerTabStrip.java @@ +58,5 @@ > + final View nextTextView = getChildAt(getChildCount() - 1); > + > + if (prevTextView == null || nextTextView == null) { > + return; > + } nit: add empty line here. @@ +63,5 @@ > + // Set up initial values for the views that will be animated. > + ViewHelper.setTranslationX(prevTextView, -INIT_OFFSET); > + ViewHelper.setAlpha(prevTextView, 0); > + ViewHelper.setTranslationX(nextTextView, INIT_OFFSET); > + ViewHelper.setAlpha(nextTextView, 0); Better not mix our framework with NineOldAndroids. Use ViewHelper from NineOldAndroid instead. @@ +83,5 @@ > +.
Attachment #8477020 - Flags: review?(lucasr.at.mozilla) → review+
(In reply to Lucas Rocha (:lucasr) from comment #39) > > +. Given the timelines, I'd consider looking at OvershootInterpolator in a follow up bug. Simplification would be better, but let's get this feature landed if the rest of the approach is OK.
Filed follow-up bug 1057569. I tried a custom interpolator at some point, but it wasn't very extensible - it was pretty similar to the Android BounceInterpolator though, but I could try again coming up with a better equation. Landed with bug 1056976.
Status: NEW → ASSIGNED
Target Milestone: --- → Firefox 34
Status: ASSIGNED → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
Comment on attachment 8477021 [details] [diff] [review] Part 2: Home page snippet Review of attachment 8477021 [details] [diff] [review]: ----------------------------------------------------------------- ::: mobile/android/components/Snippets.js @@ +416,1 @@ > loadSyncPromoBanner(); Drive-by: We don't want the sync promo at all during the first run? Home banner messages rotate, so it is possible to have both. Also, FYI, if the snippets update timer fires some time during the first run session, new snippets will get added to the rotation anyway (loadSnippetsFromCache doesn't actually do anything on first run, since there is no cache).
Oh, I didn't realize that, actually. Filed bug 1058100.
Product: Firefox for Android → Firefox for Android Graveyard | https://bugzilla.mozilla.org/show_bug.cgi?id=1021751 | CC-MAIN-2021-10 | refinedweb | 2,670 | 56.05 |
Revision history for Perl extension GIS-Distance. 0.18 2019-05-10T20:31:31Z - Switch to the GNU General Public License version 3. - Fixed pod error as reported by CPANTS. - Documentation edits. 0.17 2019-03-17T17:01:31Z - Finalize the new internal formula interface. 0.16 2019-03-16T22:19:33Z - Renamed the args and module attributes to formula_args and formula_module to make them less generic and a bit more accurate. - Fix a pretty bad math typo in the SPEED section. - Lots of other documentation edits. 0.15 2019-03-13T06:06:09Z - Support using Geo::Point as arguments to distance(). - Lots of documentation edits. - Moved TODO section into GitHub issues. - Added a benchmark to the SPEED section. - Made a bunch of improvements to the author tools. 0.14 2019-03-10T05:04:54Z - Add GIS::Distance::ALT formula. - Removed distance_km(). 0.13 2019-03-09T12:32:55Z - Add abs() to Haversine. - Added GIS::Distance::Null, the fastest formula yet. 0.12 2019-03-08T18:32:13Z - Added the distance_metal() method to GIS::Distance. - Various documentation edits, including a new SPEED section. 0.11 2019-03-07T22:23:02Z - Support the GIS_DISTANCE_PP environment variable. - Don't support older ::Formula modules, makes no sense and they wouldn't work anyways. - Declare Carp dep. - Lots and lots of documentation edits. - Recommend the newer GIS::Distance::Fast. 0.10 2019-03-07T16:28:48Z - WARNING: The GIS::Distance object is now immutable, thus the formula can no longer be set with the formula attribute! - Moved GIS::Distance::GeoEllipsoid to a separate distro. - Added the distance_km() method to GIS::Distance. - Removed Moo and Type::Tiny, all unecessary, simple OO. - Move GIS::Distance::Formula:: modules to GIS::Distance::. - Migrate build tooling from Dist::Zilla to Minilla. 0.09 2015-06-11 - Move away from Any::Moose to Moo (yay!). - Better formula loading logic. - Support single-arg (formula) GIS::Distance instantiation. 0.08 2012-03-23 - Release with Dist::Zilla. - Fix Great Circle formula to use ** instead of ^. 0.07 2010-02-02 - Use Any::Moose instead of Moose directly. - Declare namespace::autoclean dependency. 0.06 2010-01-30 - Minor build updates to include some extra info (github, etc). 0.05 2010-01-30 - Speed improvements under Moose (now uses immutable). 0.04 2009-06-29 - Fixed for latest Moose. - Fixed for latest Class::Measure. 0.03 - Fixed for latest Moose. - Reduced the README to a one-liner. - Added docs to GIS::Distance::Formula. - Fixed some documentation typos that were using GID instead of GIS. - Fixed a typo that mispelled "formula" as "formuka". - Refer to GIS::Distance::Fast in the SEE ALSO section. - Added a one liner to the SYNOPSIS showing how to used the returned distance object. - Added a TEST COVERAGE section with output from Devel::Cover. 0.02 2008-03-16 - Added Geoid to the TODO section. - Using Module::Install now. - Moved all formulas in to the GIS::Distance::Formula namespace. - Using Moose for all OO now. - Added tests (bout time!). - Added support for the up-and-coming ::Fast:: modules. - Added (BROKEN) to the abstract for the GreatCircle and Polar formulas. - Fixed POD testing. - Changed version scheme to use the simple x.xx format. 0.01001 2006-09-20 - Added basic META.yml. - Geo::Ellipsoid support. - Added a TODO section. - Various bug fixes to the Vincenty formula. - GreatCircle formula marked as broken. - Added a dev script for graphing the deviations in the formulas. 0.01000 2006-09-19 - Renamed from Geo::Distance to GIS::Distance. - Moved distance calculations in to their own modules. - Use Class::Measure::Length to handle distance return values. - Test POD syntax. Revision history for Perl extension Geo-Distance. 0.11 2005-09-01 - Fixed some errors in the documentation. - Added an highly accurate ellipsoid formula. - lon_field and lat_field were not being used by closest. (D. Hageman) 0.10 2005-07-11 - The closest() method has a changed argument syntax and no longer supports array searches. - The closest() method works! - A real gcd formula (still, hsin is much better). - Tweaked docs. - Added some tests (yay!). 0.09 2005-04-01 - Modified the todo list to include ideas for future algorithms. - Fixed the nautical mile, mile, yard, and light second units. - Added the British spellings for kilometre, metre, and centimetre. - Added the poppy seed, barleycorn, rod, pole, perch, chain, furlong, league, fathom, millimeter, and millimetre units. - The totw.pl script was written by Ren and can be used to take over the world. 0.08 2005-03-20 - Updated the README description. - Removed debug print()s. Eeek! 0.07 2005-03-16 - Intermixed documentation with code so it is easier to keep the docs up-to-date. - OO interface only - method interface completely removed. - By default no units are defined. Call default_units. - Slightly more precise measurement of the base kilometer rho. - Added "nautical mile" unit type. - Reworked the closest() function. 0.06 2004-06-29 - Optional Haversine formula. - Misc documentation tweaks. 0.05 2003-03-19 - Added a note in the documentation about the inaccuracies of using Math::Trig. - The 'mile' unit was being calculated wrong which meant it was returning very inaccurate distances. - Fixed a silly bug where a sub was being relied on that no longer exists. - Documentation tweaks as usual. 0.04 2003-02-18 - Documentation revised once again. - Added reg_unit() for adding your own unit type. - find_closest has been overhauled: - Now accepts more than one field in the field=>'' parameter. - Will now return an array reference of distances instead of attaching the distances to the locations array ref - A little more effecient. - Now accepts a count argument. - Accepts an array reference for searching. Mostly good for testing, but who knows? - Removed geo_ portion of names for exported functions. - Removed some of the input checking. Just not necessary. - Enhanced tests. Now we're actually doing some real testing. Need more tests tho. 0.03 2003-02-15 - Documentation modified. - Added find_closest() which accepts a $dbh for searching in an SQL database. - distance_dirty() can now accept locations as array refs. 0.02 2003-02-14 - Based on a suggestion by Jack D. I migrated the code to use Math::Trig for most of the distance math. - POD documentation written. - Object oriented interface created. 0.01 - First version. | http://web-stage.metacpan.org/changes/distribution/GIS-Distance | CC-MAIN-2019-35 | refinedweb | 1,038 | 63.46 |
Red […]
The post Bringing IoT to Red Hat AMQ Online appeared first on Red Hat Developer Blog.]]>
Red into this general-purpose messaging layer. And the whole reason why you need an IoT messaging layer is so you can focus on connecting your cloud-side application with the millions of devices that you have out there.
Eclipse Hono is an IoT abstraction layer. It defines APIs in order to build an IoT stack in the cloud, taking care of things like device credentials, protocols, and scalability. For some of those APIs, it comes with a ready-to-run implementation, such as the MQTT protocol adapter. For others, such as the device registry, it only defines the necessary API. The actual implementation must be provided to the system.
A key feature of Hono is that it normalizes the different IoT-specific protocols on AMQP 1.0. This protocol is common on the data center side, and it is quite capable of handling the requirements on throughput and back-pressure. However, on the IoT devices side, other protocols might have more benefits for certain use cases. MQTT is a favorite for many people, as is plain HTTP due to its simplicity. LoRaWAN, CoAP, Sigfox, etc. all have their pros and cons. If you want to play in the world of IoT, you simply have to support them all. Even when it comes to custom protocols, Hono provides a software stack to easily implement your custom protocol.
Hono requires an AMQP 1.0 messaging backend. It requires a broker and a component called “router” (which doesn’t own messages but only forwards them to the correct receiver). Of course, it expects the AMQP layer to be as scalable as Hono itself. AMQ Online is a “self-service,” messaging solution for the cloud. So it makes sense to allow Hono to run on top of it. We had this deployment model for a while in Hono, allowing the use of EnMasse (the upstream project of AMQ Online).
In a world of Kubernetes and operators, the thing that you are actually looking for is more like this:
kind: IoTProject apiVersion: iot.enmasse.io/v1alpha1 metadata: name: iot namespace: myapp spec: downstreamStrategy: managedStrategy: addressSpace: name: iot plan: standard-unlimited addresses: telemetry: plan: standard-small-anycast event: plan: standard-small-queue command: plan: standard-small-anycast
You simply define your IoT project, by creating a new custom resource using
kubectl create -f and you are done. If you have the IoT operator of AMQ Online 1.1 deployed, then it will create the necessary address space for you, and set up the required addresses.
The IoT project will also automatically act as a Hono tenant. In this example, the Hono tenant would be
myapp.iot, and so the full authentication ID of e.g.
sensor1 would be
sensor1@myapp.iot. The IoT project also holds all the optional tenant configuration under the section
.spec.configuration.
With the Hono admin tool, you can quickly register a new device with your installation (the documentation will also tell you how to achieve the same with
curl):
$ # register the new context once with 'hat' $ hat context create myapp1 --default-tenant myapp.iot(oc -n messaging-infra get routes device-registry --template='{{ .spec.host }}') $ # register a new device and set credentials $ hat reg create 4711 $ hat cred set-password sensor1 sha-512 hono-secret --device 4711
With that, you can simply use Hono as always. First, start the consumer:
$ # from the hono/cli directory $ export MESSAGING_HOST=$(oc -n myapp get addressspace iot -o jsonpath={.status.endpointStatuses[?(@.name==\'messaging\')].externalHost}) $ export MESSAGING_PORT=443 $ mvn spring-boot:run -Drun.arguments=--hono.client.host=$MESSAGING_HOST,--hono.client.port=$MESSAGING_PORT,--hono.client.username=consumer,--hono.client.password=foobar,--tenant.id=myapp.iot,--hono.client.trustStorePath=target/config/hono-demo-certs-jar/tls.crt,--message.type=telemetry
And then publish some data to the telemetry channel:
$ curl -X POST -i -u sensor1@myapp.iot:hono-secret -H 'Content-Type: application/json' --data-binary '{"temp": 5}'(oc -n enmasse-infra get routes iot-http-adapter --template='{{ .spec.host }}')/telemetry
For more detailed instructions, see: Getting Started with Internet of Things (IoT) on AMQ Online.
As mentioned before, you don’t do IoT just for the fun of it (well, maybe at home, with a Raspberry Pi, Node.js, OpenHAB, and mosquitto). But when you want to connect millions of devices with your cloud backend, you want to start working with that data. Using Hono gives you a pretty simple start. Everything you need is an AMQP 1.0 connectivity. Assuming you use Apache Camel, pushing telemetry data towards a Kafka cluster is as easy as (also see ctron/hono-example-bridge):
<route id="store"> <from uri="amqp:telemetry/myapp.iot" /> <setHeader id="setKafkaKey" headerName="kafka.KEY"> <simple>${header[device_id]}</simple> </setHeader> <to uri="kafka:telemetry?brokers={{kafka.brokers}}" /> </route>
Bringing together solutions like Red Hat Fuse, AMQ and Decision Manager makes it a lot easier to give your custom logic in the data center (your value add‑on) access to the Internet of Things.
AMQ Online 1.1 is the first version to feature IoT as a tech preview. So, give it a try, play with it, but also keep in mind that it is a tech preview.
In the upstream project EnMasse, we are currently working on creating a scalable, general purpose device registry based on Infinispan. Hono itself doesn’t bring a device registry, it only defines the APIs it requires. However, we think it makes sense to provide a scalable device registry, out of the box, to get you started. In AMQ Online, that would then be supported by using Red Hat Data Grid.
In the next months, we hope to also see the release of Eclipse Hono 1.0 and graduate the project from the incubation phase. This is a big step for a project at Eclipse but also the right thing to do. Eclipse Hono is ready, and graduating the project means that we will pay even closer attention to APIs and stability. Still, new features like LoRaWAN, maybe Sigfox, and a proper HTTP API definition for the device registry, are already under development.
So, there are lots of new features and enhancements that we hope to bring into AMQ Online 1.2.
The post Bringing IoT to Red Hat AMQ Online […]
The post IoT edge development and deployment with containers through OpenShift: Part 2 Why you should care about RISC ‘G’):
An early access RISC-V development system. Upper right is the HiFive board. Bottom is a VC707 board which provides a PCIe bridge. Middle left is a PCIe riser board. At the top is a commodity PCIe SSD card. Connections on the right: USB serial console, ethernet, power. Additional mess is optional, and at the discretion of the desk owner.
But are there down sides to choosing an open core? Well, there are considerations that anyone should be aware of when choosing any core. Here are a few:
The post Why you should care about RISC-V […]
The post Announcing AMQ Streams: Apache Kafka on OpenShift:
oc createcommand.:
We expect to release further previews as we iterate towards the general availability release, which is planned for later this year.
Please give it a try and let us know what you think.
The post Announcing AMQ Streams: Apache Kafka on OpenShift […]
The post IoT Developer Survey – Deadline March 5, 2018 output from this survey will help the open source community focus on the resources most needed by IoT developers.
The survey is organized by the Eclipse IoT Working Group, IEEE IoT Initiative, the Open Mobile Alliance, and the AGILE-IoT H2020 Research Project.
The survey deadline is this Monday, March 5, 2018.
Don’t procrastinate — take the survey now:
This survey should take you about 10 minutes to complete.
The post IoT Developer Survey – Deadline March 5, 2018 […]
The post ARM TechCon 2017 – Embedded, IoT, Networking, and more….
The post ARM TechCon 2017 – Embedded, IoT, Networking, and more… […]
The post Open IoT Challenge – CFP deadline next week to prototype. You can use any open source technology to build the IoT solution.
The Open IoT Challenge 4.0 CFP deadline is November 13, 2017. You will have until mid-March, 2018 to build your solution.
The top 10 proposals will receive a hardware development kit to help get them started.
The solutions will be judged on the following criteria:
For more information on the Open IoT Challenge:
Join the Red Hat Developer Program (it’s free) and get access to related cheat sheets, books, and product downloads.
The post Open IoT Challenge – CFP deadline next week). Use Case Description Let’s consider a Money transfer institution operating […]
The post Tutorial: Building and consuming Virtual Microdatabase with JBoss Data Virtualization).
Let’s consider a Money transfer institution operating in EMEA region. For business purposes, the institution has two relational databases:
The two databases are shipped as images on Docker Hub. In the following section, we will create an EMEA federated live view containing customer’s data from the two databases..
Source model establishes a link with the physical database we want to work with. Follow the following steps to create a source model for the MySQL database containing transactions for African segment.:
Select the Public tables: eu_customer and eu_moneytransfer.
Associate these two tables to a source model with the following configuration:
You should now be able to preview Posgres Data using the Modeling>Preview Data Menu Action on the eu_customers table.
Remember customer table store both senders and receivers details.]
Once the VDB is deployed, you can access it through various interfaces including Rest OData or Teiid JDBC..
The post Tutorial: Building and consuming Virtual Microdatabase with JBoss Data Virtualization […]
The post Jug Summer Camp 2017, Vert.x and collaborative DJ mix groups.
My talk was an introduction to reactive programming with Eclipse Vert.x, featuring demos with RxJava-based edge services as well a collaborative DJ mix session. The great thing about Vert.x is that it scales well for all kinds of distributed applications.
@dadoonet @jponge co mixing at @jugsummercamp #summercamp2017 pic.twitter.com/bz3zM0UkgZ
— ? Ph. Charrière (@k33g_org) September 15, 2017
The DJ mix demo (called Boiler Vroom) allowed attendees to connect to a “real-time” web application where they could see live the DJ actions, have the control on some of the elements (filters, sequencer patterns…) and listen to the stream. There was also a WiFi-connected RaspberryPi to provide a volume meter:
#Vert.x + #gololang + #Traktor = VU-meter with a #RPi pic.twitter.com/jktUjJzt8w
— Julien Ponge (@jponge) June 12, 2017
Vert.x shined here on several fronts:
Here are the slides:
If you understand French, here is the video of the talk:
Many thanks to the organizers and attendees for this great event.
Download the Eclipse Vert.x cheat sheet, this cheat sheet provides step by step details to let you create your apps the way you want to.
The post Jug Summer Camp 2017, Vert.x and collaborative DJ mix appeared first on Red Hat Developer Blog.]]> | https://developers.redhat.com/blog/category/iot/feed/atom/ | CC-MAIN-2019-22 | refinedweb | 1,849 | 56.15 |
Before the Big Bang, there was no time. But after the Big Bang, you can import time in Python. I know it’s a bad joke. But to practice Python projects, you can make a lot of time-related projects in Python like a digital clock, countdown timer, alarm clock, and stopwatch. Let’s make a digital clock using Python and Tkinter.
Tkinter is mainly used for UI in Python projects. Before jumping into the code, you have to do some setup for the project. So, let’s start this project with a cup of coffee.
If you are looking for a video tutorial, then it’s here:
Project Setup and Installation
As you know, making a Python project in a virtual environment is a good habit. So, start with a virtual environment.
- If you are a Linux user, then make a virtual environment using the following command. But make sure that the virtual environment is installed in your system.
virtualenv -p python3.8 clock cd clock source bin/activate
If you are a Windows user then make a virtual environment using the following command. But make sure that the virtual environment is installed in your system.
virtualenv clock clock\scripts\activate
Now, make a python file inside this “clock” directory and name it as the file.py.
Install the required package.
pip install tk
To use the “tk” library, you must have Tkinter installed in your system. Tkinter is a Python package that mainly deals with the GUI part of a Python project.
To install Tkinter in Linux, you can use the command as shown below.
sudo apt install python3.8-tk
If you are a windows user then, you can install Tkinter at the time of Python installation. You can install it manually too.
Clock using Python
Code of Digital Clock in Python
file.py
from tkinter import * from time import strftime root = Tk() root.geometry("500x500") root.resizable(0,0) root.title('Python Clock') Label(root,text = 'YOUTUBE - BUG NINZA', font ='arial 20 bold').pack(side=BOTTOM) def time(): string = strftime('%H:%M:%S %p') mark.config(text = string) mark.after(1000, time) mark = Label(root, font = ('calibri', 40, 'bold'), pady=150, foreground = 'black') mark.pack(anchor = 'center') time() mainloop()
Import the package. We are using the system’s time. To retrieve the system’s time, import strftime function.
Define a window screen. Here the dimension of the window screen is 500x500. You can’t resize the window screen. After that, set a title for the window screen. It’s optional.
I added a footer label on the window screen too. It’s also optional.
Define the time function. The format would be H: M: S.
Style the clock widget to make it attractive. The font size of the clock string is set to 40 and I added vertical padding of 150. The foreground color is set to “black”.
Add a mainloop function at the end.
Our code is complete. Now run the “file.py” file using the following command:
python file.py
After the successful run, you will see something like this on-screen.
Clock using Python
Congratulations! You successfully built a Python project.
If this article sounds informative to you, then make sure to follow and share it with your geek community.
Here are more Python projects for practice!
Make a Calculator using Python and Tkinter
YouTube Video Downloader using Python and Tkinter
Text to Speech Conversion using Python with gTTS
Capture and Process Video Footage from a Webcam using OpenCV Python
Rock, Paper, and Scissors Game using Python Programming
Random Password Generator using Python Programming
Shorten Your URL using Python and Bitly
Happy coding!
Hello, My Name is Rohit Kumar Thakur. I am open to freelancing. I build React Native projects and am currently working on Python Django. Feel free to contact me at (freelance.rohit7@gmail.com). | https://plainenglish.io/blog/make-a-digital-clock-using-python-and-tkinter | CC-MAIN-2022-40 | refinedweb | 644 | 69.28 |
Capturing stderr and exceptions from python in org-mode
Posted September 27, 2013 at 07:37 PM | categories: org-mode | tags: | View Comments
Updated September 27, 2013 at 07:47 PM
I have used org-mode extensively to create examples of using python using the code blocks. For example to illustrate the difference between integer and float division you can do this:
print 1 / 3 print 1.0 / 3.0
0 0.333333333333
There are some limitations to showing output though. For example, the code blocks do not capture anything from stderr.
import sys print >>sys.stderr, 'message to stderr'
And exceptions result in no output whatsoever. That is not helpful if you are trying to teach about exceptions!
I discovered a way around this. The key is using a python sandbox that redirects stdout, stderr and that captures anything sent to those channels. You can also capture any exceptions, and redirect them to a variable. Finally, you can construct the output anyway you see fit.
Below is the code that runs python code in a sandbox, with redirected outputs. I defined a function that temporarily redirects the output to stdout and stderr, so they can be captured. I execute the code wrapped in a try/except block to capture any exceptions that occur. Finally, I construct a string formatted in a way that lets you know what was on stdout, stderr, and what was an exception.
#!/usr/bin/env python from cStringIO import StringIO import os, sys def Sandbox(code): '''Given code as a string, execute it in a sandboxed python environment return the output, stderr, and any exception code ''' old_stdout = sys.stdout old_stderr = sys.stderr redirected_output = sys.stdout = StringIO() redirected_error = sys.stderr = StringIO() ns_globals = {} ns_locals = {} out, err, exc = None, None, None try: exec(code, ns_globals, ns_locals) except: import traceback exc = traceback.format_exc() out = redirected_output.getvalue() err = redirected_error.getvalue() # reset outputs to the original values sys.stdout = old_stdout sys.stderr = old_stderr return out, err, exc if __name__ == '__main__': content = sys.stdin.read() out, err, exc = Sandbox(content) s = '''---stdout----------------------------------------------------------- {0} '''.format(out) if err: s += '''---stderr----------------------------------------------------------- {0} '''.format(err) if exc: s += '''---Exception-------------------------------------------------------- {0} '''.format(exc) print s
To use this, we have to put this file (sandbox.py) in our PYTHONPATH. Then, we tell org-babel to run python using our new sandbox.py module. org-babel pipes the code in a src block to stdin of the python command, which will be intercepted by our sandbox module. If you put this in your init.el, or other customization location, then subsequent uses of python in org-mode will use your sandbox module. I usually only run this for a session as needed.
(setq org-babel-python-command "python -m sandbox")
Now, when we use python, we can capture output to stderr!
import sys print >>sys.stderr, 'message to stderr'
---stdout----------------------------------------------------------- ---stderr----------------------------------------------------------- message to stderr
And, we can capture exceptions!
print 1 / 0
---stdout----------------------------------------------------------- ---Exception-------------------------------------------------------- Traceback (most recent call last): File "c:\Users\jkitchin\Dropbox\blogofile-jkitchin.github.com\_blog\sandbox.py", line 20, in Sandbox exec(code, ns_globals, ns_locals) File "<string>", line 1, in <module> ZeroDivisionError: integer division or modulo by zero
There is a little obfuscation in the exception, since it technically occurs in the Sandbox, but this is better than getting no output whatsoever! I have not tested the sandbox.py code extensively, so I don't know if there will be things that do not work as expected. If you find any, please let me know!
Copyright (C) 2013 by John Kitchin. See the License for information about copying. | http://kitchingroup.cheme.cmu.edu/blog/2013/09/27/Capturing-stderr-and-exceptions-from-python-in-org-mode/ | CC-MAIN-2017-39 | refinedweb | 595 | 59.5 |
Destructors are functions which are just the opposite of constructors. In this chapter, we will be talking about destructors.
We all know that constructors are functions which initialize an object. On the other hand, destructors are functions which destroy the object whenever the object goes out of scope.
It has the same name as that of the class with a tilde (~) sign before it.
class A
{
public:
~A();
};
Here, ~A() is the destructor of class A.
When is a destructor called?
A destructor gets automatically called when the object goes out of scope. We know that a non-parameterized constructor gets automatically called when an object of the class is created. Exactly opposite to it, since a destructor is also always non-parameterized, it gets called when the object goes out of scope and destroys the object.
If the object was created with a new expression, then its destructor gets called when we apply the delete operator to a pointer to the object. We will learn more about new and delete in the chapter Dynamic Memory Allocation.
Destructors are used to free the memory acquired by an object during its scope (lifetime) so that the memory becomes available for further use.
Let's see an example of a destructor.
#include <iostream> using namespace std; class Rectangle { int length; int breadth; public: void setDimension(int l, int b) { length = l; breadth = b; } int getArea() { return length * breadth; } Rectangle() // Constructor { cout << "Constructor" << endl; } ~Rectangle() // Destructor { cout << "Destructor" << endl; } }; int main() { Rectangle rt; rt.setDimension(7, 4); cout << rt.getArea() << endl; return 0; }
28
Destructor
In this example, when the object 'rt' of class Rectangle was created, its constructor was called, no matter in what order we define it in the class. After that, its object called the functions 'setDimension' and 'getArea' and printed the area. At last, when the object went out of scope, its destructor got called.
Note that the destructor will get automatically called even if we do not explicitly define it in the class.
The difference between ordinary and extraordinary is practice.
-Vladimir Horowitz | https://www.codesdope.com/cpp-destructors/ | CC-MAIN-2022-40 | refinedweb | 345 | 63.59 |
As you know, node-sets are XPath's way of dealing with multiple nodes. For example, you can see the node-set returned by the expression //planet on our sample XML document in the XPath Visualiser in Figure 2.7. But there's more to know about node-sets.
When you're working with a node-set, XPath gives you a variety of resources that are available at any time called the XPath context . You'll see more about what's in the XPath context in the upcoming chapters; here's what in it:
The context node , which is the XML node in the XML document that the XPath expression was invoked on. In other words, XPath expressions are executed starting from the context node. We'll see how to use relative expressions in XPath soon, and such expressions are always relative to the context node.
The context position , which is a nonzero positive integer indicating the position of a node in a node-set. The first node has position 1, the next position 2, and so on.
The context size , which is also a nonzero positive integer, the context size gives the maximum possible value of the context position. (It's the same as the number of nodes in a node-set.)
A set of variables you can use variables to hold data in XSLT, and if you do, those variables are stored in the expression's context, which can be accessed in XPath.
A function library full of functions ready for you to call, such as the sum function, which returns the sum of the numbers you pass it.
The set of XML namespace declarations available to the expression.
In addition to these context items, there is also the current node, which we've already discussed. The current node is not the same as the context node . The context node is set before you start evaluating an XPath expressionit's the node the expression is invoked on. However, as the XPath processor evaluates an XPath expression, it can work on various parts of that expression piece by piece, and the node that the XPath processor is working on at the moment is called the current node.
Here's an example showing how to work with context nodes and positions . Say that you apply the XPath expression /planets/planet to our planetary data:
<?xml version="1.0"?> > . . .
The first / in /planets/planet makes the root node the context node for the rest of the expression. The planets part makes the <planets> element the context node for the rest of the expression after that point. That means that the remainder of this expression, /planet , will be evaluated with respect to the <planets> element, so the <planets> element is the context node for the /planet part of this XPath expression.
The whole expression, /planets/planet , matches and returns the three <planet> elements in a node-set. The first <planet> element will have the context position 1, the next will have context position 2, and so on. The context size of the node-set containing the three <planet> elements is three.
Here's an example showing how to work with the variables present in a node-set context. XPath doesn't let you define variables. However, you can create variables in an XSLT stylesheet with the <xsl:variable> element like this, where I'm creating a variable named myPosition with the value 3:
<xsl:variable
This new XSLT variable, myPosition , can be used in XPath expressions. For example, as we saw in Chapter 1, you can assign XPath expressions to the XSLT <xsl:value-of> element's select attribute. And in XPath, you can refer to the value in a variable by prefacing the variable's name with a $ , as you see in ch02_02.xsl in Listing 2.2.
<?xml version="1.0"?> <xsl:stylesheet <xsl:template <HTML> <xsl:apply-templates/> </HTML> </xsl:template> <xsl:variable <xsl:template <P> <xsl:value-of </P> </xsl:template> </xsl:stylesheet>
This will insert the value of myPosition into the document. This stylesheet just replaces each <planet> element with the value in myPosition , which is 3, in a <P> element, this way:
<HTML> <P> 3 </P> <P> 3 </P> <P> 3 </P> </HTML>
And we've already seen some of the XPath functions, such as the position function, which we've used like this: //planet[position()=3] , where we're using the position() function to return the current node's context position. All the XPath 1.0 functions are coming up in Chapter 4.
We've already seen that nodes have string values, and it turns out that node-sets also have string values in XPathbut a node-set's string value might surprise you. If you followed the discussion earlier about the string value of a root node, which is the concatenation of text nodes in the document, you might expect the string-value of a node-set to be made up of the concatenated string-values of all the nodes in the set.
But that's not soin XPath, the string-value of a node-set is simply the string-value of the first node in the node set only. For example, if you apply the XPath expression //planet to our planets example, ch02_01.xml , you'll get a node-set holding the three <planet> elements in that document, in document order. However, the string value of this node-set is the string value of the first element only, the Mercury element:
<planet> <name>Mercury</name> <mass units="(Earth = 1)">.0553</mass> <day units="days">58.65</day> <radius units="miles">1516</radius> <density units="(Earth = 1)">.983</density> <distance units="million miles">43.4</distance> <!--At perihelion--> </planet>
Here's the string value of this element, and therefore of the entire //planets node-set:
Mercury .0553 58.65 1516 .983 43.4
That completes our look at nodes and node-sets. The next step up in organization in XPath is to start thinking in terms of node trees . | https://flylib.com/books/en/1.256.1.31/1/ | CC-MAIN-2020-40 | refinedweb | 1,005 | 60.75 |
Eclipse Community Forums - RDF feed Eclipse Community Forums Help: JSP, EL and code assist <![CDATA[Hello, I have a question about WTP support to EL code assist: I don't kow if it exists and, if so, how is it supposed to work. Reading on the bug reports it would seem that some sort of support does exist... I mean, suppose I have a Dynamic Web Project with the following Java class: package test; public class MyClass { public String getStringProperty() { return "Hello World"; } } Now, suppose I have a JSP like this: <%@page <html> <head> <meta http- <title>Insert title here</title> </head> <body> <c:out</c:out> </body> </html> (of course, I have the JARs for JSTL in the classpath and the needed TLDs mapped in web.xml) In other words, I want to print the value of "stringProperty" property of my MyClass instance, which is previously saved in the pageContext in the provided scriptlet. Then, if I try code completion on {$myc.stringProperty}, either for the context key name or for the property name, it does not work. So, is it supposed to work? Are there other cases in which code assist support for EL is supported? Thanks in advance, Mauro.]]> Mauro Molinari 2010-11-12T10:18:47-00:00 | https://www.eclipse.org/forums/feed.php?mode=m&th=200146&basic=1 | CC-MAIN-2020-50 | refinedweb | 211 | 70.33 |
andae15,715 Points
While True clause causes an infinite loop.
Hi. In the video, Megan makes this loop. I've checked it several times to make sure it isn't a typo on my part. When she runs this loop, it shows the menu and question once and waits for input. When I run this code, the 1-5 menu options run infinitely until it throws and error, and never reaches the "choice" variable.
It actually makes more sense to me that an infinite loops would break this, and I can think of fixes, but I'm wondering why this runs the way she wants it to, but mine breaks. She starts this around 36 secs into the video, and runs the program at 1:50.
def menu(): while True: print(''' \nPROGRAMMING BOOKS \r1) Add a Book \r2) View All Books \r3) Search for a Book \r4) Book Analysis \r5) Exit ''') choice = input('What would you like to do? ') if choice in [1, 2, 3, 4, 5]: return choice else: input('Please choose an option by its number (1-5), listed above.')
2 Answers
Mel RumseyTreehouse Staff
Hey amandae You are close! Currently your
choice = input('What would you like to do? ') if choice in [1, 2, 3, 4, 5]: return choice else: input('Please choose an option by its number (1-5), listed above.')
is outside of your
while loop. You'll need to make sure this code is inside of the loop! Otherwise, your code looks good! :)
amandae15,715 Points
<facepalm>
Thank you. It seems so easy once you say it!
For anyone else who can't see it...
When Megan writes her code, the
choice = input... is a tab closer to the edge than the string above it. In my code, the two will be vertically aligned. This is simply a style difference between VSCode and PyCharm. Both are fine for the string, but it means PyCharm code will line up differently. I was moving this line of code back a tab, to match hers, while not noticing I was a tab back already. | https://teamtreehouse.com/community/while-true-clause-causes-an-infinite-loop | CC-MAIN-2021-43 | refinedweb | 347 | 91.31 |
Introducing Julia/Functions
Contents
- 1 Functions
- 1.1 Single expression functions
- 1.2 Functions with multiple expressions
- 1.3 Optional arguments and variable number of arguments
- 1.4 Keyword and positional arguments
- 1.5 Functions with variable number of arguments
- 1.6 Local variables and changing the values of arguments
- 1.7 Anonymous functions
- 1.8 Map
- 1.9 Reduce and folding
- 1.10 Functions that return functions
- 1.11 Function chaining and composition
- 2 Methods
- 3 Type parameters in method definitions) expression expression expression ....
function doublesix() return (6, 6) end doublesix (generic function with 1 method)
julia> doublesix() (6, 6)
Here you could write
6, 6 without parentheses.
Optional arguments and variable number of arguments[edit]
You can define functions with optional arguments, so that the function can use sensible defaults if specific values aren't supplied. You provide a default symbol and value in the argument list:.
function test(a,b,c) subtotal = a + b + c end function that changes its argument to 5:
function set_to_5(x) x = 5 end
julia> x = 3 3 julia> set_to_5(x) 5 julia> x 3
Although the
x inside the function is changed, the
x outside the function isn't. Variable names in functions are local to the function.
But a function can modify the contents of a container, such as an array. This function uses the
[:] syntax to access the contents of the container
x, rather than change the value of the variable
x:
function fill_with_5(x) x[:] .= 5 end one) #3 (generic function with 1 method) modify the contents of the original array.
Often, you don't have to use
map() to apply a function like
sin() to every member of an array, because many functions automatically operate "element-wise". The timings of the two different versions are similar (
sin.() has the edge perhaps, depending on the number of elements):
julia> @time map(sin, 1:10000); 0.149156 seconds (568.96 k allocations: 29.084 MiB, 2.01% gc time) julia> @time sin.(1:10000); 0.074661 seconds (258.76 k allocations: 13.086 MiB, 5.86% gc time)
and
ans is nothing (
ans == nothing is
true)., but you can buy them online.) then beaten by "sentence", but finally "containing" takes the lead, and there are no other challengers after that. If you want to see the magic happen, redefine
l like this:
julia> l(a, b) = (println("comparing \"$a\" and \"$b\""); length(a) > length(b) ? a : b) l (generic function with 1 method) julia> reduce(l, split("This is a sentence containing some very long strings")) comparing "This" and "is" comparing "This" and "a" comparing "This" and "sentence" comparing "sentence" and "containing" comparing "containing" and "some" comparing "containing" and "very" comparing "containing" and "long" comparing "containing" and "strings" "containing".
store = Int[]; reduce((x,y) -> (push!(store, x * y); y), 1), 1:4, init=256)
Now we can construct lots of exponent-making functions. First, let's build a
squarer() function:
julia> squarer = create_exponent_function(2) #8 (generic function with 1 method)
and a
cuber() function:
julia> cuber = create_exponent_function(3) #9 (generic function with 1 method)
While we're at it, let's do a "raise to the power of 4" function (called
quader, although I'm starting to struggle with the Latin and Greek naming):
julia> quader = create_exponent_function(4) #10 (generic function with 1 method)
julia> a = make_counter() #15 (generic function with 1 method) julia> a() 1 julia> a() 2 julia> a() 3 julia> for i in 1:10 a() end julia> a() 14
Function chaining and composition[edit]
Functions in Julia can be used in combination with each other.
Function composition is when you apply two or more functions to arguments. You use the function composition operator (
∘) to compose the functions. (You can type the composition operator at the REPL using
\circ). For example, the
sqrt() and
+ functions can be composed like this:
julia> (sqrt ∘ +)(3, 5) 2.8284271247461903
which adds the numbers first, then finds the square root.
This example composes three functions.
julia> map(first ∘ reverse ∘ uppercase, split("you can compose functions like this")) 6-element Array{Char,1}: 'U' 'N' 'E' 'S' 'E' 'S'
Function chaining (sometimes called "piping" or "using a pipe to send data to a subsequent function") is when you apply a function to the previous function's output:
julia> 1:10 |> sum |> sqrt 7.416198487095663
where the total produced by
sum() is passed to the
sqrt() function. The equivalent composition is:
julia> (sqrt ∘ sum)(1:10) 7.416198487095663
Piping can send data to a function that accepts a single argument. If the function requires more than one argument, you may be able to use an anonymous function:
julia> collect(1:9) |> n -> filter(isodd, n) 5-element Array{Int64,1}: 1 3 5 7 9
Methods[edit]
A function can have one or more different methods of doing a similar job. Each method usually concentrates on doing the job for a particular type.
Here is a function to check a longitude when you type in a location:
function check_longitude_1(loc) if -180 < loc < 180 println("longitude $loc is a valid longitude") else println("longitude $loc should be between -180 and 180 degrees") end end check_longitude_1 (generic function with 1 method)
The message ("generic function with 1 method") you see if you define this in the REPL, say,_1():
function check_longitude(loc::Real) if -180 < loc < 180 println("longitude $loc is a valid longitude") else println("longitude $loc should be between -180 and 180 degrees") end end
Now add(+) # 176 methods for generic function "+": [1] +(x::Bool, z::Complex{Bool}) in Base at complex.jl:276 [2] +(x::Bool, y::Bool) in Base at bool.jl:104 ... [174] +(J::LinearAlgebra.UniformScaling, B::BitArray{2}) in LinearAlgebra at /Users/osx/buildbot/slave/package_osx64/build/usr/share/julia/stdlib/v0.7/LinearAlgebra/src/uniformscaling.jl:90 [175] +(J::LinearAlgebra.UniformScaling, A::AbstractArray{T,2} where T) in LinearAlgebra at /Users/osx/buildbot/slave/package_osx64/build/usr/share/julia/stdlib/v0.7/LinearAlgebra/src/uniformscaling.jl:91 [176] +(a, b, c, xs...) in Base at operators.jl:466 work with type information in method definitions. Here's a simple example:
julia>function test(a::T) where T <: Real println("$a is a $T") end test (generic function with 1, the definition of
T was where T is a subtype of Real, so the type of T must be a subtype of the Real type (it can be any real number, but not a complex number). 'T' can be used like any other variable — in this method it's just printed out using string interpolation. (It doesn't have to be
T, but it nearly always is!)
This mechanism is useful when you want to constrain the arguments of a particular method definition to be of a particular type. For example, the type of argument
a must belong to the Real number supertype, so this
test() method doesn't apply when
a isn't a number, because then the type of the argument isn't a subtype of Real:
julia> test("str") ERROR: MethodError: no method matching test(::ASCIIString) julia> test(1:3) ERROR: MethodError: no method matching test(::UnitRange{Int64})
Here's an example where you might want to write a method definition that applies to all one-dimensional integer arrays. It finds all the odd numbers in an array:
function findodds(a::Array{T,1}) where T <: Integer filter}) Closest candidates are: findodds(::Array{T<:Integer,1}) where T<:Integer at REPL[13]:2
Note that, in this simple example, because you're not using the type information inside the method definition, you might be better off sticking to the simpler way of defining methods, by adding type information to the arguments:
function findodds(a::Array{Int64,1}) findall(isodd, a) end
But if you wanted to do things inside the method that depended on the types of the arguments, then the type parameters approach will be useful. | https://en.wikibooks.org/wiki/Introducing_Julia/Functions | CC-MAIN-2019-30 | refinedweb | 1,323 | 53.21 |
This is the mail archive of the binutils@sources.redhat.com mailing list for the binutils project.
Hi -? - FChE Index: opcodes/ChangeLog =================================================================== @@ -1,3 +1,8 @@ +2002-01-25 Frank Ch. Eigler <fche@redhat.com> + + * cgen-dis.in (print_insn_@arch@): Support HAVE_CGEN_ISA_NOT_MACH + disassemble_info flag. + Index: opcodes/cgen-dis.in =================================================================== @@ -380,13 +380,19 @@ #ifdef CGEN_COMPUTE_MACH mach = CGEN_COMPUTE_MACH (info); #else - mach = info->mach; + if (info->flags & HAVE_CGEN_ISA_NOT_MACH) + mach = 0; + else + mach = info->mach; #endif #ifdef CGEN_COMPUTE_ISA isa = CGEN_COMPUTE_ISA (info); #else - isa = 0; + if (info->flags & HAVE_CGEN_ISA_NOT_MACH) + isa = info->mach; + else + isa = 0; #endif /* If we've switched cpu's, close the current table and open a new one. */ Index: include/ChangeLog =================================================================== @@ -1,3 +1,8 @@ +2002-01-25 Frank Ch. Eigler <fche@redhat.com> + + * dis-asm.h (HAVE_CGEN_ISA_NOT_MACH): New possible bitmask for + disassemble_info flags. + Index: include/dis-asm.h =================================================================== @@ -93,6 +93,7 @@ The bottom 16 bits are for the internal use of the disassembler. */ unsigned long flags; #define INSN_HAS_RELOC 0x80000000 +#define HAVE_CGEN_ISA_NOT_MACH 0x40000000 /* .mach is really a cgen isa bitmask. */ PTR private_data; /* Function used to get bytes to disassemble. MEMADDR is the | http://sourceware.org/ml/binutils/2002-01/msg00699.html | crawl-002 | refinedweb | 182 | 50.94 |
01 August 2012 11:22 [Source: ICIS news]
TOKYO (ICIS)--Japanese producer Ube Industries said on Wednesday its first-quarter net profit declined 53% year on year decline to yen (Y) 2.34bn ($30m), partly on significant declines in caprolactam sales volumes in ?xml:namespace>
Operating profit in the three months to 30 June 2012 fell 30% year on year to Y6.08bn, while net sales increased 1.2% to Y151.3bn, the company said in a statement.
First-quarter earnings tend to be lower than the other quarters because the maintenance schedule of Ube Industries' chemical plants are concentrated in the April-to-June period, it said.
In the chemicals and plastics segment, three-month net sales rose 1.8% year on year to Y54.4bn, while operating profit declined 74% to Y1.5bn as caprolactam prices fell, Ube Industries said.
The company is a major capro producer in
($1 = Y78 | http://www.icis.com/Articles/2012/08/01/9582720/japans-ube-industries-q1-profit-falls-as-china-capro-sales-slump.html | CC-MAIN-2014-35 | refinedweb | 152 | 68.36 |
Game development HOWTO
From OLPC
This document describes how to use the Pygame library for Python for game development -- to create a new game activity for OLPC's Sugar platform. Its intention is to allow a Python programmer who wants to learn (or already knows) Pygame to integrate their Pygame application into a Sugar-hosted activity using the OLPCGames Pygame wrapper.
If you are looking to create a slightly more limited Activity, you may want to check out Pippy's Pygame capabilities.
This HOWTO is current as of December 2007. More recent notes are available at Porting pygame games to the XO (from March 2008).
[edit] Requirements
This HOWTO assumes that you know the basics of computer programming, how to navigate a file-system, and how to edit files on your machine. It also assumes that you will largely learn Pygame programming through the large number of available Pygame references and tutorials. We focus here on how to integrate your Pygame games into the Sugar environment.
[edit] Components
- Pygame -- this is a Python wrapper around the Simple Direct-media Layer (SDL) library. It is used for lots of games coded in Python and can run on most machines (including Windows, Mac and Linux). If you are running on an OLPC-XO, Pygame should already be available. If not, use your system's package manager to install the Pygame distribution.
- The slides from Noah's lecture at the start of the game jam are online at (both PDF and PowerPoint form).
- Pygame/Mac setup instructions
- OLPCGames -- the OLPC Sugar specific library which provides the glue code that lets your Pygame game run inside a Sugar activity. It also gives you access to the various "special" features in the Sugar environment, such as the mesh network and the camera. If you are on an OLPC-XO, you can download the current OLPCGames distribution and unpack it.
- Note: The OLPCGames Pygame wrapper requires at least build 432 to work for version 1.0 and at least an Update.2 build (649) for version 1.1 and above. See the reference manual at Pygame wrapper. See also Game development.
[edit] Environment
You will need a working Sugar Developer's environment. If you are working directly on an OLPC-XO, you will need to know how to use a standard text editor, such as vi or nano, which are available within the Terminal Activity in your activity toolbar.
- You'll need to set aside a few hours to learn vi before you start this HOWTO if you don't already know it and want to use it well)
- Nano is often considered easier to learn immediately because the major commands are all spelled out at the bottom of the screen (where you need to remember the vim commands yourself)
If you are working in an emulated environment, or a sugar-jhbuild environment, you can use whatever text editor you prefer to create the files we will be working on. There are many text editors with some Python support, and full IDEs are also available.
[edit] Skeleton Setup
To start, you will likely want to download the OLPCGames source package. This package includes a skeleton script that lets you generate a new OLPCGames-based Pygame Activity with a single command.
[edit] Getting the Skeleton Script
To install the package, you will need to download the .zip or .tar.gz to your machine and extract it with either of:
unzip OLPCGames-1.4.zip
or
tar -zxf OLPCGames-1.4.tar.gz
which will create a directory named OLPCGames-1.4. Change to the skeleton directory:
cd OLPCGames-1.4/skeleton
Make sure that your python file has the required permissions to be used.
chmod a+x buildskel.py
And run the command:
./buildskel.py activityname "My Activity Name"
to create a new generic activity instance.
[edit] Installing and Testing
To test that you have your environment properly configured, we'll restart sugar and attempt to run the newly created (empty) activity. Change to the new activity directory (activityname.activity) and run:
python setup.py dev
when you restart Sugar you should have a new activity in your Activity bar named "My Activity Name". Clicking on this activity should result in dark blue screen with a toolbar at the top of the window. Type Esc to exit.
[edit] Testing Outside Sugar
The run.py script in the skeleton project is where you skeleton activity currently points for its "mainloop", particularly the "main" function within it. When you are just starting you'll likely want to work within run.py to create new code and experiment. run.py is actually set up to be used as a python script via:
python run.py
which will run on a non-Sugar environment (i.e. a normal Linux, Windows or Mac desktop with Pygame installed). You may, however, have to configure your system to have the current working directory in the Python path (this is the default on Sugar systems, including emulators and sugar-jhbuild shells).
[edit] Customizing the Skeleton
Your Sugar-specific activity values are stored in two main locations; the activity.py file and the activity directory. The pydoc for the PyGameActivity class describes the various attributes/settings available for your Activity object. These include changing the file-name and method-name for your mainloop function, and changing the title of your activity.
The activity directory is used by Sugar to find things such as your svg icon, translated names and the like. See Activity Bundles for details.
[edit] Getting Started with Pygame
At this point, your OLPC Sugar activity is running as a host for a simple Pygame event loop. You should now, largely, be able to use standard Pygame code to produce graphics, play sounds, and process input.
[edit] Pygame Examples and Tutorials
Example Activities:
- Journal/File Test -- example of using the Journal to save/restore state for an activity
-
Tutorials:
- Pygame Tutorials -- a wiki-based collection of tutorials for learning Pygame programming
- 5-part Tutorial -- a fairly extensive tutorial on Pygame usage
- Pygame Documentation -- the official collection of Pygame documentation, you will need this to get any Pygame programming done
- Game templates -- serves as a starting point for creating Pygame games
- OLPCGames-based Activities can often be read to find sample code.
[edit] Reference Links
- Pygame wrapper pydoc -- the OLPCGames wrapper's reference manual and pydoc, you'll want to familiarize yourself with this to understand what's different on the OLPC platform from regular Pygame
- If you are new to game and GUI programming, you may wish to use a Pygame GUI engine to simplify creating buttons, text entry boxes and the like.
[edit] Support
If you have questions, suggestions or problems, please feel free to post to the OLPC Game Development mailing list. This is a relatively low-traffic list with lots of Pygame users on it. Alternately, IRC channels are available on freenode as #pygame, #sugar and #olpc-content if you want more conversational support. Lastly, User:Mcfletch is the current maintainer of the wrapper. Contact him if you get stuck, but be aware he tends to be spread a bit thin, the mailing list is generally a better avenue.
[edit] Reducing CPU Load
The code in run.py does some trickery to make the event loop reasonably efficient, by limiting the number of frames rendered per second using a "pygame.time.Clock()" instance. It also uses a complex iteration mechanism:
events = pausescreen.get_events() for event in events:
which allows your activity to go completely quiet if there are no pending events for a given time, but still processes all pending events in a timely manner.
You can see the code that implements this in the olpcgames.pausescreen module. None of that machinery is OLPC or Sugar specific, incidentally, it's just good practice to reduce your processing load when running on an OLPC machine.
- Note: the event iteration mechanism reduces the cpu-load from 99% to 0.7 - 4% in our tests versus a simple pygame.event.get() loop.
- Read more about Monitoring System Load
[edit] Eliminating Mouse-move Events
If your activity does not use MOUSEMOTION events it is possible to reduce the overall number of events processed (you can combine this with using a
pygame.event.wait() event loop as well. Keep in mind that with this optimization you cannot do mouse-over highlighting or the like.
An example code structure might look like this:
import sys import pygame from pygame.locals import * def main(): window = pygame.display.set_mode((400, 225)) pygame.event.set_blocked(MOUSEMOTION) pygame.init() while True: for event in [ pygame.event.wait() ] + pygame.event.get( ): print event if event.type == KEYUP: # Quit on 'q' if event.key == 113: sys.exit(0) if __name__=="__main__": main()
[edit] Extending PyGame with C++
See Extending PyGame with C++ for instructions on how to mix Python and C++ code for better performance.
[edit] Troubleshooting
Ensure you are using at least an Update.2 (first manufacturing release) Sugar environment. Also ensure you are using a recent version of OLPCGames.
Check your log files. On modern Sugar, use the Log Viewer activity to view the log for your activity. Open this activity and find your activity in the list of activity instances on the left. The numeric suffixes increase as you run your activity multiple times. | http://wiki.laptop.org/go/Game_development_HOWTO | crawl-002 | refinedweb | 1,554 | 55.34 |
Posting to the Twitter API on an admin changeApr 07, 2009 Django Tweet
For context, I've been working on a project that involves submitted messages going through a manual review process - once they've been approved in the admin, each message should be posted to Twitter. My Twitter API script is over here, for reference.
I had originally thought I'd set up a cron job to run the posting script, and I still might go back to that if the manual review from the admin starts to take too much time. Doug Hellman suggested using signals, but after some consideration I realized that Django signals would probably be overkill - it's a little disappointing because I've never actually gotten signals working in a project. I need a use case to make me figure it out, and I thought this would be it.
Instead I was browsing some old code and realized that, oh yeah, I can put a save method in my admin class ... honestly, I'm embarrassed that I didn't think of it first - it's a no-brainer.
I just wrapped all the script code in a method:
#!/usr/bin/python import MySQLdb as Database import base64, urllib, urllib2 def main():(username + ':' +() if __name__ == "__main__": main()
Then imported it as a module and called it from the relevant admin.py:
import twitterpost class TweetAdmin(admin.ModelAdmin): def save_model(self, request, obj, form, change): obj.save() if obj.approved: twitterpost.main() list_filter = ('approved',)
Note that the script is called after the obj.save() - the script only acts on records that have already been marked 'approved'.
That's it. Three extra lines of code (six if you count what I added to the script). | http://www.mechanicalgirl.com/post/posting-twitter-api-admin-change/ | CC-MAIN-2020-50 | refinedweb | 288 | 70.33 |
Red Hat Bugzilla – Bug 830862
Kernel panic at reboot (possibly NFS related).
Last modified: 2017-03-01 07:22:59 EST
Description of problem:
When I am running kernel-3.4.0-1.fc17.x86_64 my system panics and freezes
at reboot time. Power cycle is required to get the system back. One time
this system freeze left the tail end of a kernel backtrace on my screen
(but unfortunately I didn't take a picture). Subsequent attempts to reproduce
it and get something I could record have failed to capture the backtrace,
but have been 100% successful at panicing the system every time.
The reason I suggest NFS is because there were routines with names containing
nfs_ on the screen in that first panic.
Version-Release number of selected component (if applicable):
kernel-3.4.0-1.fc17.x86_64
How reproducible:
100%
Steps to Reproduce:
1.type reboot
2.
3.
Actual results:
kernel panic scrolls past too fast to see then system is frozen.
Expected results:
System reboots.
Additional info:
Created attachment 590949 [details]
Here is the list of mounted filesystems before the reboot.
Note that many of these systems I'm talking to are very old and have no idea
that such a thing as NFS version 4 exists, so I have to put a proto=udp on
many of the mount lines.
There is no panic with the same NFS mounts when I'm running
kernel-3.3.7-1.fc17.x86_64.
Now I'm about to try typing reboot back in 3.4.0 after manually un-mounting
every NFS filesystem.
Maybe this isn't NFS related after all. I still got the kernel panic at
reboot with no NFS mounts active when I rebooted.
Anyone know if there is some way to make the reboot process stop clearing the
screen? The kernel panic almost always flashes past immediately prior to the
reboot clearing the screen.
Here's the smolt page for the machine in case there is something
hardware related involved:
Based on Comment 3, I'm assuming this isn't nfs-related. Let's try checking the "Reset Assignee to default for component" box and see if that gets us anywhere....
(In reply to comment #5)
> Based on Comment 3, I'm assuming this isn't nfs-related. Let's try checking
> the "Reset Assignee to default for component" box and see if that gets us
> anywhere....
Reassinging it back based strictly on comments is fair play I guess.
Tom, please add 'pause_on_oops=30' and remove 'rhgb quiet' to the boot. That should cause the oops to pause on the screen and hopefully you get get the backtrace. I'm guessing this is at the tail end, so it isn't already stored in /var/log/messages because syslog has been shutdown?
Created attachment 591007 [details]
screen photo
I'm attaching a bunch of marginally readable photos of my screen. I did
remember correctly - there are a bunch of nfs_ routines in the backtrace.
I gotta remember that pause_on_oops parameter.
Created attachment 591008 [details]
screen photo 2
Created attachment 591009 [details]
screen photo 3
Created attachment 591010 [details]
screen photo 4
Created attachment 591011 [details]
screen photo 5
Created attachment 591012 [details]
screen photo 6
Created attachment 591013 [details]
screen photo 7
Created attachment 591014 [details]
screen photo 8
Also, when rebooting many times to try and figure out what camera settings
produced something, I found that if I reboot soon after logging in, it
doesn't panic. I have to do some random amount of work before the panic
will happen (run editor, run firefox, etc). Seems like I need it to be
up at least 5 or ten minutes before it panics on reboot.
Hand transcribed:
lockd_down+0x90/0x140 [lockd]
nlmclnt_done+0x13/0x70/0x90
free_nsproxy+0x1f/0xb0
switch_task_namespaces=0x50/0x60
exit_task_namespaces+0x10/0x20
do_exit+0x456/0xa0
do_group_exit+0x3f/0xa0
sys_exit_group+0x17/0x20
system_call_fastpath+0x16/0x1b
Tom, were all of those with or without NFS mounted?
I had NFS mounts active on all of these. I only tried the experiment of
unmounting everything once and still got the system freeze, but didn't
see anything of the walkback that time. I think I'm probably also doing
an NFS export of a couple of partitions as well, and I didn't try it with
exports disabled.
OK, thanks. Reassigning, this time with data.
Just to throw more fuel on the NFS fire, I tried another experiment this morning.
I deleted all the nfs mounts in /etc/fstab. I checked /etc/exports and found
that I wasn't exporting any filesystems (so it is empty), and I found the
only enabled service with "nfs" in the name was nfs-lock.service, so I
disabled it.
I then rebooted, power cycled through the same old panic, then after I got
the system back up, used it for a while to get it up long enough to panic.
I then typed reboot, and it rebooted just fine with no oops and no getting
frozen.
It certainly does seem to be NLM related...
(gdb) list *(lockd_down+0x90)
0x3eb0 is in lockd_down (fs/lockd/svc.c:386).
381 lockd_down(void)
382 {
383 mutex_lock(&nlmsvc_mutex);
384 if (nlmsvc_users) {
385 if (--nlmsvc_users) {
386 lockd_down_net(current->nsproxy->net_ns);
387 goto out;
388 }
389 } else {
390 printk(KERN_ERR "lockd_down: no users! task=%p\n",
It's not clear what caused the oops from the screenshots however. Usually there's a message prior to the backtrace that says something like NULL pointer dereference, but maybe that scrolled off the screen. Knowing that would help us determine where the bug actually is. Maybe nsproxy is a NULL pointer?
Yeah, this is being called from process exit context. I think that code is setting current->nsproxy to NULL and then proceeding to tear down all of the namespaces. Unfortunately, we require current->nsproxy to not be NULL in order to do that.
I think we'll need to pull in Stanislav Kinsbursky to look at this since he did most of this work upstream...
The "screen photo 2" attachment above has the last [OK] message leftover
from the initial boot right at the top, so everything the oops printed
when I rebooted should be on that photo (I don't see anything that looks
like a reason).
Ok, got a quick response from Stanislav:
---------------------[snip]----------------------
Jeff, this is known issue..
*** Bug 831194 has been marked as a duplicate of this bug. ***
(In reply to comment #23)
> Ok, got a quick response from Stanislav:
>
> ---------------------[snip]----------------------
> Jeff, this is known issue.
> Search for my patch set "NFS: callback shutdown panic fix".
>.
These two commits?
commit 9793f7c88937e7ac07305ab1af1a519225836823
Author: Stanislav Kinsbursky <skinsbursky@parallels.com>
Date: Wed May 2 16:08:38 2012 +0400
SUNRPC: new svc_bind() routine introduced
commit 786185b5f8abefa6a8a16695bb4a59c164d5a071
Author: Stanislav Kinsbursky <skinsbursky@parallels.com>
Date: Fri May 4 12:49:41 2012 +0400
SUNRPC: move per-net operations from svc_destroy()
Neither one of them are CC'd to stable...
*** Bug 829413 has been marked as a duplicate of this bug. ***
When this finishes building, could you please test:
it has the above two commits added to it.
Actually, those two commits are really just preparation; the version Stanislav posted to the mailing list also had another patch ("hard-code init_net for NFS callback transports") folded into one of them.
I have these queued up to submit to stable soon, in the for-3.4 branch of
git://linux-nfs.org/~bfields/linux-topics.
(In reply to comment #28)
> Actually, those two commits are really just preparation; the version
> Stanislav posted to the mailing list also had another patch ("hard-code
> init_net for NFS callback transports") folded into one of them.
Yeah. I pulled them off the list instead out of git directly. It was pretty clear he did something additional when he said they were backported from 3.5. The scratch build above contains the list patches.
> I have these queued up to submit to stable soon, in the for-3.4 branch of
> git://linux-nfs.org/~bfields/linux-topics.
Great!
(In reply to comment #29)
> Yeah. I pulled them off the list instead out of git directly. It was
> pretty clear he did something additional when he said they were backported
> from 3.5. The scratch build above contains the list patches.
OK, good, that's the right thing to test then, thanks.
I installed kernel-3.4.2-4.fc17.x86_64 from the build in comment #27, and
it sure seems to fix my problems. I ran for a while, referenced NFS filesystems
and let the system settle in, then did a reboot, and it rebooted with no
problems. Looks fixed to me.
OK. I committed the patches to Fedora git. They'll be in the next official build.
DOH! I rebooted this morning after running kernel-3.4.2-4.fc17.x86_64 all night
and the kernel panic at reboot was back again. Maybe it is more random now?
There are other NFS problems in 3.4 kernel:
I don't know if they are related to this one.
The walkback I got this morning looked just like the one transcribed above,
but the RIP seems to be at lockd_down+0x25/0x10 (I think, the 0x10 is
very fuzzy :-).
kernel-3.4.2-4.fc17 has been submitted as an update for Fedora 17.
kernel-3.4.2-1.fc16 has been submitted as an update for Fedora 16.
Package kernel-3.4.2-4.fc17:
* should fix your issue,
* was pushed to the Fedora 17 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing kernel-3.4.2-4.fc17'
as soon as you are able to, then reboot.
Please go to the following url:
then log in and leave karma (feedback).
Have the same problem here, installed the new kernel and see if it helps.
kernel-3.4.2-4.fc17 has been pushed to the Fedora 17 stable repository. If problems still persist, please make note of it in this bug report.
Created attachment 592622 [details]
New crash looks like the old crash
I'm getting better at taking readable pictures of the screen, but it looks like
kernel-3.4.2-4.fc17.x86_64 is still prone to these crashes, though may be
a little more random than it used to be (I have rebooted a couple of times
without this happening).
I guess I should re-open this (I'm not sure what ASSIGNED means, but that's
the only option bugzilla gives me other than CLOSED :-).
Created attachment 592735 [details]
hard-code init_net in lockd_up/down
Actually, Stanislav's patches were meant for nfs callback up/down, but I wonder if lockd needs the same treament.
Maybe something like the attached? (Totally untested.)
kernel-3.4.2-1.fc16 has been pushed to the Fedora 16 stable repository. If problems still persist, please make note of it in this bug report..
Note the previous attachment had an obvious typo (inet_net should have been init_net on both lines). With that fixed, I've submitted to stable:<20120620132738.GA30742@fieldses.org>
(In reply to comment #45)
>.
So hopefully the fix will eventually make its way in one way or another.
In the mean time any testing to confirm that patch would be helpful.
Looks like this bug got closed again? I'm not sure what the correct procedure is. Probably a new bug needs to be opened and the patch attached.
We'll just reopen this.
Still seeing the problem with kernel-devel-3.4.3-1.fc17.x86_64.
That's expected.
I'll submit to stable today if I can get a reproducer working. If somebody wants to speed the process then testing of what I intend to submit would be helpful. It's in the for-3.4-take-3 branch at git://linux-nfs.org/~bfields/linux-topics.git. Should backport easily to Fedora.
Same problem here. Contrary to what is stated here and what I believed until just now, the machine will reboot/shutdown, it just takes its sweet time (it was something like 5-10 minutes).
I will test your git branch tomorrow. Unfortunately, it takes a while to do so, because it doesn't happen right away, but only if the machine has had a few hours of activity.
I have rebooted a few times with the patches from for-3.4-take-3 applied (on top of 3.4.3-1, though). So far, without adverse effects. I will let you know in a few days how it goes.
If you're interested in the setup I'm running to trigger this: I have an OpenIndiana (151a-4) machine as a KVM guest which starts at system boot and has a fixed IP. It serves as an iSCSI target (via COMSTAR) as well as the NFS server. After the OpenIndiana guest is booted up, I mount an XFS filesystem that lives on the iSCSI target, and I also mount an NFS share.
Before shutting down the real machine, I unmount both of them, then I shut down the OpenIndiana guest. Only then do I shut down the physical machine.
Created attachment 596671 [details]
spec file patch
Created attachment 596672 [details]
screen photo of traceback
Created attachment 596673 [details]
screen photo of traceback (bottom)
Bad news:
Something similar, but slightly different happened now, during reboot. I could not scroll up any further than this, but the "Fixing recursive fault" at the bottom may be enough information to infer the cause for the traceback. As before, the system stayed in this condition for a few minutes, then proceeded with the reboot.
The exact version of the source tree I was running was based on this version (see attached patch, which is rather obvious though; the four applied patches are exactly the ones from the mentioned for-3.4-take-3 branch):;a=commit;h=50263f813155c14a2cfe6e263eab5325afe0015f
Hand transcription again:
? __schedule+0x3c7/0x7b0
nsm_create+0x8b/0xb0 [lockd]
nsm_mon_unmon+0x64/0x100 [lockd]
nsm_unmonitor+0x68/0xc0 [lockd]
nlm_destroy_host_locked+0x6b/0xc0 [lockd]
nlmclnt_release_host+0x88/0xc0 [lockd]
nlmclnt_done+0x1a/0x78/0x90
free_nsproxy+0x1f/0xb0
switch_task_namespaces+0x50/0x60
exit_task_namespaces+0x10/0x20
do_exit+0x456/0x8a0
do_group_exit+0x3f/0xa0
get_signal_to_deliver+0x1a5/0x5c0
? pollwake+0x66/0x70
do_signal+0x68/0x610
? security_file_alloc+0x16/0x20
? eventfd_ctx_read+0x58/0x190
do_notify_resume+0x65/0x80
? __audit_syscall_exit+0x3ec/0x450
int_signal+0x12/0x17
RIP [<ffffffffa0498391>] rpc_create+0x401/0x540 [sunrpc]
Fixing recursive fault but reboot is needed!
Stefan, could you, please, install and configure kdump on you machine and extract full log from kernel core?
Here is the simpliest way to do so:
makedumpfile —dump-dmesg <core> <out_text_file>
*** Bug 839515 has been marked as a duplicate of this bug. ***
Stanislav, does kdump even generate a core file for this? It's not really a crash/panic after all, just an oops (I think). At the moment, I'm having trouble activating kdump. When I trigger a crash using sysrq-trigger, it panics as usual and doesn't start the crash-kernel at all. I tested the setup in a virtual machine, where it works just fine.
First of all, you have to make sure, that CONFIG_PANIC_ON_OOPS_VALUE kernel option is not equal to zero.
After enabling kdump you have to reboot.
FYI, the 3.4.5-2.fc17 build that has been submitted for updates-testing contains the 4 patches that are in the for-3.4-take-3 branch mentioned in comment #49. One should be able to use that for debugging any further issues.
(In reply to comment #61)
> First of all, you have to make sure, that CONFIG_PANIC_ON_OOPS_VALUE kernel
> option is not equal to zero.
Thanks for this info!
Arghh, the crash kernel has at least tried to boot all along, but for some reason, the secondary kernel does not display anything when nouveau modesetting is in use :(. Now I need to get the actual dumping working, and then I'll just have to somehow manage to trigger the oops again...
Ok, I've finally managed to configure kdump. dracut has made life difficult for me. The panic on oops is a very recent addition which doesn't even exist in my kernel version, but the effect can apparently more easily be had via the "oops=panic" command line option.
Now I'll just need to wait for it to happen again...
Just to update this with my experiences with recently released kernels:
kernel-3.4.4-5.fc17.x86_64 - still got the lockd oops when I rebooted this
after an update the other day.
kernel-3.4.5-2.fc17.x86_64 - this was the update I got above, and when I
rebooted this morning, the system did not get the oops anymore (Yah!)
kernel-3.4.6-2.fc17.x86_64 - this is what I'm now running after the above
reboot, hopefully it won't get the oops anymore either, but haven't had any
reason to reboot yet.
And a status update from me as well: still waiting for the oops to occur again. It's shy. It knows I'm watching :(.
It happened again just now, but unfortunately, the kdump kernel was not booted for some reason, although I can verify that it works when I write "c" to /proc/sysrq-trigger.
I have another one on foto now. The thing is, with oops=panic, no kdump gets triggered. It just blinks the keyboard LEDs. As I said already, a "c" sysrq-trigger does in fact start up kdump and create a memory dump.
I'm still running the exact same version described in comment 56.
Anyway, the reconstructed stack trace looks like this now:
CR2: 0000000000000008
Process mysqld
Call Trace:
? __schedule+0x3c7
nsm_create+0x8b
nsm_mon_unmon+0x64
nlm_destroy_host_locked+0x6b
nlmclnt_release_host+0x88
nlmclnt_done+0x1a
nfs_destroy_server+0x24
nfs_free_server+0xce
nfs_kill_super+0x34
deactivate_locked_super+0x57
deactivate_super+0x4e
mntput_no_expire+0xcc
mntput+0x26
release_mounts+0x77
put_mnt_ns+0x78
free_nsproxy+0x1f
switch_task_namespaces+0x50
exit_task_namespaces+0x10
do_exit+0x456
do_group_exit+0x3f
sys_exit_group+0x17
system_call_fastpath+0x16
RIP rpc_create+0x401 [sunrpc]
Kernel panic - not syncing
Subjectively, it seems to happen only when I've used large amounts of RAM during the session, probably with a small amount paged out, so that during shutdown the waking up of various processes causes noticeable disk accesses.
Do you know if it is on purpose that oops=panic does not trigger a kdump?
This all looks very strange.
According to your trace, mysqld is in container with it's own mount and pid namespace (at least). And it's the init of this container. And it's have NFS mounts inside.
If so, then "killall -9 mysqld" most probably will trigger this oops.
Could you try it, please?
(In reply to comment #69)
> This all looks very strange.
> According to your trace, mysqld is in container with it's own mount and pid
> namespace (at least). And it's the init of this container. And it's have NFS
> mounts inside.
The crashing RIP is 3391 (rpc_create starts at 2f90, rpc_new_client was inlined):
3375: 0f 87 85 00 00 00 ja 3400 <rpc_create+0x470>
337b: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax
3382: 00 00
3384: 48 8b 80 38 05 00 00 mov 0x538(%rax),%rax
338b: 4c 89 e7 mov %r12,%rdi
338e: 4d 89 e6 mov %r12,%r14
=> 3391: 48 8b 70 08 mov 0x8(%rax),%rsi
3395: 48 83 c6 45 add $0x45,%rsi
3399: e8 22 e1 ff ff callq 14c0 <rpc_clnt_set_nodename>
339e: 4c 89 e7 mov %r12,%rdi
33a1: e8 aa ef ff ff callq 2350 <rpc_register_client>
33a6: e9 b4 fc ff ff jmpq 305f <rpc_create+0xcf>
33ab: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
From what I see, the crash happens here: <>, because something in the chain current->nsproxy->uts_ns->name (inside utsname()) is NULL. Did you also come to this conclusion?
> If so, then "killall -9 mysqld" most probably will trigger this oops.
> Could you try it, please?
Yes, I will try that later.
Thanks. I know where the bug is.
static struct rpc_clnt * rpc_new_client(...)
{
< snip >
rpc_clnt_set_nodename(clnt, utsname()->nodename);
<snip>
}
static inline struct new_utsname *utsname(void)
{
return ¤t->nsproxy->uts_ns->name;
}
current->nsproxy is NULL already.
Patch sent to mainline.
Topic: "SUNRPC: check current nsproxy before set of node name on client creation"
> If so, then "killall -9 mysqld" most probably will trigger this oops.
> Could you try it, please?
Nothing happened, except mysqld died.
I will update to the most recent F17 kernel and apply your new patch tomorrow.
(In reply to comment #73)
>
> Nothing happened, except mysqld died.
>
> I will update to the most recent F17 kernel and apply your new patch
> tomorrow.
This is expected. Otherwise you would experience this oops every time.
I don't know, what you or mysqld is doing, but there have to be NFS mounts on your node (mounted by mysqld or it's child) to trigger oops on "kill -9" command.
I'm not doing anything weird. If I were, I would have described it. Yes, I do have an NFS mount, but MySQL has nothing to do with that one. It is the default Fedora package in the default location on a SATA disk. I mounted the NFS before I started mysqld and unmounted it before trying to shutdown (which led to the oops).
I've upgraded to 3.5.1-1.fc17 with this [1] patch applied, but unfortunately, something much worse happens now (iscsi related, #848425), so I cannot comment on the state of affairs regarding the NFS problem.
[1]
I also saw
[443503.679360] BUG: unable to handle kernel NULL pointer dereference at 0000000000000014
[443503.679386] IP: [<ffffffffa01e20aa>] svc_destroy+0x1a/0x130 [sunrpc]
[443503.679413] PGD 0
[443503.679420] Oops: 0000 [#1] SMP
[443503.679429] CPU 11
[443503.679434] Modules linked in: nfs fscache tpm_bios igb ptp pps_core i7core_edac ioatdma edac_core dca lpc_ich coretemp kvm_intel i2c_i801 snd_hda_codec_realtek shpchp mfd_core kvm snd_hda_intel snd_hda_codec snd_hwdep snd_pcm nfsd snd_page_alloc snd_timer nfs_acl snd auth_rpcgss microcode soundcore lockd sunrpc uinput crc32c_intel ghash_clmulni_intel firewire_ohci firewire_core crc_itu_t nouveau mxm_wmi wmi video i2c_algo_bit drm_kms_helper ttm drm i2c_core [last unloaded: scsi_wait_scan]
[443503.679565]
[443503.679568] Pid: 13391, comm: rpc.nfsd Not tainted 3.5.1-1.fc17.x86_64 #1 Intel Corporation S5520SC/S5520SC
[443503.679584] RIP: 0010:[<ffffffffa01e20aa>] [<ffffffffa01e20aa>] svc_destroy+0x1a/0x130 [sunrpc]
[443503.679607] RSP: 0018:ffff880322801e58 EFLAGS: 00010246
[443503.679625] RAX: 00000000ffffff91 RBX: 0000000000000000 RCX: 0000000000000100
[443503.679633] RDX: 0000000000000100 RSI: 0000000050251f41 RDI: 0000000000000000
[443503.679642] RBP: ffff880322801e60 R08: ffff88065c1cd000 R09: 0000000000000000
[443503.679650] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000002
[443503.679659] R13: 0000000000000004 R14: ffff880322801f58 R15: 00007f2d0d0212ae
[443503.679668] FS: 00007f2d0cff4740(0000) GS:ffff88066fca0000(0000) knlGS:0000000000000000
[443503.679684] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[443503.679698] CR2: 0000000000000014 CR3: 000000065be8d000 CR4: 00000000000007e0
[443503.679739] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[443503.679796] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[443503.679837] Process rpc.nfsd (pid: 13391, threadinfo ffff880322800000, task ffff88035d701710)
[443503.679878] Stack:
[443503.679895] ffff880382fae008 ffff880322801ee0 ffffffffa0285b7f 00007f2dffffff91
[443503.679961] ffff880382fae009 ffff880322801e8f 0034fffffffffff4 0000000000000002
[443503.680023] ffff880322801ea8 ffffffff81128586 ffff880322801ee0 ffffffff811a89d5
[443503.680103] Call Trace:
[443503.680133] [<ffffffffa0285b7f>] write_ports+0x2cf/0x3e0 [nfsd]
[443503.680185] [<ffffffff81128586>] ? get_zeroed_page+0x16/0x20
[443503.680229] [<ffffffff811a89d5>] ? simple_transaction_get+0xc5/0xe0
[443503.680264] [<ffffffffa02858b0>] ? write_gracetime+0x60/0x60 [nfsd]
[443503.680295] [<ffffffffa0285027>] nfsctl_transaction_write+0x57/0x90 [nfsd]
[443503.680327] [<ffffffff811861bc>] vfs_write+0xac/0x180
[443503.680348] [<ffffffff811864ea>] sys_write+0x4a/0x90
[443503.680365] [<ffffffff8160a26d>] system_call_fastpath+0x1a/0x1f
[443503.680385] Code: 31 c0 e8 82 63 41 e1 eb be e8 23 60 e7 e0 0f 1f 00 55 48 89 e5 53 66 66 66 66 90 f6 05 6c 1c 02 00 02 48 89 fb 0f 85 ef 00 00 00 <8b> 43 14 85 c0 0f 84 ce 00 00 00 83 e8 01 85 c0 89 43 14 0f 85
[443503.680724] RIP [<ffffffffa01e20aa>] svc_destroy+0x1a/0x130 [sunrpc]
[443503.680761] RSP <ffff880322801e58>
[443503.680777] CR2: 0000000000000014
[443503.685860] ---[ end trace 7a679bc1c3066273 ]---
[root@gnu-mic-2 gcc]#
*** Bug 848740 has been marked as a duplicate of this bug. ***
(In reply to comment #77)
> I also saw
>
> [443503.679360] BUG: unable to handle kernel NULL pointer dereference at
> 0000000000000014
> [443503.679386] IP: [<ffffffffa01e20aa>] svc_destroy+0x1a/0x130 [sunrpc]
I think the above is unrelated to the other issues in this bug. It looks like svc_destroy got passed a NULL pointer. Could you open a new bug for this one?
(In reply to comment #69)
> This all looks very strange.
> According to your trace, mysqld is in container with it's own mount and pid
> namespace (at least). And it's the init of this container. And it's have NFS
> mounts inside.
I suspect that this is because of systemd's PrivateTmp feature. On the same system, this is one of the other problems that plague me: bug #851970.
Because of this, I'll have to shut down everything in a well defined order before rebooting, so I will not be able to do meaningful testing of this feature in the near future. I'm running with your latest patch applied, at least, for a while now.
Whew, this bug is a mess...
It sounds like the originally reported bug is now fixed in more recent kernels, based on comment 76.
The oops in comment #77 is likely a duplicate of bug 848867.
At this point, I'm going to go ahead and close this bug. If anyone is still suffering from the originally reported issue, then please reopen this bug. If you are suffering from a different one, then please open a new bug.
It seems that everything that went on before I first chimed in (comment 50) has gone into the released Fedora kernel. From that point on, there were more patches.
First, the 4 ones from;a=shortlog;h=refs/heads/for-3.4-take-3
They have all been taken from the mainline and have gone into the released F17 kernel since then.
OTOH, I still experienced a crash with those 4 applied (comment 56). This led to the posting of a patch in <>. The thread petered out with no patch applied. Instead, it seems to have been supplanted by 1b63a75180c6c65c71655c250a4e6b578ba7d1c0 (and probably ba9b584c1dc37851d9c6ca6d0d2ccba55d9aad04, its direct successor), both of which have not made their way into 3.6.11, and by extension, into the F17 kernel.
So technically, it's still not fixed. Practically, it's not easily triggered anymore because of a systemd upgrade described below.
Right now, it would be very time consuming for me to try and reproduce it again, but I think I know how one would do that. In the meantime, an update to systemd has happened that will not cause the code path in question to be taken, so the update from bug #851970, comment 24 needs to be taken out for this to work.
- Boot up normally
- Mount an NFS share
- Start the MySQL service (it has PrivateTmp set, which is key to this issue)
- Use up all the memory. If MySQL has not been configured to bypass the file system cache, this can be done by loading a large dump into it.
- Unmount the NFS share
- Reboot
The "use up all memory" seems to be required to trigger disk accesses during shutdown, thus causing waits and scheduler intervention. | https://bugzilla.redhat.com/show_bug.cgi?id=830862 | CC-MAIN-2018-05 | refinedweb | 4,542 | 74.29 |
The File class
Posted on March 1st, 2001 collection classes because the number of elements is fixed, and if you want a different directory listing you just create a different File object. In fact, “FilePath” would have been a better name. This section shows a complete example of the use of this class, including the associated FilenameFilter interface.
A directory lister
Suppose you’d like to see a directory listing. The File object can be listed in two ways. If you call list( ) with no arguments, you’ll get the full list that the File object contains. However, if you want a restricted list, for example, all of the files with an extension of .java, then you use a “directory filter,” which is a class that tells how to select the File objects for display.
//: DirList.java // Displays directory listing package c10; import java.io.*; public class DirList { public static void main(String[] args) { try { File path = new File("."); String[] list; if(args.length == 0) list = path.list(); else list = path.list(new DirFilter(args[0])); for(int i = 0; i < list.length; i++) System.out.println(list[i]); } catch(Exception e) { e.printStackTrace(); } } }. (Interfaces were covered in Chapter 7.) It’s useful to see how simple the FilenameFilter interface is:
public interface FilenameFilter { boolean accept(File dir, String name); }
It says that.
DirFilter shows that just because an interface contains only a set of methods, you’re not restricted to writing only those methods. (You must at least provide definitions for all the methods in an interface, however.) In this case, the DirFilter constructor is also created.( ).
To make sure that what you’re working with is only the regular expression “wildcard” matching such as “fo?.b?r*” which is much more difficult to implement.
The list( ) method returns an array. You can query this array for its length and then move through it selecting the array elements. This ability to easily pass an array in and out of a method is a tremendous improvement over the behavior of C and C++.Anonymous inner classes
This example is ideal for rewriting using an anonymous inner class (described in Chapter 7). As a first cut, a method filter( ) is created that returns a handle to a FilenameFilter:
//: DirList2.java // Uses Java 1.1 anonymous inner classes import java.io.*;) { try { File path = new File("."); String[] list; if(args.length == 0) list = path.list(); else list = path.list(filter(args[0])); for(int i = 0; i < list.length; i++) System.out.println(list[i]); } catch(Exception e) { e.printStackTrace(); } } } ///:~
Note that the argument to filter( ) must be final. This is required by the anonymous inner class so that it can use an object from outside its scope.
This design is an improvement because the FilenameFilter class is now tightly bound to DirList2. However, you can take this approach one step further and define the anonymous inner class as an argument to list( ), in which case it’s even smaller:
//: DirList3.java // Building the anonymous inner class "in-place" import java.io.*; public class DirList3 { public static void main(final String[] args) { try { File path = new File("."); String[] list; if(args.length == 0) list = path.list(); else list = path.list( new FilenameFilter() { public boolean accept(File dir, String n) { String f = new File(n).getName(); return f.indexOf(args[0]) != -1; } }); for(int i = 0; i < list.length; i++) System.out.println(list[i]); } catch(Exception e) { e.printStackTrace(); } } } ///:~
The argument to main( ) is now final, since the anonymous inner class uses args[0] directly..A sorted directory listing
Ah, you say that you want the file names sorted? Since there’s no support for sorting in Java 1.0 or Java 1.1 (although sorting is included in Java 1.2), it will have to be added into the program directly using the SortVector created in Chapter 8:
//: SortedDirList.java // Displays sorted directory listing import java.io.*; import c08.*; public class SortedDirList { private File path; private String[] list; public SortedDirList(final String afn) { path = new File("."); if(afn == null) list = path.list(); else list = path.list( new FilenameFilter() { public boolean accept(File dir, String n) { String f = new File(n).getName(); return f.indexOf(afn) != -1; } }); sort(); } void print() { for(int i = 0; i < list.length; i++) System.out.println(list[i]); } private void sort() { StrSortVector sv = new StrSortVector(); for(int i = 0; i < list.length; i++) sv.addElement(list[i]); // The first time an element is pulled from // the StrSortVector the list is sorted: for(int i = 0; i < list.length; i++) list[i] = sv.elementAt(i); } // Test it: public static void main(String[] args) { SortedDirList sd; if(args.length == 0) sd = new SortedDirList(null); else sd = new SortedDirList(args[0]); sd.print(); } } ///:~
A few other improvements have been made. Instead of creating path and list as local variables to main( ), they are members of the class so their values can be accessible for the lifetime of the object. In fact, main( ) is now just a way to test the class. You can see that the constructor of the class automatically sorts the list once that list has been created.
The sort is case-insensitive so you don’t end up with a list of all the words starting with capital letters, followed by the rest of the words starting with all the lowercase letters. However, you’ll notice that within a group of file names that begin with the same letter the capitalized words are listed first, which is still not quite the desired behavior for the sort. This problem will be fixed in Java 1.2.
Checking for and creating directories
The File class is more than just a representation for an existing directory path, file, or group of files. the remaining methods available with the File class:
//: the various file investigation methods put to use to display information about the file or directory path.
The first method that’s exercised by main( ) is renameTo( ), which allows you to rename (or move) a file to an entirely new path represented by the argument, which is another File object. This also works with directories of any length.
If you experiment with the above program, you’ll find that you can make a directory path of any complexity because mkdirs( ) will do all the work for you. In Java 1.0, the -d flag reports that the directory is deleted but it’s still there; in Java 1.1 the directory is actually deleted.
There are no comments yet. Be the first to comment! | http://www.codeguru.com/java/tij/tij0111.shtml | CC-MAIN-2017-13 | refinedweb | 1,094 | 68.67 |
#include <vtkMergeCells.h>
Designed to work with distributed vtkDataSets, this class will take vtkDataSets and merge them back into a single vtkUnstructuredGrid.
The vtkPoints object of the unstructured grid will have data type VTK_FLOAT, regardless of the data type of the points of the input vtkDataSets. If this is a problem, someone must let me know.
It is assumed the different DataSets have the same field arrays. If the name of a global point ID array is provided, this class will refrain from including duplicate points in the merged Ugrid. This class differs from vtkAppendFilter in these ways: (1) it uses less memory than that class (which uses memory equal to twice the size of the final Ugrid) but requires that you know the size of the final Ugrid in advance (2) this class assumes the individual DataSets have the same field arrays, while vtkAppendFilter intersects the field arrays (3) this class knows duplicate points may be appearing in the DataSets and can filter those out, (4) this class is not a filter.
Definition at line 56 of file vtkMergeCells.h.
Reimplemented from vtkObject.
Definition at line 59 of file vtkMergeCells.h.
Set the vtkUnstructuredGrid object that will become the union of the DataSets specified in MergeDataSet calls. vtkMergeCells assumes this grid is empty at first.
Specify the total number of cells in the final vtkUnstructuredGrid. Make this call before any call to MergeDataSet().
Specify the total number of points in the final vtkUnstructuredGrid Make this call before any call to MergeDataSet(). This is an upper bound, since some points may be duplicates.
vtkMergeCells attempts eliminate duplicate points when merging data sets. This is done most efficiently if a global point ID field array is available. Set the name of the point array if you have one.
vtkMergeCells attempts eliminate duplicate points when merging data sets. If no global point ID field array name is provided, it will use a point locator to find duplicate points. You can set a tolerance for that locator here. The default tolerance is 10e-4.
vtkMergeCells will detect and filter out duplicate cells if you provide it the name of a global cell ID array.
vtkMergeCells attempts eliminate duplicate points when merging data sets. If for some reason you don't want it to do this, than MergeDuplicatePointsOff().
We need to know the number of different data sets that will be merged into one so we can pre-allocate some arrays. This can be an upper bound, not necessarily exact.
Provide a DataSet to be merged in to the final UnstructuredGrid. This call returns after the merge has completed. Be sure to call SetTotalNumberOfCells, SetTotalNumberOfPoints, and SetTotalNumberOfDataSets before making this call. Return 0 if OK, -1 if error. | http://www.vtk.org/doc/release/5.4/html/a01007.html | crawl-003 | refinedweb | 454 | 65.22 |
2.7mm 3mm 4mm 5mm 6mm China Supplier Bulk Mirrors
US $2-3.64 / Square Meter
1 Square Meter (Min. Order)
Qinhuangdao Sunglory Glass Co., Ltd.
98.7%
Top quality bulk mirror manufacturer with CE certificates
US $3-5 / Square Meter
100 Square Meters (Min. Order)
Yantai Thriking Glass Co., Ltd.
96.9%
bulk pocket mirrors /wedding gift pocket mirror for promotion
US $0.4-0.65 / Piece
1 Piece (Min. Order)
Yiwu Lifeng Arts & Crafts Co., Ltd.
97.0%
China alloy bulk sublimation compact mirror
US $0.1-2 / Unit
1 Carton (Min. Order)
Shanghai Mejorsub Industry And Trade Co., Ltd.
67.4%
5 inch perfect cosmetics bulk pocket mirrors
US $3-5 / Box
500 Boxes (Min. Order)
Jiangmen Greenfrom Household Co., Ltd.
96.3%
Free sample 2018 hot product buy bulk compact mirror
US $0.2-0.85 / Piece
3000 Pieces (Min. Order)
Foshan Yuli Cosmetic Development Co., Ltd.
70.0%
China high quality 2mm 3mm 4mm 5mm 6mm aluminum mirror, mirror glass buy bulk mirrors
US $2.45-9.25 / Square Meter
2000 Square Meters (Min. Order)
Sinoy Mirror, Inc.
97.9%
compact mirrors wholesale purple folding pocket mirrors,cheap buy bulk mirrors
US $0.73-0.92 / Piece
3000 Pieces (Min. Order)
Ningbo Pinbo Plastic Manufactory Co., Ltd.
81.9%
Lovely Heart Shape Mini Buy Bulk Mirrors Pocket Cosmetic Mirror
US $0.1-1 / Piece
1000 Pieces (Min. Order)
Cangnan Verizon Crafts Co., Ltd.
84.2%
Sale by bulk small cheap makeup mirror
US $1.5-1.93 / Units
500 Units (Min. Order)
Shenzhen Smile Technology Co., Ltd.
86.2%
China Supplier Bulk Mirrors
US $1.85-8.8 / Square Meter
100 Square Meters (Min. Order)
Qinhuangdao Aohong Glass Limited Company
97.7%
China 3mm Silver Layer Double Coated Clear Glass Mirror, Bulk Mirrors
US $1-10 / Square Meter
500 Square Meters (Min. Order)
Qingdao Vatti Glass Co., Ltd.
96.2%
Wholesaler Hot Sale Pu Leather Bulk Square Pocket Mirror
US $0.5-0.8 / Piece
500 Pieces (Min. Order)
Shenzhen Whose Gift Co., Ltd.
88.9%
Hot Sale Competitive Price buy bulk mirrors
US $0.59-0.99 / Piece
300 Pieces (Min. Order)
Ningbo Yinzhou Shunda Plastic & Glue Co., Ltd.
90.5%
hotel used bulk sale best quality aluminum mirror wall mirror
US $2-4 / Square Meter
500 Square Meters (Min. Order)
Qingdao Yujing Fanyu Trading Co., Ltd.
81.8%
Shatterproof Small Vintage Star Sterling Magnifying Silver Target Handbag Hand Held Mirror Bulk, Men Shabby Chic Hand Mirror
US $5.42-8.93 / Piece
40 Pieces (Min. Order)
Jiangmen Greenfrom Household Co., Ltd.
92.0%
Hot Selling Fabric Covered Bulk Makeup Pocket Mirror
US $0.2-0.78 / Piece
500 Pieces (Min. Order)
Shenzhen Hengyixin Gift Co., Ltd.
88.7%
Good Reputation cheap bulk makeup mirror
US $0.39-0.65 / Piece
1000 Pieces (Min. Order)
Ningbo Hi-Tech Zone Neater Trade Co., Ltd.
93.0%
Supplies Bulk Cheap Plastic Mirror Modern Rechargeable Mirror Illuminated Dimmable Led Makeup Mirror
US $12-13 / Set
100 Sets (Min. Order)
Shenzhen Jinsanyang Technology Co., Ltd.
Hotel bulk order unbreakable shaving mirror with light
US $45-100 / Piece
50 Pieces (Min. Order)
Shanghai Divas Glass Co., Ltd.
bulk custom shape leather mirror at small size and fold up
US $0.6-0.81 / Piece
300 Pieces (Min. Order)
Shenzhen Yingbosang Crafts & Gifts Co., Ltd.
89.0%
Wood Frame Standing Bulk Craft Mirrors
500 Pieces (Min. Order)
Qufu Xinyi Picture Frame Co., Ltd.
85.0%
Professional cheap bulk makeup mirror table vanity mirror
US $10.1-28.99 / Piece
1 Piece (Min. Order)
Touchbeauty Beauty & Health (Shenzhen) Co., Ltd.
83.3%
OEM Cheap Bulk Leather Pocket Mirror
US $0.35-0.93 / Piece
100 Pieces (Min. Order)
Zhongshan Xiangda Art & Craft Co., Ltd.
82.6%
China 2mm-6mm Mirror Glass buy bulk mirrors
US $2-5 / Square Meter
1 Twenty-Foot Container (Min. Order)
Qingdao Haisen Glass Co., Ltd.
Professional cheap bulk makeup mirror table vanity mirror
US $65-155 / Piece
20 Pieces (Min. Order)
Dongguan Ouyimei Cosmetics Co., Ltd.
import decorative chinese mirror manufacturer and supplier,buy producer bulk retail store mirrors,china wall mirror factory
US $40-60 / Piece
10 Pieces (Min. Order)
Ningbo Yoho Commodity Co., Ltd.
76.2%
Buy bulk mirror for interior decoration
US $30-100 / Piece
30 Pieces (Min. Order)
Dongguan Ruijing Glass Craftworks & Hardware Co., Ltd.
77.4%
Wholesale Bulk Antique Brown Craft Circle Mirror
US $21.69-26.51 / Piece
50 Pieces (Min. Order)
Fuzhou Topwell Home Decor Co., Ltd.
100%
MDF framed tabletop mirror cheap bulk craft mirror in white 30cm*40cm
US $5.8-8.8 / Piece
200 Pieces (Min. Order)
Suzhou Refine Arts & Crafts Co., Ltd.
66.7%
hot sale beveled edge bulk decorative wall mirrors
US $20-25 / Piece
50 Pieces (Min. Order)
Tengzhou Haolong Glass Co., Ltd.
56.3%
2015 small mirrors bulk pocket mirror,MB274
US $0.6-2.0 / Piece
600 Pieces (Min. Order)
Guangdong Textiles Import & Export Cotton Manufactured Goods Company Limited
71.4%
Art Wooden Hotel Hanging Irregular Shape Buy Bulk Mirror
US $12.49-13.59 / Piece
100 Pieces (Min. Order)
Fuzhou Myee Industry & Trade Co., Ltd.
100%
Portable Wholesale Bulk Pocket Mirror Metal Foldable Glass Compact Mirror
US $0.6-1 / Piece
1200 Pieces (Min. Order)
Jinhua Baoxin Cultural Innovation Co., Ltd.
85.7%
Bulk Buy Well-designed Stylish Mirror
US $23.49-25.97 / Piece
100 Pieces (Min. Order)
Minhou Dacor Household Crafts Co., Ltd.
58.8%
New design colorful with great price buy bulk mirrors
US $95-198 / Piece
10 Pieces (Min. Order)
Dongguan OE Home Co., Ltd.
Metal make-up leopard head new style bulk pocket mirrors
US $0.5-3 / Piece
1000 Pieces (Min. Order)
Dongguan Daai Shijia Craft Gifts Co., Ltd.
54.1%
Bulk sale two side mirror compact
US $0.8-2 / Piece
500 Pieces (Min. Order)
Purple Cloud Gifts Limited (Quanzhou)
69.0%
New design colorful with great price buy bulk mirrors
US $95-198 / Piece
10 Pieces (Min. Order)
Longyan OE Home Arts Co., Ltd.
69.3%
Metal Bulk Square Wall Mirror
US $22.1-28.2 / Piece
100 Pieces (Min. Order)
Fuzhou Baodeyou Trading Co., Ltd.
42.9%
- About product and suppliers:
Alibaba.com offers 12,291 bulk mirrors products. About 3% of these are makeup mirror, 2% are mirrors, and 1% are cosmetic bags & cases. A wide variety of bulk mirrors options are available to you, such as glass, plastic, and wood. You can also choose from metal, aluminum, and stainless steel. As well as from zinc alloy, chrome, and nickel. And whether bulk mirrors is pocket mirror, desktop mirror, or wall mounted mirror. There are 12,291 bulk mirrors suppliers, mainly located in Asia. The top supplying country is China (Mainland), which supply 100% of bulk mirrors respectively. Bulk mirrors products are most popular in Domestic Market, South America, and Southeast Asia. You can ensure product safety by selecting from certified suppliers, including 6,842 with ISO9001, 4,167 with ISO14001, and 2,960 with OHSAS18001 certification. | https://www.alibaba.com/countrysearch/CN/bulk-mirrors.html | CC-MAIN-2019-04 | refinedweb | 1,167 | 71 |
This.
Output recorded during:
Deduction
On pulse
Off time
1
0
There is an obstacle
The sensor is saturated by ambient light
0
There is no obstacle
This reading is not logical (Sensor error)
You may refer here about building a simple IR sensor circuit.
The next thing is an RF receiver. Commands are constantly transmitted from the RF transmitter connected to the serial port of a PC as a form of series of bits and received by the RF receiver located on the car. I have created a simple C# application for command bits generation and for video display and replay functionalities. The SerialPort class is responsible for sending these series of bytes depending on user’s inputs. The RF transmitter accepts these data from the serial port and transmits them through its antenna with a proper modulation. I used ICs from Linx for creating an RF link between my car and PC. The transmitter and receiver data sheets with their application can be found here and here.
SerialPort
To control DC motors on my RC car, I used two H-bridges as shown before. These two H-bridges are driven by outputs of micro-controller through one of parallel port pins. A very good introduction and application about H-bridges can be found here.
The PC program has two main parts: a video display and replay, and a car control module.
The video display and replay module uses DirectShow libraries to access video data from wireless video receiver but since this software part is not my concern, I am not discussing it right now. As a reference, you may check a really nice article by Andrew Krillov on codeProject (here).
DirectShow
The car control module is implemented using a SerialPort object, seven buttons for direction control and a combo box for listing available serial ports. Just remember to include the following namespace:
using System.IO.Ports;
On form load event handler, I fetch all available ports in a combo box:
// create an array for getting available port on my PC
string[] availablePorts;
// fetch them all
availablePorts = SerialPort.GetPortNames();
// copy them to a comboBox
for (int i = 0; i < availablePorts.Length; i++)
{
cboPort.Items.Add(availablePorts[i]);
}
On each button click, a SendToSerialPort(string data) method is called with its respective string parameter as explained below:
SendToSerialPort(string data)
string
private void SendToSerialPort(string data)
{
// create an instance of serial port
SerialPort ser = new SerialPort();
byte[] command = new byte[2];
command[0] = (byte)Convert.ToByte(data, 16);
// set its properties
// i preferred the ff values
ser.BaudRate = 9600;
ser.DataBits = 8;
ser.Parity = Parity.None;
ser.StopBits = StopBits.One;
ser.PortName = cboPort.Text;
// if our command array is not empty then...
if (command != null)
{
// open it if it is closed
if (!ser.IsOpen)
ser.Open();
// write the byte
ser.Write(command, 0, 1); // this sends only a byte to the port
// then close it
ser.Close();
}
}
The motor drivers (H-bridges) are connected to the micro controller's Port1 and are arranged as follows:
Port1
The direction control is realized by controlling both DC motors' directions. For example, to turn right, we drive the left motor forward and stop the right motor. And the individual motor direction is controlled by switching the four transistors ON and OFF which is represented by the byte values in the braces.This lets us decide which series of bits is to be sent from our application to the serial port and then to the RC car. Accordingly, forward (“00100010”), backward (“00010001”), right (“00100000”), left (“00000010”), stop (“00000000”), spin clockwise (“00100001”) and spin counter clockwise (“00010010”).
Before going through the coding stuff, let us see the flow of the instructions of the whole program.
After initializing ports, the IR sensor is read as a reference for the debugging of IR sensor output and is saved as a variable. Then the IR LED is fired ON and ON state reading of IR sensor is saved as a second variable. Then come a series of comparison of these two readings and follows taking appropriate action.
If the OFF state reading is HIGH, there is either a problem of sensor’s saturation by high ambient light or a problem with the sensor itself which is to be determined by looking at the ON state reading (See table above). For each case, the corresponding debugging LEDs (High Ambient Light and Sensor Error LEDs) are set ON and the program jumps to the next step.
If the OFF state reading is LOW, we have no problem with the issues discussed before. So, if the second (ON state) reading is HIGH, there is definitely an obstacle in front and both motors shall be driven backward for a second. If the ON state reading is LOW, then there is nothing in front of the sensor so the program goes to accepting commands from RF receiver and sends them to DC motors.
This whole process is repeated indefinitely, so I put the code in a while(1) loop.
while(1)
After each reading from and writing to ports procedure is finished, I wanted to wait for some time. So, I used a 50ms delay function delay_50ms(void) in the program which is implemented using timers in the microcontroller itself. Here the microcontroller frequency is assumed to be 12MHz with 12 oscillation cycles. The definite amount of time (integral multiples of 50ms) to be waited, is given as a parameter to a wait(int sec) function.
delay_50ms(void)
wait(int sec)
void delay_50ms(void)
{
// Configure Timer 0 as a 16-bit timer
TMOD &= 0xF0; // Clear all T0 bits (T1 left unchanged)
TMOD |= 0x01; // Set required T0 bits (T1 left unchanged
ET0 = 0; // No interrupts
// Values for 50 ms delay
TH0 = 0x3C; // Timer 0 initial value (High Byte)
TL0 = 0xB0; // Timer 0 initial value (Low Byte)
TF0 = 0; // Clear overflow flag
TR0 = 1; // Start timer 0
while (TF0 == 0); // Loop until Timer 0 overflows (TF0 == 1)
TR0 = 0; // Stop Timer 0
}
The serial port (of the microcontroller) initializing function is given as follows (9600 baud rate, no parity and 1 stop bit are assumed):
// serial port initializing function
void serial_init(void)
{
TMOD = 0x20; // T1 in mode 2, 8-bit auto reload
SCON = 0x50; // 8-bit data, none parity bit, 1 stop bit
TH1 = 0xFD; //12MHz freq. 12 osc. cycle and 9600 baud rate
TL1 = 0xFD;
TR1 = 1; // Run the timer
}
Command reading task from a serial port is to be managed by the following method which returns value that is read as a char.
char
// serial port reading function
unsigned char serial_read(void)
{
bit over = 0;
while(!RI || !over)
{
wait(500);
over = 1;
RI = 0;
return SBUF;
}//wait some time till received flag is set and read the buffer
}
The whole other stuff is handled in the main function of the program and here are given the main( void ) function and the wait(int sec) function (responsible for delaying the program execution for some time set in the input parameter sec).
main( void )
wait(int sec)
// some 'sec' milliseconds wait function
void wait (int sec)
{
unsigned int i;
for ( i = 0; i < (sec / 50); i++ )
{
delay_50ms();
}
}
//here goes the main function
void main( void )
{
P0 = 0; // initialize P0
P1 = 0; // initialize P1
P2 = 0; // initialize P2
while(1)
{
unsigned char val = 0x00;
unsigned char var1 = 0x00;
unsigned char var2 = 0x00;
var1 = P2; //read IR sensor
wait(50); // delay
P2 = num[1]; //turn IR LED ON
wait(200); // delay
var2 = P2; //read IR sensor again
wait(50); // delay
P2 = num[0]; //turn IR LED OFF
if(var1 == num[2])
{
if(var2 == num[1])
P0 = num[2]; //Set sensor error flag
if(var2 == num[3])
P0 = num[1]; //Set high ambiet light flag
serial_init();
val = serial_read(); //Read the serial port
P1 = val; //Command motors
}
if(var1 == num[0])
{
if(var2 == num[3])
{
P1 = num[4]; //drive motors backward
wait(1000); //delay for a second
P1 = num[0];
}
if(var2 == num[1])
{
serial_init();
val = serial_read(); //Read the serial port
P1 = val; //Command the motors
}
P0 = num[0]; //Set the flags to zero
}
}
}
I compiled the C code above and simulated on the Proteus Simulation program with the following diagram:
For simulation purpose, I used Virtual Serial Port Driver (It has 14 days evaluation period and free trial download can be found here to create a virtual port pair and connected my PC software to one of these COM pair's end port and COMPIM serial port of the Proteus Simulation to the other end. The two H-bridges at the bottom are made using NPN and PNP BJT transistors.
That is it. If you want details, just. | https://www.codeproject.com/articles/126859/rc-car-control-programming?msg=4367391 | CC-MAIN-2017-13 | refinedweb | 1,440 | 54.36 |
First time here? Check out the FAQ!
We have a requirement which describes that our webservice namespace (we are the service provider) should be like
Is this possible to implement? In Ivy Designer (6.0.4) it looks like there are only points allowed no slashes. Is there a workaround possible
asked
06.12.2016 at 09:12
peter_muc
29●11●11●15
accept rate:
0%
edited
07.12.2016 at 10:56
Yes, there is a workaround.
First, each WS Process is backed by a Java file, which contains the configuration/annotations to provide the WS endpoint methods. This file is located in the folder [project]/src_wsproc/[fully-qualified-name].java. The fully-qualified-name could be defined in the inscription mask of the process. The file is generated by the project builder 'Ivy Web Serivce Process Class Builder', which is defined in the project properties -> builders.
[project]/src_wsproc/[fully-qualified-name].java
The generaded file is annotated with @javax.jws.WebService. This annotation has the property targetNamespace which allows to define the webserivce namespace, as asked in the question. But per default this property is not set and could not be configuration in the inscription mask.
@javax.jws.WebService
targetNamespace
Because the file gets recreaed when the process changes, in could not be changed directly. Therefore the java file has to be copied to the src-folder of your project and the file in the src_wsproc-folder has to be deleted. The version in the src-folder could be configured/changed as requested. BUT it has to be in line with the configuration of the WS Start Elements. Its now under control of the developer - means, when a WS Start Element configuration changes, the change has to be adapted in the java file too!
src
src_wsproc
Axon.ivy >= 6.5
The java file in the src_wsproc-folder gets not recreated anymore, because the project specific file in the src-folder will be recognized ;-)
Axon.ivy < 6.5
The java file in the src_wsproc-folder gets recreatd as soon the process changes. So it has to be deleted by the developer again and again. Or the corrsponding builder is disabled, but then NO WS Process Java files of this project would be generaded any more...
Example of the annotation on the corresponding class:
@javax.jws.WebService(targetNamespace="")
public class MySAPWebService extends ch.ivyteam.ivy.webservice.process.restricted.AbstractWebServiceProcess
{
...
}
answered
07.12.2016 at 16:05
Flavio Sadeghi ♦♦
1.8k●5●7●23
accept rate:
75%
Once you sign in you will be able to subscribe for any updates here
Answers
Answers and Comments
Markdown Basics
learn more about Markdown
webservice ×47
Asked: 06.12.2016 at 09:12
Seen: 1,186 times
Last updated: 07.12.2016 at 16:05
How can I evalute webservice response?
The maven build has problem with axis2
How to use MTOM in Ivy web service call
How to increase java heap space
Web service development in AXON Ivy
WebService Process: XmlElement(required=true) does not work
How can I integrate with SAP?
How can I enable webservice call SOAP request and respons logging on the eninge
Mocking SOAP web services
Import certificate for HTTPS web service calls | https://answers.axonivy.com/questions/2245/fully-qualified-name-of-webservice-including-sign | CC-MAIN-2020-29 | refinedweb | 534 | 57.06 |
In an effort to decouple the model layer from the UI layer, we've taken to implementing Zope 3 views in Zenoss. So far, we've just done the JSON-providing methods that feed the portlets, event console, etc., but ideally we would like to move the entire application to this style.
Let's say you're adding a new screen to Zenoss. This screen shows a list of components under a Device and their event pills (the actual worth of this screen is both nonexistent and irrelevant). Here's how you'd do it, the old way and the new way.
Add a method to the relevant class that assembles and delivers your data. In this case, you'd probably add a method to the
Products.ZenModel.Device.Deviceclass that walks components under self and generates an event pill for each. We'll call it
getComponentList. If your method should logically be broken up into several methods, for organization or otherwise, you'll add those to the class as well, or find a way to use nested functions.
Create a page template that calls the method and renders the data. Your template would be, say,
ZenModel/skins/zenmodel/viewDeviceComponents.pt. Surrounding the content block, you'd have something like:
<tal:block tal:...</tal:block>
Link to your template. Either by adding a tab to the Device class, or by dropping a link in another template, you're going to point to a URL that describes a Device instance and your template:
<a tal: Component List</a>
And you're done! Now, here are the problems with this approach:
You've added a method used only for the UI layer to a class in the model layer, which leads to bloated classes and a terrible time reading
dir().
Another developer will have a difficult time figuring out why the method is there, unless they grep templates for a call.
There's nothing identifying the template as being applicable to a particular class or group of classes.
If your method is applicable to another class, or if you want your template to apply to different kinds of objects, you either need to define the same method on the other classes, or create a mixin and modify your classes to inherit from it. In the first case, you've got to (remember to) update methods in two places if changes are ever desired. In the second case, you add to the already terrible Zope class inheritance tree (plus, where do you draw the line? Should we really have forty-seven mixins for a class if only the UI demands it?).
Calling your template on another object will get you a traceback. Not a 404, a traceback.
Create a
BrowserViewclass to contain logic and load the template. Instead of inflating model classes with
viewmethods, make yourself a
BrowserView, which will adapt the context to add logic you need to render the template. That is, when a view is the result of traversal, the view class will be instantiated, passing the context into the constructor (it will be available on the view instance as self.context; the request object will be
self.request).
You'll put something like this in
ZenModel/browser/DeviceViews.py(
browseris a convention):
from Products.Five.browser import BrowserView from Products.Five.pagetemplatefile import ViewPageTemplateFile class ComponentListView(BrowserView): __call__ = ViewPageTemplateFile('viewDeviceComponents.pt') def getComponentList(self): ... do things with self.context and self.request ...
BrowserViewsare called when they're the result of a traversal, so that's your hook.
ViewPageTemplateFile()is a callable, so the assignment is fine. If, instead of rendering a template, you just wanted to return some text (for example, JSON), you could do:
from Products.Five.browser import BrowserView from Products.Five.pagetemplatefile import ViewPageTemplateFile class ComponentListView(BrowserView): def __call__(self): ... do things with self.context and self.request ... return results
Create a page template that calls the method and renders the data. This is the same as the Zope 2 way, except for one key difference:
viewis now a global, and that's how you can access your custom method (here is still available and still refers to the context, just as before).
<tal:block tal: ... </tal:block>
Another difference is that you don't render the template by traversing to a template against a context; instead, you traverse to a
BrowserView, which knows which template to use. This is great, especially when you want to use the same template for radically different contexts; as long as you have two
BrowserViewsthat know how to provide the methods the template wants, you're good.
Wire everything up with ZCML. This is where most people start scoffing. It's okay. It actually makes sense.
So you have a view, but you don't have a way to call that view; there isn't a URL that will resolve to an instance of your
BrowserView. To fix that, you register the view.
When Zope starts up, it looks inside every
Productfor a file called
configure.zcml. In Zenoss, most Products don't have one (though some do now). You can do a bunch of stuff with these, but we're going to ignore everything except registration of views.
You would, in this case, modify
Products/ZenModel/browser/configure.zcml(because
Deviceis in
ZenModel; it doesn't actually matter where you register the view, but you should try to keep
Productspluggable), adding the registration of your view:
<browser:page
Notice that your view is defined as being applicable only to instances of the
Deviceclass. Were you to attempt to call
componentlistagainst an
IpInterfaceinstance, for example, you'd get a 404 -- not so if
componentlistwere a mere template. Also notice the relative import in the class attribute;
.DeviceViewswill look for the
DeviceViewsmodule in the current package, that is,
ZenModel.browser.
So, the whole request workflow progresses thusly:
Someone asks for
/zport/dmd/Devices/devices/
mydevice/componentlist
Zope resolves
mydevice; that's the context in which it'll attempt to resolve
componentlist
Zope attempts to resolve
componentlistas an attribute of
mydevice, then a method of
mydevice, then a dictionary key of
mydevice, then starts looking up registered views.
We find a view in the ZCML. Does it match?
name="componentlist": Check.
Context class="Products.ZenModel.Device.Device": Check.
We want the view
DeviceViews.ComponentListView.
Zope makes sure the user has
zope2.Viewin this context. We'll assume they do; if not, kicked out to login screen.
Zope instantiates
ComponentListView(mydevice), then calls it, which renders the template file.
The template is rendered, using
viewand
here, and returned as the response.
So much better! No bloated classes; no ridiculous class inheritance; great code organization. Define a method in one place, then adapt objects to provide it, instead of modifying many classes with the same method. If you want to see the screens available for a Device, just go look in the ZCML -- no need to remember which page templates are applicable to which objects. Also, you can adapt many different objects for the same template with different views.
There are a few other things that could be mentioned, but they all require a discussion of interfaces, which will deferred to a later section. Briefly, the Zope Component Architecture, and its aspect-oriented approach, saves a lot of hackery. Also it's the rules now. | http://community.zenoss.org/docs/DOC-10091 | CC-MAIN-2014-15 | refinedweb | 1,221 | 65.01 |
Created on 2018-02-13 13:38 by cheryl.sabella, last changed 2018-06-15 05:40 by terry.reedy.
In tkinter, after_cancel has a call to after info:
data = self.tk.call('after', 'info', id)
Since this is a supported command, there should be a function to access it directly.
What is the use case for this method? How it could be used?
I was working on the tests for issue32831. One of the methods was `__del__` which made sure timer events were canceled with `after_cancel`. In the test, to assert that the after events no longer existed after calling `__del__` and after reading the Tcl documentation for `after`, I tried to call `after_info` but it didn't exist. So I added a call to `self.tk.call('after', 'info', id)` directly to assert that the after events no longer existed.
I don't know if there is a general need to know whether timer or idle events exist, but this command gives that information.
I'm not sure what after_info(id) should return.
I've made a pull request. I understand that you may not want to add this functionality, but perhaps the docstring will answer your questions. I took it from the Tcl docs page.
I am in favor of exposing all of tk where it makes sense to do so, and I think it does here.
After 'translating' the tk after_info entry into tkinter-ese, I would expect and want that
root.after_info(root.after(f, 100000))[0] is f
be true. the same as would be true of the tcl equivalent. (This could even be a test.) It appears that the current patch instead returns a (python) reference to the tcl wrapper of f. The fact that python callbacks get wrapped as tcl callbacks is currently transparent to tkinter users and should remain so.
Serhiy, I presume that this is what you were uncertain about. I am presuming above that f can be recovered.
Returning the function kind as 'timer' or 'idle' is fine. In other contexts, an enumeration would be a possibility, but this does not seem to fit tkinter.
I presume a bad id results in TclError. Do other tkinter functions allow TclError to propagate? My impression is no. If so, it should be replaced here in a way consistent with other tkinter practice.
On one side, the first item of the list returned by the Tcl command `after info $id` is a name of the Tcl command generated by Tkinter. It is internal, it was not exposed to Tkinter users before, and the API for restoring the original Python callable is private.
On other side, `after info` can return not only events created by Tkinter, but events created by Tcl (either by direct execution of Tcl script, this is still can be useful with programming with Tkinter, or created by the Tcl standard library or third-party Tcl libraries). In that case a Python callable can't be returned.
This is what I was uncertain about. Maybe after_info() should return a Python callable if possible, and keep the original result otherwise? This complicates its implementation and definition.
TclError is legal and expected. In some methods it is caught, either because the method is purposed to be called at widget destroying stage, when the order of different cleanup procedures is not specified, and Tcl names can be destroyed before destroying Tkinter wrappers, or because the method was implemented differently in the past, and catching TclError is needed for backward compatibility. after_info() is not the case.
Note that in the tests for issue32831 you need to use call('after', 'info') if you want to backport them.
>>> It is internal, it was not exposed to Tkinter users before, and the API for restoring the original Python callable is private.
I thought `bind(sequence)` also returned these internal Tcl function names? For example, if I do a print on the set_breakpoint_here text widget in IDLE, it prints :
if {"[140358823678376set_breakpoint_here %# %b %f %h %k %s %t %w %x %y %A %E %K %N %W %T %X %Y %D]" == "break"} break
In order for it to return the python function name, I think the `after` function would need to write to a dictionary of `name: func` where name is currently used as the Tcl function name that is registered? Is there something in tkinter that does this now that it could be modeled from? Since events are removed from Tcl once that are invoked, how would the dictionary be cleaned up? Would after_info need to be polled every once in a while to clean up the dictionary or would it just exist until the object is destroyed?
A person who can create a tcl callback with tk.call can inquire with tk.call('after', 'info', id). That does not cover callbacks created by tcl or extensions thereof, but references to such callbacks are unlikely to be useful to anyone who does not know any tcl.
I see these choices for after_info(id):
A. Return the tcl script reference even when it wraps a python function. I don't like this, as the tcl reference is useless to most people.
B. Convert the reference to a Python function if possible but return it if not. This is a bit awkward to document and any use requires a type check. Having a function return such different types, depending on the input, is frowned upon.
C. Convert the reference to a function if possibe and raise TypeError or ValueError is not. This is incomplete but easier for a pure Python programmer to deal with. The documentation could specify how those who want a tcl reference can get it.
D. Don't implement after_info(id), at least not now, and just after_info(). Information about the current existence of a callback is contained in the list returned by after_info(). Each of the following pairs should be equivalent:
assertIn(id, after_info())
assertEqual(len(after_info(id)), 2)
assertNotIn(id, after_info())
assertRaises(TclError, after_info, id)
(For testing after_info(), assertIn and assertNotIn avoid assuming that tcl does not add any internal callbacks.)
> Since events are removed from Tcl once that are invoked, how would the dictionary be cleaned up? Would after_info need to be polled every once in a while to clean up the dictionary or would it just exist until the object is destroyed?
Good question. Currently the reference to a callable is kept in the dict until the object is destroyed. This can be considered as a bug (see issue1524639).
I agreed with Cheryl's conclusion that likely after_cancel() had been called with None. The comments about 8.4 is wrong, and the solution in issue763637 is not correct. The current code code deletes the script for the first event if pass None to after_cancel(). Do you ming to open a PR for proper solving issue763637 Cheryl?
I created issue32857 for the after_cancel issue. Thanks!
A few questions about returning the Python function name (specifically, how to derive it). This doesn't address the open issue with what to do about a Tcl command not tied to a Python function.
1. Serhiy wrote "and the API for restoring the original Python callable is private." What is that API?
2. In the _register method, the Tcl command name is the callback ID + the function name:
f = CallWrapper(callback, None, self._root).__call__
cbname = repr(id(f))
try:
callback = callback.__func__
except AttributeError:
pass
try:
cbname = cbname + callback.__name__
except AttributeError:
pass
So, with the values returned from tk.call('after', 'info', id) as (script, type), the Python function should be the same as script.lstrip('0123456789'). I'm not sure if that would be the best way to get the name back.
3. In tkinter, there is a list created/added to during _register:
self._tclCommands.append(cbname)
where cbname is the Tcl command name (as defined by the code in q2 above). Would it be possible to change _tclCommands to a dict mapping Tcl command name to Python function name? _tclCommands already has some logic around it, including .remove functions, so I think a dictionary would be more efficient for the exisitng purposes. Since it's semi-private, is there a fear with backward compatibility if it changes from a list to a dict? Is it better to add a new dict variable?
Thanks!
Real use case for after_info() (with not arg): #33855 is about minimally testing all IDLE modules. At least import the module and create class instances when easily possible. For test_editor, I started with
def test_init(self): # Temporary.
e = Editor(root=self.root)
self.assertEqual(e.root, self.root)
and got in Shell
warning: callback failed in WindowList <class '_tkinter.TclError'> : invalid command name ".!menu.windows"
and in the console
invalid command name "119640952recolorize"
while executing
"119640952recolorize"
("after" script)
invalid command name "119872312timer_event"
while executing
"119872312timer_event"
("after" script)
invalid command name "119872440config_timer_event"
while executing
"119872440config_timer_event"
("after" script)
Perhaps this is why I previously omitted something so obvious (it add 24% to coverage).
I added e._close(), which tries to cleanup, and the messages, in console only, are reduced to
bgerror failed to handle background error.
Original error: invalid command name "115211704timer_event"
Error in bgerror: can't invoke "tk" command: application has been destroyed
bgerror failed to handle background error.
Original error: invalid command name "115211832config_timer_event"
Error in bgerror: can't invoke "tk" command: application has been destroyed
I would like to know what _close misses, but it is hard to track them down.
print(self.root.tk.call('after', 'info')) after the close returned ('after#4', 'after#3', 'after#1', 'after#0'). Adding
for id in cls.root.tk.call('after', 'info'):
self.root.after_cancel(id)
before cls.root.destroy() in shutDownClass stops the messages.
--
For test_del in #32831, I think the following might work, and be much shorter than the current code.
n = len(self.root.tk.call('after', 'info')
self.cc.__del__()
self.assertEqual(len(self.root.tk.call('after', 'info')), n-2) | https://bugs.python.org/issue32839 | CC-MAIN-2019-09 | refinedweb | 1,663 | 65.42 |
Use Abstract Class And Interface Class?Jan 29, 2010
When to use Abstract class and when to use Interface class.View 10 Replies
When to use Abstract class and when to use Interface class.View 10 Replies
Error 1279 Cannot create an instance of the abstract class or interface 'System.Web.Mvc.FileResult'
[Code]....
I am using MVC 2. The same code works in my onather application. I have no idea about this error.
From the following URL i got some doubts about the Recommendations for using Abstract class vs interfaces
[URL]
1.. { Is there any example for this t ounderstand throughly ?}. { Is there any example for this t ounderstand throughly?
I am trying to compile the following code and i am getting the error:
Cannot create instance of abstract class .
m_objExcel = new Excel.Application();
m_objBooks = (Excel.Workbooks)m_objExcel.Workbooks;
m_objBook = (Excel._Workbook)(m_objBooks.Add(m_objOpt));
m_objSheets = (Excel.Sheets)m_objBook.Worksheets;
m_objSheet = (Excel._Worksheet)(m_objSheets.get_Item(1));
// Create an array for the headers and add it to cells A1:C1.
object[] objHeaders = {"Order ID", "Amount", "Tax"};
m_objRange = m_objSheet.get_Range("A1", "C1");
m_objRange.Value = objHeaders;
m_objFont = m_objRange.Font;
m_objFont.Bold=true;
// Create an array with 3 columns and 100 rows and add it to
// the worksheet starting at cell A2.
object[,] objData = new Object[100,3];
Random rdm = new Random((int)DateTime.Now.Ticks);
double nOrderAmt, nTax;
for(int r=0;r<100;r++)
{
objData[r,0] = "ORD" + r.ToString("0000");
nOrderAmt = rdm.Next(1000);
objData[r,1] = nOrderAmt.ToString("c");
nTax = nOrderAmt*0.07;
objData[r,2] = nTax.ToString("c");
}
m_objRange = m_objSheet.get_Range("A2", m_objOpt);
m_objRange = m_objRange.get_Resize(100,3);
m_objRange.Value = objData;
// Save the Workbook and quit Excel.
m_objBook.SaveAs(m_strSampleFolder + "Book2.xls", m_objOpt, m_objOpt,
m_objOpt, m_objOpt, m_objOpt, Excel.XlSaveAsAccessMode.xlNoChange,
m_objOpt, m_objOpt, m_objOpt, m_objOpt);
m_objBook.Close(false, m_objOpt, m_objOpt);
m_objExcel.Quit();
I just want to know that how can I utilize the concept of Abstract class, virtual class etc. in my shopping cart website. I have read the tutorial out there on internet and I saw some examples too, but those examples are so general that they dosen't fit into real world scenerio like I am searching for a shopping website. Same questions again and again comes to my mind that why to made a class only to give the declaration of methods and property.View 4 Replies
We all know that, we cannot create the object of Abstract class.
But why it is so?
I mean, can anyone explain it with real time example?
what is the function of abstract class, why design pattern need to build BLL and DAL
anyone give an example for my reference as I am getting strat to build my web-based project
Why do we use the reference of abstract class (or base class) to create object of it's sub-class. eg: TextWriter is the abstract class for StreamWriter & StreamWriter.
TextWriter writer = new StreamWriter();
why can't we simply use :
StreamWriter writer = new StreamWriter();.View 3 Replies
What is the use of abstract class design in real time projects
and how to use abstract clases and interfaces
and inheritense concepts
like big shoping portals or content managment,blogs etc
I am developing a couple of small ASP.NET application and would like to know what pattern. approach do you use in your projects.
My projects involve databases, using Data access and Business logic layers.
The data-access approach that I was using so far is the following(I read in some book and liked it):
For DAL layer:
Creating an abstract class that will define all database manipulation methods to implement.
The abstract class will contain a static "Instance" property, that will load (if instance == null) an instance (Activator.CreateInstance) of the needed type (a class that implements it).
Creating a classes that implement this abstract class, the implementation will be according to the databases (SQL, mySQL and etc) in use.
With this I can create different implementation according to database in use.
For BLL layer:
A class that encapsulates all all retrieved fields , and static methods that will call the DAL classes.
I am attempting to bind a Repeater (but it could be a GridView or ListView) to a list of objects. The List's type is an abstract type, which has two different classes derived from it, both with different properties. Because they have different properties, I cannot just have one ItemTemplate. If I bind a control to a property of one type of class and the other type doesn't have it, it throws an error.
Here's where I'm at:
I cannot use <% if (whatever) { %> some stuff <% } else { %> some other stuff <% } %> because I cannot access the databound item to make the choice based on its type. I cannot use the <%# %> syntax, which lets me use the databound information, because you cannot code logic like if...then...else. I cannot (rather not) call a function and return a string with the code because what I want to render is complex and contains further nested databound controls. Has anyone found an ingenious way of doing if it is this type of object, display these controls, else display these other controls? getting this build error on the following line of code, and do not find anyhting wrong there.
public partial class _Default : System.Web.UI.Page
{
}
am following the Nerd Dinner tutorial as I'm learning ASP.NET MVC, and I am currently on Step 3: Building the Model. One part of this section discusses how to integrate validation and business rule logic with the model classes. All this makes perfect sense. However, in the case of this source code, the author only validates one class: Dinner.
What I am wondering is, say I have multiple classes that need validation (Dinner, Guest, etc). It doesn't seem smart to me to repeatedly write these two methods in the partial class:
[code]....
This doesn't "feel" right, but I wanted to check with SO to get opinions of individuals smarter than me on this. I also tested it out, and it seems that the partial keyword on the OnValidate method is causing problems (understandably so). This doesn't seem possible to fix (but I could very well be wrong). am getting the same error (as this post:) as expected class,delegate,enum,interface or struct and also type or namespace definition or end of file expected.Below is the code:
public
partial
class
[code]... | http://asp.net.bigresource.com/-use-Abstract-class-and-Interface-class--CnSLUJCbY.html | CC-MAIN-2018-47 | refinedweb | 1,076 | 66.74 |
qbitval.3qt man page
QBitVal — Internal class, used with QBitArray
Synopsis
All the functions in this class are reentrant when Qt is built with thread support.</p>
#include <qbitarray.h>
Public Members
QBitVal ( QBitArray * a, uint i )
operator int ()
QBitVal & operator= ( const QBitVal & v )
QBitVal & operator= ( bool v )
Description
The QBitVal class is an internal class, used with QBitArray.
The QBitVal is required by the indexing [] operator on bit arrays. It is not for use in any other context.
See also Collection Classes.
Member Function Documentation
QBitVal::QBitVal ( QBitArray * a, uint i )
Constructs a reference to element i in the QBitArray a. This is what QBitArray::operator[] constructs its return value with.
QBitVal::operator int ()
Returns the value referenced by the QBitVal.
QBitVal & QBitVal::operator= ( const QBitVal & v )
Sets the value referenced by the QBitVal to that referenced by QBitVal v.
QBitVal & QBitVal::operator= ( bool v )
This is an overloaded member function, provided for convenience. It behaves essentially like the above function.
Sets the value referenced by the QBitVal tobitval.3qt) and the Qt version (3.3.8).
Referenced By
QBitVal.3qt(3) is an alias of qbitval.3qt(3). | https://www.mankier.com/3/qbitval.3qt | CC-MAIN-2017-26 | refinedweb | 190 | 51.95 |
I'm trying to get a transparent overlay sliding down in an app, pretty much like this here (all/filter-by):
So far I found react-native-slider and react-native-overlay. I modified the slider to work from top to bottom, but it always moves down the ListView as well. If using react-native-overlay, the overlay is static and I can't move it.
I added some demo code from the original react-native tutorial in this gist. When clicking the button, the content should stick, and the menu should overlay. The transparency is not that important right now but would be awesome.
What would be the smartest solution?
The key to your ListView not moving down, is to set the positioning of the overlay to
absolute. By doing so, you can set the position and the width/height of the view manually and it doesn't follow the flexbox layout anymore. Check out the following short example. The height of the overlay is fixed to 360, but you can easily animate this or make it dynamic.
'use strict'; var React = require('react-native'); var Dimensions = require('Dimensions'); // We can use this to make the overlay fill the entire width var { width, height } = Dimensions.get('window'); var { AppRegistry, StyleSheet, Text, View, } = React; var SampleApp = React.createClass({ render: function() { return ( <View style={styles.container}> <Text style={styles.welcome}> Welcome to the React Native Playground! </Text> <View style={[styles.overlay, { height: 360}]} /> </View> ); } }); var styles = StyleSheet.create({ container: { flex: 1, justifyContent: 'center', alignItems: 'center', backgroundColor: '#F5FCFF', }, welcome: { fontSize: 20, textAlign: 'center', margin: 10, }, // Flex to fill, position absolute, // Fixed left/top, and the width set to the window width overlay: { flex: 1, position: 'absolute', left: 0, top: 0, opacity: 0.5, backgroundColor: 'black', width: width } }); AppRegistry.registerComponent('SampleApp', () => SampleApp); module.exports = SampleApp;
I think you can try this component. I mean you should just set
position: 'absolute' to view that is pulled down and it could work.
Can you share the styles for the overlay ? I guess you can make it possible if you use position absolute for the Overlay instead of the regular position flex. | http://www.dlxedu.com/askdetail/3/8af118680ff65009f248f3cfc5288c6a.html | CC-MAIN-2018-43 | refinedweb | 356 | 58.28 |
This tutorial explains how to use the foreach loop to go through each element in an array. A foreach loop is simpler and easier to use than a for loop if you want to loop through each and every element in an array. There is no need to use a counter, specify an increment, or a condition. The foreach loop will simply loop through every element in the array.
Foreach loops are written in a way that is easy to understand, for example:
foreach (string item in itemsList){ Console.WriteLine(item); }
Watch the video below and then scroll down for the sample code.
Sample code
using System; namespace MyCSharpProject { class Program { static void Main(string[] args) { // Create an array of string type string[] names = {"Jim","Kate","Sam","Sally"}; // Store length of names array in variable int arrayLength = names.Length; // Go through each name in names array and display on new line foreach (string name in names) { Console.WriteLine(name); } // Wait for user input before quitting program Console.ReadLine(); } } } | https://www.codemahal.com/video/foreach-loops-in-c-sharp/ | CC-MAIN-2018-47 | refinedweb | 168 | 62.07 |
BBC micro:bit
Temperature Sensor
Introduction
If the temperature reading from the chip surface doesn't satisfy you, you can connect a tmp36 temperature sensor. This is a relatively cheap component. The only real advantage of using this is that you can move the reading away from the main circuit board.
Circuit
It's a very simple circuit that you need for this sensor.
Programming
We need to use the datasheet for the tmp36 to be able to convert readings into temperatures. This graph is the key item,
The first thing we need to do is work out what voltage we are reading from the sensor. Our power output on the micro:bit is 3.3V and, when we read our inputs, we get a figure from 0 - 1023. To convert this into millivolts, we need to multiply our reading by 3300/1024.
The graph shows us that, at 0°, the voltage reading should be 0.5V or 500mv. We subtract this number from our reading.
Finally, we can see on the graph that each volt in the reading represents 100°. Dividing by 10 on our figures will gives us a celsius reading.
from microbit import * while True: x = pin0.read_analog() x = x * (3300/1024) print("Voltage" + str(x)) c = (x-500) /10 print("TempC:" + str(c)) sleep(100)
This test program will output the information on the serial port. You need to press the REPL button if you are using the Mu application or use a terminal emulator if not.
Challenges
As with the built-in sensor, you could find a much more interesting way to display this information than this program. You could also use the temperature and its changes over time as a trigger for some other action on the micro:bit. That might be to turn a light on or off, or to sound an alarm. Alternatively, work out a set of suitable 5x5 icons to display according to the temperature reading you are getting at any time. | http://www.multiwingspan.co.uk/micro.php?page=tmp | CC-MAIN-2019-09 | refinedweb | 332 | 73.17 |
Home > Products >
women belts Suppliers
Home > Products >
women belts Suppliers
> Compare Suppliers
Company List
Product List
< Back to list items
Page:1/1
We have over 16 years of experience in manufacturing stylish accessories including men and women's belts, belt buckles and mini-hardware,LED lam...
Taiwan
We are a special manufacturer and exporter of badges/pins
in China, which produces custom products to match the needs
and designs of our custome...
China (mainland)
Established in 1995, Yiwu Xingni Handicraft Ornaments Co.,
Ltd. is located nearby Yiwu Airport and Ningbo, one of the
main port cities in China....
1982: Step into the headwear industry by setting up the 1st cap factory ( Chi Hsing Caps) in Taichung, Taiwan
1996: Chi Hsing Vietnam facility ...
Guangzhou Gold Fox Leather Co., Ltd. is specializing in
leather products, and gathering design, production,
packing, trade with a whole, 110 ski...
Yatai fashion accessories is a professional manufacturer,
designer and exporter of innovative and trendy fashion
accessories. Currently our prod...
We specialize in the design and export of women's and men's
belts, as well as other fashion accessories. Wenzhou lonson
maintains close relation...
We are a belt manufacture in Wenzhou of China. We mainly
product women and men, pet belt made of PU, PVC, artificial
leather, genuine leather be...
Yiwu Pingzhan Weaving belt Co., Ltd is located in Yiwu
which is one of the biggest commodity market in the world.
Our company is specializing in...
The H & Z International Trade Co. is dealing all kinds of
export work. We have professional employees to help you to
purchase, translate, introd...
We are a professional exporter of innovative and trendy
fashion accessories. We mainly export men belts and women
fashion belts to foreign count...
Our company is specialized in belt produce, we locate at
Yiwu, China.
Our company is a well repudiated export and manufacture. We
specialize in women?s fashion accessories such as
belts, bags, jewelry, scarves &glo...
Shanghai Transcheer Industry Co., Ltd. as a professional
import and export company, established in 2003. The company
has more than ten departmen...
Guangdong Han Belt Ltd. (HBL) was found in August 2005
and our history can be traced back to 2001. Now has
developed to be professional manufact...
Since 2003, Eliya Ornaments (HK) Ltd. professionally
specialize in offering high-end quality finished metal
buckle of fashion trend designs and ...
Our company professionally develops and produces all kinds
of men's and women's belts, Specializes in various belts,
such as cotton belt, leathe...
We are Wenzhou yicktak trade co.,ltd specializing in the
manufacture and export of belts.Our company is led by a
professional with rich expensiv...
Guangzhou Daiyating Fashion Co.,Ltd was reestablished in
1998.We are specialize in producing all kinds of
underwears with high quality. After ...
Established in 1996,Shanghai Youren Leather Co., Ltd. is a
professional manufacture enterprise that intergrates
designing,manufacturing,processi...
old wheel Leather Co.,Ltd Guangzhou during the 10-year R&D
and production has been working hard on casual fashion belt
for men and women to the ...
Established in 1996, Shanghai Youren Leather Co., Ltd. is a
professional manufacture enterprise that intergrates
designing, manufacturing, proce...
Elaine fashion Inc has 12 years of factoryproduction experience supplying innovative and trendy accessories. These include Fashion belts, PU Bel...
We are professional manufactory of belt, our company set up on 2002, we have 10 years production experience, we own ourself design department, w...
China wholesale & exporter of Men leather belts: men dress belt, men office belt, men casual belt, mens golf belts, mens plaited belt, mens brai...
This is a set belt design, production, sales, which integrates enterprise. Specializing in the production of all kinds of men and women fashion ...
Dongguan Zhongtang Bisheng Belt Factory is equipped with powerful technical force, and advanced and complete manufacturing equipment. Our produc...
Show:
20
30
40
Showing 1 - 27 items
Supplier Name
Year Established
Business Certification
Tacamoli Co., Ltd.
16th
Raybaud Group Ltd.
Yiwu Xingni Handicraft Ornaments Co., Ltd.
Max Succeed Enterprises Co., Ltd.
Guangzhou Gold Fox Leather Co., Ltd.
Wenzhou Yatai Fashion Accessories Co., Ltd.
Wenzhou Lonson Enterprise Co Ltd
Yongjia Chaolei Leather Co., Ltd.
Yiwu Pingzhan Weaving Factory
H & Z International Trade Co.
Hangzhou Fangjia Import & Export Co., Ltd.
Yiwu Gaoge Arts & Crafts Co., Ltd.
Wenzhou New Ocean Imp. & Exp. Co., Ltd.
Shanghai Transcheer Industry Co., Ltd.
Guangdong Han Belt Ltd.
Eliya Ornaments (HK) Ltd.
S.kingdom belt co.,ltd.
Wenzhou Yicktak Trade Co.,Ltd
Guangzhou Daiyating Fashion Co., Ltd.
Shanghai Youren Leather Manufacture Co., Ltd.
Old Wheel Leather Co., Ltd.
Shanghai Youren Leather Co., Ltd
Yiwu Elaine Leather Co.,Ltd
Yiwu City JBJ Leather Co.,Ltd
China Bessen Leather Products Limited
Wenzhou Wensen Leather Co., Ltd
Bisheng Belt Factory | https://www.ttnet.net/hot-search/suppliers-women-belts-.html | CC-MAIN-2018-51 | refinedweb | 787 | 61.43 |
Chemical Markup Language
October 2, 1997
Chemical Markup Language
A Simple introduction to Structured Documents
Peter Murray-Rust
Abstract
Structured documents in XML are capable of managing complex documents with many separate information components. In this article, we describe the role of the XML-LANG specification in supporting this. Examples are supplied explaining how components can be managed and how documents can be processed, with an emphasis on scientific and technical publishing. We conclude that structured documents are sufficiently powerful to allow complex searches simply through the use of their markup.
Historical Overview
Originally published as an HTML file, this paper was part of the CDROM e-publication ECHET96 ("Electronic Conference on Heterocyclic Chemistry"), run by Henry Rzepa, Chris Leach, and others at Imperial College, London, U.K. The CDROM was sponsored by the Royal Society of Chemistry, who (along with Cambridge, Leeds, and IC) are participants in the CLIC project. This is one of the projects under E-Lib, a U.K.-based program to promote electronic publishing. CLIC makes substantial use of SGML and Chemical Markup Language (CML). As part of this project I have been developing CML, one of the first applications of XML. CML, and its associated software JUMBO, probably represented one of the first complete XML applications (authoring tools, documents, and browser) in any discipline. Although the CML component was essentially a proof-of-concept, it was robust enough to be distributed as a standalone Java-based XML application. A wide variety of examples could therefore be viewed using JUMBO running under a Java-enabled browser.[1]
The audience for this paper need not be acquainted with SGML or XML; it serves as an introduction to the concept of document structure. As such, we assume no knowledge about markup languages, other than a familiarity with HTML. Though some parts may be trivially obvious to some readers, they may still find it useful as a tutorial aid for their colleagues. It is primarily aimed at those who are interested in authoring or browsing documents with the next generation of markup languages, especially those created with XML. CML [1] is part of the portfolio of the Open Molecule Foundation [2], which is a newly constituted open body to promote interoperability in molecular sciences. The latest versions of JUMBO can be found under the Virtual School of Molecular Sciences [3], which has also recently run a virtual course on Scientific Information Components using Java and XML [4].
The paper alludes to various software tools, but does not cover their operation or implementation. However, with the exception of stylesheets, most of the operations described here for CML have already been implemented as a prototype using the JUMBO browser and processor. The paper does not require any knowledge of chemistry or specific understanding of CML.
Finally, I should emphasize that SGML can be used in many ways; my approach does not necessarily do justice to the most common use, which is the management and publication of complex (mainly textual) documents. Projects in this area often involve many megabytes of data and industrial strength engines. I hope, however, that the principles described here will generally be of use.
Introduction
Two years ago I had never heard of structured documents, and have since come to see them as one of the most effective and cheapest ways to manage information. Though the basic idea is simple, when I first came across it I failed to see its importance. This paper is written as a guide to what is now possible. In particular, it explains XML--the simple new language being developed by a working group (WG) of the W3 Consortium. I have used this language as the basis for a markup language in technical subjects (Technical Markup Language, TecML) and particularly molecular sciences (Chemical Markup Language, CML).
The paper was originally written as a simple structured document, using HTML, although it could have been written in CML. I shall slant it towards those who wish to carry precise, possibly nontextual, information arranged in (potentially quite complex) data structures. While I use the term document, this could represent a piece of information without conventional text, such as a molecule. Moreover, documents can have a very close relation to objects; if you are comfortable with object-oriented languages you may like to substitute "object" for "document." In practice, XML documents can be directly and automatically transformed into objects, although the reverse may not always be quite so easy.
The markup I describe essentially uses the same syntax as HTML; it is the concepts, rather than the syntax that may be new. Although this paper is written in the context of document delivery over networks, markup is also ideally suited to the management of "traditional" documents. Markup languages are often seen as key tools in making them "future-proof" and interchangeable between applications (interoperability). the components and what behavior a machine is expected to perform). This is a much more challenging area than people realize,. Thirty years later we have most of the tools that are required to get the best information in the minimum quantity in the shortest time, from the people who are producing the information to the people who want it, whether they know they want it or not.[2]
Many scientists are unaware of the research during the last thirty years into the management of information.[3] In this review, (domain ontologies). For that reason, complex systems such as natural language processing (NLP) are required to extract implicit information from the documents, and they rely on having appropriate text to analyze. Automatic extraction of numerical and other nontextual information will be much more difficult.
Structure and Markup
We often take for granted the power of the human brain in extracting implicit information from documents. We have been trained over centuries to realize that documents have structure (Table Of Contents [TOCs], Indexes, Chapters with included Sections, and so on). It probably seems "obvious" to you that you are reading the fourth section ("Structure and Markup") in the paper ("A Simple Introduction to Sructured Documents"). The HTML language and rendering tools that you are using to read [the online version] provide a simple but extremely effective set of visual clues; for instance, "Chapter" is set in larger type. However, the logical structure of the document is simply:
HTML HEAD TITLE BODY H1 (Chapter) H2 (Section) H3 (Subsection) H3 H2 P (Paragraph) P P P P H2 P P P H2 P P ... and so on ... ADDRESS
where I have used the convention of indentation to show that one component includes another. This is a common approach in many TOCs, and human readers will implicitly deduce a hierarchy from the above diagram. But a machine could not unless it had sophisticated heuristics, and it would also make mistakes.
The formal structure in this document is quite limited, and that is one of the reasons that HTML has been so successful but also increasingly insufficient. Humans can author documents (the SGML/XML term is #PCDATA). There is a formal set of rules in HTML for which elements can contain which other Elements and where they can occur. Thus, it's not formally allowed to have TITLE in the BODY of your document. These rules, which are primarily for machines and SGML gurus to read, are combined in a Document Type Definition (DTD).
Note preceding H2. It would be quite natural to use phrases like "the second sentence of the second paragraph in the section called Introduction." Although humans can do this easily, it's common to get lost in large documents. The important news is that XML now makes it possible for machines to do the same sort of thing with simple rules and complete precision. The Text Encoding Initiative (a large international project to mark up the world's literature) has developed tools for doing this, and they will be available to the XML community.
In HTML there are no formal conventions for what constitutes a Chapter or Section, and no restriction as to what elements can follow others. Therefore, you can't rely on analyzing an arbitrary HTML document in the way I've outlined. This highlights the need for more formal rules, agreements, and guidelines. In XML we are likely to see communities such as users of CML develop their own rules, which they enforce or encourage as they see fit. For example, there is no restriction on what order Elements can occur in a CML document, but there is a requirement that ATOMS can only occur within a MOL (molecule Element). (In CML I use the term "ChemicalElement" to avoid confusion!)
In the Schatz reference that is footnoted earlier, you will probably "know automatically" what the components are. The thing in brackets must be the year, "pp." is short for "pages," the bold type must be the volume, and the italics are the journal title. But this is not obvious to a machine;, which are well understood and largely agreed within the bibliographic community, are a good example of something that can be enhanced by markup. Markup is the process of adding information to a document that is not part of the content but adds information about the structure or elements. Using the Schatz>
A scientist never having seen markup before would implicitly understand this information. The advantage is that it's also straightforward to parse it by machine. If the tags (<...>) and their content are ignored, then the remainder (content) is exactly the same as it was earlier (except for punctuation and rendering). It's often useful to think of markup as invisible annotations on your document. Many modern systems do not mark up the document itself, but provide a separate document with the markup. For example, you may not be allowed to edit a document but can still point to, and comment on, a phrase, section, chapter, etc. This is a feature of hypermedia systems, and one of the goals of XML is to formalize this through the development of linking syntax and semantics in XML-LINK (XLL), but this is outside the scope of this paper.
What is so remarkable about this? In essence we have made it possible for a machine to capture some of those things that a human takes for granted.
- Punctuation and other syntax are no longer a problem, as there are extremely carefully defined rules in XML. If your markup characters are <...>, how do you actually send < and > characters without them being mistaken for markup? One way is to encode them as < and >.
- Character encoding and other character entities have received a huge amount of attention and many entity sets have been developed, some by ISO. For example, the copyright symbol (©) is number 169 in ISO-Latin
©. It also has a symbolic representation (
©). XML itself has only a very few built-in character entities, but will support Unicode and other approaches to encoding characters. Most browsers do not yet support a wide range of glyphs for entities, but this is likely to change very rapidly, especially since languages like Java have addressed the problem.
- The role of information elements is defined. In the previous example, you can see what the precise components are and what their extent is. Note how the AUTHOR element is divided into three components. What you do with this information is the remit of semantics, and XML separates syntax precisely from semantics in a way that very few other non-SGML systems can do.
- Documents can be reliably restructured or filtered by machine. An author might enter the LASTNAME, FIRSTNAME, and INITIAL sequentially, but the machine could be asked to sort them into a different order. This may not appear very important, but to those implementing programs it is an enormous help. If the house style was initials-only, the program could easily turn Bruce into B.
- Documents can be transformed, merged, and edited automatically. This is a great advance in information management. For example, it would be straightforward to write a citation analyzer that found all BIB elements in a document and abstracted parts of them by JOURNAL or YEAR.
- It's easy to convert from one structured document to another. The bibliographic example above is not in strict CML, but it's very easy to convert it to CML, without losing any information.
- All information in a document can be precisely identified. The above example is marked down to the granularity of a single character (the INITIAL). It is conceptually easy to extend this to markup of numbers, formulae, and parts of things such as regions in diagrams or atoms in molecules.
Rules, Meta-Languages, and Validity
I started writing Chemical Markup Language because I wanted to transfer molecules precisely using ATOMS, BONDS, and related information. It was always clear that "chemistry" involved that enable markup languages to be written;, which uses a more flexible set of rules (but is also harder to parse or read by machine). Another rule is: all attribute values must occur within quotes (" or '). Writing a markup language is somewhat in a DTD. This is usually a separate file, but part or all can be included in the document itself. An example of a validity criterion in HTML is that LI (a ListItem) must occur within a UL or OL container. Well-formedness is a less strict criterion and requires primarily that the document can be automatically parsed without the DTD. The result can be represented as a tree structure. The bibliographic example above is well-formed, but without a DTD, it may not be valid. It might have been an explicit rule, like "the author must include an element describing the language that the article was written in, such as <LANGUAGE>EN</LANGUAGE>"; in this case, the document fragment would be invalid.
The importance of validity will depend on the philosophy of the community using XML. In molecular science all *.cml documents will be expected to be valid and this is ensured by running them through a validating parser such as NXP.[4] If a browser or other processing application such as a search engine can assume that a certified document was valid (perhaps from a validation stamp) there would be no need to write a validating parser. Being valid doesn't mean the contents are necessarily sensible; further processing may be needed for that purpose..
Processing Documents
At this stage it's useful to think about how an XML document might be created and processed. At its simplest level a document can be created with any text editor; this computers.
XML documents can be created, processed, and displayed in many ways. The schematic diagram in Figure 1 (which emphasizes the tree structure) shows some of the possible operations.
The lefthand module shows parts of the editing process. Legacy documents can be imported and converted on the fly, and the tree can be edited. There will normally also be a module for editing text. The editor may have access to a DTD and can therefore validate the document as it is created. An important aspect of XML-LINK is that editors should be able to create hyperlinks, either internally or to external files.
The complete document will then be mounted on a server. This will associate it with stylesheets, Java classes, the DTD, entities, and other linked components. The packaged documents are then delivered to the client where the application requires an XML parser. If the client wishes to validate the document the DTD is required.
Many XML applications will then hold the parsed document in memory as a tree (or grove) which can then be further processed. A frequent method will be the delivery of DSSSL stylesheets with the document (or provided client-side), or other transformation tools (perhaps written in Perl). Alternatively, the components of the document may be associated with Java classes either for display or transformation (as in the JUMBO browser). All of these methods may involve semantic validation (such as "does the document contain sensible information?").
Some of the operations required in processing XML are now explained in more detail:
Authoring
- One of the hardest problems is to write the authoring tools for an SGML/XML system. A good tool has to provide a natural interface for authors, most of whom won't know the principles of markup languages. It may also have to enforce strict and complex rules, possibly after every keystroke. Many current authoring tools are therefore tailored to a limited number of specific applications, one of the most versatile of which is an SGML add-on to Emacs. Sometimes a customer will approach an SGML house and, after agreeing on a DTD, a specific tool will be built. For some common document types--such as military contracts--there is enough communality that commercial tools are available.
In some cases authoring involves conversion of legacy documents; if these are well understood, conventional programs can be written in Perl or similar languages. Where the XML documents represent database entries or the output from programs, the authoring process is particularly simple--many CML applications will fall in that category. XML makes it particularly easy to reuse material either by "cut-and-paste" of sections, or preferably through entities. Classes written for JUMBO can already convert 15 different types of legacy files into CML.
Figure 1
- Editing and merging affects the structure of the document and therefore may require validation. To write programs that do this on the fly is again difficult; and it may be useful, where possible, to divide documents into "chunks" or entities. SGML has a very powerful concept or entities and can describe documents whose components are distributed over a network. For example, if I have an address, it is extremely useful to refer to that chunk by a symbolic name, such as
&pmraddress;. With appropriate software I can include this at appropriate places and the software will include the full content of the entity. (If the entity contains references to other entities, they are also expanded, and so on.)
- The server has a vital role to play in many XML applications. It is possible to mount sophisticated SGML systems that retrieve document components and assemble them on the fly into XML documents. Alternatively, the components could be retrieved from databases, as with chemical and biological molecules or data, and converted into XML files. Since XML maps onto object storage, it is particularly attractive for those developing object-based systems such as CORBA. Whether the complete document is assembled at the server or the addresses of the fragments are sent to the client will depend on bandwidth, the preference of the community, the availability of software, and many other considerations.
- Parsing is the process of syntactic analysis and validation. It normally produces a standardized output either on file or in memory. Whether you need to validate documents when you receive them will depend on your community's requirements. For example, if I receive a database entry from a major molecular data center I can rely on its validity, but a publisher getting a hand-edited XML manuscript will probably want to validate it. A validating parser requires that the document be valid against a specified DTD. Finding this DTD normally requires interpretation of the DOCTYPE statement at the head of an XML document. Some authors/servers are prepared to distribute the DTDs when documents are downloaded. While this adds precision in that the correct DTD is used, it can add to the burden of server maintenance and can increase bandwidth. If a community agrees on a DTD, they may find it useful to distribute it with the browsing software. The result of parsing is usually a parse-tree. If this is an unfamiliar concept, think of it as a table of contents with every Element corresponding to a chapter or (sub. . .sub) section. Trees are easy to manipulate and display; JUMBO displays the tree as a TOC. There are already two freely available XML parsers written in Java (NXP and Lark)[5] and I have used both. Lark creates a parse tree in memory that can be subclassed, while NXP produces it on the output stream.
- Most documents require at least some postprocessing, and many need a lot. Most users of XML applications will think of "browsers" or "plug-ins" as the obvious tools to use on a document. This will probably be true, but because it's machine processable XML is so powerful that many completely new applications will be developed. An XML document might consist of an airline reservation and the postprocessor could decide to order a taxi to the airport. A chemical reaction in a CML document could trigger the supply of checmicals and interrogate the safety databases.
- An XML document carries no semantics with it, and there has to be an explicit or implicit agreement between the author and reader. Most authors understand roughly the same thing by the TITLE in HTML documents, although they might try and use them in different ways. TITLE is valuable for indexers such as AltaVista, which abstract their content separately from the body of the document. This emphasizes the value of structural markup. However, some widely used element names are ambiguous (A is variously used in different DTDs for author, anchor, etc.), and for some, such as LINK, it's unclear what their role is. Clarifying this for each DTD requires semantics. Traditionally, semantics have been carried in documentation: if this is not done clearly then implementers may provide different actions for the same Element. The XML project is actively investigating formal automatic ways of delivering semantics, such as stylesheets and Java classes.
- The DTD/validating-parser cannot deal with some aspects of validation, which must be tackled by a conventional program/application. Common examples of validation are content ("is this number in the allowed range?"), and occurrence counts ("no more than five sections per chapter"). This is likely to need special coding for each application, and will be most important where high precision and low flexibility is the intention.
- Stylesheets are sets of rules that accompany a document.[6] They can be used to filter or restructure the document ("as in extract all footnotes and put them at the end of a section"). Their most common use is in formatting or providing typesetting instructions ("all subsections must be indented by x mm and typeset in this font"). ISO has produced a standard for the creation of stylesheets (DSSSL), which allows their description in Scheme (a derivative of LISP). Stylesheets are generally written to produce a transformed document, rather than to create an object in memory; Java classes are more suitable for this. I expect to see the technologies converge--which is used will depend on the application and the community using it. There are at least four ways that stylesheets might be used; the technology exists for each one. Which overrides which is a matter of politics, not technology.
- By the author. If an author wishes to impart a particular style to a document, he can attach or include a stylesheet. This can be invoked at the postprocessor level, unless it has been overridden.
- By the server. If an organization such as publishing house is running the server, it may impose a particular style, such as for bibliographic references. XML would give the author the freedom to prepare them in a standard way (e.g., using CML), while the journals could transform this by sending their stylesheets to the reader.
- By the client software (browser). The software manufacturer has an interest in providing a common look-and-feel to the display. It reduces training and documentation costs and might provide a competitive market edge.
- By the reader. She may have personal preferences concerning the presentation of material, perhaps because of her education. Alternatively, her employer may require a common house style to facilitate training and internal communication.
- Every Element can be thought of as an object and have methods (or behavior) associated with it. Thus, a LIST object might count and number the items it contains. Most elements will have a
display()method, which could be implemented differently from object to object. Thus, in JUMBO,
MOLNode.display()brings up a rotatable screen display of the molecule, while BIB.display() displays each citation in a mixture of fonts. As with stylesheets, Java classes can be specified at any of the four places listed above, and the appropriate one downloaded from a Web site if required. One of the problems the XML-WG is tackling and solving is how to locate Java classes. Because Java is a very powerful programming language with full WWW support, it offers almost unlimited scope for XML applications. A document need not be passive, but could awake the client to take a whole series of actions--mailing people, downloading other data, and updating the local database are examples.
- Most XML "documents" will consist of several physical files or streams, and these may be distributed over more than one server. An important attraction of XML is that common document components such as citations, addresses, boilerplate, etc. can be reused by many authors. Packaging these components is a challenge that the W3C and others are tackling. It involves:
- Methods of locating components. XML uses URLs or their future evolution (such as URNs).
- Labeling a file with its type. XML has provision for NOTATION, which may be linked to a reference URL or a MIME type.
- Creating a manifest of all the components required in a package (perhaps through a Java archive file [*.jar]).
Attributes
So far I have used only Element names (sometimes called GIs) to carry the markup. XML also provides attributes as another way of modulating the element. Attributes occur within start-tags, and well-known examples:
- Describing the type of information (e.g., what language the Element is written in)
- Adding information about the document or parts of it (who wrote it, what its origins are)
- Suggestions for rendering, such as recommended sizes for pictures
- Help for the postprocessor (e.g., the wordcount in a paragraph)
In XML-LINK attributes are extensively used to provide the target, type, and behavior of links.
Flexibility and Meta-DTDs
As discussed earlier, when developing an XML application, the author has to decide whether precision and standardization. Because this is a major effort and cost, careful planning of the DTD is necessary.
If flexibility is more important, either because the field is evolving or because it is very broad, a rigid DTD may restrict development. In that case a more general DTD is useful, with flexibility being added through attributes and their values.
In TecML I created an Element type, XVAR, for a scalar variable. Attributes are used to tune the use and properties of XVAR, and it's possible to make it do "almost anything"! For example, it can be given a TYPE such as STRING, FLOAT, DATE, an important feature of XML; the precise syntax is being developed in XML-LINK. CML uses DICTNAME to refer to an entry in a specified glossary that defines what "Melting Point" is. This entry could have further links to other resources, such as world collections of physical data. Similarly, UNITS is used to specify precisely what scale of temperature is used. Again, this is provided by a glossary in which SI[7].
Note
In the preceding example the links are implicit; later versions of CML will probably use the explicit links provided by XML-LINK.
The TecML DTD uses very few Element types, and these have been carefully chosen to cover most of the general concepts that mark up most technical scientific publications. There has to be general agreement about the semantics of the markup, of course, but this is a great advance compared with having no markup at all.
Entities and Information Objects
When documents have identifiable components it is often useful to put them into ENTITYs in separate files or resources. For example, although a citation might be used by many documents, only one copy is needed as long as all documents can address it. Chapters in an anthology might all be held as separate entities, allowing each to be edited independently. If the entity is updated (it might be an address, for example) all references to the entity will automatically point to the correct information. Entities in XML can be referenced through URLs allowing truly global hyperdocuments.
Many documents involve more than one basic discipline. For example, a scientific paper may include text, images, vector graphics, mathematics, molecules, bibliography, and glossaries. All of these are complex objects and most have established SGML conventions. Authors of these documents would like to reuse these existing conventions without having to write their own (very complicated) DTDs. The XML community is actively creating the mechanisms for doing this. If components are mixed within the same document, their namespaces must be identified (e.g., "this component obeys the MathML DTD and that one obeys CML"). For example, all the mathematical equations could be held in separate entities, and so could the molecular formulae. This would also support another method of combining components through XML-LINK, where the components are accessed through the HREF syntax.
Searching
Realizing the power of structured documents (SD) for carrying information was a revelation for me. In many disciplines, data needed to hold XML-like documents, and suppliers now offer SGML interfaces. (For any particular application, of course, there may be a choice between RDBs and ORDBs.) The attraction of objects overRDBs is that it is much easier to design the data architecture with objects.
In many cases simply creating well marked-up documents may be all that is required for their use in the databases of the future. The reason for this confident statement is that SDs provide a very rich context for individual Elements. Thus we can ask questions like:
- "Find all MOLECULEs which contain MOLECULEs." (e.g., ligands in proteins)
- "Which DATASET contains one MOLECULE and one SPECTRUM whose attribute TYPE has a value of nmr?"
- "Find all references to journals not published by the Royal Society of Chemistry."
Despite their apparent complexity, these can all be managed with standard techniques for searching structured documents. Because of this power, a special language (Structured Document Query Language--SDQL) has been developed and will interoperate with XML. If simple application-specific tools are developed then queries like the following are possible:
- "Find all XVARs whose DICTNAME value is Melting Point; retrieve the value of the UNITS attribute and use it to convert the content to a floating point number representing a temperature on the Celsius scale. Then include all data with values in the range 150-170."
The XML-LINK specification has borrowed the syntax of extended pointers (XPointers) from the Text Encoding Initiative (TEI). Although primarily intended to access specific components within an XML document, the syntax is quite a powerful query language. The first two queries might be represented as:
ROOT,DESCENDANT(1,MOLECULE) DESCENDANT(1,MOLECULE) ROOT,DESCENDANT(DATASET)CHILD(1,MOLECULE)ANCESTOR(1,DATASET)CHILD(1,SPECTRUM,TYPE,"nmr")
The first finds the first MOLECULE, which is a descendant of the root of the document, and then the first MOLECULE, which is somewhere in the subtree from that. The second is more complex, and requires the MOLECULE and SPECTRUM to be directly contained within the DATASET element. (The details of TEI Xpointers in XML may still undergo slight revision and are not further explained here.)
Summary, and the Next Phase
This document has described only part of what XML can offer to a scientific or publishing community. XML has three phases; only the first has been covered here in any depth. XML-LINK defines a hyperlinking system and XML-STYLE defines how stylesheets will be used. Hyperlinking can range from the simple, unverified link (as in HTML's HREF attribute for Anchors) to a complete database of typed and validated links over thousands of documents. XML-LINK required components- plug-ins, and perhaps there will be more autonomous tools that are capable of independent action. It's an excellent approach to managing legacy documents rather than writing a specific helper for each type.
I hope enough tools will be available for XML to provide the same creative and expressive opportunities as HTML provided in the past. However, it's important to realize that freely available software is required--any tools for structured document management, especially in Java, will be extremely welcome. The accompanying paper describes my own contribution through the JUMBO browser.
-
-
-
-
- Robin Cover's SGML Home page:
- FAQ for XML run by Peter Flynn:
About the Author
- Peter Murray-Rust
- Virtual School of Molecular Sciences
- Nottingham University, UK
- pazpmr@unix.ccc.nottingham.ac.uk
Peter Murray-Rust is the Director of the Virtual School of Molecular Sciences at the University of Nottingham, where he is participating in a new venture in virtual education and communities. Peter is also a visiting professor at the Crystallography Department at Birkbeck College, where he set up the first multimedia virtual course on the WWW (Principles of Protein Structure).
Peter's research interests in molecular informatics include participation in the Open Molecule Foundation--a virtual community sharing molecular resources; developing the use of Chemical MIME for the electronic transmission of molecular information; creating the first publicly available XML browser, JUMBO; and developing the Virtual HyperGlossary--an exploration of how the world community can create a virtual resource in terminology.
[1] An accompanying article by Peter Murray-Rust, "JUMBO: An Object-based XML Browser," is included in this issue as well. The JUMBO paper is more technical, and describes novel work in relating XML document structure to Java classes.
[2] Bernal's words, quoted in Sage, Maurice Goldsmith, p. 219.
[3] A recent and valuable review is, "Information Retrieval in Digital Libraries: Bringing Search to the Net," Bruce R. Schatz, Science, 275, pp. 327-334 (1997). (I shall comment on the format of the last sentence shortly.)
[4] Norbert Mikula's validating XML parser at.
[5] See the article entitled "An Introduction to XML Processing with Lark," by Tim Bray.
[6] For more information on stylesheets, and particularly on W3C's cascading stylesheets, see the article entitled "XML and CSS" (Culshaw, Leventhal, and Maloney) in this issue. Also see the Winter 1997 issue of the W3J for the CSS1 specification as well as an implementation guide to the spec by Norman Walsh.
[7] Systèm Internationale: the international standard for scientific units. | https://www.xml.com/pub/a/w3j/s3.rustintro.html | CC-MAIN-2018-34 | refinedweb | 5,749 | 53.1 |
Write a parser and writer for a new structure text format targetted at non-programmers
Budget $20-100 USD
Senior Executive Summary
Write a parser and writer for a new structure text format targetted at non-programmers
Executive Summary
Elegant Technologies Structured Text (ETST) is yet another structured text initiative to create a structured text format, but this one is being specifically designed to be used by non-programmers. The intended uses are:
1 a default text format for non-programmers, so when writing copy for a web site using Microsoft Word, they can created their own basic formatting.
2 a human readable format to store formatted text, such as when entered via rich html client (aka WYSIWYG in-browser editors) on a website, like earthlink’s webmail client, instead of storing in, say, RTF or HTML.
See the uploaded file for more details.
This will problably require at least three classes, and they must be changable intermingling at run-time (see the Design Pattern 'Builder').
If you are not familiar with the term "Design Pattern", then this project probably isn't right for you.
I'll provide a some test files before program kick-off.
Please allow some time for refinement.
You have have to record total hours worked record them at the end of each week on our phpdotproject site - but you will be paid by the job.
Please give an estimated total labor hours required to accomplish the job.
Links:
[url removed, login to view]
[url removed, login to view]
[url removed, login to view]
[url removed, login to view]
The Job:
Create a data storage class. Create a reader class that parses text files and inputs data into the storage class. Create a reader class that reads the text string that are stored in a MySQL database and inputs the info into the storage class. Create a class that interfaces with the storage class and translates it to pretty HTML. Create a class that interfaces with the storage class and translates it to a file.
You must be able to quickly parse and retrieve 500 pages of text from a text file in less than ½ sec on a Athlon 1.2 Ghz CPU with 512MG ram running Apache 2 and PHP 4.3.9. The code must also run under PHP 5.0 It is also acceptable for the importer code run in Python, but I need a PHP reader.
History
I recently finished writing a large-ish handbook for the web. I worked with several very knowledgable subject matter experts on the subject, but I was faced with the task of inputting all of their information into a website. The book was going to be about 400 web-pages long, and since everyone was working in their own version of MS Word, I had asked then to write there text in structured text (specifically, Python ReStructured Text, aka RST). I gained a lot of respect for for the structured text community, but this experience, combined with trying to roll out a WIKI, which used a different structured text format, at my company pursuaded me that non-programmers needed something a little simplier.
Principles
There is only one right way to do things. I don’t want people to get confused.
It should always be obvious where the special instructions are.
Simplier is better than complex. This language is for writers, not website designers – we aren’t doing anything too special here.
All advanced features are encapsulated in '[]'s
Authors can generate copy in MS Word 97 through 2003 – perhaps after turning off auto-completion.
It shall be implemented in PHP 4.3.9 & 5.0 using known Design Patterns.
It shall have PHP 5.0 unit tests.
When completed, a one page instruction sheet that neatly fits onto a web page AND a two page PDF cheat sheet, suitable for putting into a manual will be made. The front page will be new users, the back page will a reference for all advanced features.
Some standard emoticons and symbol shortcuts will be used.
A lot of special formatting, like acronyms, will be left up to custom output classes, and will not be part of the ETST spec.
Syntax errors are not possible, except inside brackets ([]). See, I spent a lot of time debugging my RST text. I mean it – I was actually debugging the text. This would have been impossible for a non-programmer.
TABS don’t exist, they are treated like five spaces.
English only. Sorry, I just don’t enough about other languages.
Each formatting addition must be modular so that we can add new stuff in the future.
The Spec
- Headers
- Header 1 (headers levels are designated by the number in dashes “-“ infront of some text)
-Header 1
-- Header 2
--Header 2
-- Header 3
---Header 3
...
---------- Header 10
----------Header 10
Headers can optionally have a line of dashes or equal signs under them. They will be ignored, but it makes text files easier to read:
-Application Programmer's Interface
=====================================
-- Overview
---------------------------------------
- Paragraphs
A paragraph starts on the next line.
Another paragraph begins after an carraige return (this is a paragraph with no extra separation).
One or more blank line between paragraphs designates an extra separation between the paragraphs
Two blanks lines, like this new paragraph, is no different than the preceeding paragraph.
One or more spaces or tabs incidates that this paragraph should be indented.
This is another indented paragraph.
Here is a third indented paragraph. These will all be indented the same amount. We just care that it was indented.
-- Justification
Paragraphs are, by default, left justified. You can, however, specify a paragraph to be centered, right justified, or justified. Once specifying a justification for a paragraph, then all subsequent paragraphs have the same formatting. Indenting issues for the first line of a paragraph are ignored for all formatting except left justified paragraphs. Bullets are always left justified
>> Here is a right justified paragrapaph.
Here is another right justified paragraph.
>< Here is a centered paragraph.
Here is a justified paragraph
<< here is a left justified paragraph.
Here is a justified paragraph.
- Bullets
Here are a bunch of bullets. The first level of bullet should be indented. Since it is generally pretty obvious that line starting with a ‘*’ is a bullet, then even if the formatting isn’t perfect, the parser will take its best guess at what the author intended, and must always rewrite the formating correctly. To guess at the author’s formatting, the parser base everything one uninterupted cluster of bullets at a time (a new paragraph breaks a cluster). Determine how many distinct levels of indentation there are, and group by indentation, so, like python, indentation counts.
* Here is a bulletted item
* Here is a second item
*Here is another bullet, and the end of the first bullet cluster.
Here is a new paragraph
*)
Now, here is the same cluster of bullets nicely formatted.
*)
-- Numbered Lists
For numbered list, the formatting is the same as for bullets, except that instead of a ‘*’, either a number or a ‘#’ can be used. Decorations immediately after the number are ignored and removed when parsed. When it is rewritten, the items are nicely numbered as ‘X. Some text.’ Here are some examples:
1. First, get a chicken
2. Find a big pot
5.) Find a seclude clearing
3 Get a bushel of wheat.
# I forget what comes next
# some other items
(9) lastly, enjoy your meal!
These are all legal, and produce the following nicely formatted text.
1. First, get a chicken
2. Find a big pot
3. Find a seclude clearing
4. Get a bushel of wheat.
5. I forget what comes next
6. some other items
7. lastly, enjoy your meal! | https://www.freelancer.com/projects/php-python/write-parser-writer-for-new/ | CC-MAIN-2017-43 | refinedweb | 1,291 | 65.01 |
#include <limits.h>
#include <stdbool.h>
#include <stdint.h>
#include <stdlib.h>
#include "nvim/api/buffer.h"
#include "nvim/api/deprecated.h"
#include "nvim/api/extmark.h"
#include "nvim/api/private/defs.h"
#include "nvim/api/private/helpers.h"
#include "nvim/api/vim.h"
#include "nvim/api/vimscript.h"
#include "nvim/extmark.h"
#include "nvim/lua/executor.h"
Retrieves a line range from the buffer
Deletes a buffer line
Removes a buffer-scoped (b:) variable
Gets a buffer line
Inserts a sequence of lines to a buffer at a certain index
Sets a buffer line
Replaces a line range on the buffer
Sets a buffer-scoped (b:) variable
@warning It may return nil if there was no previous value or if previous value was `v:null`.
Clears highlights and virtual text from namespace and range of lines
Gets the buffer number
Set the virtual text (annotation) for a buffer line.
The text will be placed after the buffer text. Virtual text will never cause reflow, rather virtual text will be truncated at the end of the screen line. The virtual text will begin one cell (|lcs-eol| or space) after the ordinary text.
Namespaces are used to support batch deletion/updating of virtual text. To create a namespace, use |nvim_create_namespace()|. Virtual text is cleared using |nvim_buf_clear_namespace()|. The same
ns_id can be used for both virtual text and highlights added by |nvim_buf_add_highlight()|, both can then be cleared with a single call to |nvim_buf_clear_namespace()|. If the virtual text never will be cleared by an API call, pass
ns_id = -1.
As a shorthand,
ns_id = 0 can be used to create a new namespace for the virtual text, the allocated id is then returned.
Removes a tab-scoped (t:) variable
Sets a tab-scoped (t:) variable
@warning It may return nil if there was no previous value or if previous value was `v:null`.
v:null.
Removes a window-scoped (w:) variable
Sets a window-scoped (w:) variable
@warning It may return nil if there was no previous value or if previous value was `v:null`. | https://neovim.io/doc/dev/deprecated_8c.html | CC-MAIN-2022-21 | refinedweb | 342 | 67.86 |
This document contains information about rewriting JavaScript editor. The umbrella task for the rewrite is issue #205870.
As of July 10th, the new JavaScript editor is now a part of the main trunk codeline, can be tried out in the daily builds. The code is in two new modules: javascript2.editor and javascript2.kit.
The editor rewrite contains mainly these tasks:
Next big task is JSON support.
Document about progress of rewriting JS editor and comparing new and old support can be watched here. See also the following blog entries about the new JavaScript editor: first, second and third.
"For now we are supporting all JsDoc2 tags" [1]
ie used by the language model for code completion and documentation [2]
* @class
* @constructor
* @constructs
* @deprecated
* @private
* @public
* @static
* @returns
* @type
* @property
* @param
* @since
* @author
* @description
* @example
* @fileOverview
* @throws
netbeans should correctly parse all other tags, but they're not used,
ie neither connected nor used by the documentation. connecting a tag to the language model has an impact on performance, so the developers are evaluating them on a case by case basis. to request support for a tag, enter an issue in the bug tracker
* @augments - support request
* @borrows _that_ as _this_
* @constant
* @default
* @event
* @field
* @function
* @ignore
* @inner
* @lends - support request
workaround #1: the lending object can be assigned to the class prototype directly
workaround #2: the lending object can be initially named as the class
* {@link ...}
* @memberOf
* @name
* @namespace
* @requires
* @see
* @version
* @argument - Deprecated synonym for @param
* @extends - Synonym for @augments | http://wiki.netbeans.org/JavaScript2 | CC-MAIN-2018-17 | refinedweb | 252 | 50.36 |
Zato—Agile ESB, SOA, REST and Cloud Integrations in Python
Laying Out the Services
The first thing you need is to diagram the integration process, pull out the services that will be implemented and document their purpose. If you need a hand with it, Zato offers its own API's documentation as an example of how a service should be documented (see and):
Zato's scheduler is configured to invoke a service (update-cache) refreshing the cache once in an hour.
update-cache, by default, fetches the XML for the current month, but it can be configured to grab data for any date. This allows for reuse of the service in other contexts.
Client applications use either JSON or simple XML to request long-term rates (get-rate), and responses are produced based on data cached in Redis, making them super-fast. A single SIO Zato service can produce responses in JSON, XML or SOAP. Indeed, the same service can be exposed independently in completely different channels, such as HTTP or AMQP, each using different security definitions and not interrupting the message flow of other channels.
Figure 1. Overall Business Process
Implementation
The full code for both services is available as a gist on GitHub, and only the most interesting parts are discussed.
linuxjournal.update-cache
Steps the service performs are:
Connect to treasury.gov.
Download the big XML.
Find interesting elements containing the business data.
Store it all in Redis cache.
Key fragments of the service are presented below.
When using Zato services, you are never required to hard-code network addresses. A service shields such information and uses human-defined names, such as "treasury.gov"; during runtime, these resolve into a set of concrete connection parameters. This works for HTTP and any other protocol supported by Zato. You also can update a connection definition on the fly without touching the code of the service and without any restarts:
1 # Fetch connection by its name 2 out = self.outgoing.plain_http.get('treasury.gov') 3 4 # Build a query string the backend data source expects 5 query_string = { 6 '$filter':'month(QUOTE_DATE) eq {} and year(QUOTE_DATE) eq {}'.format(month, year) 7 } 8 9 # Invoke the backend with query string, fetch # the response as a UTF-8 string 10 # and turn it into an XML object 11 response = out.conn.get(self.cid, query_string)
lxml is a very good Python library for XML processing and is used in the example to issue XPath queries against the complex document returned:
1 xml = etree.fromstring(response) 2 3 # Look up all XML elements needed (date and rate) using XPath 4 elements = xml.xpath('//m:properties/d:*/text()', ↪namespaces=NAMESPACES)
For each element returned by the back-end service, you create an entry
in the Redis cache in the format specified by
REDIS_KEY_PATTERN—for instance,
linuxjournal:rates:2013:09:03 with a value of 1.22:
1 for date, rate in elements: 2 3 # Create a date object out of string 4 date = parse(date) 5 6 # Build a key for Redis and store the data under it 7 key = REDIS_KEY_PATTERN.format( 8 date.year, str(date.month).zfill(2), ↪str(date.day).zfill(2)) 9 self.kvdb.conn.set(key, rate) 10 12 # Leave a trace of our activity 13 self.logger.info('Key %s set to %s', key, rate)
linuxjournal.get-rate
Now that a service for updating the cache is ready, the one to return the data is so simple yet powerful that it can be reproduced in its entirety:
1 class GetRate(Service): 2 """ Returns the real long-term rate for a given date 3 (defaults to today if no date is given). 4 """ 5 class SimpleIO: 6 input_optional = ('year', 'month', 'day') 7 output_optional = ('rate',) 8 9 def handle(self): 10 # Get date needed either from input or current day 11 year, month, day = get_date(self.request.input) 12 13 # Build the key the data is cached under 14 key = REDIS_KEY_PATTERN.format(year, month, day) 15 16 # Assign the result from cache directly to response 17 self.response.payload.rate = self.kvdb.conn.get(key)
A couple points to note:
SimpleIO was used—this is a declarative syntax for expressing simple documents that can be serialized to JSON or XML in the current Zato version, with more to come in future releases.
Nowhere in the service did you have to mention JSON, XML or even HTTP at all. It's all working on a high level of Python objects without specifying any output format or transport method.
This is the Zato way. It promotes reusability, which is valuable because a generic and interesting service, such as returning interest rates, is bound to be desirable in situations that cannot be predicted.
As an author of a service, you are not forced into committing to a particular format. Those are configuration details that can be taken care of through a variety of means, including a GUI that Zato provides. A single service can be exposed simultaneously through multiple access channels each using a different data format, security definition or rate limit independently of any other. | http://www.linuxjournal.com/content/zato%e2%80%94agile-esb-soa-rest-and-cloud-integrations-python?page=0,1 | CC-MAIN-2014-52 | refinedweb | 852 | 51.89 |
Tuesday 24 September 2013
We are very happy to announce the RC3 release of Scala 2.10.3! If no serious blocking issues are found this will become the final 2.10.3 version.
The release is available for download from scala-lang.org or from Maven Central.
The Scala team and contributors fixed 50 issues since 2.10.2!
In total, 63 RC1 pull requests, 19 RC2 pull requests and 2 RC3 pull requests were opened on GitHub of which 72 were merged after having been tested and reviewed.
Known Issues
Before reporting a bug, please have a look at these known issues.
Scala IDE for Eclipse
The Scala IDE with Scala 2.10.3-RC3 built right in is available through the following update-site:
- for Eclipse 4.2/4.3 (Juno/Kepler)
Have a look at the getting started guide for more info.
New features in the 2.10 series
Since 2.10.3 is strictly a bug-fix release, here’s an overview of the most prominent new features and improvements as introduced in 2.10.0:
Value Classes
A class may now extend
AnyValto <- WS.url(restApiUrl).get()) yield (req.json \ "users").as[List[User]](uses play!)
Dynamic and applyDynamic
x.foobecomes
x.applyDynamic("foo")if
x’s type does not define a
foo, but is a subtype of
Dynamic
Dependent method types:
def identity(x: AnyRef): x.type = x// the return type says we return exactly what we got
New ByteCode emitter based on ASM
Can target JDK 1.5, 1.6 and 1.7
Emits 1.6 bytecode by default
Old 1.5 backend is deprecated
A new Pattern Matcher
rewritten from scratch to generate more robust code (no more exponential blow-up!)
code generation and analyses are now independent (the latter can be turned off with
-Xno-patmat-analysis)
Scaladoc Improvements
Implicits (-implicits flag)
Diagrams (-diagrams flag, requires graphviz)
Groups (-groups)
Modularized Language features
Get on top of the advanced Scala features used in your codebase by explicitly importing them.
Fixes in immutable
TreeSet/
TreeMap
Improvements to PartialFunctions
- Addition of
???and
NotImplementedError
- Addition of
IsTraversableOnce+
IsTraversableLiketype classes for extension methods
Deprecations and cleanup
Floating point and octal literal syntax deprecation
Removed scala.dbc
Experimental features. | http://www.scala-lang.org/announcement/2013/09/24/release-notes-v2.10.3-RC3.html | CC-MAIN-2017-47 | refinedweb | 373 | 58.79 |
://
Thanks. Hi Soniya,
We can use oracle too in struts...Hi Hi friends,
must for struts in mysql or not necessary... know it is possible to run struts using oracle10g....please reply me fast its
hi.......
for such a programme... plz help me...
Hi Friend,
Try this:
import java.awt....hi....... i've a project on railway reservation... i need to connect... enter in my frame should reach mysql and should get saved in a database which we've
Hi - Struts
please help me. its very urgent Hi friend,
Some points to be remember...Hi Hi Friends,
I want to installed tomcat5.0 version please help me i already visit ur site then i can't understood that why i installed
Thanks - Java Beginners
Thanks Hi Rajnikant,
Thanks for reply.....
I am not try for previous problem becoz i m busy other work...
please tell me what... and analyze you got good scenario about Interface
Thanks
Rajanikant
Hi
Hi
Hi Hi
How to implement I18N concept in struts 1.3?
Please reply to me
Hi
Hi I want import txt fayl java.please say me...
Hi,
Please clarify your problem!
Thanks
hi
hi how i get answer of my question which is asked by me for few minutes ago.....rply
hi
in my frame should reach mysql and should get saved in a database which we've... for such a programme... plz help me... i hope for a speedy reply from ur side
hi
in my frame should reach mysql and should get saved in a database which we've... for such a programme... plz help me... i hope for a speedy reply from ur side
Hi... - Struts
Hi... Hi,
If i am using hibernet with struts then require... of this installation Hi friend,
Hibernate is Object-Oriented mapping tool... more information,tutorials and examples on Struts with Hibernate visit!!!!!!!!!!!!!!!!!!!!!
HI!!!!!!!!!!!!!!!!!!!!! import java.awt.*;
import java.sql....+"')");
JOptionPane.showMessageDialog(null,"Thanks for creating an account.");
}
catch(Exception e...();
}
}
CAN ANYONE HELP ME TO DESIGN A FRAME FOR THIS PROGRAMME??
thanks - Java Beginners
thanks Sir , i am very glad that u r helping me a lot in all... to understood...can u please check it and tell me..becoz
it says that any two... it to me .....thank you vary much
Hi... - Struts
Hi... Hello,
I want to chat facility in roseindia java expert please tell me the process and when available experts please tell me Firstly you open the browser and type the following url in the address bar
to get radio button value - Struts
me for this.
Thanking you. Hi Friend,
Please post your full code.
Thanks...to get radio button value hello friend, i have a problem regarding>
Thanks - Java Beginners
Thanks Hi,
Thanks ur sending url is correct..And fullfill requirement..
I want this..
I have two master table and form vendor... and send me...
Thanks once again...for sending scjp link
Reply Me - Struts
Reply Me Hi Friends,
Please write the code using form element...because i got error in textbox null value Hi Soniya
Would you... to provide a better solution for you..
Thanks
Thanks - Java Beginners
and send me...
Thanks once again...for sending scjp link Hi friend...Thanks Hi,
Thanks ur sending url is correct..And fullfill... to visit :
Thanks
hi - Hibernate
hi hi all,
I am new to hibernate.
could anyone pls let me know... as possible.
thanks
Hi friend,
Read for more information... {
Userin instance = (Userin) getSession().get("org.Userin", id);
return instance
hi - Ajax
);
xmlObj.open("GET", url)
alert("3");
xmlObj.onreadystatechange = function... me know what kind of error u getting in Ajax.
if possible come online through googletalk
Thanks
Rajanikant
rajanikant.misra@gmail.com
struts
struts i have no any idea about struts.please tell me briefly about struts?**
Hi Friend,
You can learn struts from the given link:
Struts Tutorials
Thanks
Hi .Again me.. - Java Beginners
Hi .Again me.. Hi Friend......
can u pls send me some code......
REsponse me.. Hi friend,
import java.io.*;
import java.awt....://
Thanks. I am sending running
.... explain and details send to me. Hi friend,
Please add struts.jar
Hi.... - Java Beginners
Hi....
I hv 290 data and very large form..In one form.......I am analyse this idea but please send me sample of code this query
Hi friend,
Some points to be member to solve the problem :
When
struts - Struts
and search you get the jar file Hi friend,
struts.config.xml : Struts has...struts hi,
what is meant by struts-config.xml and wht are the tags.../struts/
Thanks
explaination and example? thanks in advance. Hi Friend,
It is not thread...://
Thanks...Struts Is Action class is thread safe in struts? if yes, how
HI.
HI. hi,plz send me the code for me using search button bind the data from data base in dropdownlist
Struts - Struts
Struts Hi All,
Can we have more than one struts-config.xml in a web-application?
If so can u explain me how with an example?
Thanks in Advance.. Yes we can have more than one struts config files..
Here we
hi....
hi.... plzz sent me d code for counting vowels in a string... gui programme
Hi
Hi Hi this is really good example to beginners who is learning struts2.0
thanks
Hi
Hi how to read collection obj from jsp to servlet and from jsp - jsp?
Hi Friend,
Please visit the following link:
Thanks
hi
hi My program is not find java.io.File; why? help me please...++)
System.out.print ("* ");
System.out.println();
}
}
}
Thanks
hi
hi int A[]=new int{2,5,4,1,3};
what is the error in declaration of this
Hi Friend,
This is the wrong way of declaring an array.
Array...
Thanks
thank u
hi!
hi! how can i write aprogram in java by using scanner when asking... to to enter, like(int,double,float,String,....)
thanx for answering....
Hi...);
System.out.println(s);
}
}
Thanks
Hi
Hi The thing is I tried this by seeing this code itself.But I;m facing a problem with the code.Please help me in solving me the issue.
HTTP Status 500 -
type Exception report
description The server encountered
Hi
Hi I have got this code but am not totally understanding what the errors. Could someone Please help. Thanks in advance!
import java.util.Random;
import java.util.Scanner;
private static int nextInt() {
public class
Answer me ASAP, Thanks, very important
Answer me ASAP, Thanks, very important Sir, how to fix this problem in mysql
i have an error of "Too many connections" message from Mysql server,, ASAP please...Thanks in Advance
hi..
hi.. I want upload the image using jsp. When i browse the file...
BRwowse File!!twad9315.jpg
show me
BROWSE FILE...);
System.out.println("show me");
// Pattern
String browsefile.......
hi....... import java.awt.;
import java.sql.*;
import javax.swing....+"',"+t6+",'"+t7+"','"+t8+"')");
JOptionPane.showMessageDialog(null,"Thanks...){
}
}
can anyone tell wts wrong with this code??
Hi,
Check it:
import
hi..
hi.. <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
"">
<%@ page contentType="text/html...);
System.out.println("show me");
// Pattern
String browsefile= browsefile1.replace
hi - Hibernate
hi Hi ,
Thanks for response.
Could anyone tel me what... process.
pls suggest me the apporopriate code using the above method .
pls help me
help - Struts
studying on struts2.0 ,i have a error which can't solve by myself. Please give me...
Hi friend,
Do Some changes in struts.xml... :
Thanks
Hi...doubt on Packages - Java Beginners
..Explain me. Hi friend,
Package javax.mail
The Java Mail API allows...Hi...doubt on Packages Does import.javax.mail.* is already Existing.... In this section we will introduce you with the Java Mail API.
The Java Mail API
Struts Validation - Struts
Struts Validation Hi friends.....will any one guide me to use the struts validator...
Hi Friend,
Please visit the following links:
http - SQL
sir,plz tell me
ThanQ Hi Friend,
Run...hi hi sir,my table is this,
SQL> desc introducer;
Name... that it works.
Thanks
hi - SQL
hi hi sir,i want to create a database in oracle,not in my sql sir,plz tell me how to create a database. Hi Friend,
Try the following...){}
}
}... as validation disabled
plz give me reply as soon as possible. Hi friend..._struts_validator.shtml
Thanks
Reply - Struts
Reply
Thanks For Nice responce Technologies::--JSP
please write the code and send me....when click "add new button" then get the null value...its urgent... Hi
can u explain in details about your project
j2ee - Struts
j2ee How to get values from a database which is stored in form1 in to form2 with in a listbox. pls help me. Hi bhaskar,
You have....
Thanks
Sreenivas
best Struts material - Struts
best Struts material hi ,
I just want to learn basic Struts.Please send me the best link to learn struts concepts
Hi Manju...://
Thanks Hi i am new to struts. I don't know how to create struts please in eclipse
Reply Me - Java Beginners
jsp file and one java file
Thanks
Rajanikant Hi friend,
MVC : M...Reply Me Hi Rajnikant,
I know MVC Architecture but how can use this i don't know...
please tell me what is the use
Advertisements
If you enjoyed this post then why not add us on Google+? Add us to your Circles | http://www.roseindia.net/tutorialhelp/comment/126 | CC-MAIN-2015-40 | refinedweb | 1,553 | 78.65 |
Download | JavaDoc | Source | Forums | Support
Since Camel became to be Apache Top Level Project for a while, and all the request to the camel old web site () will be redirect to the new site (). We updated the camel schemas' target namespace to refect the URL change in Camel 2.0. But we don't change the schemas namespace of Camel 1.x to keep the compatibility.
We recently released Camel 2.0-ml, and the site schema maintenance script always copys the latest released Camel version's schema to the default schema without version number. You may meet the schema validation problem if your application spring configuration is still using the old Camel 1.x schema namespace and using the web site schema for validations.
To walk around this issue, you just need to specify the schema version in your Spring configuration's schemaLocation attribute. | http://camel.apache.org/how-to-validate-the-camel-1x-context-xml-from-apache-camel-web-site.html | CC-MAIN-2014-49 | refinedweb | 145 | 72.97 |
How to get the underlying file name when using get_images() and not get_attachments()?
- halloleooo
A new chapter with my Image Share Extension:
I started using
appex.get_attachmentsinstead of
appex.get_images, because with
get attachmentsI get the path name and thus can check for HEIC images with the same name as JPGs. This is great for the Photos app.
However in the Files app I get via
get_attachmentsduplicates for the same file path!
get_images(image_tyepe='ui')still gives me only single files.
Here some debug output with two files selected in the Files app when calling the Share Extension.
get_attachments list: _Leo.jpg /private/var/mobile/Library/Mobile Documents/com~apple~CloudDocs/Journals/Diary/_assets/Aubergine-menu_MAR2021_Leo.jpg get_images list <_ui.Image object at 0x1162F7240> <_ui.Image object at 0x1162F7550>
So I clearly prefer using
get_images! But then I do not know the name and extension of the underlying file (for the HEIC deduping). - Any way to get this?
Many thanks for pointers in advance!
Running an attachment through the console, you can take it step by step, so based on your argument for get_images(), run through each item via for loop, and access the .name property. Without your argument it would return a PIL object, and you can access it via .filename or .fp.name
Got it (finally, same as @ts first post)
import appex imgs = appex.get_images(image_type='ui') print(len(imgs)) print(dir(imgs[0])) for img in imgs: print(img.name)
- halloleooo
@cvp Yes, I noticed , because I got “AttributeError: '_ui.Image' object has no attribute 'format'”.
I’ll try your new suggestion. Thanks!
@halloleooo yes, sorry, my first (deleted and purged) answer was using format but only valid with pil | https://forum.omz-software.com/topic/7061/how-to-get-the-underlying-file-name-when-using-get_images-and-not-get_attachments/2 | CC-MAIN-2022-21 | refinedweb | 284 | 60.92 |
Silverlight bitmap effects are easy to use but subtle. Before moving on to consider custom effects we take a careful look at what is provided as standard.
Bitmap effects are a strange topic because they mix the simple with the complex.
In Silverlight 4 WPF bitmap effects with a very similar set of characteristics are supported but there is one very big difference. Under WPF bitmap effects are rendered in hardware if possible - they make use of the GPU if one is available. In the case of Silverlight the effects are rendered using software only. This means that they run slower but in many cases they are still fast enough to be useful.
Using the supplied effects is very easy but custom effects rely on the use of "pixel shaders" written using HLSL - High Level Shader Language. Even though the effects are rendered in software, the use of HLSL means you have to pretend that the GPU hardware is going to be used. This makes things more complex than need be. On the other hand if you know how to implement a shader for WPF then you can make use of it under Silverlight with only a loss of performance. In addition, one day the Silverlight designers might work out how to let a web application have access to the GPU without compromising security and then bitmap effects will work just as fast as under WPF.
This article is based on a similar one describing WPF bitmap effects. The major difference between WPF and Silverlight supplied bitmap effects is speed of implementation and there are also some differences concerning how they interact with the other elements of the graphics system.
If you want to know more about creating custom bitmap effects and using HLSL then you will be pleased to know that these aspects are covered in future articles.
If you know the WPF BitmapEffect class you will wonder where it has gone? The first thing to say that this class is obsolete in WPF and it has never existed in Silverlight. The Effect class and classes derived from it are the only way to apply effects in Silverlight. The Silverlight Effect class and the classes derived from it do work in much the same way that they do in WPF and so your knowledge of the classes generalises. However, when it comes to custom effects, Silverlight only supports Pixel Shader 2.0 and this makes it less clear cut that WPF custom shaders, which support Pixel Shader 3.0, will work under Silverlight.
However all of this said it is worth knowing that a Silverlight out of browser application does make use of the GPU. Out of browser applications is a topic we will return to soon.
There also is no support for the WPF RenderCapability class. As this is used to tell if an effect will be rendered using software or hardware it is clearly not needed in Silverlight, which always uses software rendering.
To see an effect in action let's try the BlurEffect.
An effect can be assigned to any object that has an Effect property.
For example, if we start a new Silverlight project and use the Add,Existing item to add a JPEG image to the project then we can load into a BitmapImage.
The bitmap file test.jpg is ready to use.
To make sure that the bitmap is packaged into the XAP file and therefore accessible to the Silverlight application when running on a remote machine you have to set the file's properties Build Action to Content and Copy to Output Directory to Do not copy.
Make sure that the bitmap file is set to Resource
We also need to add to the start of the file:
using System.Windows.Media.Effects;using System.Windows.Media.Imaging;
Now that we have the bitmap file packaged in the assembly DLL we can load it into a BitmapImage object:
Uri uri = new Uri(@"/test.jpg", UriKind.Relative);BitmapImage BMI = new BitmapImage();BMI.UriSource = uri;
Next229799 >
<ASIN:1430224258 >
<ASIN:0470650923 >
<ASIN:1847199844 >
<ASIN:143022455X > | https://www.i-programmer.info/programming/silverlight/884-silverlight-bitmap-effects.html | CC-MAIN-2019-22 | refinedweb | 680 | 69.82 |
Hi i have been trying to compile a simple C program
and i seem to have 2 errors, i tried to use the debugger
but it didnt help if you can see the problem please tell me
heres my program:
#include "stdafx.h"
int main(int argc, char* argv[])
{
printf("Integers of 2 up 2^n!\n");
return 0;
}
#include "stdafx.h"
#define N 1600
int main(void) {
int n; /* The current exponent */
int val = 1; /* The current power of 2 */
printf("\t n \t 2^n\n");
printf("\t====================\n");
for (n=0; n<=N; n++) {
printf("\t%3d \t %600d\n", n, val);
val = 2*val;
}
return 0;
}
this program is meant to have a title "integers of 2"
and then list the integers of it until 2^n.
BTW i tried to compile this program in not only Emacs but also Microsoft VC++ 6.0 Pro which i got for about $0.60 american dollars from a romanian gentleman, no luck thier way!
thanks for your help. | http://www.linuxforums.org/forum/programming-scripting/need-help-print-2217.html | CC-MAIN-2017-51 | refinedweb | 170 | 77.98 |
1. Caching set cookies
Caching an object with a Set-Cookie header can have devastating effects, as any client requesting the object will get that same cookie set. This can potentially lead to a session transfer. In general we recommend avoiding the use of return (deliver) in vcl_fetch, to stay safe against this. If you really do need a return (deliver), be careful and check for the presence of Set-Cookie first. By default, Varnish will of course not cache responses with this header set.
2. Varying on User-Agent
Many content management systems will issue a “Vary: User-Agent”. This will more or less render the cache useless as finding two users with the exact same user-agent string is pretty hard. Normalize the string and your cache hit rates will increase dramatically.
3. Setting high TTLs without a proper plan
The higher the TTL the better the speed of the website and the better the user experience. Setting a high TTL also reduces the load on the backend which can save you a lot of money. However, if you plan on setting a high TTL you should also have a way to invalidate the contents in the cache as the content changes (such as Varnish Enhanced Cache Invalidation).
4. Believing everything you read about Varnish Cache online
There are a lot of tuning tips out there, both for Varnish Cache and Linux kernel itself. We’ve seen multiple installations with more or less random settings that we’ve traced back to blog posts where people have been testing various settings. For instance, there are options that you can enable that will work very well on a local area network but will break when clients are accessing the website across the internet. Read the documentation and be careful changing settings without understanding the implications. Get yourself up to speed by downloading The Varnish Book.
5. Failure to monitor Varnish Cache’s ‘Nuked Counter’
Monitoring varnishstat’s n_lru_nuked counter will tell us how many times Varnish Cache had to forcefully evict an object from the cache in order to fit new objects. Monitoring this counter will let you know if your cache is starved for storage. If you see an elevated value here, it means your working set does not fit in the configured storage and you will benefit from adding more space.
6. Not using custom error messages
In the case that the origin server has fallen over, and Varnish Cache finds it does not have a suitable candidate object to serve to the client, Varnish Cache will respond with the dreaded “Guru Meditation” error response. We recommend you customize this error message (this can be done by editing the response in the VCL subroutine vcl_error) to be more in line with the look and feel of your website. You can have a look at for some inspiration. You can even embed images inline in the HTML markup.
7. Messing with accept-encoding
Varnish 3.0 and later has native support for gzip. This means that there is no longer a need to manually mangle the Accept-Encoding request header in order to cache both compressed and uncompressed versions of responses. Varnish 3.0 and later will handle this automatically, by uncompressing content on the fly when needed. This isn’t the deepest pitfall out there, but maintaining a short, sweet and well readable VCL is always a good thing.
8. Failure to understand hit-for-pass
Hit-for-pass is not an intuitive concept. Many users fail to understand how it works and may misconfigure Varnish Cache. Varnish Cache will coalesce multiple requests into one backend request. If that response then does something funny, like doing a Set-Cookie, Varnish Cache will create a hit-for-pass object in order to remember that requests to this URL should not be put on the waiting list and simply sent straight to the backend. The default TTL for these objects is 120 seconds. If you set the TTL for hit-for-pass reponses to 0 you’ll force serialized access to that URL. With some traffic on that URL access will be slow as molasses.
9. Misconfiguration of memory
If you give Varnish Cache too much memory you run the risk of running out of memory. This might be painful, especially on Linux which has some issues with paging performance. In addition many users fail to realize there is a per-object memory overhead.
On the other hand giving Varnish too little memory will most likely result in a very low cache hit rate, giving your users a bad user experience. When we onboard new customers this is one of the things we pay a lot of attention to as the consequences of running out of memory are pretty dire.
10. Failure to monitor sys log
Varnish Cache runs as two separate processes; the management process and the child process. The management process is responsible for keeping the child running, and various other tasks, while the child process does the actual heavy lifting. In the event of a crash, the child process will automatically be started back up again—often so quickly that the downtime is not noticeable.
We recommend monitoring syslog in order to catch these events. A different possibility is to pay attention to varnishstat’s uptime counter—if that resets to 0, it means the child process has been restarted. In Varnish 4.0 the management process has its own counters (MGT.*) that can also be used to monitor child restarts.
Still learning about Varnish Cache and how to use it optimally? Download the Varnish Book to learn more tips and tricks.
Ready for the single, most useful and informative hour of your Varnish life?
Register to our on demand webinar to watch Varnish Technical Evangelist Thijs Feryn and Senior Engineer Arianna Aondio for a run down of the 'Top 10 Dos and Don’ts' when it comes to caching with Varnish. | https://info.varnish-software.com/blog/10-varnish-cache-mistakes-and-how-avoid-them | CC-MAIN-2021-31 | refinedweb | 995 | 61.06 |
Overview
IBM InfoSphere Streams is high-performance real-time event processing middleware. Its unique strength lies in its ability to ingest structured and unstructured data from a variety of data sources for performing real-time analytics. It does this through a combination of an easy-to-use application development language called Streams Processing Language (SPL) and a distributed runtime platform. This middleware also provides a flexible application development framework to integrate code written in C++ and Java into Streams applications. In addition to C++ and Java, many developers who build real-world IT assets also use dynamic programming languages. With its strength in system integration capabilities, Python is a viable option for many companies to quickly build solutions. For those with existing assets written in Python, there is a way to integrate Python code inside Streams applications. This article explains the details about doing that through a simple Streams application example.
This article assumes familiarity with InfoSphere Streams and its SPL programming model. Working knowledge in C++ and Python is also needed to understand the programming techniques. For in-depth details about InfoSphere Streams and Python, see Resources.
InfoSphere Streams is a key component in IBM's big data platform strategy. Many of IBM's current and prospective customers with Python assets and skills can take advantage by mixing it with InfoSphere Streams. This article is targeted at readers whose technical focus is big data applications, including application designers, developers, and architects.
Example scenario
In order to explain the nitty-gritty technical details involved in calling Python code from a Streams application, we will stick to a simple example. This scenario involves reading the names of a few web addresses from an input CSV file and calling a simple user-written Python function that will return the following details as its result. We will then write the result for each web address into a separate output CSV file:
- Primary hostname of the URL
- List of alternate hostnames for the URL
- List of IP addresses for the URL
- Company name specified in the URL string
Prerequisites
Code snippets used below explain the implementation details for the scenario explained above. This example code can also be downloaded so you can run it on your own IBM InfoSphere Streams installation. The example code was tested in the following environment:
- RedHat Enterprise Linux 6.1 or above (or an equivalent CentOS version)
- gcc version 4.4.5 20110214 (Red Hat 4.4.5-6) (GCC)
- Python 2.6.6 (r266:84292, Apr 11 2011, 15:50:32, shipped with RHEL6)
- /usr/lib/libpython2.6.so
- /usr/include/python2.6 directory with Python.h and other include files
- IBM InfoSphere Streams 3.x configured with a working Streams instance
The same techniques could work in slightly different environments (e.g., RHEL 5.8 and Streams 2.0.0.4) with some tweaks to the code or environment setup.
High-level application components
In our simple example scenario, there are three major components. Each component is independent enough to be in its own project because of the natural separation by the programming language used in each of them:
- UrlToIpAddress Python script
- StreamsToPythonLib C++ project
- streams-to-python SPL project
UrlToIpAddress is a Python script with simple logic that uses Python APIs to get IP address and hostname information for a given web address. This script can be tested independently using the Python interpreter. This tiny script plays a major part in this article in demonstrating how to call functions in a Python script from a Streams application.
StreamsToPythonLib is a C++ project. Inside of it, source code for the SPL native function logic is included. Primarily, source code here uses the Python/C API to embed Python code during the execution of C++ code. Embedding Python in C++ code is well described as part of the Python documentation. This project contains a Wrapper include (.h) file, which is an important one and this file provides an entry point for a Streams SPL application to call any C++ class method. All the C++ logic in this project will be compiled into a shared object library (.so) file and made available to the SPL application.
streams-to-python is a Streams SPL project. Inside of it, we
have a basic SPL flow graph to make a call chain
(
SPL<-->C++<-->Python). This SPL code reads URLs from an input
file in the data directory, calls the C++ native function to execute the
Python code, receives the results, and writes it to an output file in the
data directory. Inside the SPL project directory, a native function model
XML file outlines the meta information needed to directly call a C++ class
method from SPL. This detail covers the C++ wrapper include file name, C++
namespace containing the wrapper functions, C++ wrapper function prototype
expressed using SPL syntax/types, name of the shared object library
created from the C++ project, location of the shared object library,
location of the wrapper include file, etc.
In the following sections, we will dive deep into each of these three application components and explain the Python, C++, and SPL code in a detailed manner.
Python logic
Listing 1 shows the Python code. This is the business logic we want to call from Streams.
Listing 1. UrlToIpAddress.py
import re, sys, socket def getCompanyNameFromUrl(url): # Do a regex match to get just the company/business part in the URL. # Example: In "", it will return "ibm". escapedUrl = re.escape(url) m = re.match(r'www\.(.*)\..{3}', url) x = m.group(1) return (x) def getIpAddressFromUrl(url): # The following python API will return a triple # (hostname, aliaslist, ipaddrlist) # hostname is the primary host name for the given URL # aliaslist is a (possibly empty) list of alternative host names for the same URL # ipaddrlist is a list of IPv4 addresses for the same interface on the same host # # aliaslist and ipaddrlist may have multiple values separated by # comma. We will remove such comma characters in those two lists. # Then, return back to the caller with the three comma separated # fields inside a string. This can be done using the Python # list comprehension. return(",".join([str(i).replace(",", "") for i in socket.gethostbyname_ex(url)])) if ((__name__ == "__main__") and (len(sys.argv) >= 2)): url = sys.argv[1] # print("url=%s" % (url, )) print "IP address of %s=%s" % (url, getIpAddressFromUrl(url)) print "Company name in the URL=%s" % repr(getCompanyNameFromUrl(url)) elif ((__name__ == "__main__") and (len(sys.argv) < 2)): sys.exit("Usage: python UrlToIpAddress.py")
It is evident from Listing 1 that the Python code
is deliberately kept simple for clarity. This has two Python functions
followed by a code snippet that will run when the Python script is
executed using a Python interpreter. To verify that the code works as
expected, this script can be run from a shell window:
python UrlToIpAddress.py.
At the top of the file, Python modules, such as regular expression and
socket, are imported. The first function is
getCompanyNameFromUrl, which
takes a web address as input. It does a regular expression match to parse
the company name from the web address and returns the company name to the
caller. Next function is
getIpAddressFromURL. It also takes a web address
as input. It calls a Python socket API to get the IP address of the given
web address. In particular, this Python API (gethostbyname) returns a
tuple with three elements in it. These three elements provide hostname of
the server for the given web address, alternate hostnames if any, and one
or more IP addresses for that server. Instead of returning the tuple type
to the caller, this function flattens the three tuple elements into a
Python string by inserting a comma after each element. Then it returns
the result as a string to the caller.
The purpose of this example is to learn about calling those two Python script functions from within a Streams application. We will focus on that in the following sections.
C++ logic
InfoSphere Streams allows for the inclusion of code written in C++ in two ways. One way is to build primitive Streams operators in C++, thereby incorporating the business logic written in C++. The other option is to execute any arbitrary C++ class methods directly from SPL as native functions. In this exercise, we will use the native function approach. To do that, we will create a separate C++ project named StreamsToPythonLib, in which we will write the necessary code to call the Python functions we covered in the previous section. Then we will create a shared object (.so) library to make this C++ code available to the Streams SPL application.
Table 1 shows the contents of the StreamToPythonLib C++ project directory.
Table 1. StreamsToPythonLib C++ project directory
Listing 2. StreamsToPython.h
#ifndef STREAMS_TO_PYTHON_H_ #define STREAMS_TO_PYTHON_H_ using namespace std; // To avoid a redefinition compiler error, undefine the following. #undef _POSIX_C_SOURCE #undef _XOPEN_SOURCE // This should be the first include file (according to Python documentation) #include "Python.h" // Include files that defines SPL types and functions. #include "SPL/Runtime/Function/SPLFunctions.h" #include <SPL/Runtime/Utility/Mutex.h> #include <SPL/Runtime/Type/ValueHandle.h> // Include standard C++ include files. #include <sys/time.h> #include <pthread.h> #include <unistd.h> #include <stdlib.h> #include <sstream> // This will allow us to access the types and functions from SPL namespace. using namespace SPL; // Your #define constant definitions go here. // Class definition follows here. namespace calling_python_from_streams { class GlobalStreamsToPythonSession { private: // This member variable tells us if a global // streams to Python caller handle already // exists for a given PE/process. boolean streamsToPythonHandleExists; // Following member variables are required for // calling Python C APIs. static boolean pyInitialized; static boolean importFailed; PyObject* pFunc1; PyObject* pFunc2; public: GlobalStreamsToPythonSession(); virtual ~GlobalStreamsToPythonSession(); // This method establishes StreamsToPython handle for a given PE/process. int32 initializeStreamsToPython(); // This method gets the IP address of a given URL. boolean getIpAddressFromUrl(rstring const & url, rstring & primaryHostName, rstring & alternateHostNames, rstring & ipAddressList, rstring & companyName); // Get the global (Singleton) Streams to Python session object. static GlobalStreamsToPythonSession & getGlobalStreamsToPythonSession(); }; } #endif /* STREAMS_TO_PYTHON_H_ */
Listing 2 shows that it is a C++ interface class. It starts off with an inclusion of Python.h, which is a must for our task of calling into the native Python code. It includes standard library header files along with SPL include files. It is important to note that by including SPL header files and by using the SPL namespace, we can access SPL data types inside C++. Many of the primitive and collection data types in SPL are representations of equivalent C++ built-in data types. Inside the namespace and class sections, member variables and member methods are declared. There are a few Python object related member variables, which we will cover later. There are prototypes declared for class constructor, destructor, and the business logic method that will get called from SPL. In the end, there is a static method, getGlobalStreamsToPythonSession, that provides a singleton access to this C++ class from the SPL code. We will see more details about all of this shortly.
Listing 3. StreamsToPython.cpp
#include "StreamsToPython.h" #include <dlfcn.h> namespace calling_python_from_streams { // Initialize the static member variables in this class. boolean GlobalStreamsToPythonSession::pyInitialized = false; boolean GlobalStreamsToPythonSession::importFailed = false; ...
See Listing 3 (StreamsToPython.cpp) in full.
Listing 3 is the implementation class. It opens with include statements for the corresponding interface class and the dynamic library loader. Python allows extension and embedding. Inside the Python code, one can extend it to invoke C functions. Similarly, inside C++ code, one can embed Python code. The main focus in Listing 3 is to use Python/C API to invoke native Python code. Our implementation class has five C++ methods. The following commentary takes a deeper look into each of those methods.
Constructor:— The following three major tasks are done when this class is instantiated:
- Set the Python path to the current directory.
- Initialize the Python interpreter as it must be done before using any Python/C API functions.
- Load the libPython shared library dynamically into our process space. Even though the Python shared library would get loaded automatically by the dynamic loader, we have to load it via dlopen so our Python script can link properly with other Python modules implemented as shared object libraries.
Destructor: The following cleanup activities are done when this class object goes out of scope:
- Reset the member variable that holds the handle needed for singleton class access.
- Clear the handles obtained for both of our Python functions.
getGlobalStreamsToPythonSession: It can be seen from Listing 2 that this method is declared as a static method. This is the entry point into this class when a Streams native function is called. Since we want to have only one instance of this C++ class per Streams processing element (PE), it is necessary to maintain a singleton object of this C++ class. Hence, when this particular method is called, a static object of this class is created and returned to the caller. That is how a Streams application can get a static handle to a C++ object and can arbitrarily call any C++ class method using the static handle.
initializeStreamsToPython: Since we maintain a singleton object of this C++ class per process, it is possible for this class to maintain state variables that can be used and shared across multiple invocations of the methods here. Even though this particular application doesn't store state, this is an important design aspect to keep in mind. A Streams application that uses C++ native functions can use such a method to initialize the state variables. Opening a database connection and storing the connection handle for subsequent database access is a good use of this approach. The application described in this article simply ensures that only the very first call made to this C++ method initializes the global handle indicating that the singleton object of this class has been created.
getIpAddressFromUrl: This is a much longer method in this C++ class
and it contains the business logic necessary to call a Python function and
fetch return values. The Python framework provides a
comprehensive set of C APIs to embed Python code within a C or C++
application. Having initialized the Python interpreter in the constructor
method using Py_Initialize, we can use the other Python/C APIs in this
method. Callers of this method will pass a web address as a method
argument (e.g.,; note that the
http:// part of a URL should
not be included). This method also accepts four other string references as
arguments, in which the result will be returned to the caller. Since we
are using the SPL namespace in this C++ class, we are allowed to access
SPL data types such as rstring, uint32, list etc. Many of the SPL data
types are derived from C++ data types such as std::string, int, vector,
etc.
The very first task in this C++ method is to get valid pointers to the two native Python functions we want to call. When this method is called for the first time, we want to get pointers to the two Python functions and store them in the member variables pFunc1 and pFunc2. That will allow us to reuse them in subsequent calls. In order to get pointers to the Python functions, we must first import the Python module containing those two functions. A Python module in this case is nothing but the filename of the Python script minus the .py extension. We have to use PyString_FromString to get a Python string object from a C++ string object holding the Python module name. Then, a call to PyImport_Import will get a handle to our Python module. On an error from any of the Python/C APIs, we will set a member variable called importFailed and return from this method. Subsequent calls to this C++ method will proceed only if importing the Python module succeeded earlier. Such Python/C API errors can be detected and logged using PyErr_Occurred and PyErr_Print APIs. It is also time now to introduce SPLAPPTRC, which is an SPL C++ Macro API that allows us to log application debug or trace information into the Streams logging system. It takes three arguments: log level, a C++ string object containing the log message, and an aspect that can be used for application-specific log filtering.
Having imported our UrlToIpAddress Python module, we now check that
the functions we want to call really exist inside the Python module using
PyObject_HasAttrString API by passing the Python function name. After
validating the availability of the Python functions inside the Python
module, we can get a pointer to that function by using the
PyObject_GetAttrString API. Once we have a valid pointer to the Python
function, it is necessary to check if that is indeed callable by using the
PyCallable_Check API. After performing these steps successfully, our two
C++ member variables (pFunc1 and pFunc2) will point to valid and callable
user written Python functions. Now, we can call PyObject_CallFunction API
to execute the function by passing the pFunc1 or pFunc2 member variable
along with a list of expected function arguments. In our case, we pass a
string (web address) as an argument to the Python function. Hence, the
second argument is
s to indicate that the
argument is in string format,
and the third argument is the actual web address represented as a regular
C string. Since both of our functions return string as a result, we use
the PyString_AsString API to convert the returned Python string object to
a regular C string. We store the result strings from both the Python
functions into our own rstring local variables. As explained in Example scenario, our first Python
function returns the result as a string with three comma-separated parts.
To parse the CSV fields, we can call standard SPL toolkit function
named csvTokenize and assign the returned values directly to the C++
method argument references passed by the caller. That is what is involved
in calling Python functions from C++.
In this C++ implementation class, there are two other important things to highlight. When we used the PyImport_Import API to import our UrlToIpAddress.py module, how does it know the physical location of the Python script file? If we can refer back to the C++ constructor, there is a call made to a standard POSIX API that sets the PYTHONPATH environment variable to current directory via a period character. That is a key reason why the PyImport_Import API is able to locate the Python script and import it. In a Streams application, current working directory is always set to the /data subdirectory available within the SPL project directory. Hence, it is a must that our Python script is copied to the /data subdirectory. Otherwise, PyImport_Import API will not be able to locate and import our Python script. Another important thing to notice in this C++ implementation class is the liberal use of Py_DECREF API. All Python objects have a reference count that counts how many places there are that have a reference to an object. When a reference count becomes zero, that object is deallocated. In Python, reference counts are always manipulated explicitly. Hence, in our code, whenever we no longer need a valid Python object, we must make a call to the Py_DECREF API.
Listing 4. StreamsToPythonWrappers.h
#ifndef STREAMS_TO_PYTHON_WRAPPERS_H_ #define STREAMS_TO_PYTHON_WRAPPERS_H_ // Include the file that contains the class definition. #include "StreamsToPython.h" namespace calling_python_from_streams { // Establish a handle to the StreamsToPython to be // accessed within a PE. inline int32 initializeStreamsToPython(void) { return GlobalStreamsToPythonSession:: getGlobalStreamsToPythonSession().initializeStreamsToPython(); } // Get the IP address of a given URL. inline boolean getIpAddressFromUrl(rstring const & url, rstring & primaryHostName, rstring & alternateHostNames, rstring & ipAddressList, rstring & companyName) { return GlobalStreamsToPythonSession:: getGlobalStreamsToPythonSession(). getIpAddressFromUrl(url, primaryHostName, alternateHostNames, ipAddressList, companyName); } } #endif /* STREAMS_TO_PYTHON_WRAPPERS_H_ */
Listing 4 is a Streams-specific extension file in
the StreamsToPtyhonLib C++ project. As discussed earlier, in order for a
Streams application to be able to call any method
in a C++ class, we need to do something extra. And that extra work is done
in this wrapper include file that contains inline functions. This file
begins by including the C++ class interface file that we saw in Listing 2. These wrapper functions are defined
within the same namespace scope as our actual C++ class in the
StreamsToPythonLib project. A Streams application can call any of the
inline functions specified in this wrapper include file. Every inline
function gets a singleton object of the intended C++ class by calling the
static method
getGlobalStreamsToPythonSession.
The first call made to
this static method does a static instantiation of the C++ class. That
static object reference is returned every time this static method is
called. By getting a reference to the singleton object, a given inline
wrapper function can now call any C++ method available in that object and
pass any return values back to the Streams SPL application. This technique
will come in handy in your other real-world Streams projects.
SPL logic
After learning about the Python and the C++ components used in this example, it is time for a basic Streams application that can tie everything together. We will write a short and sweet SPL application to have a flow graph with three Streams operators available readily in the SPL standard toolkit.
Table 2 shows the contents of the streams-to-python SPL project directory.
Table 2. streams-to-python SPL project directory
Listing 5. streams_to_python.spl
namespace python.wrapper.example; composite streams_to_python { // Define input and output schema for this application. type InputSchema = tuple<rstring url>; OutputSchema = tuple<rstring url, rstring primaryHostName, rstring alternateHostNames, rstring ipAddressList, rstring companyName>; graph // Read from an input file all the URLs for which we need to // get the corresponding IP addresses. stream<InputSchema> UrlInput = FileSource() { param file: "UrlInput.csv"; initDelay: 4.0; } // In the custom operator below, we will call python code to get the // primary host name, alternative host names, and IP addresses. stream<OutputSchema> IpAddressOfUrl = Custom(UrlInput) { logic onTuple UrlInput: { mutable rstring _primaryHostName = ""; mutable rstring _alternateHostNames = ""; mutable rstring _ipAddressList = ""; mutable rstring _companyName = ""; // Call the C++ native function that in turn will call Python functions. boolean result = getIpAddressFromUrl(UrlInput.url, _primaryHostName, _alternateHostNames, _ipAddressList, _companyName); if (result == true) { mutable OutputSchema _oTuple = {}; _oTuple.url = UrlInput.url; _oTuple.primaryHostName = _primaryHostName; _oTuple.alternateHostNames = _alternateHostNames; _oTuple.ipAddressList = _ipAddressList; _oTuple.companyName = _companyName; submit(_oTuple, IpAddressOfUrl); } } } // Write the results to a file using FileSink. () as FileWriter1 = FileSink(IpAddressOfUrl) { param file: "UrlToIpAddress-Result.csv"; } }
Listing 5 is the SPL flow graph that starts by
defining a namespace. It is followed by a definition of an SPL main
composite. In the types section, two tuple data types are defined for
input and output of this application. Then, a basic graph clause is
filled with three Streams operators available in the SPL standard toolkit.
The first operator is a FileSource, which reads the rows from an
input CSV file from the default location (the data subdirectory of the SPL
project). Tuples emitted by the FileSource operator are consumed by a
Custom operator, which calls an SPL native function (
getIpAddressFromUrl) written in C++. As we saw, that C++ code in turn
executes the Python functions to return the results for a given web
address. Those result values are assigned to an output tuple and submitted
from the Custom operator. Finally, a FileSink operator consumes the output
tuples from the Custom operator and writes the results to an output CSV
file. It is important to note that the C++ native function code is
compiled into a shared object (.so) library as explained below.
Function model
Listing 6. function.xml
<?xml version="1.0" encoding="UTF-8"?> <functionModel xmlns: <functionSet> <headerFileName>StreamsToPythonWrappers.h</headerFileName> <cppNamespaceName>calling_python_from_streams</cppNamespaceName> <functions> <function> <description>Initialize the Streams to Python module</description> <prototype>public int32 initializeStreamsToPython()</prototype> </function> <function> <description>Get the IP addresses for a given URL</description> <prototype>public boolean getIpAddressFromUrl(rstring url, mutable rstring primaryHostName, mutable rstring alternateHostNames, mutable rstring ipAddressList, mutable rstring companyName)</prototype> </function> </functions> <dependencies> <library> <cmn:description>Streams to Python Shared Library</cmn:description> <cmn:managedLibrary> <cmn:lib>StreamsToPythonLib</cmn:lib> <cmn:libPath>../../impl/lib</cmn:libPath> <cmn:includePath>../../impl/include</cmn:includePath> <cmn:command>../../impl/bin/archLevel</cmn:command> </cmn:managedLibrary> </library> <library> <cmn:description/> <cmn:managedLibrary> <cmn:lib>python2.6</cmn:lib> <cmn:libPath>/usr/lib64</cmn:libPath> <cmn:includePath>/usr/include/python2.6</cmn:includePath> </cmn:managedLibrary> </library> </dependencies> </functionSet> </functionModel>
Listing 6 is the native function model XML file. In Listing 5 for the SPL code, we saw a C++ native function being called inside the Custom operator. How does SPL code find out about the location of that C++ code? The native function model XML file is the glue between the SPL code and the C++ code. When compiling the SPL code, the Streams compiler resolves the C++ function name through the information we provide in this XML file. At the start of this XML file, we specify the name of the C++ wrapper include file that contains the inline native functions covered in Listing 4. Then we indicate the C++ namespace in which the inline C++ native functions are defined. That is followed by an XML segment, where we declare the prototype for the C++ native functions. It is important to note that the prototype declarations are specified using the SPL types that correspond to the C++ data types. If a C++ native function expects a function argument to be passed as a reference, that function argument should be declared mutable in the function prototype. If the C++ native function logic is made available via a shared object (.so) file, a library XML segment should be included. In that, we have to specify the library name. (The first three letters of a Linux library are typically 'lib' and those three letters should be omitted, while specifying the library name. Similarly, the .so extension is also not required.) The location of the .so file and the location of the include file for the shared library must be specified. It is good practice to bundle the shared object library file and its include files as part of the SPL project directory so that it is easy to ship them across different Streams installations.
As shown above in Table 2, the SPL project directory has the impl/lib and impl/include subdirectories suitable for this purpose. In the native function model XML file, it is indicated as ../../impl/lib and ../../impl/include (../../ is a relative path to the impl directory that can be resolved from the location of the function model XML file). If your application is supported on multiple versions of Linux® and on 32- and 64-bit CPUs, it is necessary to provide different versions of the libraries in separate directories. To make it easier to automate this, this example uses a shell script (../../impl/bin/archLevel) that will automatically select the correct library location based on the Linux version and CPU (32-bit vs. 64-bit). If you read the archLevel script, you will understand how that is done. Finally, we have a library section to indicate our dependency on libpython2.6.so by specifying its name, the location of the library, and its include files.
Building the example
This article includes the full source code for the example discussed here (see Downloads). A given Streams application can be compiled in two modes (stand-alone and distributed). In stand-alone mode, the entire SPL main composite is compiled into a single Linux executable. In the distributed mode, the SPL main composite is compiled as distributed components configured to run on one or more machines. If you have a test environment that meets the prerequisites, you can follow the instructions below to build the example:
- Obtain the streams-to-python.zip file (see Downloads).
- Unzip the file to your home directory on your Linux machine that has Streams installed.
- Change directory to ~/workspace1/StreamsToPythonLib C++ project directory.
- You are going to create the .so shared library by running the ./mk script.
- The previous command creates and copies the .so file to ../../impl/lib/x86_64.RHEL6 directory and copies the include files to ../../impl/include.
- Change directory to ~/workspace1/streams-to-python SPL project directory.
- Create a stand-alone mode application by running the ./build-standalone.sh script.
- Create a distributed mode application by running the ./build-distributed.sh script.
- You should now see ~/workspace1/streams-to-python/output directory with stand-alone and distributed executables.
Running the example
A great feature in Streams allowed us to build both stand-alone and distributed applications without making changes to the source code. We can now run both as described below.
Stand-alone: This kind of Streams application is a single Linux executable that can be run without the need to start and stop a Streams runtime instance:
- Change directory to ~/workspace1/streams-to-python SPL project directory.
- Run the ./run-standalone.sh script.
Distributed: This kind of Streams application contains the Streams operators specified in the SPL flow graph compiled into many Processing Elements (PEs). These processing elements are distributed as individual Linux processes to make use of multiple CPU cores and a cluster of machines. In order to run a distributed mode Streams application, it is required to start a Streams instance, submit the application as a job on that Streams instance, collect the results, and stop the Streams instance:
- Ensure that you have already created a Streams instance.
- Change directory to ~/workspace1/streams-to-python SPL project directory.
- Run this script with a command-line argument:
./run-distributed.sh -i YOUR_STREAMS_INSTANCE_NAME.
- You should give your Streams instance name as an argument to the script in the previous step.
- Since it is a very simple application, it will finish quickly. Wait 60 seconds.
- You can stop the Streams instance now by running this script:
./stop-streams-instance.sh -i YOUR_STREAMS_INSTANCE_NAME.
Verifying the results: Regardless of whether you ran a stand-alone or a distributed application, our SPL program logic reads the web addresses from an input CSV file (data/UrlInput.csv) one line at a time. It calls the C++ native function to get network details about a given web address and writes the results into an output CSV file (data/UrlToIpAddress-Result.csv). Following are the web addresses already stored in the input CSV file of this example:
-
-
-
-
-
-
If the stand-alone/distributed application worked correctly, you should see the results in the data/UrlToIpAddress-Result.csv file. Your results should look similar to the ones packaged with this example from a test run made at the time of this writing (data/Expected-UrlToIpAddress-Result-Feb2013.csv). Expected results are as shown below. Result for a given web address contains five comma-separated fields with this format: WebAddress, PrimaryHostName, AlternateHostNames, IPAddresses, CompanyName.
"","www-int.ibm.com.cs186.net","['']","['129.42.58.158']","ibm"
"","www-v6.stanford.edu","['']","['171.67.215.200']","stanford"
"","cnn-lax-tmp.gslb.vgtf.net","['' '']","['157.166.240.11' '157.166.240.13' '157.166.241.10' '157.166.241.11']","cnn"
"","e1630.c.akamaiedge.net","['' '']","['72.247.70.198']","ieee"
"","star.c10r.facebook.com","['']","['66.220.158.27']","facebook"
"","ds-any-fp3-real.wa1.b.yahoo.com","['' 'fd-fp3.wg1.b.yahoo.com' 'ds-fp3.wg1.b.yahoo.com' 'ds-any-fp3-lfb.wa1.b.yahoo.com']","['98.139.183.24']","yahoo"
Conclusion
Python has evolved nicely over the past two decades. As a dynamic programming language, it has an avid group of followers from academia to world-renowned companies. Its ease of use and programmer productivity gains are often cited as the key reasons for its success among the other top languages, such as C++, PHP, and the Java programming language.
IBM InfoSphere Streams is a market-leading event-processing platform that offers superior capabilities for big data analytics. It packs a powerful, flexible, and extensible programming model via its Streams Processing Language (SPL) supporting out-of-the-box integration features with business logic written in C++ and the Java language.
This article focused on bringing together the best capabilities of two worlds (SPL and Python). It summed up a way for you to seamlessly mix analytics code written in Python in the Streams applications to take advantage of its unparalleled features in scaling and distributed processing. In addition to educating you about Streams+Python integration, this article introduced the mechanisms involved in calling any arbitrary methods in a C++ class directly from the SPL code.
In summary, we covered how to make a round-trip call chain between three languages (SPL<-->C++<-->Python). This article proved the concepts with fully working sample code (see Downloads). You can use the example to run as a stand-alone Linux application or as a distributed Streams application.
Download
Resources
Learn
- Learn more about IBM InfoSphere Streams.
- Manage and analyze massive volumes of structured and unstructured data at rest with IBM InfoSphere BigInsights, IBM's mature Hadoop distribution for big data analytics.
- Speed up your SPL learning through beginner examples.
- View the complete Streams product documentation.
- Read the IBM Redbooks® publication titled "IBM InfoSphere Streams: Assembling Continuous Insight in the Information Revolution."
- Learn more about IBM big data analytics.
- Visit Python.org.
- Learn more about the Python/C API.
- Learn more about Embedding Python in another application.
- Explore Python on developerWorks.
- Learn more about big data in the developerWorks big data content area. Find technical documentation, how-to articles, education, downloads, product information, and more.
- Stay current with developerWorks technical events and webcasts.
- Follow developerWorks on Twitter.
Get products and technologies
-
- Participate in the discussion forum.
- Check out new things exchanged by the Streams developer community.
- Visit the SPL blog.
-. | http://www.ibm.com/developerworks/library/bd-pythonstreams/index.html | CC-MAIN-2015-35 | refinedweb | 5,596 | 55.44 |
referring to Burkhard Hassel post this is his program]
Also remove the lineX: label as you are only allowed to label loops.
Change the > to <.
Change i++ to i-- and change list.add(i); to list.add(new Integer(i)); int != java.lang.Integer
Saikrishna Cinux wrote: quote:referring to Burkhard Hassel post this is his program] No, you changed it a bit, see here. And, I also posted in the original thread: "I know, what's wrong with this code snippet. So you don't need to post it here!" My intention was something else. Fortunately, Harshad Khasnis just wrote it (also on the original thread).
Originally posted by saikrishna cinux: referring to Burkhard Hassel post this is his program
import java.util.*;
public class Test{
public static void main(Stringargs[]) {
LinkedList<Integer> list = new LinkedList<Integer>();
int i =0;
for (i= 5; i > 0; i++)
line X: list.add(i);
System.out.println(list);
System.out.println(list.get(i));
}
}
i ran this program but unable to get the output (and also runtime exception) if i remove lineX from this program it runs correctly..(displays the ouput) thanks
Originally posted by Burkhard Hassel: True, but as you can see from the use of generics, the code must be java 5, not earlier. And in java 5, autoboxing would do that job. | http://www.coderanch.com/t/259702/java-programmer-SCJP/certification/fooled-programs | CC-MAIN-2014-10 | refinedweb | 223 | 66.54 |
Need to edge out the competition for your dream job? Train for certifications today.
Submit
public class Foo
{
System.Data.SqlClient.SqlConnection connection;
System.Data.SqlClient.SqlCommand command;
}
Select all
Open in new window
using System.Data.SqlClient;
public class Foo
{
SqlConnection connection;
SqlCommand set a reference through a right click on the References entry in the Solution Explorer.
-----
Using directives are a way to simplify your code by making it possible to specify the namespace for classes that you use often.
As an exempla, if you want to connect to SQL Server, you normally need to declare objects with the following syntax:
Open in new windowIf you add a using directive at the top of the file however, you do not need to specify the namespace:
Open in new window
Assembly references simply tell the compiler which assemblies (dll's, e.g.) to include during compilation/linking (where the namespaces are manifested).
Assemble Reference :
There are n' numbr of DLL's in .Net Framework, having all those in your Solution makes no-sense.
hence we have assembly reference, By Default Solution when drawn from Template take all the required
assemble, So Windows App might require different set of assembly and Web App the another set of Assemblies
These set of assemblies are minimum required assemblies.
Over and above that if you require any other assemble from .Net framework itself or and third pary DLL's,
You add the reference to that assembly in your solution
Using directive :
It is to allow the use of types in a namespace so that you do not have to qualify the use of a type in that namespace
and scope of a using directive is limited to the file in which it appears.
Thanks!
(Well, I suppose my statement actually IS basically correct - but I hadn't pointed out [and had forgotten, actually] the important distinction that the mechanism is only necessary when you'd like to avoid fully qualifying the namespace at each instance.) | https://www.experts-exchange.com/questions/27382746/using-directive-or-assembly-reference.html | CC-MAIN-2018-26 | refinedweb | 332 | 50.06 |
Are you sure?
This action might not be possible to undo. Are you sure you want to continue?
ns-3 project feedback: ns-developers@isi.edu
28 January 2010
This is an ns-3 tutorial. Primary documentation for the ns-3 project is available in four forms: • ns-3 Doxygen/Manual: Documentation of the public APIs of the simulator • Tutorial (this document) • Reference Manual: Reference Manual • ns-3 wiki.
Table of Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 1.2 1.3 For ns-2 Users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Contributing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Tutorial Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2
Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1 2.2 2.3 2.4 2.5 The Web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mercurial . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Socket Programming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 3 4 4
3
Getting Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.1 Downloading ns-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.1.1 Downloading ns-3 Using Mercurial . . . . . . . . . . . . . . . . . . . . . . . . . 6 3.1.2 Downloading ns-3 Using a Tarball . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.2 Building ns-3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.2.1 Building with build.py . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.2.2 Building with Waf. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 3.3 Testing ns-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.4 Running a Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4
Conceptual Overview . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.1 Key Abstractions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.1 Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.2 Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.3 Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.4 Net Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1.5 Topology Helpers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 A First ns-3 Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Boilerplate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Module Includes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3 Ns3 Namespace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4 Logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.5 Main Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6 Topology Helpers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6.1 NodeContainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6.2 PointToPointHelper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6.3 NetDeviceContainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6.4 InternetStackHelper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6.5 Ipv4AddressHelper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.7 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 15 15 16 16 17 17 17 18 18 19 19 20 20 20 21 22 22 22
ii 4.2.7.1 UdpEchoServerHelper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.7.2 UdpEchoClientHelper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.8 Simulator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.9 Building Your Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Ns-3 Source Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 24 24 25 26
5
Tweaking ns-3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.1 Using the Logging Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.1 Logging Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2 Enabling Logging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.3 Adding Logging to your Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Using Command Line Arguments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 Overriding Default Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Hooking Your Own Values. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Using the Tracing System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1 ASCII Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.1.1 Parsing Ascii Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 PCAP Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.1 Reading output with tcpdump . . . . . . . . . . . . . . . . . . . . . . . 5.3.2.2 Reading output with Wireshark . . . . . . . . . . . . . . . . . . . . . 28 28 29 33 33 33 37 38 39 40 41 42 42
6
Building Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
6.1 6.2 Building a Bus Network Topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Building a Wireless Network Topology. . . . . . . . . . . . . . . . . . . . . . . . . 52
7
The Tracing System . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
7.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.1 Blunt Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 A Simple Low-Level Example. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1.1 Callbacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1.2 Example Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Using the Config Subsystem to Connect to Trace Sources . . 7.2.3 How to Find and Connect Trace Sources, and Discover Callback Signatures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.4 What Trace Sources are Available? . . . . . . . . . . . . . . . . . . . . . . . 7.2.5 What String do I use to Connect? . . . . . . . . . . . . . . . . . . . . . . . . 7.2.6 What Return Value and Formal Arguments? . . . . . . . . . . . . . . 7.2.6.1 Take my Word for It . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.6.2 The Hard Way . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.7 What About TracedValue? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 A Real Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.1 Are There Trace Sources Available? . . . . . . . . . . . . . . . . . . . . . . 7.3.2 What Script to Use? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.3 A Common Problem and Solution . . . . . . . . . . . . . . . . . . . . . . . . 7.3.4 A fifth.cc Walkthrough . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3.4.1 How Applications are Started and Stopped . . . . . . . . . . 63 63 65 66 66 67 69 72 72 73 74 75 75 79 80 80 81 81 82 84
. .3. . . . . . . . . . . . . . . . . .2 The MyApp Application . . . . . . . . . . . . . . . . .4. . . . . . . .iii 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . 93 Closing . . . . . . . . . . . . . . . . . . . . . . . . .1 8. . . . . . . . . . . . . . . 85 87 88 90 8 Closing Remarks . . .3.4. . . .3 The Trace Sinks . . . . . . 94 . . . .4 The Main Program . . . . . . 7. . . 93 8. .cc . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . .5 Running fifth. . . . . . . . . . . . . . . . . . . . . . . . . . . . 7. . . . . . . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 Futures . . . . . 93 Index . . . . . . . . . . . . . . . . . . . . 7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 1: Introduction
1
1 Introduction
The ns-3 simulator is a discrete-event network simulator targeted primarily for research and educational use. The ns-3 project, started in 2006, is an open-source project developing ns-3. Primary documentation for the ns-3 project is available in four forms: • • • • ns-3 Doxygen/Manual: Documentation of the public APIs of the simulator Tutorial (this document) Reference Manual: Reference Manual ns-3 wiki
The purpose of this tutorial is to introduce new ns-3 users to the system in a structured way. It is sometimes difficult for new users to glean essential information from detailed manuals and to convert this information into working simulations. In this tutorial, we will build several example simulations, introducing and explaining key concepts and features as we go. As the tutorial unfolds, we will introduce the full ns-3 documentation and provide pointers to source code for those interested in delving deeper into the workings of the system. A few key points are worth noting at the onset: • Ns-3 is not an. • Ns-3 is open-source, and the project strives to maintain an open environment for researchers to contribute and share their software.
1.1 For ns-2 Users
For those familiar with ns-2, the most visible outward change when moving to ns-3 is the choice of scripting language. Ns-2 is scripted in OTcl and results of simulations can be visualized using tutorial, we will first concentrate on scripting directly in C++ and interpreting results via trace files. But there are similarities as well (both, for example, are based on C++ objects, and some code from ns-2 has already been ported to ns-3). We will try to highlight differences between ns-2 and ns-3 as we proceed in this tutorial. A question that we often hear is "Should I still use ns-2 or move to ns-3?" The answer is that it depends. ns-3 does not have all of the models that ns-2 currently has, but on the
Chapter 1: Introduction
2
other hand, ns-3 does have new capabilities (such as handling multiple interfaces on nodes correctly, use of IP addressing and more alignment with Internet protocols and designs, more detailed 802.11 models, etc.). ns-2 models can usually be ported to ns-3 (a porting guide is under development). There is active development on multiple fronts for ns-3. The ns-3 developers believe (and certain early users have proven) that ns-3 is ready for active use, and should be an attractive alternative for users looking to start new simulation projects.
1.2 Contributing
Ns-3 is a research and educational simulator, by and for the research community. It will rely on the ongoing contributions of the community to develop new models, debug or maintain existing ones, and share results. There are a few policies that we hope will encourage people to contribute to ns-3 like they have for ns-2: • Open source licensing based on GNU GPLv2 compatibility; • wiki; • Contributed Code page, similar to ns-2’s popular Contributed Code page; • src/contrib directory (we will host your contributed code); • Open bug tracker; • Ns-3 developers will gladly help potential contributors to get started with the simulator (please contact one of us). We realize that if you are reading this document, contributing back to the project is probably not your foremost concern at this point, but we want you to be aware that contributing is in the spirit of the project and that even the act of dropping us a note about your early experience with ns-3 (e.g. "this tutorial section was not clear..."), reports of stale documentation, etc. are much appreciated.
1.3 Tutorial Organization.
Chapter 2: Resources
3
2 Resources
2.1 The Web
There.
2.2 Mercurial.
2.3 Waf, and will only have to understand a tiny and intuitively obvious subset of Python in order to extend the system in most cases. For those interested in the gory details of Waf, the main web site can be found at.
However. It can cause your Cygwin or MinGW DLLs to die in mysterious ways and often prevents debuggers from running.4 Development Environment As mentioned above. There are an almost unimaginable number of sources of information on C++ available on the web or in print. The ns-3 system uses several components of the GNU “toolchain” for development.5 Socket Programming We will assume a basic facility with the Berkeley Sockets API in the examples used in this tutorial. For a quick review of what is included in the GNU toolchain see. A working knowledge of C++ and object-oriented concepts is assumed in this document. although some of the project maintainers to work with it).edu/~donahoo/practical/CSockets/. scripting in ns-3 is done in C++ or Python. Search for “Logitech” and read the FAQ entry. Typically an ns-3 author will work in Linux or a Linux-like environment.stackdump file when I try to compile my source code.baylor. We use Waf for these functions. If you are new to sockets. For instance. so we do expect a basic command of the language. and gdb. Beware of Logitech software when using Cygwin.cygwin. ns-3 uses gcc. GNU binutils. As of ns-3. this tutorial. we do not use the GNU build system tools.org/wiki/GNU_toolchain.. idioms and design patterns as they appear. If you do use Cygwin or MinGW. It can. you may want to find a tutorial.com/ for details on downloading (MinGW is presently not officially supported. however. we will save you quite a bit of heartburn right off the bat and encourage you to take a look at the MinGW FAQ. “why does make often crash creating a sh. and use Logitech products. See. If you are new to C++. We don’t want this tutorial to devolve into a C++ tutorial. most of the ns-3 API is available in Python. For those running under Windows.” Believe it or not. and sometimes interactions with other Windows software can cause problems. but the models are written in C++ in either case. we recommend reviewing the API and some common usage cases.2. The ns-3 project supports development in the Cygwin environment for these users. there do exist environments which simulate the Linux environment to various degrees. though. . Another alternative to Cygwin is to install a virtual machine environment such as VMware server and install a Linux virtual machine. the Logitech Process Monitor insinuates itself into every DLL in the system when it is running. which you can find at:. For a good overview of programming TCP/IP sockets we recommend TCP/IP Sockets in C. A software toolchain is the set of programming tools available in the given environment.or cookbook-based book or web site and work through at least the basic features of the language before proceeding. 2. neither make nor autotools.Chapter 2: Resources 4 2. sometimes be problematic due to the way it actually does its emulation.wikipedia. Cygwin provides many of the popular Linux system commands. We will take some time to review some of the more advanced concepts or possibly unfamiliar language features. There is an associated web site that includes source for the examples in the book.
.Chapter 2: Resources 5 If you understand the first four chapters of the book (or for those who do not have access to a copy of the book. There is a similar book on Multicast Sockets. that covers material you may need to understand if you look at the multicast examples in the distribution. Multicast Sockets. the echo clients and servers shown in the website above) you will be in good shape to understand the tutorial.
py dist. If you go to the following link:. you should see something like the following displayed.1 Downloading ns-3 From this point forward. destination directory: ns-3-allinone requesting all changes adding changesets adding manifests adding file changes added 31 changesets with 45 changes to 7 files 7 files updated. 0 files merged. We recommend that you begin your ns-3 adventures in this environment as it can really simplify your life at this point. We are also going to assume that you have Mercurial and Waf installed and running on the target system as described in the Getting Started section of the ns-3 web site:. you will see a number of repositories.nsnam. you can get a copy of ns-3-allinone by typing the following into your Linux shell (assuming you have installed Mercurial): cd mkdir repos cd repos hg clone* download.org/ns-3-allinone As the hg (Mercurial) command executes. The repositories .html. Hint: we will assume you do this later in the tutorial. This is a set of scripts that manages the downloading and building of various subsystems of ns-3 for you.) and has the GNU toolchain installed and verified.org. we are going to assume that the reader is working in Linux or a Linux emulation environment (Linux. 0 files unresolved After the clone command completes.py Notice that you really just downloaded some Python scripts. We recommend using Mercurial unless there’s a good reason not to.1 Downloading ns-3 Using Mercurial One practice is to create a directory called repos in one’s home directory under which one can keep local Mercurial repositories. 0 files removed. etc. or you can work with repositories using Mercurial.org/releases/.py* constants. See the end of this section for instructions on how to get a tarball release.Chapter 3: Getting Started 6 3 Getting Started 3.org/. If you adopt that approach. you should have a directory called ns-3-allinone under your ~/repos directory. Cygwin.1. The next step will be to use those scripts to download and build the ns-3 distribution of your choice.nsnam.py* README util. the contents of which should look something like the following: build. You can also download a tarball release at. 3. The simplest way to get started using Mercurial repositories is to use the ns-3-allinone environment.nsnam. Many are the private repositories of the ns-3 development team. The ns-3 code is available in Mercurial repositories on the server.
nsnam. Official releases of ns-3 will be numbered as ns-3. Since the release numbers are going to be changing.nsnam. # # Get NS-3 # Cloning ns-3 branch => hg clone. Go ahead and type the following into your shell (remember you can substitute the name of your chosen release number instead of ns-3-dev – like "ns-3. For each release.Chapter 3: Getting Started 7 of interest to you will be prefixed with “ns-3”. working states but they are in a development area with unreleased code present.2.6-ref-traces) in the text below. but you can replace the string “ns-3-dev” with your choice of release (e. .nsnam. For example. For example. We provide this example to illustrate how to specify alternate repositories. The developers attempt to keep these repository in consistent.org/ns-3-dev/ and the associated reference traces may be found at.. We are now going to use the download.g.6 and ns-3.nsnam.org/ns-3-dev ns-3-dev requesting all changes adding changesets adding manifests adding file changes . so you may want to consider staying with an official release if you do not need newly. This is a good idea to do at least once to verify everything has built correctly.py As the hg (Mercurial) command executes.9. We have had a regression testing framework in place since the first release. you should see something like the following. in.<release>. It is crucial to keep these files consistent if you want to do any regression testing of your repository. You will also find a separate repository named ns-3. a second hotfix to a still hypothetical release nine of ns-3 would be numbered as ns-3.1-reftraces that holds the reference traces for the ns-3.1 release.6-reftraces" if you want to work with a stable release).1 which is the first stable release of ns-3. You can find the latest version of the code either by inspection of the repository list or by going to the “Getting Started” web page and looking for the latest release identifier. ns-3.py -n ns-3-dev -r ns-3-dev-ref-traces Note that the default for the -n option is ns-3-dev and the default for the -r option is ns-3-dev-ref-traces and so the above is actually redundant. I will stick with the more constant ns-3-dev here in the tutorial. .py script to pull down the various pieces of ns-3 you will be using.introduced features. In order to download ns-3-dev you can actually use the defaults and simply type. Go ahead and change into the ns-3-allinone directory you created when you cloned that repository. The current development snapshot (unreleased) of ns-3 may be found at" and "ns-3.<hotfix>.org/ you will find a repository named ns-3. These known good output files are called reference traces and are associated with a given release by name. a set of output files that define “good behavior” are saved.
0. Next. => bzr checkout -rrevno:640. => hg clone. on most platforms. 0 files unresolved This is output by the download script as it fetches the actual ns-3 code from the repository.nz/mercurial/nsc nsc requesting all changes adding changesets adding manifests adding file changes added 273 changesets with 17565 changes to 15175 files 10622 files updated. This was the download script getting the Python bindings generator for you. 0 files removed. However. 0 files unresolved This is the download script fetching the reference trace files for you. # # Get PyBindGen # Required pybindgen version: 0. this will fail if no network connection is available. 0 files unresolved This part of the process is the script downloading the Network Simulation Cradle for you. 0 files merged. On your platform you may not see some of these pieces come down.nz/mercurial/nsc => hg clone. Hit Ctrl .10. The download script is smart enough to know that on some platforms various pieces of ns-3 are not supported. you should see something like.wand.nsnam. 0 files merged.5.0 Retrieving nsc from 3: Getting Started 8 added 4634 changesets with 16500 changes to 1762 files 870 files updated. Next you should see (modulo platform variations) something along the lines of. 0 files merged. # # Get the regression traces # Synchronizing reference traces using Mercurial.net.org/ns-3-dev-ref-traces ns-3-dev-ref-traces requesting all changes adding changesets adding manifests adding file changes added 86 changesets with 1178 changes to 259 files 208 files updated. the process should continue with something like.net/pybindgen pybindgen Fetch was successful. 0 files removed.net. # # Get NSC # Required NSC version: nsc-0. 0 files removed.640 Trying to fetch pybindgen.wand.
6-ref-traces/ nsc-0.py You are now ready to build the ns-3 distribution. you should have several new directories under ~/repos/ns-3-allinone: build. 3.0./build.Chapter 3: Getting Started 9 After the clone command completes.tar. You should see something like the following there: AUTHORS bindings/ CHANGES.bz2 tar xjf ns-allinone-3.2.py RELEASE_NOTES samples/ scratch/ src/ utils/ VERSION waf* waf.py You are now ready to build the ns-3 distribution.py* download. download it and decompress it.1/ pybindgen-0.py The first time you build the ns-3 project you should build using the allinone environment.6.py* constants. One could also keep a tarballs directory. As mentioned above. If you downloaded using Mercurial you should have a directory called ns-3-allinone under your ~/repos directory.py util.12.py ns-3.6 under your ~/tarballs directory.bat* wscript wutils. you can get a copy of a release by typing the following into your Linux shell (substitute the appropriate version numbers.2 Downloading ns-3 Using a Tarball The process for downloading ns-3 via tarball is simpler than the Mercurial process since all of the pieces are pre-packaged for you. 3.6.pyc dist.6 you should see a number of files: build.5.py .bz2 If you change into the directory ns-allinone-3. If you downloaded using a tarball you should have a directory called something like ns-allinone-3.nsnam.pyc Go ahead and change into ns-3-dev under your ~/repos/ns-3-allinone directory.1 Building with build. If you adopt the tarballs directory approach. This will get the project configured for you in the most commonly useful way.py constants.2 Building ns-3 3. You just have to pick a release.html doc/ examples/ LICENSE ns3/ README regression/ regression. Take a deep breath and type the following: .py* constants.6/ ns-3. Hint: the tutorial will assume you downloaded into a repos directory.700/ README util. one practice is to create a directory called repos in one’s home directory under which one can keep local Mercurial repositories. Change into the directory you created in the download section above.py* ns-3-dev/ ns-3-dev-ref-traces/ nsc/ pybindgen/ README util.org/releases/ns-allinone-3.tar. of course): cd mkdir tarballs cd tarballs wget. so remember the placekeeper.1.
h : ok high precision time implementation : 128-bit integer header stdint.2 Building with Waf We use Waf to configure and build the ns-3 project. To explain to Waf that it should do optimized builds you will need to execute the following command. Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking Checking for for for for for for for for for for for for for for for for for for for for for for for program g++ : ok /usr/bin/g++ program cpp : ok /usr/bin/cpp program ar : ok /usr/bin/ar program ranlib : ok /usr/bin/ranlib g++ : ok program pkg-config : ok /usr/bin/pkg-config regression reference traces : ok ./ns-3-dev-ref-traces (guessed) -Wno-error=deprecated-declarations support : yes -Wl. ./waf -d optimized configure This runs Waf out of the local directory (which is provided as a convenience for you). You got what you needed from them and will now interact directly with Waf and we do it in the ns-3-dev directory. It’s not strictly required at this point.. the ns-3-allinone scripts. As the build system checks for various dependencies you should see output that looks similar to the following.h : ok header pthread.h : not found library rt : ok header netpacket/packet.h : ok pkg-config flags for GSL : ok header linux/if_tun. not in the ns-3-allinone directory. Let’s tell the project to do make an optimized build. but it will be valuable to take a slight detour and look at how to make changes to the configuration of the project.586s) Once the project has built you can say goodbye to your old friends. By default you have configured your project to build the debug version.h : ok header inttypes.Chapter 3: Getting Started 10 You will see lots of typical compiler output messages displayed as the build script builds the various pieces you downloaded.h : ok pkg-config flags for GTK_CONFIG_STORE : ok pkg-config flags for LIBXML2 : ok library sqlite3 : ok .h : ok header sys/inttypes. Go ahead and change into the ns-3-dev directory (or the directory for the appropriate release you downloaded. Eventually you should see the following magic words: Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (2m30. cd ns-3-dev 3.2.--soname=foo support : yes header stdlib.h : ok header signal. Probably the most useful configuration change you can make will be to build the optimized version of the code.
/pybindgen (guessed) Checking for Python module pybindgen : ok Checking for pybindgen version : ok 0.9.Summary of optional NS-3 features: Threading Primitives : enabled Real Time Simulator : enabled Emulated Net Device : enabled GNU Scientific Library (GSL) : enabled Tap Bridge : enabled GtkConfigStore : enabled XmlIo : enabled SQlite stats data output : enabled Network Simulation Cradle : enabled Python Bindings : enabled Python API Scanning Support : enabled Use sudo to set suid bit : not enabled (option --enable-sudo not selected) Build examples and samples : enabled Static build : not enabled (option --enable-static not selected) ’configure’ finished successfully (2.10.Chapter 3: Getting Started 11 Checking for NSC location : ok .870s) Note the last part of the above output.” Now go ahead and switch back to the debug build..5./waf -d debug configure The build system is now configured and you can build the debug versions of the ns-3 programs by simply typing.0 must be found on the system../nsc (guessed) Checking for library dl : ok Checking for NSC supported architecture x86_64 : ok Checking for program python : ok /usr/bin/python Checking for Python version >= 2.0 Checking for program sudo : ok /usr/bin/sudo Checking for program hg : ok /usr/bin/hg Checking for program valgrind : ok /usr/bin/valgrind ---.h : ok Checking for -fvisibility=hidden support : yes Checking for pybindgen location : ok . This is not enabled by default and so this feature is reported as “not enabled.3 : ok 2. Some ns-3 options are not enabled by default or require support from the underlying system to work properly.5 Checking for program gccxml : ok /usr/local/bin/gccxml Checking for gccxml version : ok 0.0. For instance. .9. Note further that there is a feature to use the progarm sudo to set the suid bit of certain programs.5-config : ok /usr/bin/python2. to enable XmlTo.5 : ok Checking for program python2. If this library were not found.5-config Checking for header Python./waf .640 Checking for Python module pygccxml : ok Checking for pygccxml version : ok 0.2 Checking for library python2. the corresponding ns-3 feature would not be enabled and a message would be displayed. the library libxml-2. .
3 Testing ns-3 You can run the unit tests of the ns-3 distribution by running the “. There are many other configure. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (1. You will also see output from the test runner and the output will actually look something like. This turns out to be a configuration-time command.py -c core” script.. You should eventually see a report saying that./test. sorry./test. . . and so you could reconfigure using the following command . PASS: TestSuite attributes PASS: TestSuite config PASS: TestSuite global-value PASS: TestSuite command-line PASS: TestSuite basic-random-number PASS: TestSuite object PASS: TestSuite random-number-generators 47 of 47 tests passed (47 passed. type: .. 0 failed./waf -d debug --enable-sudo configure If you do this. 0 valgrind errors) This command is typically run by users to quickly verify that an ns-3 distribution has built correctly. Okay.and build-time options available in waf.Chapter 3: Getting Started 12 Some waf commands are meaningful during the build phase and some commands are valid in the configuration phase. 0 crashed. 47 of 47 tests passed (47 passed. if you wanted to use the emulation features of ns-3 you might want to enable setting the suid bit using sudo as described above. For example. 0 valgrind errors) This is the important message. waf will have run sudo to change the socket creator programs of the emulation code to run as root.py -c core These tests are run in parallel by waf. 0 failed. I made you build the ns-3 part of the system twice.799s) PASS: TestSuite ns3-wifi-interference PASS: TestSuite histogram PASS: TestSuite sample PASS: TestSuite ipv4-address-helper PASS: TestSuite devices-wifi PASS: TestSuite propagation-loss-model ./waf --help We’ll use some of the testing-related commands in the next section. To explore these options. but now you know how to change the configuration and build optimized code. 0 crashed. 3.
/download. you can do the following: cd build/debug/regression/traces/second./. If they are identical. Regression testing summary: PASS: 22 of 22 tests passed Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (25. you provide Waf with the regression flag..2 and ns-3./waf --regression You should see messages indicating that many tests are being run and are passing. The content of these trace files are compared with the reference traces. (Warning: The ns-3. To run the regression tests. . Remember to cd back into the top-level ns-3 directory after you are done: cd . If a regression test fails you will see a FAIL indication along with a pointer to the offending trace file and its associated reference trace file along with a suggestion on diff parameters and options in order to see what has gone awry./..ref tcpdump -nn -tt -r second-2-0..py process above. Some regression tests may be SKIPped if the required support is not present. .. We’ll have much more to say on pcap files later in this tutorial. During regression testing Waf will run a number of tests that generate what we call trace files. Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ [647/669] regression-test (test-csma-bridge) [648/669] regression-test (test-csma-broadcast) [649/669] regression-test (test-csma-multicast) [650/669] regression-test (test-csma-one-subnet) PASS test-csma-multicast [651/669] regression-test (test-csma-packet-socket) PASS test-csma-bridge .Chapter 3: Getting Started 13 You can also run our regression test suite to ensure that your distribution and toolchain have produced binaries that generate output that is identical to known-good reference output files.3 releases do not use the ns-3-allinone environment and require you to be online when you run regression tests because they dynamically synchronize the reference traces directory with an online repository immediately prior to the run). Note that the regression tests are also run in parallel and so the messages may be interleaved.826s) If you want to take a look at an example of what might be checked during a regression test.. it will be useful to convert the pcap files to text using tcpdump prior to comparison. If the error was discovered in a pcap file./.. the regression tests report a PASS status.pcap The output should be clear to anyone who is familiar with tcpdump or net sniffers.. You downloaded these reference traces to your machine during the ./.
Hello Simulator Congratulations. simply use the --run option in Waf./waf --run hello-simulator Waf first checks to make sure that the program is built correctly and executes a build if required. Waf then executes the program. Let’s run the ns-3 equivalent of the ubiquitous hello world program by typing the following: . see this wiki entry. . To run a program.Chapter 3: Getting Started 14 3. which produces the following output. If you want to run programs under another tool such as gdb or valgrind. This allows the build system to ensure that the shared library paths are set correctly and that the libraries are available at run time.4 Running a Script We typically run scripts under the control of Waf. You are now an ns-3 user.
Instead. network. We do. not specifically an Internet simulator. Just as software applications run on computers to perform tasks in the “real world. 4. we use a more generic term also used by other simulators that originates in Graph Theory — the node. but have a specific meaning in ns-3. In ns-3 there is no real concept of operating system and especially no concept of privilege levels or system calls. have the idea of an application. Often. but we recommend taking the time to read through this section just to ensure you are starting on a firm foundation. The Node class provides methods for managing the representations of computing devices in simulations.1. In ns-3 the basic computing device abstraction is called the node. protocol stacks and peripheral cards with their associated drivers to enable the computer to do useful work. In ns-3 the basic abstraction for a user program that generates some activity to be simulated is the application. 4.1 Key Abstractions In this section. however. a computing device that connects to a network is called a host or sometimes an end system.. processor cycles. In this tutorial.Chapter 4: Conceptual Overview 15 4 Conceptual Overview The first thing we need to do before actually starting to look at or write ns-3 code is to explain a few core concepts and abstractions in the system.1 Node In Internet jargon. the line of separation between system and application software is made at the privilege level change that happens in operating system traps. we’ll review some terms that are commonly used in networking. One adds things like applications. Much of this may appear transparently obvious to some. As you might expect. We use the same basic model in ns-3. etc. these applications compose a client/server application set used to generate and echo simulated network packets . System software usually does not use those resources to complete tasks that directly benefit a user. This abstraction is represented in C++ by the class Application.1. Because ns-3 is a network simulator. A user would typically run an application that acquires and uses the resources controlled by the system software to accomplish some goal.” ns-3 applications run on ns-3 Nodes to drive simulations in the simulated world. 4. The Application class provides methods for managing the representations of our version of user-level applications in simulations. You should think of a Node as a computer to which you will add functionality. we will use specializations of class Application called UdpEchoClientApplication and UdpEchoServerApplication. System Software organizes various computer resources such as memory. Developers are expected to specialize the Application class in the object-oriented programming sense to create new applications.2 Application Typically. disk. computer software is divided into two broad classes. according to some computing model. This abstraction is represented in C++ by the class Node. we intentionally do not use the term host since it is closely associated with the Internet and its protocols.
The Channel class provides methods for managing communication subnetwork objects and connecting nodes to them. and may be specialized by developers in the object-oriented programming sense. The CsmaChannel. and WifiNetDevice in this tutorial. The net device abstraction is represented in C++ by the class NetDevice. If the peripheral card implemented some networking function. models a version of a communication subnetwork that implements a carrier sense multiple access communication medium. you are connecting your computer to an Ethernet communication channel. Here the basic communication subnetwork abstraction is called the channel and is represented in C++ by the class Channel.1. The NetDevice class provides methods for managing connections to Node and Channel objects. you had to buy a specific kind of network cable and a hardware device called (in PC terminology) a peripheral card that needed to be installed in your computer. . a Node may be connected to more than one Channel via multiple NetDevices. A NIC will not work without a software driver to control the hardware. When you connect your Ethernet cable to the plug in the wall. or NICs. one connects a Node to an object representing a communication channel. 4. the PointToPointNetDevice is designed to work with a PointToPointChannel and a WifiNetNevice is designed to work with a WifiChannel. The specialized Channel can also model things as complicated as a large Ethernet switch. This gives us Ethernet-like functionality. for example. In the simulated world of ns-3. Devices are controlled using device drivers.4 Net Device It used to be the case that if you wanted to connect a computers to a network. they were called Network Interface Cards. We will use specialized versions of the Channel called CsmaChannel. Today most computers come with the network interface hardware built in and users don’t see these building blocks. A net device is “installed” in a Node in order to enable the Node to communicate with other Nodes in the simulation via Channels. A Channel specialization may model something as simple as a wire.1.3 Channel In the real world. In ns-3 the net device abstraction covers both the software driver and the simulated hardware. We will use the several specialized versions of the NetDevice called CsmaNetDevice.Chapter 4: Conceptual Overview 16 4. Channels may also be specialized by developers in the object oriented programming sense. the CsmaNetDevice is designed to work with a CsmaChannel. PointToPointNetDevice. In Unix and Linux you refer to these net devices by names such as eth0. one can connect a computer to a network. PointToPointChannel and WifiChannel in this tutorial. Just as an Ethernet NIC is designed to work with an Ethernet network. or three-dimensional space full of obstructions in the case of wireless networks. a piece of peripheral hardware is classified as a device. Just as in a real computer. In Unix (or Linux). Often the media over which data flows in these networks are called channels. and network devices (NICs) are controlled using network device drivers collectively known as net devices.
You should see a file named first. NetDevices to Channels. indent-tabs-mode:nil. just get used to the look and feel of ns-3 code and adopt this standard whenever you are working with our code.Mode:C++. well.py scratch/ waf* wutils. assigning IP addresses. you will eventually have to conform to the ns-3 coding standard as described in the file doc/codingstd. Let’s take a look at that script line by line.bat* Change into the examples/tutorial directory. The emacs mode line above makes it easier to get the formatting correct if you use the emacs editor. /* . 4.py build/ LICENSE regression. Since connecting NetDevices to Nodes. are such common tasks in ns-3. add a MAC address. /* -*. so go ahead and open first.pyc src/ waf.5 Topology Helpers In a real network. The ns-3 simulator is licensed using the GNU General Public License.2 A First ns-3 Script If you downloaded the system as was suggested above. All of the development team and contributors have done so with various amounts of grumbling. you will find host computers with added (or built-in) NICs.cc located there. and you should find a directory structure something like the following: AUTHORS doc/ README RELEASE_NOTES utils/ wscript bindings/ examples/ regression/ samples/ VERSION wutils. If you want to contribute your code to the project. This is a script that will create a simple point-to-point link between two nodes and echo a single packet between the nodes. like most large projects. it may take many distinct ns-3 core operations to create a NetDevice. This tells emacs about the formatting conventions (coding style) we use in our source code.html ns3/ regression.Chapter 4: Conceptual Overview 17 4. In ns-3 we would say that you will find Nodes with attached NetDevices. In a large simulated network you will need to arrange many connections between Nodes. We provide topology helper objects that combine those many distinct operations into an easy to use model for your convenience. Often you will see a copyright notice for one of the institutions involved in the ns-3 project above the GPL text and an author listed below. -*. and then connect the NetDevice to a Channel. has adopted a coding style to which all contributed code must adhere.txt or shown on the project web page here. Change into that release directory. etc. c-file-style:’’gnu’’.2. 4. For example.1. so we might as well get it out of the way immediately. configure the node’s protocol stack. you will have a release of ns-3 in a directory called repos under your home directory.1 Boilerplate The first line in the file is an emacs mode line.. Even more operations would be required to connect multiple devices onto multipoint channels and then to connect individual networks together into internetworks.pyc CHANGES. we provide what we call topology helpers to make this as easy as possible.cc in your favorite editor. You will see the appropriate GNU legalese at the head of every file in the ns-3 distribution. The ns-3 project. install that net device on a Node. We recommend that you.*/ This is always a somewhat controversial subject. NetDevices and Channels.
Suite 330. So now if you look in the directory . You will also have done a .h" To help our high-level script users deal with the large number of include files present in the system. you will already have done a ./waf -d debug configure in order to configure the project to perform debug builds. 59 Temple Place.2. of course. We provide a single include file that will recursively load all of the include files used in each module. write to the Free Software * Foundation. Boston. When you do a build.h" #include "ns3/helper-module.cc script is a namespace declaration.2 Module Includes The code proper starts with a number of include statements. Each of the ns-3 include files is placed in a directory called ns3 (under the build directory) during the build process to help avoid include file name collisions. Inc. * but WITHOUT ANY WARRANTY. and possibly have to get a number of dependencies right./waf to build the project..h file corresponds to the ns-3 module you will find in the directory src/core in your downloaded release distribution.2. * * This program is distributed in the hope that it will be useful. if not.. Since you are. we give you the ability to load a group of files at a large granularity. Rather than having to look up exactly what header you need. MA 02111-1307 USA */ 4. This is not the most efficient approach but it certainly makes writing scripts much easier.Chapter 4: Conceptual Overview 18 * This program is free software./build/debug/ns3 you will find the four module include files shown above. See the * GNU General Public License for more details. Waf will place public header files in an ns3 directory under the appropriate build/debug or build/optimized directory depending on your configuration. Waf will also automatically generate a module include file to load all of the public header files. 4../. following this tutorial religiously. we group includes according to relatively large modules. you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation.h" #include "ns3/simulator-module. * * You should have received a copy of the GNU General Public License * along with this program. . You can take a look at the contents of these files and find that they do include all of the public include files in their respective modules.3 Ns3 Namespace The next line in the first.h" #include "ns3/node-module. The ns3/coremodule. without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. #include "ns3/core-module. If you list this directory you will find a large number of header files.
2. but to summarize. If you look at the project web site. Now. you will find a graphical representation of the structure of the documentation. go ahead and take a look at the specific NS_LOG_COMPONENT_DEFINE documentation. char *argv[]) { This is just the declaration of the main function of your program (script). 4. you will find a link to “Doxygen (ns-3-dev)” in the navigation bar. You can either scroll down or select the “More.” link under the collaboration diagram to do this. This is a fancy way of saying that after this declaration. I won’t duplicate the documentation here. you need to define a main function that will be the first function run. int main (int argc. so go ahead and expand that documentation node. this line declares a logging component called FirstScriptExample that allows you to enable and disable console message logging by reference to the name. please consult almost any C++ tutorial and compare the ns3 namespace and usage here with instances of the std namespace and the using namespace std. expand the Debugging book and then select the Logging page. statements you will often find in discussions of cout and streams. The next two lines of the script are used to enable two logging components that are built into the Echo Client and Echo Server applications: . It turns out that the ns-3 logging subsystem is part of the core module. ns-3 project. The C++ using statement introduces the ns-3 namespace into the current (global) declarative region. Your ns-3 script is just a C++ program. In the list of #defines at the top of the page you will see the entry for NS_LOG_COMPONENT_ DEFINE. which we hope will help with integration with other code. Just as in any C++ program. You should now be looking at the Doxygen documentation for the Logging module. NS_LOG_COMPONENT_DEFINE ("FirstScriptExample"). Before jumping in. If you are unfamiliar with namespaces. Along the left side. The concept of module here ties directly into the module include files discussed above. If you expand Modules you will see a list of ns-3 module documentation. 4.Chapter 4: Conceptual Overview 19 using namespace ns3. it would probably be good to look for the “Detailed Description” of the logging module to get a feel for the overall operation..2. If you select this link. There is also a link to “Doxygen (stable)” that will take you to the documentation for the latest stable release of ns-3. This groups all ns-3related declarations in a scope outside the global namespace. you will be taken to our documentation page for the current development release. Once you have a general idea of what is going on. A good place to start is the NS-3 Modules “book” in the ns-3 navigation tree. you will not have to type ns3:: scope resolution operator before all of the ns-3 code in order to use it.5 Main Function The next lines of the script you will find are.. The ns-3 project is implemented in a C++ namespace called ns3. There is nothing at all special here.4 Logging The next line of the script is the following. We will use this statement as a convenient place to talk about our Doxygen documentation system.
2. nodes. The nodes as they stand in the script do nothing. The first line above just declares a NodeContainer which we call nodes. and. manage and access any Node objects that we create in order to run a simulation. . The simplest form of network we support is a single point-to-point link between two nodes. We use the topology helper objects to make this job as easy as possible. If you still have the Doxygen handy. in a pattern which will become quite familiar to you.6. go ahead and select it to go to the documentation for the class. for example.2. LOG_LEVEL_INFO). looking for ns3::NodeContainer. Ethernet devices and wireless channels. we use a topology helper object to do the low-level work required to put the link together. Typically these two things are intimately tied together and one cannot expect to interchange. You should see a new set of tabs appear. When you find the class. Under that tab you will see a list of all of the ns-3 classes. These two lines of code enable debug logging at the INFO level for echo clients and servers. The NodeContainer topology helper provides a convenient way to create.6 Topology Helpers 4. Scroll down. Let’s find the documentation for the NodeContainer class before we continue. The second line calls the Create method on the nodes object and asks the container to create two nodes. the container calls down into the ns-3 system proper to create two Node objects and stores pointers to those objects internally.Create (2).1 NodeContainer The next two lines of code in our script will actually create the ns-3 Node objects that will represent the computers in the simulation. LogComponentEnable("UdpEchoServerApplication". As described in the Doxygen. Our Topology Helpers follow this intimate coupling and therefore you will use a single PointToPointHelper to configure and connect ns-3 PointToPointNetDevice and PointToPointChannel objects in this script. Now we will get directly to the business of creating a topology and running a simulation. Another way to get into the documentation for a given class is via the Classes tab in the Doxygen pages.2. applications and peripheral cards. LOG_LEVEL_INFO).Chapter 4: Conceptual Overview 20 LogComponentEnable("UdpEchoClientApplication". just scroll up to the top of the page and select the Classes tab. We’ll construct one of those links here.2 PointToPointHelper We are constructing a point to point link. The next step in constructing a topology is to connect our nodes together into a network. these terms correspond roughly to peripheral cards and network cables. Recall that two of our key abstractions are the NetDevice and the Channel. If you have read over the Logging component documentation you will have seen that there are a number of levels of logging verbosity/detail that you can enable on each component. one of which is Class List. 4. NodeContainer nodes. You may recall that one of our key abstractions is the Node. This will result in the application printing out messages as packets are sent and received during the simulation. This represents a computer to which we are going to add things like protocol stacks.6. In the real world. 4.
4. The final line.SetChannelAttribute ("Delay". pointToPoint. A PointToPointChannel is created and the two PointToPointNetDevices are attached. When objects are created by the PointToPointHelper. will finish configuring the devices and channel. StringValue ("5Mbps")). a NetDeviceContainer is created. tells the PointToPointHelper object to use the value “5Mbps” (five megabits per second) as the “DataRate” when it creates a PointToPointNetDevice object. so we use a NetDeviceContainer to hold them just as we used a NodeContainer to hold the nodes we created. Most user-visible ns-3 objects have similar lists of Attributes. From a high-level perspective the next line. pointToPoint. If you look at the Doxygen for class ns3::PointToPointNetDevice and find the documentation for the GetTypeId method. StringValue ("2ms")). The first line.SetChannelAttribute ("Delay". We have a PointToPointHelper that is primed and ready to make PointToPointNetDevices and wire PointToPointChannel objects between them. PointToPointHelper pointToPoint. For each node in the NodeContainer (there must be exactly two for a point-to-point link) a PointToPointNetDevice is created and saved in the device container. the Attributes previously set in the helper are used to initialize the corresponding Attributes in the created objects.3 NetDeviceContainer At this point in the script. Similar to the “DataRate” on the PointToPointNetDevice you will find a “Delay” Attribute associated with the PointToPointChannel. Internally. NetDeviceContainer devices. pointToPoint. .6. StringValue ("2ms")). instantiates a PointToPointHelper object on the stack. we will ask the PointToPointHelper to do the work involved in creating. configuring and installing our devices for us. Among these is the “DataRate” Attribute. From a more detailed perspective. we have a NodeContainer that contains two nodes. Just as we used the NodeContainer topology helper object to create the Nodes for our simulation. The first line declares the device container mentioned above and the second does the heavy lifting. pointToPoint. StringValue ("5Mbps")). the string “DataRate” corresponds to what we call an Attribute of the PointToPointNetDevice. tells the PointToPointHelper to use the value “2ms” (two milliseconds) as the value of the transmission delay of every point to point channel it subsequently creates.Chapter 4: Conceptual Overview 21 The next three lines in the script are. The Install method of the PointToPointHelper takes a NodeContainer as a parameter.SetDeviceAttribute ("DataRate". PointToPointHelper pointToPoint. We will need to have a list of all of the NetDevice objects that are created. We use this mechanism to easily configure simulations without recompiling as you will see in a following section.2.Install (nodes). you will find a list of Attributes defined for the device. The following two lines of code. devices = pointToPoint.SetDeviceAttribute ("DataRate".
Just as we sometimes need a list of net devices created by a helper for future reference we sometimes need a list of Ipv4Interface objects.1. UDP.Chapter 4: Conceptual Overview 22 After executing the pointToPoint. When it is executed.Install (nodes) call we will have two nodes. "255.1. with stacks installed and IP addresses assigned.7 Applications Another one of the core abstractions of the ns-3 system is the Application. by the way). so the first address allocated from this base will be 10. Ipv4AddressHelper address. In this script we use two specializations of the core ns-3 class Application called UdpEchoServerApplication and UdpEchoClientApplication. The only user-visible API is to set the base IP address and network mask to use when performing the actual address allocation (which is done at a lower level inside the helper). it will install an Internet Stack (TCP. etc.255.255.5 Ipv4AddressHelper Next we need to associate the devices on our nodes with IP addresses.0". We provide a topology helper to manage the allocation of IP addresses. The next two lines of code in our example script. we use helper objects to help configure and manage the underlying .6. 4. 4.4 InternetStackHelper We now have nodes and devices configured. The low level ns-3 system actually remembers all of the IP addresses allocated and will generate a fatal error if you accidentally cause the same address to be generated twice (which is a very hard to debug error.1. but we don’t have any protocol stacks installed on our nodes. declare an address helper object and tell it that it should begin allocating IP addresses from the network 10.0").2. IP. Ipv4InterfaceContainer interfaces = address. 4. etc. first. What we need at this point are applications to generate traffic. The InternetStackHelper is a topology helper that is to internet stacks what the PointToPointHelper is to point-to-point net devices. The next line of code.1. By default the addresses allocated will start at one and increase monotonically. Just as we have in our previous explanations.cc.1. Now we have a point-to-point network built.0 using the mask 255. In ns-3 we make the association between an IP address and a device using an Ipv4Interface object.255. The Install method takes a NodeContainer as a parameter.0 to define the allocatable bits.1.2. stack. Both devices will be configured to transmit data at five megabits per second over the channel which has a two millisecond transmission delay.2.2. InternetStackHelper stack. The next two lines of code will take care of that.1.Assign (devices). address.Install (nodes).6. performs the actual address assignment.1. each with an installed point-to-point net device and a single point-to-point channel between them. followed by 10.) on each of the nodes in the node container. The Ipv4InterfaceContainer provides this functionality.255.SetBase ("10.1.
0)). so you can’t always just assume that parameters will be happily converted for you. In this case. Install will return a container that holds pointers to all of the applications (one in this case since we passed a NodeContainer containing one node) created by the helper.0)). it is an object used to help us create the actual applications.0 and convert it to an ns-3 Time object using a Seconds cast.0)). By virtue of the . One of our conventions is to place required Attributes in the helper constructor. Applications require a time to “start” generating traffic and may take an optional time to “stop”. If you are ever at a loss to find a particular method signature in C++ code that compiles and runs just fine.Start (Seconds (1. As usual. If you want.Install is going to install a UdpEchoServerApplication on the node found at index number one of the NodeContainer we used to manage our nodes. will cause the echo server application to Start (enable itself) at one second into the simulation and to Stop (disable itself) at ten seconds into the simulation. and C++ has its own rules. the Install method takes a NodeContainter as a parameter just as the other Install methods we have seen.7. you can set the “Port” Attribute to another value later using SetAttribute. first.0)).Start (Seconds (1. 4. This is actually what is passed to the method even though it doesn’t look so in this case. These methods take Time parameters. In this case. simply does a SetAttribute with the passed value.Stop (Seconds (10. we require the port number as a parameter to the constructor. we use UdpEchoServerHelper and UdpEchoClientHelper objects to make our lives easier. The first line of code in the above snippet declares the UdpEchoServerHelper. There is a C++ implicit conversion at work here that takes the result of nodes.Stop (Seconds (10.Get (1) (which returns a smart pointer to a node object — Ptr<Node>) and uses that in a constructor for an unnamed NodeContainer that is then passed to Install.Chapter 4: Conceptual Overview 23 objects. serverApps. The constructor. Be aware that the conversion rules may be controlled by the model author.2. serverApps. We provide both. serverApps. Here. we use an explicit C++ conversion sequence to take the C++ double 1. The two lines. serverApps. the helper can’t do anything useful unless it is provided with a port number that the client also knows about. These times are set using the ApplicationContainer methods Start and Stop. It is the execution of this method that actually causes the underlying echo server application to be instantiated and attached to a node. Rather than just picking one and hoping it all works out. this isn’t the application itself.1 UdpEchoServerHelper The following lines of code in our example script. UdpEchoServerHelper echoServer (9). We now see that echoServer. look for these kinds of implicit conversions. ApplicationContainer serverApps = echoServer.Install (nodes. in turn. the UdpEchoServerHelper object has an Install method.Get (1)). are used to set up a UDP echo server application on one of the nodes we have previously created.cc. Interestingly. Similar to many other helper objects.
and the “PacketSize” Attribute tells the client how large its packet payloads should be. UintegerValue (1)). . Just as in the case of the echo server.2. UintegerValue (1024)). echoClient. clientApps. Simulator::Run (). but here we start the client one second after the server is enabled (at two seconds into the simulation). however.Chapter 4: Conceptual Overview 24 fact that we have declared a simulation event (the application stop event) to be executed at ten seconds. serverApps. UdpEchoClientHelper echoClient (interfaces. echoClient. For the echo client. Recall that we used an Ipv4InterfaceContainer to keep track of the IP addresses we assigned to our devices. we need to set five different Attributes. echoClient. We pass parameters that are used (internally to the helper) to set the “RemoteAddress” and “RemotePort” Attributes in accordance with our convention to make required Attributes parameters in the helper constructors.SetAttribute ("PacketSize".SetAttribute ("MaxPackets". The zeroth interface in the interfaces container is going to correspond to the IP address of the zeroth node in the nodes container.8 Simulator What we need to do at this point is to actually run the simulation. We also tell it to arrange to send packets to port nine. in the first line of code (from above)..Get (0)). The “Interval” Attribute tells the client how long to wait between packets.GetAddress (1). There is an underlying UdpEchoClientApplication that is managed by an UdpEchoClientHelper. This is done using the global function Simulator::Run.2 UdpEchoClientHelper The echo client application is set up in a method substantially similar to that for the server.0)). we are telling the client to send one 1024-byte packet. TimeValue (Seconds (1.0)). clientApps. When we previously called the methods. 4. we tell the echo client to Start and Stop. clientApps..0)).Stop (Seconds (10. So. the simulation will last at least ten seconds.Install (nodes. . The “MaxPackets” Attribute tells the client the maximum number of packets we allow it to send during the simulation. serverApps.Start (Seconds (2.Start (Seconds (1. The first interface in the interfaces container corresponds to the IP address of the first node in the nodes container.0)).))). 9). 4. ApplicationContainer clientApps = echoClient.2. With this particular combination of Attributes.7.0)).Start (Seconds (2.Stop (Seconds (10. we are creating the helper and telling it so set the remote address of the client to be the IP address assigned to the node on which the server resides.SetAttribute ("Interval". The first two Attributes are set during construction of the UdpEchoClientHelper.
0 seconds which will start the echo client application. the chain of events triggered by that single client echo request will taper off and the simulation will go idle.0 seconds.cc -> build/debug/scratch/myfirst_3.2.cc Now build your first example script using waf: . When Simulator::Run is called.Chapter 4: Conceptual Overview 25 clientApps. First it will run the event at 1. Let’s try it. since we only send one packet (recall the MaxPackets Attribute was set to one).o -> build/debug/scratch/myfirst Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (2. As the helper functions (or low level ns-3 code) executed. Eventually. All you have to do is to drop your script into the scratch directory and it will automatically be built if you run Waf. do just that: Simulator::Destroy (). The simulation is then complete.cc. The ns-3 system took care of the hard part for you. This is done by calling the global function Simulator::Destroy. the system will begin looking through the list of scheduled events and executing them. return 0. first. When these events are executed.357s) You can now run the example (note that if you build your program in the scratch directory you must run it out of the scratch directory): . Once this happens. Then it will run the event scheduled for t=2.cc scratch/myfirst. cd . there are no further events to process and Simulator::Run returns. schedule many other events). The remaining lines of our first ns-3 script.0 seconds and two events at 10. Copy examples/tutorial/first. The start event implementation in the echo client application will begin the data transfer phase of the simulation by sending a packet to the server. they arranged it so that hooks were inserted in the simulator to destroy all of the objects that were created. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ [614/708] cxx: scratch/myfirst.. You did not have to keep track of any of these objects yourself — all you had to do was to call Simulator::Destroy and exit. we actually scheduled events in the simulator at 1. in turn. this event may schedule many more events. Again.0)). which will enable the echo server application (this event may.9 Building Your Script We have made it trivial to build your simple scripts./waf You should see messages reporting that your myfirst example was built successfully. All that remains is to clean up.0 seconds.o [706/708] cxx_link: build/debug/scratch/myfirst_3. 2.0 seconds.cc into the scratch directory after changing back into the top level directory. cp examples/tutorial/first.Stop (Seconds (10. The act of sending the packet to the server will trigger a chain of events that will be automatically scheduled behind the scenes and which will perform the mechanics of the packet echo according to the various timing parameters that we have set in the script. } 4. the remaining events will be the Stop events for the server and the client.
/waf --run scratch/myfirst You should see some output: Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.2009-07-01 12:47 +0200 30961 CHANGES.2009-07-01 12:47 +0200 1276 AUTHORS file | revisions | -rw-r--r-. There. The echo server silently echoes the packet and you see the echo client log that it has received its packet back from the server. You also see the logging component on the echo server say that it has received the 1024 bytes from 10.1 Received 1024 bytes from 10. The most recent code can be browsed on our web server at the following link: 12:47 +0200 17987 LICENSE file | revisions | -rw-r--r-. summary | shortlog | changelog | graph | tags | files Go ahead and select the files link. you will see a number of links.2 Here you see that the build system checks to make sure that the file has been build and then runs it. you will see the Mercurial summary page for our ns-3 development tree.2009-07-01 12:47 +0200 1886 .bat file | revisions | annotate annotate annotate annotate annotate annotate annotate annotate annotate annotate annotate .html file | revisions | -rw-r--r-.2.hgtags file | revisions | -rw-r--r-.org/ns-3-dev.Chapter 4: Conceptual Overview 26 . You see the logging component on the echo client indicate that it has sent one 1024 byte packet to the Echo Server on 10.2009-07-01 12:47 +0200 10946 regression. 4.2009-07-01 12:47 +0200 16171 RELEASE_NOTES file | revisions | -rw-r--r-.1.py file | revisions | -rwxr-xr-x 2009-07-01 12:47 +0200 88110 waf file | revisions | -rwxr-xr-x 2009-07-01 12:47 +0200 28 waf.2009-07-01 12:47 +0200 6 VERSION file | revisions | -rw-r--r-.2009-07-01 12:47 +0200 560 .nsnam.1. This is what the top-level of most of our repositories will look: drwxr-xr-x [up] drwxr-xr-x bindings python files drwxr-xr-x doc files drwxr-xr-x examples files drwxr-xr-x ns3 files drwxr-xr-x regression files drwxr-xr-x samples files drwxr-xr-x scratch files drwxr-xr-x src files drwxr-xr-x utils files -rw-r--r-. At the top of the page.1.1.hgignore file | revisions | -rw-r--r-.1.1.1.2 Received 1024 bytes from 10.1.1.3 Ns-3 Source Code Now that you have used some of the ns-3 helpers you may want to have a look at some of the source code that implements that functionality.418s) Sent 1024 bytes to 10.1.2009-07-01 12:47 +0200 3742 README file | revisions | -rw-r--r-.1.
you will be taken to the listing of the src subdirectories.cc. You can view source code either by clicking on the directory name or by clicking on the files link to the right of the directory name.h link.cc you will find the code you just walked through.h which contains useful macros for exiting scripts if abnormal conditions are detected. Feel free to poke around in the directory tree to get a feel for what is there and the style of ns-3 programs. If you click on the abort. If you click on first.Chapter 4: Conceptual Overview 27 -rw-r--r-. If you click on examples you will see a list of files. The first file you will find (as of this writing) is abort.h.py file | revisions | annotate Our example scripts are in the examples directory. you will find a list of files. you will be sent to the source file for abort. . If you click on the src directory.2009-07-01 12:47 +0200 7673 wutils. One of the files in that directory is first. The source code is mainly in the src directory. The source code for the helpers we have used in this chapter can be found in the src/helper directory.2009-07-01 12:47 +0200 35395 wscript file | revisions | annotate -rw-r--r-. If you then click on core subdirectory.
• NS LOG FUNCTION — Log a message describing each function called.cc example script you have already built. Logging should be preferred for debugging information.Chapter 5: Tweaking ns-3 28 5 Tweaking ns-3 5. . You should understand that we do provide a general purpose mechanism — tracing — to get data out of your models which should be preferred for simulation output (see the tutorial section Using the Tracing System for more details on our tracing system). There are currently seven levels of log messages of increasing verbosity defined in the system. and it provides selectable verbosity levels. let’s use some of that knowledge to get some interesting information out of the scratch/myfirst. the logging system has Doxygen documentation and now would be a good time to peruse the Logging Module documentation if you have not done so. irrespective of logging levels or component selection. 5. • NS LOG ALL — Log everything. As was seen earlier in the tutorial. and ns-3 is not an exception. In some cases. Ns-3 takes the view that all of these verbosity levels are useful and we provide a selectable. • NS LOG UNCOND – Log the associated message unconditionally. warnings. logging facilities are used to output debug messages which can quickly turn the output into a blur. • NS LOG INFO — Log informational messages about program progress.1 Using the Logging Module We have already taken a brief look at the ns-3 logging module while going over the first. warning messages may be output as well as more detailed informational messages. multi-level approach to message logging. Logging can be disabled completely. error messages. • NS LOG ERROR — Log error messages.1 Logging Overview Many large systems support some kind of message logging facility. relatively easy to use way to get useful information out of your simulation. • NS LOG WARN — Log warning messages. In other systems. Now that you have read the documentation in great detail. only error messages are logged to the “operator console” (which is typically stderr in Unix. and logging can be set up using a shell environment variable (NS LOG) or by logging system function call. enabled on a component-by-component basis. We also provide an unconditional logging level that is always displayed. ad-hoc debugging messages. In some cases.cc script.1.based systems). • NS LOG DEBUG — Log relatively rare. • NS LOG LOGIC – Log messages describing logical flow within a function. We will now take a closer look and see what kind of use-cases the logging subsystem was designed to cover. or any time you want to easily get a quick message out of your scripts or models. The ns-3 log module provides a straightforward. or enabled globally. Each level can be requested singly or cumulatively.
NS_LOG_DEBUG.1 Received 1024 bytes from 10. go ahead and run the last script just as you did previously. We can ask the client application. for example. and the right hand side is the flag we want to use. just to get our bearings. LogComponentEnable("UdpEchoClientApplication". we have enabled NS_LOG_INFO. UdpEchoClientApplication=level_all The left hand side of the assignment is the name of the logging component we want to set.1. then you will have to convert my examples to the “setenv VARIABLE value” syntax required by those shells. In this case. the UDP echo client application is responding to the following line of code in scratch/myfirst.1.1 .1. we are going to turn on all of the debugging levels for the application. .1. In this case.2 Enabling Logging Let’s use the NS LOG environment variable to turn on some more logging./waf --run scratch/myfirst You should see the now familiar output of the first ns-3 example program Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.cc.2 Received 1024 bytes from 10. LOG_LEVEL_INFO). NS_LOG_WARN and NS_LOG_ERROR.1. Right now.Chapter 5: Tweaking ns-3 29 5. to print more information by setting its logging level via the NS LOG environment variable.1. This line of code enables the LOG_LEVEL_INFO level of logging.2 Received 1024 bytes from 10.1. we are actually enabling the given level and all lower levels.404() Sent 1024 bytes to 10. If you run the script with NS LOG set this way.1.1. the ns-3 logging system will pick up the change and. but first. We can increase the logging level and get more information without changing the script and recompiling by setting the NS LOG environment variable like this: export NS_LOG=UdpEchoClientApplication=level_all This sets the shell environment variable NS_LOG to the string.413s) Sent 1024 bytes to 10. I am going to assume from here on that you are using an sh-like shell that uses the“VARIABLE=value” syntax.2 It turns out that the “Sent” and “Received” messages you see above are actually logging messages from the UdpEchoClientApplication and UdpEchoServerApplication. If you are using a csh-like shell.1.1. When we pass a logging level flag.
Chapter 5: Tweaking ns-3 30 UdpEchoClientApplication:HandleRead(0x6241e0. this will generally be the class name but you should understand that it is not actually a class name.1. The name is not actually a class name. This shows every time a function in the application is called during script execution.1 UdpEchoClientApplication:HandleRead(0x6241e0.1.1.2” comes from.417() UdpEchoClientApplication:Send(): Sent 1024 bytes to 10. Note that there are no requirements in the ns-3 system that models must support any particular logging functionality. 0x624a20) Received 1024 bytes from 10. 0x624a20) UdpEchoClientApplication:HandleRead(): Received 1024 bytes from 10. Try doing the following. export ’NS_LOG=UdpEchoClientApplication=level_all|prefix_func’ Note that the quotes are required since the vertical bar we use to indicate an OR operation is also a Unix pipe connector. The decision regarding how much information is logged is left to the individual model developer.1.1. if you run the script you will see that the logging system makes sure that every message from the given log component is prefixed with the component name.2 UdpEchoClientApplication:StopApplication() UdpEchoClientApplication:DoDispose() UdpEchoClientApplication:~UdpEchoClient() .1. When there is a one-to-one correspondence between a source file and a class. You can now see a log of the function calls that were made to the application. In the case of the echo applications. If you look in the text above. This is intentional. It turns out that in some cases. If you look closely you will notice a single colon between the string UdpEchoClientApplication and the method name where you might have expected a C++ scope operator (::). Now.1. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0. You can resolve this by OR’ing the prefix_func level into the NS_LOG environment variable. and there is a single colon there instead of a double colon to remind you in a relatively subtle way to conceptually separate the logging component name from the class name. it can be hard to determine which method actually generates a log message.1.1. a good deal of log output is available.2 UdpEchoClientApplication:StopApplication() UdpEchoClientApplication:DoDispose() UdpEchoClientApplication:~UdpEchoClient() The additional debug information provided by the application is from the NS LOG FUNCTION level.2 Received 1024 bytes from 10. you may wonder where the string “Received 1024 bytes from 10. it is a logging component name.1.
1. export ’NS_LOG=UdpEchoClientApplication=level_all|prefix_func|prefix_time: UdpEchoServerApplication=level_all|prefix_func|prefix_time’ Again. You can do this by ORing in the prefix time bit..2” is now clearly identified as coming from the echo client application. We can enable that component by entering a colon separated list of components in the NS LOG environment variable. 0x625160) UdpEchoClientApplication:HandleRead(): Received 1024 bytes from 10.1.1. The message “Received 1024 bytes from 10. The remaining message must be coming from the UDP echo server application.Chapter 5: Tweaking ns-3 31 You can now see all of the messages coming from the UDP echo client application are identified as such.1 UdpEchoServerApplication:HandleRead(): Echoing packet UdpEchoClientApplication:HandleRead(0x624920. You may see that this can be very useful in debugging problems. you will have to remove the newline above.1. If you run the script now. export ’NS_LOG=UdpEchoClientApplication=level_all|prefix_func: UdpEchoServerApplication=level_all|prefix_func’ Warning: You will need to remove the newline after the : in the example text above which is only there for document formatting purposes. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.1.1. if you run the script you will see all of the log messages from both the echo client and server applications.2 UdpEchoServerApplication:StopApplication()() It is also sometimes useful to be able to see the simulation time at which a log message is generated. Now UdpEchoServerApplication:HandleRead(): Received 1024 bytes from 10.1.
called at one second. I won’t reproduce the output here (as of this writing it produces 1265 lines of output for the single packet echo) but you can redirect this information into a file and look through it with your favorite editor if you like. Try setting the NS_LOG variable to the following.1.00369s UdpEchoServerApplication:HandleRead(): Received 1024 bytes from 10.1.1. in fact. You can also see that the echo client application is started at a simulation time of two seconds as we requested in the script. 0x624ad0) 2.00737s UdpEchoClientApplication:HandleRead(): Received 1024 bytes from 10. .2 2. I transition into a debugger for a fine-grained examination of the problem.2 10s UdpEchoServerApplication:StopApplication() 10s() You can see that the constructor for the UdpEchoServer was called at a simulation time of 0 seconds.00369s UdpEchoServerApplication:HandleRead(): Echoing packet 2. This will turn on all of the logging in all of the components used in the simulation. You can now see that the StartApplication method of the server is. I can follow the progress of the code quite easily without having to set breakpoints and step through code in a debugger. You can very easily follow the entire process by turning on all of the logging components in the system. This is actually happening before the simulation starts. export ’NS_LOG=*=level_all|prefix_func|prefix_time’ The asterisk above is the logging component wildcard.69 milliseconds./waf --run scratch/myfirst > log.1. after another channel delay. When I have a general idea about what is going wrong.1.cc script started the echo server application at one second into the simulation.out 2>&1 I personally use this extremely verbose version of logging when I am presented with a problem and I have no idea where things are going wrong. and see things happening that I don’t expect. you see the echo client receive the echoed packet in its HandleRead method. This kind of output can be especially useful when your script does something completely unexpected. Note that the elapsed time for the packet to be sent across the point-to-point link is 3. There is a lot that is happening under the covers in this simulation that you are not seeing as well. I can just edit up the output in my favorite editor and search around for things I expect.00737s UdpEchoClientApplication:HandleRead(0x624290. but the time is displayed as zero seconds. Recall that the scratch/first. You see the echo server logging a message telling you that it has echoed the packet and then. The same is true for the UdpEchoClient constructor message.1. You can now follow the progress of the simulation from the ScheduleTransmit call in the client that calls Send to the HandleRead callback in the echo server application. If you are stepping using a debugger you may miss an unexpected excursion completely.Chapter 5: Tweaking ns-3 32 2. Logging the excursion makes it quickly visible. .
1.2 Using Command Line Arguments 5.3 Adding Logging to your Code You can add new logging to your simulations by making calls to the log component via several macros.Create (2). You now know that you can enable all of the logging for this component by setting the NS_LOG environment variable to the various levels. .1 Received 1024 bytes from 10.1. you can enable it by. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.2 Received 1024 bytes from 10. if you run the script.2 5. We provide a mechanism to parse command line arguments and automatically set local and global variables based on those arguments. .1. NodeContainer nodes. NS_LOG_INFO ("Creating Topology").” This is done as in this code snippet.Chapter 5: Tweaking ns-3 33 5.1.1 Overriding Default Attributes Another way you can change how ns-3 scripts behave without editing and building is via command line arguments. If you just want to see this particular level of logging. Open scratch/myfirst.404s) Creating Topology Sent 1024 bytes to 10. Let’s do so in the myfirst. In order to see your message you will have to enable the FirstScriptExample logging component with a level greater than or equal to NS_LOG_INFO. export NS_LOG=FirstScriptExample=info If you now run the script you will see your new “Creating Topology” log message.1. right before the lines. Let’s go ahead and add some logging to the script./waf export NS_LOG= Now.1.1. nodes.cc in your favorite editor and add the line. Go ahead and add one (just before we start creating the nodes) that tells you that the script is “Creating Topology.cc script we have in the scratch directory.2./waf --run scratch/myfirst you will not see your new message since its associated logging component (FirstScriptExample) has not been enabled. Now build the script using waf and clear the NS_LOG variable to turn off the torrent of logging we previously enabled: . The macro used to add an informational level log message is NS_LOG_INFO. Recall that we have defined a logging component in that script: NS_LOG_COMPONENT_DEFINE ("FirstScriptExample").
The help listing says that we should provide a TypeId. This is done quite simply (in your main program) as in the following code. but ask the script for help in the following way. . The quotes are required to sort out which program gets which argument. pointToPoint. --PrintTypeIds: Print all TypeIds. It opens the door to the ns-3 global variable and Attribute systems. In this case it will be ns3::PointToPointNetDevice. . This corresponds to the class name of the class to which the Attributes belong. pointToPoint.. int main (int argc. } This simple two line snippet is actually very useful by itself. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.SetChannelAttribute ("Delay". Go ahead and add that two lines of code to the scratch/myfirst. We looked at the following lines of code. We have already hinted at the ns-3 Attribute system while walking through the first.cc script at the start of main.SetDeviceAttribute ("DataRate"..Chapter 5: Tweaking ns-3 34 The first step in using the command line argument system is to declare the command line parser. --PrintGroup=[group]: Print all TypeIds of group. cmd. --PrintGlobals: Print the list of globals. Let’s focus on the --PrintAttributes option.. The command line parser will now see the --PrintHelp argument and respond with. Let’s use the command line argument parser to take a look at the Attributes of the PointToPointNetDevice..Parse (argc.413s) TcpL4Protocol:TcpStateMachine() CommandLine:HandleArgument(): Handle arg name=PrintHelp value= --PrintHelp: Print this help message. Let’s go ahead and type in. --PrintAttributes=[typeid]: Print all attributes of typeid.cc script. CommandLine cmd. PointToPointHelper pointToPoint./waf --run "scratch/myfirst --PrintHelp" This will ask Waf to run the scratch/myfirst script and pass the command line argument --PrintHelp to the script. argv). . StringValue ("5Mbps")). Go ahead and build the script and run it./waf --run "scratch/myfirst --PrintAttributes=ns3::PointToPointNetDevice" . StringValue ("2ms")). and mentioned that DataRate was actually an Attribute of the PointToPointNetDevice. --PrintGroups: Print the list of groups. char *argv[]) { .
Chapter 5: Tweaking ns-3
35
The system will print out all of the Attributes of this kind of net device. Among the Attributes you will see listed is, --ns3::PointToPointNetDevice::DataRate=[32768bps]: The default data rate for point to point links This is the default value that will be used when a PointToPointNetDevice is created in the system. We overrode this default with the Attribute setting in the PointToPointHelper above. Let’s use the default values for the point-to-point devices and channels by deleting the SetDeviceAttribute call and the SetChannelAttribute call from the myfirst.cc we have in the scratch directory. Your script should now just declare the PointToPointHelper and not do any set operations as in the following example, ... echo server application and turn on the time prefix. export ’NS_LOG=UdpEchoServerApplication=level_all|prefix_time’ If you run the script, you should.405s) 0s UdpEchoServerApplication:UdpEchoServer() 1s UdpEchoServerApplication:StartApplication() Sent 1024 bytes to 10.1.1.2 2.25732s Received 1024 bytes from 10.1.1.1 2.25732s Echoing packet Received 1024 bytes from 10.1.1.2 10 at 2.00369 seconds. 2.00369s UdpEchoServerApplication:HandleRead(): Received 1024 bytes from 10.1.1.1 Now it is receiving the packet at 2.25732 seconds. This is because we just dropped the data rate of the PointToPointNetDevice down to its default of 32768 bits per second from five megabits per second.
Chapter 5: Tweaking ns-3
36
If we were to provide a new DataRate using the command line, we could speed our simulation up again. We do this the result? It turns out that in order to get the original behavior of the script back, we will have to set the speed-of-light delay of the channel as well. We can ask the command line system to print out the Attributes of the channel just like ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.417s) 0s UdpEchoServerApplication:UdpEchoServer() 1s UdpEchoServerApplication:StartApplication() Sent 1024 bytes to 10.1.1.2 2.00369s Received 1024 bytes from 10.1.1.1 2.00369s Echoing packet Received 1024 bytes from 10.1.1.2 10s UdpEchoServerApplication:StopApplication() UdpEchoServerApplication:DoDispose() UdpEchoServerApplication:~UdpEchoServer() Note that the packet is again received by the server at 2.00369 seconds. We could actually set any of the Attributes used in the script in this way. In particular we could set the UdpEchoClient Attribute MaxPackets to some other value than one. How would you go about that? Give it a try. Remember you have to comment out the place we override the default Attribute and explicitly set MaxPackets in the script. Then you have to rebuild the script. You will also have to find the syntax for actually setting the new default atribute value using the command line help facility. Once you have this figured out you should be able to control the number of packets echoed from the command line. Since we’re nice"
Chapter 5: Tweaking ns-3
37
5.2.2 Hooking Your Own Values
You can also add your own hooks to the command line system. This is done quite simply by using the AddValue method to the command line parser. Let’s use this facility to specify the number of packets to echo in a completely different way. Let’s add a local variable called nPackets to the main function. We’ll initialize it to one to match our previous default behavior. To allow the command line parser to change this value, we need to hook the value into the parser. We do this by adding a call to AddValue. Go ahead and change the scratch/myfirst.cc script to start with the following code, int main the variable nPackets instead of the constant 1 as is shown below. echoClient.SetAttribute ("MaxPackets", UintegerValue (nPackets)); Now if you run the script and provide the --PrintHelp argument, you should see your new User Argument listed in the help display. Try, ./waf --run "scratch/myfirst --PrintHelp" Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.403s) -. User Arguments: --nPackets: Number of packets to echo If you want to specify the number of packets to echo, you can now do so by setting the --nPackets argument in the command line, ./waf --run "scratch/myfirst --nPackets=2" You should now see Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.404s)
1 3. 5. } You could even use the logging module to add a little structure to your solution.. and a uniform mechanism for connecting sources to sinks. • Intermediate users must be able to extend the tracing system to modify the output format generated. The basic goals of the ns-3 tracing system are: • For basic tasks. without modifying the core of the simulator.25732s Echoing packet Received 1024 bytes from 10.2 Sent 1024 bytes to 10. or to insert new tracing sources. the tracing system should allow the user to generate standard tracing for popular tracing sources.1. .25732s Echoing packet Received 1024 bytes from 10. you can add new variables to your scripts and hook them into the command line system quite painlessly. Pretty easy. If you are a model author.1.1.1.. There are many well-known problems generated by such approaches and so we have provided a generic event tracing subsystem to address the issues we thought were important.2 10s UdpEchoServerApplication:StopApplication() UdpEchoServerApplication:DoDispose() UdpEchoServerApplication:~UdpEchoServer() You have now echoed two packets. and the ns-3 tracing system is a primary mechanism for this.. Since ns-3 is a C++ program. The ns-3 tracing system is built on the concepts of independent tracing sources and tracing sinks.1.2 2.1.1.Chapter 5: Tweaking ns-3 38 0s UdpEchoServerApplication:UdpEchoServer() 1s UdpEchoServerApplication:StartApplication() Sent 1024 bytes to 10. and to customize which objects generate the tracing.3 Using the Tracing System The whole point of simulation is to generate output for further study.1 2. • Advanced users can modify the simulator core to add new tracing sources and sinks. you can add new Attributes to your Objects and they will automatically be available for setting by your users through the command line system.1. Trace sources are . int main () { .1. isn’t it? You can see that if you are an ns-3 user. std::cout << "The value of x is " << x << std::endl..25732s Received 1024 bytes from 10.1..25732s Received 1024 bytes from 10. standard facilities for generating output from C++ programs could be used: #include <iostream> ..1. If you are a script author.1.2 3. you can use the command line argument system to control global values and Attributes.
"-"./waf --run scratch/myfirst Just as you have seen many times before. The last line of code in the snippet above tells ns-3 that you want to enable ASCII tracing on all point-to-point devices in your simulation. For example.tr generated by many scripts. Let’s just jump right in and add some ASCII tracing output to our scratch/myfirst. in the example above. this type of trace is analogous to the out. "d". they must be “connected” to other pieces of code that actually do something useful with the information provided by the sink. If you enable this functionality.tr”. You can now build the script and run it from the command line: . PointToPointHelper::EnableAsciiAll (ascii). one could create a trace sink that would (when connected to the trace source of the previous example) print out interesting parts of the received packet. you will see output in a ASCII files — thus the name. we will walk through some pre-defined sources and sinks and show how they may be customized with little user effort.tr"). See your favorite C++ tutorial if you are unfamiliar with this code.Chapter 5: Tweaking ns-3 39 entities that can signal events that happen in a simulation and provide access to interesting underlying data. add the following lines of code: std::ofstream ascii. 5.3. The rationale for this explicit division is to allow users to attach new types of sinks to existing tracing sources. and you want the (provided) trace sinks to write out information about packet movement in ASCII format to the stream provided. the traced events are equivalent to the popular trace points that log "+". Trace sources are not useful by themselves. right before the call to Simulator::Run ().1 ASCII Tracing Ns-3 provides helper functionality that wraps the low-level tracing system to help you with the details involved in configuring some easily understood packet traces. . a trace source could indicate when a packet is received by a net device and provide access to the packet contents for interested trace sinks. The first thing you need to do is to add the following include to the top of the script just after the GNU GPL comment: #include <fstream> Then. For example. For those familiar with ns-2.open ("myfirst. The first two lines are just vanilla C++ code to open a stream that will be written to a file named “myfirst.cc script. In this tutorial. and "r" events. a user could define a new tracing sink in her script and attach it to an existing tracing source defined in the simulation core by editing only the user script. Thus. you will see some messages from Waf and then “’build’ finished successfully” with some number of messages from the running program. Trace sinks are consumers of the events and data provided by the trace sources. ascii. See the ns-3 manual or how-to sections for information on advanced tracing configuration including extending the tracing namespace and creating new tracing sources. For those familiar with ns-2 output. without requiring editing and recompilation of the core of the simulator.
Note that each line in the trace file begins with a lone character (has a space after it). This corresponds to a container managed in the ns-3 core code that contains all of the nodes that are created in a script.1 Parsing Ascii Traces There’s a lot of information there in a pretty dense form. You can think of the tracing namespace somewhat like you would a filesystem namespace.tr in your favorite editor. The second line (reference 01) is the simulation time expressed in seconds.1. typically because the queue was full.Chapter 5: Tweaking ns-3 40 When it ran. • -: A dequeue operation occurred on the device queue. Let’s take a more detailed view of the first line in the trace file. It may be difficult to see this clearly unless you widen your window considerably. We have a + character. 5. Each line in the file corresponds to a trace event.1. If you want to control where the traces are saved you can use the --cwd option of Waf to specify this.1. Just as a filesystem may have directories under the root. I’ll break it down into sections (indented for clarity) with a two digit reference number on the left side: 00 01 02 03 04 05 06 07 08 09 10 + 2 /NodeList/0/DeviceList/0/$ns3::PointToPointNetDevice/TxQueue/Enqueue ns3::PppHeader ( Point-to-Point Protocol: IP (0x0021)) ns3::Ipv4Header ( tos 0x0 ttl 64 id 0 protocol 17 offset 0 flags [none] length: 1052 10. indeed. You may recall that we asked the UdpEchoClientApplication to start sending packets at two seconds.tr. Here we see confirmation that this is. The root of the namespace is the NodeList. Because of the way that Waf works. In this case we are tracing events on the transmit queue present in every point-to-point net device in the simulation. thus we need to change into the top level directory of our repo and take a look at the ASCII trace file myfirst. the program will have created a file named myfirst. it is created at the top-level directory of the repository by default. we may have node numbers in the NodeList. the file is not created in the local directory. • r: A packet was received by the net device. happening. so this corresponds to an enqueue operation on the transmit queue.2) ns3::UdpHeader ( length: 1032 49153 > 9) Payload (size=1024) The first line of this expanded trace event (reference number 00) is the operation. but the first thing to notice is that there are a number of distinct lines in this file.1.1 > 10. We have not done so.3. The transmit queue is a queue through which every packet destined for a point-to-point channel must pass. • d: A packet was dropped. The string /NodeList/0 therefore refers to . The next line of the example trace (reference 02) tell us which trace source originated this event (expressed in the tracing namespace). This character will have the following meaning: • +: An enqueue operation occurred on the device queue.1.
The most popular program that can read and display this format is Wireshark (formerly called Ethereal).1 and is destined for 10. This time should be familiar as you have seen it before in a previous section.1. References 08-09 show that this packet has a UDP header and. we concentrate on viewing pcap traces with tcpdump.3.1.1.1.pcap format.2) 06 ns3::UdpHeader ( 07 length: 1032 49153 > 9) 08 Payload (size=1024) Notice that the trace operation is now r and the simulation time has increased to 2. 5. Recall that the operation + found at reference 00 meant that an enqueue operation happened on the transmit queue of the device.1. and is actually an API that includes the definition of a .25732 02 /NodeList/1/DeviceList/0/$ns3::PointToPointNetDevice/MacRx 03 ns3::Ipv4Header ( 04 tos 0x0 ttl 64 id 0 protocol 17 offset 0 flags [none] 05 length: 1052 10.25732 seconds. The trace source namespace entry (reference 02) has changed to reflect that this event is coming from node 1 (/NodeList/1) and the packet reception trace source (/MacRx).Chapter 5: Tweaking ns-3 41 the zeroth node in the NodeList which we typically think of as “node 0”. . The next line in the trace file shows the same packet being dequeued from the transmit queue on the same node. However.1 > 10. The code used to enable pcap tracing is a one-liner. reference 10 shows that the payload is the expected 1024 bytes.1. In this tutorial. 00 r 01 2.2. You can see that this trace event comes from DeviceList/0 which is the zeroth device installed in the node. References 05-07 show that the packet has an IP version four header and has originated from IP address 10. The next string. there are many traffic trace analyzers that use this packet format. References 03-04 indicate that the packet is encapsulated in the point-to-point protocol.2 PCAP Tracing The ns-3 device helpers can also be used to create trace files in the . The Third line in the trace file shows the packet being received by the net device on the node with the echo server. This list appears next in the namespace. The remaining lines in the trace should be fairly intuitive.1. $ns3::PointToPointNetDevice tells you what kind of device is in the zeroth position of the device list for node zero. If you have been following the tutorial steps closely this means that you have left the DataRate of the net devices and the channel Delay set to their default values. I have reproduced that event below. In each node there is a list of devices that have been installed. This is reflected in the final segments of the “trace path” which are TxQueue/Enqueue.pcap file format. finally. It should be quite easy for you to follow the progress of the packet through the topology by looking at the rest of the traces in the file. The acronym pcap (usually written in lower case) stands for packet capture. We encourage users to exploit the many tools available for analyzing pcap traces.1.
length 1024 2. Go ahead and insert this line of code after the ASCII tracing code we just added to scratch/myfirst. we will eventually see files named “myfirst-0-0.pcap. You see the packet being echoed back at 2. myfirst-0-0.1.49153: UDP.1.514648 IP 10.257324 IP 10. 5.2.1.1.257324 IP 10.pcap” and “myfirst1-0.1 Reading output with tcpdump The easiest thing to do at this point will be to use tcpdump to look at the pcap files.pcap and myfirst-1-0.1. If you have Wireshark available.cc.pcap reading from file myfirst-0-0. you should now see three log files: myfirst.2 Reading output with Wireshark If you are unfamilar with Wireshark. The file names will be built using the prefix.1. Once you have added the line of code to enable pcap tracing.pcap) you can see that packet being received at 2.1. Wireshark is a graphical user interface which can be used for displaying these trace files.2.257324 seconds in the second dump.2. If you look at the second dump (myfirst-1-0.2.pcap are the new pcap files we just generated. Notice that we only passed the string “myfirst.9 > 10. respectively.wireshark. not a complete file name.257324 seconds.pcap.49153 > 10.1.514648 seconds.1.2./waf --run scratch/myfirst If you look at the top level directory of your distribution.000000 IP 10. you can open each of the trace files and display the contents as if you had captured the packets using a packet sniffer.49153 > 10. length 1024 tcpdump -nn -tt -r myfirst-1-0.9: UDP.1.pcap (the client device) that the echo packet is sent at 2 seconds into the simulation. you can run the script in the usual way: . tcpdump -nn -tt -r myfirst-0-0.1.1. the node number. link-type PPP (PPP) 2.org/.tr is the ASCII trace file we have previously examined. In our example script. link-type PPP (PPP) 2. and finally.” and not “myfirst.1. you see the packet being received back at the client in the first dump at 2.9 > 10.pcap” suffix.1.1.1. the device number and a “.pcap” which are the pcap traces for node 0-device 0 and node 1-device 0. .1.1. there is a web site available from which you can download programs and documentation: reading from file myfirst-1-0.Chapter 5: Tweaking ns-3 42 PointToPointHelper::EnablePcapAll ("myfirst"). 5. length 1024 2.1. This is because the parameter is a prefix.1.pcap” or something similar. length 1024 You can see in the dump of myfirst-0-0. The helper will actually create a trace file for every point-to-point device in the simulation.3.49153: UDP.3.9: UDP.
This script builds on the first. Ns-3 provides a net device and channel we call CSMA (Carrier Sense Multiple Access).2. A real Ethernet uses CSMA/CD (Carrier Sense Multiple Access with Collision Detection) scheme with exponentially increasing backoff to contend for the shared transmission medium.cc example. Just as we have seen point-to-point topology helper objects when constructing pointto-point topologies. Go ahead and open examples/tutorial/second.0 Then the ns-3 namespace is used and a logging component is defined.cc script and adds a CSMA network to the point-to-point simulation we’ve already considered. This is all just as it was in first. The appearance and operation of these helpers should look quite familiar to you.h" One thing that can be surprisingly useful is a small bit of ASCII art that shows a cartoon of the network topology constructed in the example. we will see equivalent CSMA topology helpers in this section. The ns-3 CSMA device models a simple network in the spirit of Ethernet. there will be a total of two nodes on the LAN (CSMA channel) — one required node and one “extra” node.n1 n2 n3 n4 // point-to-point | | | | // ================ // LAN 10. We provide an example script in our examples/tutorial directory. By default there are three “extra” nodes as seen below: // Default Network Topology // // 10. .1.1 Building a Bus Network Topology In this section we are going to expand our mastery of ns-3 network devices and channels to cover an example of a bus network.cc example (and in all ns-3 examples) the file begins with an emacs mode line and some GPL boilerplate.Chapter 6: Building Topologies 43 6 Building Topologies 6.h" #include "ns3/simulator-module. Just as in the first. In this case.cc. Notice that this is the default network topology since you can actually vary the number of nodes created on the LAN. You will have already seen enough ns-3 code to understand most of what is going on in this example. The actual code begins by loading module include files just as was done in the first. If you set nCsma to one.1. but we will go over the entire script and examine some of the output.cc in your favorite editor. You will find a similar “drawing” in most of our examples.0 // n0 -------------.1. you can see that we are going to extend our point-to-point example (the link between the nodes n0 and n1 below) by hanging a bus network off of the right side.h" #include "ns3/helper-module. The ns-3 CSMA device and channel models only a subset of this. #include "ns3/core-module.h" #include "ns3/node-module. so there is nothing new yet.
CommandLine cmd.AddValue (‘‘nCsma’’. csmaNodes. csmaNodes. .Create (nCsma). the number of “extra” nodes means the number nodes you desire in the CSMA section minus one. verbose). The main program begins with a slightly different twist. The NodeContainer is used to do this just as was done in first. LOG_LEVEL_INFO). cmd.Chapter 6: Building Topologies 44 using namespace ns3. LogComponentEnable(‘‘UdpEchoServerApplication’’. NodeContainer csmaNodes. The next step is to create two nodes that we will connect via the point-to-point link. NodeContainer p2pNodes. ‘‘Tell echo applications to log if true’’. We then create a number of “extra” nodes that compose the remainder of the CSMA network. The next line of code Gets the first node (as in having an index of one) from the point-topoint node container and adds it to the container of nodes that will get CSMA devices. The node in question is going to end up with a point-to-point device and a CSMA device. we declare another NodeContainer to hold the nodes that will be part of the bus (CSMA) network. cmd. nCsma). This flag defaults to true (the logging components are enabled) but allows us to turn off logging during regression testing of this example. } nCsma = nCsma == 0 ? 1 : nCsma. First. bool verbose = true.cc. if (verbose) { LogComponentEnable(‘‘UdpEchoClientApplication’’. p2pNodes. NS_LOG_COMPONENT_DEFINE ("SecondScriptExample"). we just instantiate the container object itself. uint32_t nCsma = 3. cmd. We use a verbose flag to determine whether or not the UdpEchoClientApplication and UdpEchoServerApplication logging components are enabled. We did something similar when we allowed the number of packets sent to be changed in the section on command line arguments. ‘‘Number of \"extra\" CSMA nodes/devices’’. You will see some familiar code that will allow you to change the number of devices on the CSMA network via command line argument.Add (p2pNodes.Create (2).argv). The code consists of variations of previously covered API so you should be entirely comfortable with the following code at this point in the tutorial.Parse (argc.Get (1)). Next. The last line makes sure you have at least one “extra” node. LOG_LEVEL_INFO). Since we already have one node in the CSMA network – the one that will have both a pointto-point and CSMA net device.AddValue (‘‘verbose’’.
Install (p2pNodes. but it creates and connects CSMA devices and channels. Just as we created a NetDeviceContainer to hold the devices created by the PointToPointHelper we create a NetDeviceContainer to hold the devices created by our CsmaHelper. First we use the network 10.1. We call the Install method of the CsmaHelper to install the devices into the nodes of the csmaNodes NodeContainer. and the next lines introduce them. p2pDevices = pointToPoint.1.Get (0)). stack.0 to create the two addresses needed for our two point-to-point devices.SetChannelAttribute ("Delay". We mentioned above that you were going to see a helper for CSMA devices and channels. We now have our nodes. The CsmaHelper works just like a PointToPointHelper.Install (p2pNodes).SetDeviceAttribute ("DataRate". we are going to use the Ipv4AddressHelper to assign IP addresses to our device interfaces. but we have no protocol stacks present. csmaDevices = csma.Chapter 6: Building Topologies 45 The next bit of code should be quite familiar by now. NetDeviceContainer p2pDevices. NetDeviceContainer csmaDevices.cc example script. stack.Install (csmaNodes). Recall that we took one of the nodes from the p2pNodes container and added it to the csmaNodes container. pointToPoint. We instantiate a PointToPointHelper and set the associated default Attributes so that we create a five megabit per second transmitter on devices created using the helper and a two millisecond delay on channels created by the helper. devices and channels created. We first set the data rate to 100 megabits per second. InternetStackHelper stack.Install (csmaNodes). for example. we will use the InternetStackHelper to install these stacks. CsmaHelper csma.SetChannelAttribute ("DataRate". Just as in the first.cc script. TimeValue (NanoSeconds (6560))). csma. csma. StringValue ("5Mbps")). In the case of a CSMA device and channel pair. and then set the speed-of-light delay of the channel to 6560 nano-seconds (arbitrarily chosen as 1 nanosecond per foot over a 100 meter segment). StringValue ("100Mbps")). and all of the nodes in the csmaNodes container to cover all of the nodes in the simulation. . PointToPointHelper pointToPoint. We then instantiate a NetDeviceContainer to keep track of the point-to-point net devices and we Install devices on the point-to-point nodes. Just as in the first.SetChannelAttribute ("Delay". StringValue ("2ms")). Thus we only need to install the stacks on the remaining p2pNodes node. Notice that you can set an Attribute using its native data type. pointToPoint. notice that the data rate is specified by a channel Attribute instead of a device Attribute. This is because a real CSMA network does not allow one to mix. 10Base-T and 100Base-T devices on a given channel.
"255. UdpEchoServerHelper echoServer (9). UintegerValue (1)).255.1.GetAddress (nCsma). UintegerValue (1024)).Stop (Seconds (10. Now we have a topology built. First. . TimeValue (Seconds (1.Install (p2pNodes. Ipv4InterfaceContainer csmaInterfaces. echoClient. address. Recall that this port can be changed later using the SetAttribute method if desired.255.2. "255.0 in this case. p2pInterfaces = address. Recall that the csmaNodes NodeContainer contains one of the nodes created for the point-to-point network and nCsma “extra” nodes.SetBase ("10.cc but we are going to instantiate the server on one of the nodes that has a CSMA device and the client on the node having only a point-to-point device. is if we create one “extra” CSMA node. as seen below.0"). By induction. The zeroth entry of the csmaNodes container will be the point-to-point node.cc example script. We now need to assign IP addresses to our CSMA device interfaces.0"). except we now are performing the operation on a container that has a variable number of CSMA devices — remember we made the number of CSMA devices changeable by command line argument.1. we set up the echo server. csmaInterfaces = address. We create a UdpEchoServerHelper and provide a required Attribute value to the constructor which is the server port number.0". echoClient. We tell the client to send packets to the server we just installed on the last of the “extra” CSMA nodes. serverApps. UdpEchoClientHelper echoClient (csmaInterfaces.SetAttribute ("PacketSize". echoClient. serverApps. Again. The CSMA devices will be associated with IP addresses from network number 10.0)).Get (0)).Start (Seconds (1. if we create nCsma “extra” nodes the last one will be at index nCsma. The easy way to think of this.Get (nCsma)). then it will be at index one of the csmaNodes container.))).SetAttribute ("MaxPackets". We install the client on the leftmost point-to-point node seen in the topology illustration.SetBase ("10. address.SetAttribute ("Interval".2.Install (csmaNodes.1. but we require it to be provided to the constructor.0".Assign (p2pDevices). then. Recall that we save the created interfaces in a container to make it easy to pull out addressing information later for use in setting up the applications. What we want to get at is the last of the “extra” nodes. we provide required Attributes to the UdpEchoClientHelper in the constructor (in this case the remote address and port).Chapter 6: Building Topologies 46 Ipv4AddressHelper address. This section is going to be fundamentally similar to the applications section of first.Assign (csmaDevices). 9).0)). ApplicationContainer serverApps = echoServer.255. You see this exhibited in the Get of the first line of code. Ipv4InterfaceContainer p2pInterfaces.1. The client application is set up exactly as we did in the first. but we need applications. The operation works just as it did for the point-to-point case. ApplicationContainer clientApps = echoClient.255.
Global routing takes advantage of the fact that the entire internetwork is accessible in the simulation and runs through the all of the nodes created for the simulation — it does the hard work of setting up routing for you without having to configure routers. In this case. That single device then “sniffs” the network for all packets and stores them in a single pcap file. we specify the device using csmaDevices.cc . This is how tcpdump. clientApps. csmaDevices. If you were on a Linux machine you might do something like tcpdump -i eth0 to get the trace. works. Each node generates link advertisements and communicates them directly to a global route manager which uses this global information to construct the routing tables for each node. ns-3 provides what we call global routing to help you out. Simulator::Run (). return 0.Start (Seconds (2. cp examples/tutorial/second. Next we enable pcap tracing. Another way is to pick one of the devices and place it in promiscuous mode. In this example. That final parameter tells the CSMA helper whether or not to arrange to capture packets in promiscuous mode.Get(1). One way is to create a trace file for each net device and store only the packets that are emitted or consumed by that net device.cc example.cc example script into the scratch directory and use waf to build just as you did with the first. which selects the first device in the container. PointToPointHelper::EnablePcapAll ("second"). for example.Get (1). Setting up this form of routing is a one-liner: Ipv4GlobalRoutingHelper::PopulateRoutingTables (). we are going to select one of the devices on the CSMA network and ask it to perform a promiscuous sniff of the network. we need some form of internetwork routing. Setting the final parameter to true enables promiscuous captures. Basically.0)).cc example.cc scratch/mysecond. If you are in the top-level directory of the repository you just type. copy the second. CsmaHelper::EnablePcap ("second". The last section of code just runs and cleans up the simulation just like the first. Each of these endpoints has a net device associated with it.Stop (Seconds (10./waf . thereby emulating what tcpdump would do. Since we have actually built an internetwork here. what happens is that each node behaves as if it were an OSPF router that communicates instantly and magically with all other routers behind the scenes. The second line enables pcap tracing in the CSMA helper and there is an extra parameter you haven’t encountered yet. } In order to run this example.0)). Simulator::Destroy (). There are two basic alternatives to gathering trace information from such a network. true). This means that there can (and are in this case) multiple endpoints on a shared medium. The CSMA network is a multi-point-to-point network. The first line of code to enable pcap tracing in the pointto-point helper should be familiar to you by now.Chapter 6: Building Topologies 47 clientApps.
pcap.1.1.2. If you are following the tutorial religiously (you are. The second message. “Received 1024 bytes from 10.2.4.49153: UDP. aren’t you) you will still have the NS LOG variable set. you will find three trace files: second-0-0. “Sent 1024 bytes to 10. and the file second-2-0.1.2.cc suggested above.1. the server is on a different network (10. export NS_LOG= . tcpdump -nn -tt -r second-0-0.” is the UDP echo client sending a packet to the server.415s) Sent 1024 bytes to 10. <name>-<node>-<device>.1.007602 IP 10.9: UDP.2. You will see that node two is the first “extra” node on the CSMA network and its device zero was selected as the device to capture the promiscuous-mode trace.cc as one of our regression tests to verify that it works exactly as we think it should in order to make your tutorial experience a positive one. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0.2. the first file in the listing is second-00. They all have the same form.” is from the UDP echo server.pcap is the pcap trace for device zero on node one. In this case. you will see that node zero is the leftmost node of the point-to-point link and node one is the node that has both a point-to-point device and a CSMA device. First. also a point-to-point net device.1. so go ahead and clear that variable and run the program.4 Received 1024 bytes from 10. The final message.1.cc. generated when it receives the echo packet.0). The file second-1-0.1.1.1.49153 > 10. “Received 1024 bytes from 10.1. If you now go and look in the top level directory.pcap Let’s take a moment to look at the naming of these files. please do the renaming to mysecond.1. device zero. length 1024 .pcap second-2-0. For example. If you refer back to the topology illustration at the start of the section.1.2. you will see similar output when you run the script.4.pcap which is the pcap trace from node zero.000000 IP 10.1. link-type PPP (PPP) 2./waf --run scratch/mysecond Since we have set up the UDP echo applications to log just as we did in first. This is the point-to-point net device on node zero.pcap.1. This means that an executable named second already exists in the project.1.1. Now.9 > 10. do a tcpdump of the trace file for the leftmost point-to-point node — node zero.pcap second-1-0.pcap You should see the contents of the pcap file displayed: reading from file second-0-0.1 Received 1024 bytes from 10. length 1024 2.pcap is the pcap trace for device zero on node two. indicating that it has received its echo back from the server.1.Chapter 6: Building Topologies 48 Warning: We use the file second.” is from the echo client.4 Recall that the first message.2.4.4. To avoid any confusion about what you are executing. let’s follow the echo packet through the internetwork.
2.003707 arp reply 10.1.1. length 1024 As you can see.1.9 > 10.1 (that was sent at 2.1.1.2.4.2. Node one knows it needs to send the packet to IP address 10.9: UDP. This is because we initialized global routing and it has figured all of this out for us.1.9 > 10.1 (ff:ff:ff:ff:ff:ff) tell 10.4 (ff:ff:ff:ff:ff:ff) tell 10. length 1024 The server receives the echo request and turns the packet around trying to send it back to the source.4 2.49153 > 10.2.1.2. link-type PPP (PPP) 2.4.003811 arp who-has 10.000000 seconds) headed toward IP address 10.1. the rightmost node replies saying it is at MAC address 00:00:00:00:00:06.003801 IP 10.4 is-at 00:00:00:00:00:06 2.1.2. 2.4.pcap.1 headed for IP address 10.1. but is sniffing the network and reporting all of the traffic it sees.2.1.4.4.4.49153 > 10.1. tcpdump -nn -tt -r second-2-0.1. The server knows that this address is on another network that it reaches via IP address 10.1.1. the echo server node doesn’t know the MAC address of the first CSMA node.pcap and see if its there.4 (the rightmost CSMA node).pcap You should now see the promiscuous dump of node two.1 2.003707 arp reply 10.1 2.4 is-at 00:00:00:00:00:06 Then node one. the packet will be forwarded to the CSMA interface and we should see it pop out on that device headed for its ultimate destination.49153: UDP.1. You see the packet from IP address 10. You then see the echo packet leaving node zero via the device associated with IP address 10.1.9: UDP.1.1. Remember that we selected node 2 as the promiscuous sniffer node for the CSMA network so let’s then look at second-2-0.003822 arp reply 10.Chapter 6: Building Topologies 49 The first line of the dump indicates that the link type is PPP (point-to-point) which we expect.2.1.1.1.2. length 1024 Here we see that the link type is also PPP as we would expect.1.1.1.1.1. This packet will move over the point-to-point link and be received by the point-to-point net device on node one. It broadcasts on the CSMA network (ff:ff:ff:ff:ff:ff) asking for the device that has IP address 10.2.2.1.003801 IP 10.4 (ff:ff:ff:ff:ff:ff) tell 10.pcap You should now see the pcap trace output of the other side of the point-to-point link: reading from file second-1-0.2. but it doesn’t know the MAC address of the corresponding node. In this case.1.1. 2.2.1. Now.pcap.4.1.003686 IP 10.49153 > 10.2. so it has to ARP for it just like the first CSMA node had to do.2.1. length 1024 2. Let’s take a look: tcpdump -nn -tt -r second-1-0.1.1.2. the Address Resolution Protocol.1.4. Something new has appeared.003696 arp who-has 10.2.2.9: UDP. length 1024 2.003915 IP 10. though.2.1 is-at 00:00:00:00:00:03 2. But. device zero: reading from file second-2-0.1.1. internally to this node. device one goes ahead and sends the echo packet to the UDP echo server at IP address 10. The bus network needs ARP.2. link-type EN10MB (Ethernet) 2. Note that node two is not directly involved in this exchange.1. the link type is now “Ethernet”.4 appear on this interface.003696 arp who-has 10.003915 IP 10.49153: UDP.1. . This exchange is seen in the following lines.
1.49153: UDP. length 1024 Finally.9 > 10. link-type PPP (PPP) 2.2.003915 IP 10.405s) Sent 1024 bytes to 10.1 Received 1024 bytes from 10.1. which is 10. 10. We know that we want to create a pcap file with the base name "second" and we also know that the device of interest in both cases is going to be zero. You can do this fairly easily/ Let’s take a look at scratch/mysecond.1 (ff:ff:ff:ff:ff:ff) tell 10.007602 IP 10.2.1.1.Chapter 6: Building Topologies 50 2.4. .1 is-at 00:00:00:00:00:03 The server then sends the echo back to the forwarding node.1.1.2.5 instead of the default case.pcap You can now see the echoed packet coming back onto the point-to-point link as the last line of the trace dump.pcap and see that the echoed packet arrives back at the source at 2.1.1.4. CsmaHelper::EnablePcap ("second". Go ahead and replace the EnablePcap calls with the calls below.2. tcpdump -nn -tt -r second-1-0. csmaNodes.1.49153 > 10.9 > 10. link-type PPP (PPP) 2.1.4./waf --run "scratch/mysecond --nCsma=4" You should now see.1.1. you can look back at the node that originated the echo tcpdump -nn -tt -r second-0-0.4.4.9: UDP.49153: UDP.1. false). 0). Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0. 2. length 1024 2.007602 seconds.1. reading from file second-1-0. so those parameters are not really interesting. reading from file second-0-0.1.1. You may really want to get a trace from a single device and you may not be interested in any other traffic on the network.cc and add that code enabling us to be more specific.1.4. ns-3 helpers provide methods that take a node number and device number as parameters. Try running the program with the number of “extra” devices set to four: .2.pcap.Get (nCsma-1)->GetId (). recall that we added the ability to control the number of CSMA devices in the simulation by command line argument. 0.cc example.1.2. CsmaHelper::EnablePcap ("second".1.1.4 2.2.1.2. PointToPointHelper::EnablePcap ("second".Get (nCsma)->GetId (). 0.1.1.2. length 1024 Lastly.2.000000 IP 10.1.9 > 10. length 1024 Looking back at the rightmost node of the point-to-point link. p2pNodes.003822 arp reply 10.1.pcap. length 1024 2.1.5 Notice that the echo server has now been relocated to the last of the CSMA nodes.1.49153 > 10. csmaNodes.2. It is possible that you may not be satisfied with a trace file generated by a bystander in the CSMA network.003811 arp who-has 10.49153: UDP.5 Received 1024 bytes from 10.9: UDP. false).003686 IP 10.1.Get (0)->GetId ().2. You can change this argument in the same way as when we looked at changing the number of packets echoed in the first.003915 IP 10.1.
/waf --run "scratch/mysecond --nCsma=100" you will.pcap rm *.1. . The Node Object has a method called GetId which will return that node’s ID. If you take a look at the network topology illustration at the beginning of the file.2. rm *.pcap is the “leftmost” point-to-point device which is the echo packet source. we also requested a non-promiscuous trace for the next-to-last node.tr If you build the new script and run the simulation setting nCsma to 100. Go ahead and take a look at the tcpdump for second-100-0.2. is to realize that the NodeContainers contain pointers to ns-3 Node Objects. nodes are numbered in a monotonically increasing fashion starting from zero in the order in which you created them.Chapter 6: Building Topologies 51 In order to get the node number.101 Note that the echo server is now located at 10. you will be taken to the detailed documentation for the method. You may have noticed that the final parameter on the call to enable pcap tracing on the echo server node was false.101 Received 1024 bytes from 10. Go to the Doxygen documentation for your release (recall that you can find it on the project web site).pcap second-101-0.pcap second-100-0. Using the GetId method can make determining node numbers much easier in complex topologies. Select ns3::Node and you will be taken to the documentation for the Node class.101 which corresponds to having 100 “extra” CSMA nodes with the echo server on the last one. The file second-101-0. One way to get a node number is to figure this number out “manually” by contemplating the order of node creation. we did this for you and you can see that the last CSMA node is going to be node number nCsma + 1. Let’s go take a look at the Doxygen for the Node and locate that method.1. This means that the trace gathered on that node was in non-promiscuous mode. If you now scroll down to the GetId method and select it.1. This approach can become annoyingly difficult in larger simulations. Let’s clear the old trace files out of the top-level directory to avoid confusion about what is going on.1. An alternate way.pcap The trace file second-0-0.2. you have two choices: first. which is further down in the ns-3 core code than we’ve seen so far. which we use here. which is the node number we seek. You can get to the Node documentation by looking through at the “Classes” tab and scrolling down the “Class List” until you find ns3::Node.1 Received 1024 bytes from 10.pcap corresponds to the rightmost CSMA device which is where the echo server resides. If you list the pcap files in the top level directory you will see. . but sometimes you have to search diligently for useful things. To illustrate the difference between promiscuous and non-promiscuous traces.407s) Sent 1024 bytes to 10. second-0-0.pcap.
101 (ff:ff:ff:ff:ff:ff) tell 10.2. Similarly.1. Ns-3 provides a set of 802.11 models that attempt to provide an accurate MAC-level implementation of the 802.101.1. Just as in the second.003822 arp reply 10. it will give you a number of “extra” CSMA nodes.2. reading from file second-100-0.1.1. we will see equivalent Wifi topology helpers in this section.1.1.1 (ff:ff:ff:ff:ff:ff) tell 10.pcap. The appearance and operation of these helpers should look quite familiar to you.pcap You can now see that node 100 is really a bystander in the echo exchange.2. tcpdump -nn -tt -r second-101-0.2.cc example (and in all ns-3 examples) the file begins with an emacs mode line and some GPL boilerplate. Notice that this is a default network topology since you can actually vary the number of nodes created on the wired and wireless networks.1. so we will go over the entire script and examine some of the output. Take a look at the ASCII art (reproduced below) that shows the default network topology constructed in the example.1.cc script case.1.1. By default there are three “extra” CSMA nodes and three wireless STA nodes.1.cc in your favorite editor.1.49153: UDP.2. link-type EN10MB (Ethernet) 2.003822 IP 10.003696 arp who-has 10. Just as we have seen both point-to-point and CSMA topology helper objects when constructing point-to-point topologies.1 (ff:ff:ff:ff:ff:ff) tell 10.9: UDP.101. This script builds on the second.2. The only packets that it receives are the ARP requests which are broadcast to the entire CSMA network.1.1.2.2.003696 arp who-has 10.101 Now take a look at the tcpdump for second-101-0. . if you change nCsma.1.11 specification and a “not-so-slow” PHY-level model of the 802. you can set nWifi to control how many STA (station) nodes are created in the simulation.2.1.003801 IP 10.11a specification.2. You will have already seen enough ns-3 code to understand most of what is going on in this example.cc script and adds a Wifi network.1.2 Building a Wireless Network Topology In this section we are going to further expand our knowledge of ns-3 network devices and channels to cover an example of a wireless network. reading from file second-101-0.1 2.2. but there are a few new things. length 1024 6.Chapter 6: Building Topologies 52 tcpdump -nn -tt -r second-100-0.9 > 10.1 is-at 00:00:00:00:00:03 2.101 (ff:ff:ff:ff:ff:ff) tell 10. You can see that we are going to further extend our example by hanging a wireless network off of the left side. Just as in the second.003811 arp who-has 10.pcap.1.101 2. link-type EN10MB (Ethernet) 2. There will always be one AP (access point) node on the wireless network.49153 > 10.101 is-at 00:00:00:00:00:67 2. length 1024 2. We provide an example script in our examples/tutorial directory.2.pcap You can now see that node 101 is really the participant in the echo exchange.003801 arp who-has 10.pcap.1 2.003696 arp reply 10. Go ahead and open examples/tutorial/third.
cc by adding some command line parameters for enabling or disabling logging components and for changing the number of devices created.0 // n5 n6 n7 n0 -------------.AddValue (‘‘nWifi’’. #include #include #include #include #include #include "ns3/core-module.0 network as shown on the left side of the illustration. The main program begins just like second. ‘‘Number of \"extra\" CSMA nodes/devices’’. After the illustration.cc example.2. uint32_t nWifi = 3. if (verbose) { LogComponentEnable(‘‘UdpEchoClientApplication’’.1. ‘‘Number of wifi STA devices’’.0 You can see that we are adding a new network device to the node on the left side of the point-to-point link that becomes the access point for the wireless network. A number of wireless STA nodes are created to fill out the new 10. CommandLine cmd.h" "ns3/mobility-module. the ns-3 namespace is used and a logging component is defined. cmd. nWifi).h" "ns3/wifi-module. cmd.Parse (argc.AddValue (‘‘nCsma’’.Chapter 6: Building Topologies 53 The code begins by loading module include files just as was done in the second. NS_LOG_COMPONENT_DEFINE ("ThirdScriptExample"). . uint32_t nCsma = 3.1.0 // AP // * * * * // | | | | 10. verbose). cmd.n1 n2 n3 n4 // point-to-point | | | | // ================ // LAN 10.1.h" "ns3/node-module. cmd.3. There are a couple of new includes corresponding to the Wifi module and the mobility module which we will discuss below.h" The network topology illustration follows: // Default Network Topology // // Wifi 10.AddValue (‘‘verbose’’.1. LOG_LEVEL_INFO).1. ‘‘Tell echo applications to log if true’’.argv). using namespace ns3. nCsma). This should all be quite familiar by now. bool verbose = true.3.h" "ns3/helper-module.h" "ns3/simulator-module.
Create (2).Chapter 6: Building Topologies 54 LogComponentEnable(‘‘UdpEchoServerApplication’’. StringValue ("100Mbps")). StringValue ("2ms")). StringValue ("5Mbps")).Install (p2pNodes). and we are going to use the “leftmost” node of the point-to-point link as the node for the access point. CsmaHelper csma.Get (1)). We are going to create a number of “station” nodes as specified by the command line argument. p2pDevices = pointToPoint.Add (p2pNodes. The next line of code Gets the first node (as in having an index of one) from the point-topoint node container and adds it to the container of nodes that will get CSMA devices.SetChannelAttribute ("Delay". NetDeviceContainer p2pDevices.SetChannelAttribute ("DataRate". The node in question is going to end up with a point-to-point device and a CSMA device.Get (0). the next step is to create two nodes that we will connect via the point-to-point link. PointToPointHelper pointToPoint. pointToPoint. TimeValue (NanoSeconds (6560))). csmaNodes. NodeContainer wifiApNode = p2pNodes. p2pNodes. LOG_LEVEL_INFO). We create a NetDeviceContainer to keep track of the created CSMA net devices and then we Install CSMA devices on the selected nodes. We then create a number of “extra” nodes that compose the remainder of the CSMA network. Next. First. } Just as in all of the previous examples. csma. We then Intall the devices on the nodes and the channel between them. we configure the PHY and channel helpers: . NodeContainer csmaNodes. we declare another NodeContainer to hold the nodes that will be part of the bus (CSMA) network. We instantiate a PointToPointHelper and set the associated default Attributes so that we create a five megabit per second transmitter on devices created using the helper and a two millisecond delay on channels created by the helper. Next. NodeContainer p2pNodes. We then instantiate a CsmaHelper and set its Attributes as we did in the previous example.Create (nCsma).SetDeviceAttribute ("DataRate".Install (csmaNodes).SetChannelAttribute ("Delay".Create (nWifi). we are going to create the nodes that will be part of the Wifi network. csmaNodes. Next. csma. wifiStaNodes. NetDeviceContainer csmaDevices. we see an old friend. NodeContainer wifiStaNodes. The next bit of code constructs the wifi devices and the interconnection channel between these wifi nodes. pointToPoint. csmaDevices = csma.
SsidValue (ssid).Chapter 6: Building Topologies 55 YansWifiChannelHelper channel = YansWifiChannelHelper::Default (). We begin this process by changing the default Attributes of the NqosWifiMacHelper to reflect the requirements of the AP. the “ActiveProbing” Attribute is set to false. the SSID of the infrastructure network we want to setup and make sure that our stations don’t perform active probing: Ssid ssid = Ssid ("ns-3-ssid"). The SetRemoteStationManager method tells the helper the type of rate control algorithm to use. This code first creates an 802. both at the MAC and PHY layers. . Once the PHY helper is configured. TimeValue (Seconds (2. that is. "Ssid". staDevices = wifi.SetChannel (channel. of course. mac. This means that probe requests will not be sent by MACs created by this helper. BooleanValue (false)). Here we choose to work with non-Qos MACs so we use a NqosWifiMacHelper object to set MAC parameters. we create a channel object and associate it to our PHY layer object manager to make sure that all the PHY layer objects created by the YansWifiPhyHelper share the same underlying channel. "Ssid". wifiStaNodes). This means that the MAC will use a “non-QoS station” (nqsta) state machine. Next. YansWifiPhyHelper phy = YansWifiPhyHelper::Default (). we configure the type of MAC. Once all the station-specific parameters are fully configured. available in Doxygen. WifiHelper wifi = WifiHelper::Default (). this code uses the default PHY layer configuration and channel models which are documented in the API doxygen documentation for the YansWifiChannelHelper::Default and YansWifiPhyHelper::Default methods. we can focus on the MAC layer. and now we need to configure the AP (access point) node. Here. BooleanValue (true).5))). we can invoke our now-familiar Install method to create the wifi devices of these stations: NetDeviceContainer staDevices. wifi. they share the same wireless medium and can communication and interfere: phy. mac.SetRemoteStationManager ("ns3::AarfWifiManager"). it is asking the helper to use the AARF algorithm — details are.Create ()). Once these objects are created.Install (phy. "BeaconInterval".SetType ("ns3::NqstaWifiMac". For simplicity. We have configured Wifi for all of our STA nodes. NqosWifiMacHelper mac = NqosWifiMacHelper::Default (). mac.11 service set identifier (SSID) object that will be used to set the value of the “Ssid” Attribute of the MAC layer implementation. "ActiveProbing". SsidValue (ssid).SetType ("ns3::NqapWifiMac". The particular kind of MAC layer is specified by Attribute as being of the "ns3::NqstaWifiMac" type. "BeaconGeneration". Finally.
0). wandering around inside a bounding box.1. the NqosWifiMacHelper is going to create MAC layers of the “ns3::NqapWifiMac” (Non-Qos Access Point) type. 50. stack. We use the MobilityHelper to make this easy for us. DoubleValue (10. 50))). First. Just as in the second.0 to create the . "DeltaY". "GridWidth".Install (wifiStaNodes).SetMobilityModel ("ns3::ConstantPositionMobilityModel"). -50.0). mobility. Feel free to explore the Doxygen for class ns3::GridPositionAllocator to see exactly what is being done. but now we need to tell them how to move.0). We want the access point to remain in a fixed position during the simulation. We want the STA nodes to be mobile. and we want to make the AP node stationary. Now. InternetStackHelper stack. MobilityHelper mobility. "LayoutType". DoubleValue (0.Install (wifiApNode). devices and channels created. mobility. This code tells the mobility helper to use a two-dimensional grid to initially place the STA nodes.SetMobilityModel ("ns3::RandomWalk2dMobilityModel". UintegerValue (3). RectangleValue (Rectangle (-50.5 seconds. We have arranged our nodes on an initial grid. we will use the InternetStackHelper to install these stacks. we are going to use the Ipv4AddressHelper to assign IP addresses to our device interfaces. DoubleValue (0. We now have our nodes.Chapter 6: Building Topologies 56 In this case. wifiApNode). "Bounds".Install (wifiStaNodes). The next lines create the single AP which shares the same set of PHY-level Attributes (and channel) as the stations: NetDeviceContainer apDevices.Install (wifiApNode). "MinX". "DeltaX".cc example script. We now tell the MobilityHelper to install the mobility models on the STA nodes.Install (csmaNodes). StringValue ("RowFirst")). We set the “BeaconGeneration” Attribute to true and also set an interval between beacons of 2. we are going to add mobility models. Just as we have done previously many times. We accomplish this by setting the mobility model for this node to be the ns3::ConstantPositionMobilityModel: mobility. DoubleValue (5.Install (phy. mobility. First we use the network 10. mac. and mobility models chosen for the Wifi nodes.0). but we have no protocol stacks present. apDevices = wifi.SetPositionAllocator ("ns3::GridPositionAllocator". stack. "MinY". We choose the RandomWalk2dMobilityModel which has the nodes move in a random direction at a random speed around inside a bounding box. stack.1. we instantiate a MobilityHelper object and set some Attributes controlling the “position allocator” functionality. mobility.
SetBase ("10.cc example script. and this will result in simulator events being scheduled into the future indefinitely. "255.Assign (p2pDevices). Ipv4InterfaceContainer p2pInterfaces. echoClient. TimeValue (Seconds (1. The following line of code tells the simulator to stop so that we don’t simulate beacons forever and enter what is essentially an endless loop. ApplicationContainer serverApps = echoServer. serverApps.3.2. address.Assign (csmaDevices).0". "255.3.0)).0 to both the STA devices and the AP on the wireless network.1.Assign (staDevices).GetAddress (nCsma).0)). csmaInterfaces = address. pointing it to the server on the CSMA network. UdpEchoClientHelper echoClient (csmaInterfaces.0").SetBase ("10.Install (wifiStaNodes.1. We have done this before.SetBase ("10. serverApps. address.2. 9). We have also seen similar operations before.0"). address. ApplicationContainer clientApps = echoClient.255. clientApps.0)).Chapter 6: Building Topologies 57 two addresses needed for our two point-to-point devices.1.Stop (Seconds (10. address.Assign (apDevices).Start (Seconds (2. One thing that can surprise some users is the fact that the simulation we just created will never “naturally” stop. UintegerValue (1)). echoClient.255.))).Get (nWifi . Simulator::Stop (Seconds (10.255.0". .Start (Seconds (1.0 to assign addresses to the CSMA network and then we assign addresses from network 10.255.Stop (Seconds (10.1. This is because we asked the wireless access point to generate beacons.Get (nCsma)). UdpEchoServerHelper echoServer (9).Install (csmaNodes.255.1.0)). p2pInterfaces = address.1)). We put the echo server on the “rightmost” node in the illustration at the start of the file.0".SetAttribute ("PacketSize".1. UintegerValue (1024)). Ipv4GlobalRoutingHelper::PopulateRoutingTables ().255. Since we have built an internetwork here. Then we use network 10.0)). And we put the echo client on the last STA node we created. address.0"). Ipv4InterfaceContainer csmaInterfaces. It will generate beacons forever. clientApps. Ipv4AddressHelper address. echoClient. we need to enable internetwork routing just as we did in the second. "255.SetAttribute ("Interval". so we must tell the simulator to stop even though it may have beacon generation events scheduled.SetAttribute ("MaxPackets".
you will find four trace files from this simulation. two from node zero and two from node one: third-0-0.4 Received 1024 bytes from 10. Finally. “Received 1024 bytes from 10.2.cc scratch/mythird.” is from the echo client.Chapter 6: Building Topologies 58 We create just enough tracing to cover all three networks: PointToPointHelper::EnablePcapAll ("third"). This will let us see all of the traffic with a minimum number of trace files. If you now go and look in the top level directory.pcap” will be the promiscuous (monitor mode) trace from the Wifi network and the file “third-1-1. true). Simulator::Destroy (). return 0.1.pcap The file “third-0-0.1. indicating that it has received its echo back from the server. phy.1.cc example script into the scratch directory and use Waf to build just as you did with the second.1. generated when it receives the echo packet.3. cp examples/third.1. csmaDevices. In this case. you will see similar output.2. } In order to run this example./waf . and will start a promiscuous trace on the CSMA network. the client is on the wireless network (10.4.3 Received 1024 bytes from 10.3.Get (0). .2. These three lines of code will start pcap tracing on both of the point-to-point nodes that serves as our backbone.407s) Sent 1024 bytes to 10. The file “third-1-0.4 Recall that the first message.” is the UDP echo client sending a packet to the server.pcap” will be the promiscuous trace from the CSMA network. since we have set up the UDP echo applications just as we did in the second.pcap third-1-1.2. apDevices./waf --run scratch/mythird Again.cc script. The final message. If you are in the top-level directory of the repository you would type.1. The file “third-0-1. Can you verify this by inspecting the code? Since the echo client is on the Wifi network.” is from the UDP echo server.pcap third-1-0. Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone/ns-3-dev/build’ ’build’ finished successfully (0. clean up and then exit the program.4. “Sent 1024 bytes to 10. you have to copy the third. CsmaHelper::EnablePcap ("third". Let’s take a look at the promiscuous (monitor mode) trace we captured on that network. Simulator::Run ().EnablePcap ("third". will start a promiscuous (monitor) mode trace on the Wifi network.3.cc example.3.1. we actually run the simulation.0).cc .Get (0)). “Received 1024 bytes from 10.pcap” corresponds to the point-to-point device on node one – the right side of the “backbone”. The second message.pcap third-0-1. let’s start there.pcap” corresponds to the point-to-point device on node zero – the left side of the “backbone”.
2.0 54.0 12.000487 arp reply 10.0* 9.0 48.0 24.500000 Beacon () [6. We leave it as an exercise to completely parse the trace dump.4 2.0* 9.0 54.11 as you would expect.000357 Assoc Response AID(0) :: Succesful 0.010029 arp reply 10.011767 Acknowledgment RA:00:00:00:00:00:0a 2.3.0 48.000112 arp who-has 10.0 48.0 12.500000 Beacon () [6.000748 Assoc Request () [6. length 1024 This is the echo packet going from left to right (from Wifi to CSMA) and back again across the point-to-point link. you should see some familiar looking contents: .4.002169 IP 10. look at the pcap file of the right side of the point-to-point link.3.11) 0.1.0 Mbit] IBSS 5.3.0 54.49153: UDP.002169 IP 10.0 12. length 1024 2.002185 Acknowledgment RA:00:00:00:00:00:09 2.pcap Again.3.3. tcpdump -nn -tt -r third-1-0.0 9.3.0 18.000263 Assoc Request () [6.0 12.000128 Acknowledgment RA:00:00:00:00:00:09 2.0 9.1.2. look at the pcap file of the right side of the point-to-point link. You can probably understand what is going on and find the IP echo request and response packets in this trace.3.0 54.0* 9.000842 Assoc Response AID(0) :: Succesful 0.3.0 Mbit] 0.0 36.0 18.1.3.001242 Assoc Request () [6.1.1.0 36.3.0 Mbit] IBSS 7.000000 Beacon () [6.2.000206 arp who-has 10.pcap You should see some wifi-looking contents you haven’t seen here before: reading from file third-0-1.0* 9. you should see some familiar looking contents: reading from file third-0-0.1.0 24. length 1024 2.3 2.0 54.4.1.4 (ff:ff:ff:ff:ff:ff) tell 10.000986 Acknowledgment RA:00:00:00:00:00:0a 0.3.0 18.3.0 12.49153 > 10.0 54.0 Mbit] IBSS 0.000764 Acknowledgment RA:00:00:00:00:00:08 0.0 48.9: UDP.pcap.0 12.3.0 48.4 (ff:ff:ff:ff:ff:ff) tell 10.0 18.0 18.49153: UDP.0 36.9 > 10.010231 IP 10.0 24.4.3.0 9.0 36.010045 Acknowledgment RA:00:00:00:00:00:09 2. Now.001480 Acknowledgment RA:00:00:00:00:00:0a 2.3.pcap Again.0 12.1.3. tcpdump -nn -tt -r third-0-0.49153 > 10.0 36. link-type IEEE802_11 (802.1.000659 Acknowledgment RA:00:00:00:00:00:0a 2.Chapter 6: Building Topologies 59 tcpdump -nn -tt -r third-0-1.4.0 54.0 Mbit] 0.0 24.2.000501 Acknowledgment RA:00:00:00:00:00:0a 0.0 36.000279 Acknowledgment RA:00:00:00:00:00:07 0.0 36.0 18.1.0 18.3 is-at 00:00:00:00:00:09 2.4 is-at 00:00:00:00:00:0a 2.0 Mbit] 0.1.0 48.009771 arp who-has 10.000025 Beacon () [6.9: UDP.1. link-type PPP (PPP) 2. length 1024 2.0 48.001336 Assoc Response AID(0) :: Succesful 0. Now.009771 IP 10.001258 Acknowledgment RA:00:00:00:00:00:09 0.0 24.0 24.0 Mbit] IBSS You can see that the link type is now 802.1.1.1.1.pcap.3 (ff:ff:ff:ff:ff:ff) tell 10.3 2.0 24.9 > 10.
006084 IP 10.2. Now. Ptr<const MobilityModel> model) { Vector position = model->GetPosition ().005877 arp reply 10. } This code just pulls the position information from the mobility model and unconditionally logs the x and y position of the node. NS_LOG_UNCOND (context << " x = " << position.1.2.49153 > 10.1 (ff:ff:ff:ff:ff:ff) tell 10.3.y).1. link-type PPP (PPP) 2. length 1024 2.2. y = " << position.cc.2.1.3.2.Get (nWifi . we spent a lot of time setting up mobility models for the wireless network and so it would be a shame to finish up without even showing that the STA nodes are actually moving around during the simulation.3.3. Let’s do this by hooking into the MobilityModel course change trace source. length 1024 2. . Just before the main program of the scratch/mythird. If you’ve forgotten. length 1024 This is also the echo packet going from left to right (from Wifi to CSMA) and back again across the point-to-point link with slightly different timings as you might expect.1. As mentioned in the “Tweaking ns-3” section.1.3.005877 IP 10. We do this using the Config::Connect function.4 2.1)->GetId () << "/$ns3::MobilityModel/CourseChange".3. but this seems a very nice place to get an example in. add the following function: void CourseChange (std::string context. length 1024 This should be easily understood.3.cc script.005980 arp who-has 10. This is just a sneak peek into the detailed tracing section which is coming up. We are going to arrange for this function to be called every time the wireless node with the echo client changes its position.49153 > 10.1 2.2. We will use the mobility model predefined course change trace source to originate the trace events. oss << "/NodeList/" << wifiStaNodes. let’s look at the promiscuous trace there: tcpdump -nn -tt -r third-1-1. The echo server is on the CSMA network.1.pcap You should see some familiar looking contents: reading from file third-1-1.pcap. and we provide functions to connect the two.1.x << ".9: UDP.006084 IP 10.1.4 is-at 00:00:00:00:00:06 2.2.9 > 10.2.pcap.2.1.49153: UDP.1. go back and look at the discussion in second. the ns-3 tracing system is divided into trace sources and trace sinks.1.4. link-type EN10MB (Ethernet) 2.1. This is the same sequence.005980 arp reply 10.3. Despite its reputation as being difficult.4.2.005855 IP 10.1.4.4 (ff:ff:ff:ff:ff:ff) tell 10.1.Chapter 6: Building Topologies 60 reading from file third-1-0.1 is-at 00:00:00:00:00:03 2.005855 arp who-has 10.4.49153: UDP. Add the following lines of code to the script just before the Simulator::Run call.9 > 10. We will need to write a trace sink to connect to that source that will display some pretty information for us. it’s really quite simple.9: UDP. std::ostringstream oss.
which will in turn print out the new position.18915. If you now run the simulation.89186.2. y = 1.3.52738.3 Received 1024 bytes from 10.17821 4. y = 1.7572. y = -2. y = 0.4 = = = = = = = 10. y = -1. y = 3. y = 1. this turns out to be node seven and the tracing namespace path to the mobility model would look like. y = -1. MakeCallback (&CourseChange)).4 Received 1024 bytes from 10. y = -2. y = -1. every course change event on node seven will be hooked into our trace sink. you may infer that this trace path references the seventh node in the global NodeList. First.46199.42888. What we do here is to create a string containing the tracing namespace path of the event to which we want to connect. We make a connection between the trace source in node seven with our trace sink by calling Config::Connect and passing this namespace path. y = -0.51981.73934.29596 .6835.51044 7.70014 7.2.811313 8.1. you will see the course changes displayed as they happen. y = -0. The dollar sign prefix implies that the MobilityModel is aggregated to node seven.22677 6. The last component of the path means that we are hooking into the “CourseChange” event of that model.556238 4. y = 0 9.Chapter 6: Building Topologies 61 Config::Connect (oss. y = 2.27897. y = 0.str (). /NodeList/7/$ns3::MobilityModel/CourseChange Based on the discussion in the tracing section.54934 5.25785 4.53175.1.91654 6.59219 6.46869 6. y = 2.98503 5.70932.01523 7.62404. Once this is done.74127. It specifies what is called an aggregated object of type ns3::MobilityModel.41539.67099. y = -2.90077 6.18521. In the case of the default number of CSMA and wireless nodes.81046.48729 6.1.34588. Build finished successfully (00:00:01) Sent 1024 bytes to 10.91689 x x x x x x x x x x x x x x x x = = = = = = = = = = = = = = = = 5. y = 1.40519.11303 7.14268 4. y = -1.48576 4.434856 4. y = 2. y = -1. y = 1.58021. y = 1.58121. we have to figure out which node it is we want using the GetId method as described earlier.45166 7.
56579 1.00393. 7.30328 3. y y y y y y y = = = = = = = 2.18682.05492.47732 1.33503.66873 .35768 2. 7. 8.25054 1.96865.Chapter 6: Building Topologies 62 = = = = = = = 7.46617. 7.29223 2. 7. 7.00968.
} Nobody is going to prevent you from going deep into the core of ns-3 and adding print statements.. Often. If you need to add some tidbit of information to the pre-defined bulk mechanisms.1. Second. std::cout << "The value of x is " << x << std::endl. If you desire... your output can be formatted directly into a form acceptable by gnuplot. Using pre-defined bulk output mechansims has the advantage of not requiring any changes to ns-3. First. . this approach fails. after all. this can certainly be done. you can control the format of the output directly so you avoid the postprocessing step with sed or awk script. or somehow developing an output mechanism that conveys exactly (and perhaps only) the information wanted. you have complete control of your . It has several important advantages.Chapter 7: The Tracing System 63 7 The Tracing System 7. the whole point of running an ns-3 simulation is to generate output for study.1 Blunt Instruments There are many ways to get information out of a program. we believe that the ns-3 Tracing system is the best way to get information out of a simulation and is also therefore one of the most important mechanisms to understand in ns-3. pcap or NS LOG output messages are gathered during simulation runs and separately run through scripts that use grep. if you use this method. . you may get your code added as a contribution. Programs must be written to do the transformation.. so this does not come for free. You can add hooks in the core which can then be accessed by other users. #include <iostream> . ns-3 provides another mechanism. 7. for example. you can reduce the amount of data you have to manage by only tracing the events of interest to you.. but which will produce no information unless explicitly asked to do so. that avoids some of the problems inherent in the bulk output mechanisms. called Tracing. void SomeFunction (void) { uint32_t x = SOME_INTERESTING_VALUE. For these reasons. if the information of interest in does not exist in any of the pre-defined output mechanisms. You have two basic strategies to work with in ns-3: using generic pre-defined bulk output mechanisms and parsing their content to extract interesting information. sed or awk to parse the messages and reduce and transform the data to a manageable form.1 Background As mentioned in the Using the Tracing System section. and if you use one of the ns-3 mechanisms. Of course. but it does require programming. as in. This is insanely easy to do and. The most straightforward way is to just directly print the information to the standard output..
. this is certainly better than adding your own print statements since it follows ns-3 coding conventions and could potentially be useful to other people as a patch to the existing core.Chapter 7: The Tracing System 64 own ns-3 branch. As the number of print statements increases in your programs.cc) you could just add a new message down in the implementation.. perhaps by turning on and off certain categories of prints.. case ACK_TX: SendEmptyPacket (TcpHeader::ACK). We mentioned above that one way to get information out of ns-3 is to parse existing NS LOG output for interesting information. If you wanted to add more logging to the ns-3 TCP socket (tcp-socket-impl. the task of dealing with the large number of outputs will become more and more complicated. If you continue down this path you may discover that you have re-implemented the NS_LOG mechanism. In order to avoid that. SendEmptyPacket (TcpHeader::ACK). break. switch (a) { case NO_ACT: NS_LOG_LOGIC ("TcpSocketImpl " << this <<" Action: NO_ACT"). . changing the code from: bool TcpSocketImpl::ProcessAction (Actions_t a) { // These actions do not require a packet or any TCP Headers NS_LOG_FUNCTION (this << a). or increasing or decreasing the amount of information you want. Now. If you discover that some tidbit of information you need is not present in existing log output. . Eventually. This will probably not turn out to be very satisfactory in the long term.. . break. Notice that in TcpSocketImpl::ProcessAction() there is no log message for the ACK_TX case. you could edit the core of ns-3 and simply add your interesting information to the output stream. Let’s pick a random example. case ACK_TX: NS_LOG_LOGIC ("TcpSocketImpl " << this << " Action: ACK_TX"). break. one of the first things you might consider is using NS_LOG itself. break. you may feel the need to control what information is being printed in some way. switch (a) { case NO_ACT: NS_LOG_LOGIC ("TcpSocketImpl " << this << " Action: NO_ACT"). to add a new NS_LOG_LOGIC in the appropriate case statement: bool TcpSocketImpl::ProcessAction (Actions_t a) { // These actions do not require a packet or any TCP Headers NS_LOG_FUNCTION (this << a). though. You could simply add one.
you will also have to live with the output that every other developer has found interesting. By using the tracing system. we consider prints to std::cout and NS LOG messages simple ways to get more information out of ns-3. You may find that in order to get the small amount of information you need. but they are really unstable and quite blunt instruments. Your code looking for trace events from a particular piece of core code could happily coexist with other code doing something entirely different from the same information. sed or awk scripts) to parse the log output in order to isolate your information. This is because even though you have some control over what is output by the logging system. you may also discover that pieces of log output on which you depend disappear or change between releases. This explicit division allows for large numbers of trace sources to be scattered around the system in places which model authors believe might be useful. If you depend on the structure of the output. along with a uniform mechanism for connecting sources to sinks. Trace sources are not useful by themselves. For example. Trace sources are generators of events and trace sinks are consumers. A trace source might also indicate when an iteresting state change happens in a model. For example. the congestion window of a TCP model is a prime candidate for a trace source. You may be forced to save huge log files to disk and process them down to a few lines whenever you want to do anything.Chapter 7: The Tracing System 65 This may seem fairly simple and satisfying at first glance. Unless a user connects a trace sink to one of these sources. you may have to wade through huge amounts of extraneous messages that are of no interest to you. they must be connected to other pieces of code that actually do something useful with the information provided by the source. If you are adding code to an existing module. It is desirable to be able to do this without having to change and recompile the core system. One can think of a trace source as a kind of point-to-multipoint information link. 7. There can be zero or more consumers of trace events generated by a trace source. Since there are no guarantees in ns-3 about the stability of NS_LOG messages. The entities that consume trace information are called trace sinks. you only have control down to the log component level. The ns-3 tracing system is designed to work along those lines and is well-integrated with the Attribute and Config subsystems allowing for relatively simple use scenarios.2 Overview The ns-3 tracing system is built on the concepts of independent tracing sources and tracing sinks. a trace source could indicate when a packet is received by a net device and provide access to the packet contents for interested trace sinks. For these reasons. nothing is output. It is desirable to have a stable facility using stable APIs that allow one to reach into the core system and only get the information required. but something to consider is that you will be writing code to add the NS_LOG statement and you will also have to write code (as in grep. both you and other people at the same trace source are getting exactly . Even better would be a system that notified the user when an item of interest changed or an interesting event happened so the user doesn’t have to actively go poke around in the system looking for things. you may find other messages being added or deleted which may affect your parsing code. Trace sources are entities that can signal events that happen in a simulation and provide access to interesting underlying data.
a trace source is a callback.2. If you happen to add a trace source. In C the canonical example of a pointer-to-function is a pointer-to-function-returninginteger (PFI). For a PFI taking one int parameter. Basically. 7. however.1 A Simple Low-Level Example Let’s take a few minutes and walk through a simple tracing example. Instead of just making one indirect call. a trace source may . people take advantage of the fact that the compiler knows what is going on and will just use a shorter form. Neither of you are impacting any other user by changing what information is output by the system. it adds a Callback to a list of Callbacks internally held by the trace source. Conceptually. without making any changes to the ns-3 core. the trace source invokes its operator() providing zero or more parameters. you can initialize the variable to point to your function: pfi = MyFunction. This ultimately means you need some kind of indirection – you treat the address of the called function as a variable. int result = (*pfi) (1234). Typically. so we have to take a small detour right away. This is suggestive since it looks like you are dereferencing the function pointer just like you would dereference any pointer. this could be declared like. This looks like you are calling a function named “pfi. What you get from this is a variable named simply “pfi” that is initialized to the value 0.1. If you want to initialize this pointer to something meaningful. The important difference that the tracing system adds is that for each trace source there is an internal list of Callbacks. int result = pfi (1234). you have to have a function with a matching signature. We are going to need a little background on Callbacks to understand what is happening in the example. This variable is called a pointer-to-function variable. The relationship between function and pointer-to-function pointer is really no different that that of object and pointer-to-object.” but the compiler is smart enough to know to call through the variable pfi indirectly to the function MyFunction.2. You can then call MyFunction indirectly using the more suggestive form of the call. The operator() eventually wanders down into the system and does something remarkably like the indirect call you just saw. this is almost exactly how the tracing system will work.Chapter 7: The Tracing System 66 what they want and only what they want out of the system. your work as a good open-source citizen may allow other users to provide new utilities that are perhaps very useful overall. When an interesting event happens. When a trace sink expresses interest in receiving trace events. int MyFunction (int arg) {} If you have this target.1 Callbacks The goal of the Callback system in ns-3 is to allow one piece of code to call a function (or method in C++) without any specific inter-module dependency. int (*pfi)(int arg) = 0. In this case. It provides zero or more parameters (the call to “pfi” above passed one parameter to the target function MyFunction. 7. you could provide a function that looks like.
Most of this code should be quite familiar to you.h" "ns3/uinteger. you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. See the * GNU General Public License for more details.2 Example Code We have provided some code to implement what is really the simplest example of tracing that can be assembled. * * You should have received a copy of the GNU General Public License * along with this program. the trace system makes heavy use of the Object and Attribute systems. . The first two includes above bring in the declarations for those systems explicitly. not an address.Chapter 7: The Tracing System 67 invoke any number of Callbacks. if not.1. operator==.h" USA #include <iostream> using namespace ns3. c-file-style:"gnu". but this illustrates how simple this all really is.h" "ns3/trace-source-accessor. In order to use value semantics at all you have to have an object with an associated copy constructor and assignment operator available. indent-tabs-mode:nil. Suite 330. value semantics just means that you can pass the object around.cc.h brings in the required declarations for tracing of data that obeys value semantics. In general. traced-value. etc.Mode:C++.*/ /* * This program is free software. without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. Boston. Let’s walk through it. operator—. You can find this code in the tutorial directory as fourth. * but WITHOUT ANY WARRANTY. 7. Operator=. The file. it basically just arranges to add its own function to the callback list. What this all really means is that you will be able to trace changes to a C++ object made using those operators. MA 02111-1307 */ #include #include #include #include "ns3/object. We extend the requirements to talk about the set of operators that are pre-defined for plain-old-data (POD) types. When a trace sink expresses interest in notifications from a trace source. * * This program is distributed in the hope that it will be useful.2. -*.. feel free to peruse the Callback section of the manual. operator++. You could use the core module header. As mentioned above. write to the Free Software * Foundation.h" "ns3/traced-value. so you will need to include them. operator+. /* -*. Inc. If you are interested in more details about how this is actually arranged in ns-3. 59 Temple Place.
The next code snippet declares and defines a simple Object we can work with. It corresponds directly to a callback function. myObject->m_myInt = 1234. void IntTrace (int32_t oldValue.AddTraceSource ("MyInteger". above. int32_t newValue) { std::cout << "Traced " << oldValue << " to " << newValue << std::endl. MakeTraceSourceAccessor (&MyObject::m_myInt)) . } MyObject () {} TracedValue<int32_t> m_myInt. there must be an ns-3 Object for the trace source to live in. The TracedValue declaration provides the infrastructure that overloads the operators mentioned above and drives the callback process. What remains is code to connect the source to the sink. We have now seen the trace source and the trace sink. } This is the definition of the trace sink.AddConstructor<MyObject> () . The two important lines of code. int main (int argc. } Here we first create the Object in which the trace source lives.".SetParent (Object::GetTypeId ()) . return tid. MakeCallback(&IntTrace)). class MyObject : public Object { public: static TypeId GetTypeId (void) { static TypeId tid = TypeId ("MyObject") . char *argv[]) { Ptr<MyObject> myObject = CreateObject<MyObject> (). "An integer value to trace. . The . this function will be called whenever one of the overloaded operators of the TracedValue is executed. Once it is connected. with respect to tracing are the . }. and Attributes work with Objects.AddTraceSource and the TracedValue declaration of m_myInt. myObject->TraceConnectWithoutContext ("MyInteger".Chapter 7: The Tracing System 68 Since the tracing system is integrated with Attributes.AddTraceSource provides the “hooks” used for connecting the trace source to the outside world through the config system.
The function IntTrace then printed this to the standard output. Let’s ignore the bit about context for now. non-trivial. To summarize..2. Finally. but the essence is that you are arranging for something that looks just like the pfi() example above to be called by the trace source. the trace source fired and automatically provided the before and after values to the trace sink.) that will use the operator() to actually invoke the Callback with the desired parameters. That is exactly the function signature for the callback function we provided — IntTrace. This function does the magic required to create the underlying ns-3 Callback object and associate it with the function IntTrace. The declaration of the TracedValue<int32_t> m_myInt. the TraceConnectWithoutContext. After this association is made. etc. More typically. myObject->m_myInt = 1234. in essence. TraceConnect makes the association between your provided function and the overloaded operator() in the traced variable referred to by the “MyInteger” Attribute. which is specified by Attribute name. of course. a trace source is. a variable that holds a list of callbacks. myObject->m_myInt = 1234. No problem. Notice the MakeCallback template function. the trace source will “fire” your provided callback function. The .AddTraceSource performs the magic to connect the Callback to the Config system.2 Using the Config Subsystem to Connect to Trace Sources The TraceConnectWithoutContext call shown above in the simple example is actually very rarely used in the system. forms the connection between the trace source and the trace sink. If you now build and run this example. We saw an example of this . It turns out that this operator is defined (by TracedValue) to execute a callback that returns void and takes two integer values as parameters — an old value and a new value for the integer in question. A trace sink is a function used as the target of a callback. and TraceConnectWithoutContext performs the magic to connect your function to the trace source. ./waf --run fourth you will see the output from the IntTrace function execute as soon as the trace source is hit: Traced 0 to 1234 When we executed the code. The Attribute and object type information systems are used to provide a way to connect trace sources to trace sinks. should be interpreted as an invocation of operator= on the member variable m_myInt with the integer 1234 passed as a parameter.Chapter 7: The Tracing System 69 The next step. the Config subsystem is used to allow selecting a trace source in the system using what is called a config path. the line. The act of “hitting” a trace source is executing an operator on the trace source which fires callbacks. —. The code to make all of this happen is. This results in the trace sink callbacks registering interest in the source being called with the parameters provided by the source. in the Object itself performs the magic needed to provide the overloaded operators (++. 7.
y = " << position.Get (nWifi . oss << "/NodeList/" << wifiStaNodes.str (). MakeCallback (&CourseChange)). if you had a pointer to the Object that has the “CourseChange” Attribute handy. For the purposes of discussion. Each segment of a path corresponds to an Object Attribute. You know by now that we typically store pointers to our nodes in a NodeContainer. The Config functions take a path that represents a chain of Object pointers. theObject->TraceConnect ("CourseChange".1). The last segment is the Attribute . Ptr<Object> theObject = wifiStaNodes. the path above turns out to be. Let’s try and make some sense of what is sometimes considered relatively mysterious code. we actually want an additional “context” to be delivered along with the Callback parameters (which will be explained below) so we could actually use the following equivalent code.1)->GetId () << "/$ns3::MobilityModel/CourseChange". In the third. It should now be a lot more clear to you what this function is doing. In fact.cc. we used what is called a “Config Path” to specify the source when we arranged a connection between the pre-defined trace source and the new trace sink: std::ostringstream oss.cc example. we used this container to get a Ptr<Node> which we used to call GetId() on. In the third.Chapter 7: The Tracing System 70 in the previous section where we hooked the “CourseChange” event when we were playing with third.cc example. "/NodeList/7/$ns3::MobilityModel/CourseChange" The last segment of a config path must be an Attribute of an Object. } When we connected the “CourseChange” trace source to the above trace sink.x << ". Recall that we defined a trace sink to print course change information from the mobility models of our simulation. MakeCallback (&CourseChange)). It turns out that the internal code for Config::ConnectWithoutContext and Config::Connect actually do find a Ptr<Object> and call the appropriate TraceConnect method at the lowest level. the Nodes of interest are stored in the wifiStaNodes NodeContainer. In fact. you could write this just like we did in the previous example. NS_LOG_UNCOND (context << " x = " << position. Config::Connect (oss.Get (nWifi . theObject->TraceConnectWithoutContext ("CourseChange". MakeCallback (&CourseChange)). void CourseChange (std::string context.y).Get (nWifi . Ptr<const MobilityModel> model) { Vector position = model->GetPosition (). In this case. while putting the path together. We could have used this Ptr<Node> directly to call a connect method directly: Ptr<Object> theObject = wifiStaNodes.1). assume that the node number returned by the GetId() is “7”.
The MobilityModel class is designed to be a base class providing a common interface for all of the specific subclasses. If you search down to the end of the file. The leading “/” character in the path refers to a so-called namespace. we have asked the system to do the following: Ptr<MobilityModel> mobilityModel = node->GetObject<MobilityModel> () We are now at the last Object in the path. or associate. When you add the “$” you are asking for another Object that has presumably been previously aggregated. Each Object in an Aggregation can be reached from the other Objects. } . and prior segments must be typed to contain or find Objects. ‘‘The value of the position and/or velocity vector changed’’. This allows us to form an association between different Objects without any programming. The MobilityModel class defines an Attribute called “CourseChange. This reference is actually a Ptr<Node> and so is a subclass of an ns3::Object.Chapter 7: The Tracing System 71 of interest.AddTraceSource (‘‘CourseChange’’. a mobility model to each of the wireless Nodes. so we turn our attention to the Attributes of that Object. MakeTraceSourceAccessor (&MobilityModel::m_courseChangeTrace)) which should look very familiar at this point. so “/NodeList/7” refers to the eighth node in the list of nodes created during the simulation.h you will find TracedCallback<Ptr<const MobilityModel> > m_courseChangeTrace. you will see a method defined called NotifyCourseChange(): void MobilityModel::NotifyCourseChange (void) const { m_courseChangeTrace(this). It then interprets the last segment as an Attribute on the last Object it found while walking the path. As described in the Object Model section of the ns-3 manual. The Config functions then call the appropriate TraceConnect or TraceConnectWithoutContext method on the final Object. The next path segment being walked begins with the “$” character. If you are familiar with GetObject. The Config code parses and “walks” this path until it gets to the final segment of the path. .” You can see this by looking at the source code in src/mobility/mobility-model. One of the predefined namespaces in the config system is “NodeList” which is a list of all of the nodes in the simulation. You can think of this as switching pointers from the original Ptr<Node> as specified by “/NodeList/7” to its associated mobility model — which is of type “$ns3::MobilityModel”. Let’s see what happens in a bit more detail when the above path is walked.cc and searching for “CourseChange” in your favorite editor. This indicates to the config system that a GetObject call should be made looking for the type that follows. we support Object Aggregation. You should find. The type declaration TracedCallback identifies m_courseChangeTrace as a special list of Callbacks that can be hooked using the Config functions described above. Items in the list are referred to by indices into the list. It turns out that the MobilityHelper used in third. If you look for the corresponding declaration of the underlying traced variable in mobility-model.cc arranges to Aggregate.
There needs to be some way to identify which trace source is actually the one that fired the Callback.2. “okay. invoke all of the registered Callbacks.22677 The first part of the output is the context.” Go ahead and select that link.27897. You will see.4 What Trace Sources are Available? The answer to this question is found in the ns-3 Doxygen.3 How to Find and Connect Trace Sources. Therefore. calling all of the trace sinks that have registered interest in the trace source by calling a Config function. whenever a course change is made in one of the RandomWalk2dMobilityModel instances installed. This method invokes operator() on the underlying m_courseChangeTrace. Now. in the third. In the example. I found a trace source.cc example we looked at. expand the “Core” book in the tree by clicking its “+” box. “okay. the only code registering an interest was the code that provided the config path. calls any registered trace sinks. Go to the ns-3 web site “here” and select the “Doxygen (stable)” link “Documentation” on the navigation bar to the left side of the page. I typed that all in and got this incredibly bizarre error message. The final piece of the puzzle is the “context.cc: /NodeList/7/$ns3::MobilityModel/CourseChange x = 7. a list of all of the trace sources available in the ns-3 core. which in turn.Chapter 7: The Tracing System 72 Derived classes will call into this method whenever they do a course change to support tracing. I know that there must be trace sources in the simulation core. “okay. It is simply the path through which the config code located the trace source. the CourseChange function that was hooked from Node number seven will be the only Callback called. what in the world does it mean”? 7. In the case we have been looking at there can be any number of trace sources in the system corresponding to any number of nodes with mobility models. So. I found a trace source. how do I figure out the config path to use when I connect to it”? The third question is.2. You should now see three extremely useful links: • The list of all trace sources • The list of all attributes • The list of all global values The list of interest to us here is “the list of all trace sources. in turn. and Discover Callback Signatures The first question that inevitably comes up for new users of the Tracing system is. which will. how do I figure out what the return type and formal arguments of my callback function need to be”? The fourth question is. there will be a NotifyCourseChange() call which calls up into the MobilityModel base class.” Recall that we saw an output looking something like the following from third. y = 2. As seen above. perhaps not too surprisingly. but how do I find out what trace sources are available to me”? The second question is. Expand the “Modules” book in the NS-3 documentation tree a the upper left by clicking the “+” box. “okay. this invokes operator() on m_courseChangeTrace. An easy way is to request a trace context when you Config::Connect. . 7.
return tid.AddTraceSource ("MyInteger". So open the “Class List” book in the NS-3 documentation tree by clicking its “+” box. you can find this out from the ns-3 Doxygen. "An integer value to trace.AddConstructor<MyObject> () . } As mentioned above. from the third./ns-3-dev/examples/wireless/mixed-wireless. MakeCallback (&CourseChangeCallback)). 7. .2. If you cannot find any examples in the distribution. this is the bit of code that connected the Config and Attribute systems to the underlying trace source. This is also the place where you should start looking for information about the way to connect. and the information you want is now right there in front of you in the Doxygen: . -name ’*. You know that you are using (again. MakeTraceSourceAccessor (&MyObject::m_myInt)) .cc example. you will see documentation for the GetTypeId function.Chapter 7: The Tracing System 73 As an example.cc has something just waiting for you to use: Config::Connect (‘‘/NodeList/*/$ns3::MobilityModel/CourseChange’’. You should always try to copy someone else’s working code before you start to write your own.cc example) an ns3::RandomWalk2dMobilityModel. You are looking at the same information for the RandomWalk2dMobilityModel. in this case. Let’s assume that you have just found the “CourseChanged” trace source in “The list of all trace sources” and you want to figure out how to connect to it. Scroll down until you see the entry for ns3::RandomWalk2dMobilityModel and follow that link. You will now see a list of all of the classes in ns-3. You will find an entry for CourseChange: The value of the position and/or velocity vector changed You should recognize this as the trace source we used in the third.5 What String do I use to Connect? The easiest way to do this is to grep around in the ns-3 codebase for someone who has already figured it out.cc’ | xargs grep CourseChange | grep Connect and you may find your answer along with working code. Try something like: find .” If you now scroll down to the “Member Function Documentation” section. Perusing this list will be helpful. scroll down to ns3::MobilityModel. For example.SetParent (Object::GetTypeId ()) . You constructed one of these in the simple tracing example above: static TypeId GetTypeId (void) { static TypeId tid = TypeId ("MyObject") . It will probably be simplest just to walk through the “CourseChanged” example.". You should now be looking at the “ns3::RandomWalk2dMobilityModel Class Reference.
for $ns3::RandomWalk2dMobilityModel. You will find. The interesting thing this bit of Doxygen tells you is that you don’t need that extra cast in the config path above to get to the concrete class. is used to “cast” the base class to the concrete imlementation class. -name ’*. . since the trace source is actually in the base class. For example. The documentation shows both of these operations for you.Chapter 7: The Tracing System 74 This object is accessible through the following paths with Config::Set and Config::Connect /NodeList/[i]/$ns3::MobilityModel/$ns3::RandomWalk2dMobilityModel The documentation tells you how to get to the RandomWalk2dMobilityModel Object. as a result of your grep. The first./ns-3-dev/examples/wireless/mixed-wireless... . Look further down in the GetTypeId doxygen. for $ns3::MobilityModel will query the aggregation for the base class. The trace source of interest is found in ns3::MobilityModel (which you knew anyway). Compare the string above with the string we actually used in the example code: "/NodeList/7/$ns3::MobilityModel" The difference is due to the fact that two GetObject calls are implied in the string found in the documentation.cc’ | xargs grep CourseChange | grep Connect and you may find your answer along with working code. TraceSources defined in parent class ns3::MobilityModel: CourseChange: The value of the position and/or velocity vector changed Reimplemented from ns3::MobilityModel This is exactly what you need to know. Therefore the additional GetObject is not required and you simply use the path: /NodeList/[i]/$ns3::MobilityModel which perfectly matches the example path: /NodeList/7/$ns3::MobilityModel 7. No TraceSources defined for this type. You will find Config::Connect (‘‘/NodeList/*/$ns3::MobilityModel/CourseChange’’. You should always try to copy someone else’s working code.cc has something just waiting for you to use. MakeCallback (&CourseChangeCallback)). Try something like: find . in this case. Ptr<const MobilityModel> model) { . there is: static void CourseChangeCallback (std::string path.2. The MakeCallback should indicate to you that there is a callback function there which you can use.6 What Return Value and Formal Arguments? The easiest way to do this is to grep around in the ns-3 codebase for someone who has already figured it out. The second implied GetObject call. It turns out that the actual Attribute you are going to be looking for is found in the base class as we have seen. Sure enough.
There is a one-to-one correspondence between the template parameter list in the declaration and the formal arguments of the callback function.. you can add the keyword static and come up with: static void CourseChangeCallback (std::string path.Chapter 7: The Tracing System 75 } 7. Before embarking on a walkthrough of the code. Here. Ptr<const MobilityModel> model) { .2.. this is in mobility-model. let’s figure out what signature of callback function is required for the “CourseChange” Attribute.. } which is exactly what we used in the third. especially for those unfamiliar with the details of templates. Ptr<const MobilityModel> model) { . challenging to actually figure out from the source code. I’ll be kind and just tell you a simple way to figure this out: The return value of your callback will always be void.2 The Hard Way This section is entirely optional.. well.. This is going to be painful.h. you need to Config::Connect and use a Callback function that takes a string context.1 Take my Word for It If there are no examples to work from. if you get through this.2. 7. However. then the required argument.cc example. It is going to be a bumpy ride. but you only need to do this once. The formal parameter list for a TracedCallback can be found from the template parameter list in the declaration. This tells you that you need a function that returns void and takes a a Ptr<const MobilityModel>.6. where we have previously found: TracedCallback<Ptr<const MobilityModel> > m_courseChangeTrace. this can be. } If you want to ensure that your CourseChangeCallback is only visible in your local file. After you get through this. which is a Ptr<const MobilityModel>. again. . you will have a very good handle on a lot of the ns-3 low level idioms. Recall that for our current example. } That’s all you need if you want to Config::ConnectWithoutContext.. void CourseChangeCallback (Ptr<const MobilityModel> model) { . void CourseChangeCallback (std::string path. there is one template parameter.6. So. For example. you will be able to just look at a TracedCallback and understand it. If you want a context.
.T3.T4. TracedCallback<T1.T3.T7. TracedCallback<T1..T5. TracedCallback<T1.T3. Just after this comment.T5..T5.T7.T4.T2. In either case. T2 a2 .T8>::operator() (T1 a1.T2.T6.Chapter 7: The Tracing System 76 The first thing we need to look at is the declaration of the trace source.T2. TracedCallback<T1. It turns out that all of this comes from the header file traced-callback.T8>::operator() (T1 a1.T8>::operator() (T1 a1) const .T7. we know this declaration is going to have to be in some kind of header file.. that’s not really a lot. -name ’*. Then.T6..T4.T6.T2. Although that may seem like it. TracedCallback<T1.h and see that there is a line which confirms this hunch: #include "ns3/traced-callback. typename T2 = empty.T2.T6.. TracedCallback<T1.T5.T2..T6.T7. the next step is to take a look at src/core/traced-callback.T8>::ConnectWithoutContext (c .h and inferring that this must be the file you want.T5. The template parameter is inside the angle-brackets... T2 a2 . you will see some very suspiciously template-looking stuff. T2 a2 .T7.T2.T3.T5. T2 a2 . TracedCallback<T1. Just pipe the output through more and start scanning through it.T3.T5.T6.T6. so just grep for it using: find . you will find.T8>::operator() (void) const . .T4.h and noticing the include of traced-callback.T6.h" Of course. so we are really interested in finding out what that TracedCallback<> is.T3.T4.T8>::DisconnectWithoutContext ..T4.T8>::operator() (T1 a1. You can then take a look at mobility-model.T6..T5.T8>::operator() (T1 a1.T3.T8>::operator() (T1 a1.T5..T7.T3.. TracedCallback<T1.h’ | xargs grep TracedCallback You’ll see 124 lines fly by (I piped this through wc to see how bad it was).T3. where we have previously found: TracedCallback<Ptr<const MobilityModel> > m_courseChangeTrace. Recall that this is in mobility-model..T4..T2..T4.T7.T7.T2.. you could have gone at this from the other direction and started by looking at the includes in mobility-model.T6.T4. template<typename T1 = empty.T6. This should sound very familiar and let you know you are on the right track.T4.T8>::operator() (T1 a1.T7.h in your favorite editor to see what is happening.T5. TracedCallback<T1. TracedCallback<T1..T2. T2 a2 .T6..T3.. This declaration is for a template.T8>::TracedCallback () TracedCallback<T1. If you have absolutely no idea where this might be found.T8>::Connect (const CallbackB . You will see a comment at the top of the file that should be comforting: An ns3::TracedCallback has almost exactly the same API as a normal ns3::Callback but instead of forwarding calls to a single function (as an ns3::Callback normally does).T2..T3. TracedCallback<T1..T2. it forwards calls to a chain of ns3::Callback. so first change into the src directory.T2.T7.T4. We are probably going to be interested in some kind of declaration in the ns-3 source. On the first page..T4.T3. T2 a2 .T3.h which sounds very promising. TracedCallback<T1.T7.T7..T5.T5.T4. grep is your friend.h.T5.T7.T8>::Disconnect (const Callba .T6.
It has eight possible type parameters with default values. When the template is instantiated for the declaration above. typename T6. cb. typename T7 = empty. typename T5 = empty. m_callbackList.T3. typename T2.T2. Go back and compare this with the declaration you are trying to understand: TracedCallback<Ptr<const MobilityModel> > m_courseChangeTrace.T7. you will see a lot of probably almost incomprehensible template code.T4. } You can now see the implementation of everything we’ve been talking about. typename T8 = empty> class TracedCallback { . typename T3. All of the other type parameters are left as defaults. The code then adds the Callback to the list of Callbacks for this source. This tells you that TracedCallback is a templated class. typename T4 = empty. void TracedCallback<Ptr<const MobilityModel>::ConnectWithoutContext .push_back (cb). typename T8> void TracedCallback<T1. cb { Callback<void../core/callback. Looking at the constructor really doesn’t tell you much. m_callbackList. typename T7. typename T4. Ptr<const MobilityModel> > cb. If you scroll down. though...push_back (cb).T8> cb. } You are now in the belly of the beast. cb. the compiler will replace T1 with Ptr<const MobilityModel>.T5.T4. You will eventually come to some Doxygen for the Callback template class.T5. This is the equivalent of the pfi = MyFunction we discussed at the start of this section.h is the one we need to look at.T3. Using the same grep trick as we used to find TracedCallback. The one place where you have seen a connection made between your Callback function and the tracing system is in the Connect and ConnectWithoutContext functions.T6. The typename T1 in the templated class declaration corresponds to the Ptr<const MobilityModel> in the declaration above. Fortunately. there is some English: .Assign (callback).T1.T8>::ConnectWithoutContext .Chapter 7: The Tracing System 77 typename T3 = empty. If you look down through the file. { Callback<void.T2.T6. you will be able to find that the file .T7. The code creates a Callback of the right type and assigns your function to it. typename T6 = empty.Assign (callback). you will see a ConnectWithoutContext method here: template<typename T1.... typename T5. The only thing left is to look at the definition of Callback.
We are trying to figure out what the Callback<void. Ptr<const MobilityModel> > cb. .Chapter 7: The Tracing System 78 This class template implements the Functor Design Pattern.cc example.the fifth optional template argument represents the type of the fourth argument to the callback. If you want a context. You’ll need: void CourseChangeCallback (std::string path. Ptr<const MobilityModel> represents the first argument to the callback. you need to Config::Connect and use a Callback function that takes a string context. . } which is exactly what we used in the third. Ptr<const MobilityModel> model) { . It is used to declare the type of a Callback: . From this you can infer that you need a function that returns void and takes a Ptr<const MobilityModel>.the sixth optional template argument represents the type of the fifth argument to the callback. declaration means. . This is because the Connect function will provide the context for you. .the third optional template argument represents the type of the second argument to the callback. ... Perhaps you should now go back and reread the previous section (Take My Word for It). you can add the keyword static and come up with: static void CourseChangeCallback (std::string path. The second (non-optional) parameter. Now we are in a position to understand that the first (non-optional) parameter. The Callback in question is your function to receive the trace events.the first non-optional template argument represents the return type of the callback. For example.the fourth optional template argument represents the type of the third argument to the callback. } That’s all you need if you want to Config::ConnectWithoutContext.. void.. represents the return type of the Callback. void CourseChangeCallback (Ptr<const MobilityModel> model) { . Ptr<const MobilityModel> model) { ..the second optional template argument represents the type of the first argument to the callback. . } If you want to ensure that your CourseChangeCallback is only visible in your local file..
. TracedCallback<T. Here you see that the TracedValue is templated. Recall that the callback target of a TracedCallback always returns void. we presented a simple piece of code that used a TracedValue<int32_ t> to demonstrate the basics of the tracing code. in my opinion. we will just point you at the correct file. int32_t newValue) { .. We just glossed over the way to find the return type and formal arguments for the TracedValue.. and the second being the new value being set. the typename is int32 t. T> which will correspond to a TracedCallback<int32_t. m_v = v.2. This means that the member variable being traced (m_v in the private section of the class) will be an int32_t m_v... void Set (const T &v) { if (m_v != v) { m_cb (m_v. a quite elegant thing.7 What About TracedValue? Earlier in this section. The Set method will take a const int32_t &v as a parameter. of course. It is. 7. Rather than go through the whole exercise. private: T m_v. } It probably won’t surprise you that this is exactly what we provided in that simple example we covered so long ago: .. src/core/traced-value. v). }. Therefore the callback will need to have a function signature that looks like: void MyCallback (int32_t oldValue. feel free to take a look at the ns-3 manual. They are one of the most frequently used constructs in the low-level parts of ns-3.T> m_cb. Further recall that there is a one-to-one correspondence between the template parameter list in the declaration and the formal arguments of the callback function. In the simple example case at the start of the section. The callback. You should now be able to understand that the Set code will fire the m_cb callback with two parameters: the first being the current value of the TracedValue.Chapter 7: The Tracing System 79 If you are interested in more details regarding the implementation of Callbacks. } } . m_cb is declared as a TracedCallback<T. int32_t> when the class is instantiated.h and to the important piece of code: template <typename T> class TracedValue { public: .
“TCP/IP Illustrated. //Congestion window You should now understand this code completely. we can TraceConnect to the “CongestionWindow” trace source if we .SetParent<TcpSocket> () . ‘‘The TCP connection’s congestion window’’. int32_t newValue) { std::cout << "Traced " << oldValue << " to " << newValue << std::endl. MakeTraceSourceAccessor (&TcpSocketImpl::m_cWnd)) . Richard Stevens is a classic.1 Are There Trace Sources Available? The first thing to think about is how we want to get the data out.” by W. Stevens calls this. } This should tell you to look for the declaration of m_cWnd in the header file src/internet-stack/tcp-socket-impl. If you open src/internet-stack/tcp-socket-impl. } 7. the following declarations: TypeId TcpSocketImpl::GetTypeId () { static TypeId tid = TypeId(‘‘ns3::TcpSocketImpl’’) .cc’ | xargs grep -i tcp You will find page after page of instances of tcp pointing you to that file. you will eventually find: ns3::TcpSocketImpl CongestionWindow: The TCP connection’s congestion window It turns out that the ns-3 TCP implementation lives (mostly) in the file src/internetstack/tcp-socket-impl. 7.10.AddTraceSource (‘‘CongestionWindow’’. you can use the recursive grep trick: find . -name ’*. return tid. you will see right up at the top of the file. Value of cwnd and send sequence number while data is being transmitted. If you scroll through the list.h. If you open this file in your favorite editor.Chapter 7: The Tracing System 80 void IntTrace (int32_t oldValue. Recall that this is found in the ns-3 Doxygen in the “Core” Module section. If we have a pointer to the TcpSocketImpl.” Let’s just recreate the cwnd part of that plot in ns-3 using the tracing system and gnuplot.3 A Real Example Let’s do an example taken from one of the best-known books on TCP around. What is it that we need to trace? The first thing to do is to consult “The list of all trace sources” to see what we have to work with. “Figure 21.cc. Volume 1: The Protocols.cc in your favorite editor. you will find: TracedValue<uint32_t> m_cWnd. I just flipped the book open and ran across a nice plot of both the congestion window and sequence numbers versus time on page 366. If you don’t know this a priori.3.
If you look at this function. we have provided the file that results from porting this test back to a native ns-3 script – examples/tutorial/fifth. We now know that we need to provide a callback that returns void and takes two uint32_ t parameters. so this is probably a very good bet. You will typically find that test code is fairly minimal.3. We haven’t visited any of the test code yet. so we can just pull it out and wrap it in main instead of in DoRun.3 A Common Problem and Solution The fifth.cc and src/test/ns3tcp/ns3tcp-cwnd-test-suite. } 7. The second time period is sometimes called . Rather than walk through this..Chapter 7: The Tracing System 81 provide an appropriate callback target. except that we are talking about uint32_t instead of int32_t. but before Simulator::Run is called. so it turns out that this line of code does exactly what we want..cc’ | xargs grep CongestionWindow This will point out a couple of promising candidates: examples/tcp/tcp-largetransfer. we could TraceConnect to the “CongestionWindow” trace source. The first time period is sometimes called “Configuration Time” or “Setup Time. This should look very familiar to you. It is a script run by the test framework.” You will find.cc in your favorite editor and search for “CongestionWindow. There are three basic time periods that exist in any ns-3 script.2 What Script to Use? It’s always best to try and find working code laying around that you can modify. This is no different than saying an object must be instantiated before trying to call it. grep is your friend: find . 7. uint32_t newValue) { . We mentioned above that if we had a pointer to the TcpSocketImpl. MakeCallback (&Ns3TcpCwndTestCase1::CwndChange.” and is in force during the period when the main function of your script is running.cc. As usual. -name ’*. It turns out that is exactly what it is. the first being the old value and the second being the new value: void CwndTrace (uint32_t oldValue. Open src/test/ns3tcp/ns3tcp-cwnd-test-suite. ns3TcpSocket->TraceConnectWithoutContext (‘‘CongestionWindow’’. this)). That’s exactly what we have here.3. you will find that it looks just like an ns-3 script. Although this may seem obvious when stated this way. Let’s return to basics for a moment. by step.cc. This is the same kind of trace source that we saw in the simple example at the start of this section. step. it does trip up many people trying to use the system for the first time. rather than starting from scratch. So the first order of business now is to find some code that already hooks the “CongestionWindow” trace source and see if we can modify it.cc example demonstrates an extremely important rule that you must understand before using any kind of Attribute: you must ensure that the target of a Config command exists before trying to use it. so let’s take a look there. Let’s go ahead and extract the code we need from this function (Ns3TcpCwndTestCase1::DoRun (void)).
c-file-style:’’gnu’’. indent-tabs-mode:nil.cc in your favorite editor.h" "ns3/helper-module. * * This program is distributed in the hope that it will be useful. The answer to this issue is to 1) create a simulator event that is run after the dynamic object is created and hook the trace when that event is executed. This is to ensure that the simulation is completely configured before the app tries to do anything (what would happen if it tried to connect to a node that didn’t exist yet during configuration time). We took the second approach in the fifth. Include. and give the object to the system to use during simulation time.3. hook it then. without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. Suite 330.*/ /* * This program is free software.cc Walkthrough Now. An ns-3 Application always has a “Start Time” and a “Stop Time” associated with it. if not. you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation.cc example. the script enters what can be called “Teardown Time.h" "ns3/simulator-module.h" USA . After it completes executing the simulation. MA 02111-1307 */ #include #include #include #include #include #include <fstream> "ns3/core-module. -*. Boston. This decision required us to create the MyApp Application.h" "ns3/node-module. an ns-3 Socket is a dynamic object often created by Applications to communicate between Nodes. or 2) create the dynamic object at configuration time. When this happens. write to the Free Software * Foundation. In the vast majority of cases. * but WITHOUT ANY WARRANTY. 7.Chapter 7: The Tracing System 82 “Simulation Time” and is in force during the time period when Simulator::Run is actively executing its events. * * You should have received a copy of the GNU General Public License * along with this program..Mode:C++. let’s take a look at the example program we constructed by dissecting the congestion window test. You should see some familiar looking code: /* -*.h" "ns3/common-module.” which is when the structures and objects created during setup and taken apart and released. an Application will not attempt to create a dynamic object until its StartApplication method is called. 59 Temple Place. the entire purpose of which is to take a Socket as a parameter. In particular. Perhaps the most common mistake made in trying to use the tracing system is assuming that entities constructed dynamically during simulation time are available during configuration time.4 A fifth. Simulator::Run will return control back to the main function. See the * GNU General Public License for more details. Open examples/tutorial/fifth.
On the minus side. // // So. This has all been covered. even if we // could arrange a call after start time. so we wouldn’t be // able to hook the socket (now) at configuration time. Normally one would use an on-off application to generate a // flow. but this has a couple of problems. Second. then we pass // this socket into the constructor of our simple application which we then // install in the source node. but this is trivial. the socket is not public so we // couldn’t get at it. On the plus side we don’t need all of the complexity of the on-off // application. NS_LOG_COMPONENT_DEFINE ("FifthScriptExample").1. First. so we have to get // a little more involved in the details. we can cook up a simple version of the on-off application that does what // we want. // =========================================================================== // // node 0 node 1 // +----------------+ +----------------+ // | ns-3 TCP | | ns-3 TCP | // +----------------+ +----------------+ // | 10.1.1. We need // to crank up a flow and hook the CongestionWindow attribute on the socket // of the sender. we don’t have a helper. the socket of the on-off // application is not created until Application Start time. // // So first.Chapter 7: The Tracing System 83 using namespace ns3. we create a socket and do the trace connect on it. so we won’t rehash it. The next lines of source are the network illustration and a comment addressing the problem described above with Socket. 2 ms // // // We want to look at changes in the ns-3 TCP congestion window. The next part is the declaration of the MyApp Application that we put together to allow the Socket to be created at configuration time.1 | | 10. class MyApp : public Application { public: .2 | // +----------------+ +----------------+ // | point-to-point | | point-to-point | // +----------------+ +----------------+ // | | // +---------------------+ // 5 Mbps.1. // =========================================================================== // This should also be self-explanatory.
m_sendEvent. app->Start (startTime). You can see that this class inherits from the ns-3 Application class. m_peer.0)). m_nPackets.3. The MyApp class is obligated to override the StartApplication and StopApplication methods. schedules an event to start the Application: Simulator::Schedule (startTime. The Start method of an Application calls Application::ScheduleStart (see src/helper/application-container.Stop (Seconds (10. app->Stop (stopTime). The application container code (see src/helper/application-container. These methods are called when the corresponding base class Start and Stop methods are called during simulation time.h if you are interested) loops through its contained applications and calls.h if you are interested in what is inherited. void SendPacket (void). The most common way to start pumping events is to start an Application.. Address address.0)). &Application::StartApplication. m_dataRate. 7.1 How Applications are Started and Stopped It is worthwhile to spend a bit of time explaining how events actually get started in the system.Start (Seconds (1. virtual void StopApplication (void).4. void Setup (Ptr<Socket> socket. apps. m_running. in turn. uint32_t nPackets. m_socket. m_packetsSent. void ScheduleTx (void). Take a look at src/node/application. on each of them. Ptr<Socket> Address uint32_t uint32_t DataRate EventId bool uint32_t }. m_packetSize.cc) which. This is done as the result of the following (hopefully) familar lines of an ns-3 script: ApplicationContainer apps = . private: virtual void StartApplication (void).. . this). apps. uint32_t packetSize.Chapter 7: The Tracing System 84 MyApp (). DataRate dataRate). virtual ~MyApp().
of course: MyApp::MyApp () : m_socket (0). Address address. m_running (false). . 7. this->StartApplication (startTime). m_peer (). m_packetSize = packetSize. m_packetSize (0). Recall that we are going to create the Socket as a TcpSocket (which is implemented by TcpSocketImpl) and hook its “CongestionWindow” trace source before passing it to the Setup method. We are just initializing member variables.2 The MyApp Application The MyApp Application needs a constructor and a destructor. m_nPackets = nPackets. uint32_t nPackets. m_peer = address. m_sendEvent (). m_dataRate (0).4. this bit of code causes the simulator to execute something that is effectively like. m_packetsSent (0) { } MyApp::~MyApp() { m_socket = 0. if you have kept it all straight. } The existence of the next bit of code is the whole reason why we wrote this Application in the first place. like sending packets. The important one from the perspective of tracing is the Ptr<Socket> socket which we needed to provide to the application during configuration time. It is then expected that another event will be scheduled in the overridden StartApplication that will begin doing some application-specific function. } This code should be pretty self-explanatory. m_nPackets (0).3. m_dataRate = dataRate. DataRate dataRate) { m_socket = socket. uint32_t packetSize.Chapter 7: The Tracing System 85 Since MyApp inherits from Application and overrides StartApplication. is the pointer to the Application in the container. StopApplication operates in a similar manner and tells the Application to stop generating events. void MyApp::Setup (Ptr<Socket> socket. where the this pointer.
You can see that it does a Socket Bind operation. After the Connect. we break the chain of events that the Application is using to keep sending its Packets and the Application goes quiet. If the Event is pending execution or executing. m_socket->Connect (m_peer). By doing this. This removes the last reference to the underlying Ptr<Socket> which causes the destructor of that Object to be called. if (m_sendEvent. m_packetsSent = 0.IsRunning ()) { Simulator::Cancel (m_sendEvent).Chapter 7: The Tracing System 86 void MyApp::StartApplication (void) { m_running = true. In this code. we Cancel the event which removes it from the simulator event queue. Recall that StartApplication called SendPacket to start the chain of events that describes the Application behavior. The next bit of code explains to the Application how to stop creating simulation events. void MyApp::StopApplication (void) { m_running = false. } The above code is the overridden implementation Application::StartApplication that will be automatically called by the simulator to start our Application running. The socket is actually deleted in the destructor when the m_socket = 0 is executed. its method IsRunning will return true. It performs the required work on the local side of the connection just as you might expect. m_socket->Bind (). . SendPacket (). After we quiet the Application we Close the socket which tears down the TCP connection. It should now be clear why we need to defer a lot of this to simulation time. since the Connect is going to need a fully functioning network to complete. if IsRunning() returns true. } } Every time a simulation event is scheduled. } if (m_socket) { m_socket->Close (). If you are familiar with Berkeley Sockets this shouldn’t be a surprise. an Event is created. The following Connect will do what is required to establish a connection with the TCP at Address m peer. the Application then starts creating simulation events by calling SendPacket.
If the Applciation is running (if StopApplication has not been called) it will schedule a new event. you see that ScheduleTx does exactly that. &MyApp::SendPacket.GetBitRate ( m_sendEvent = Simulator::Schedule (tNext. if (++m_packetsSent < m_nPackets) { ScheduleTx (). this). 7. The data rate of an Application is just that.3.3 The Trace Sinks The whole point of this exercise is to get trace callbacks from TCP indicating the congestion window has been updated. } } Here. You can probably imagine that you could load the resulting output into . if you know Berkeley Sockets.4. It does not take into account any overhead for the various protocols or channels that it uses to transport the data. so the next lines call ScheduleTx to schedule another transmit event (a SendPacket) until the Application decides it has sent enough. The alert reader will spot something that also trips up new users. so we won’t dwell on the details. The next piece of code implements the corresponding trace sink: static void CwndChange (uint32_t oldCwnd. you see that SendPacket does just that. uint32_t newCwnd) { NS_LOG_UNCOND (Simulator::Now (). It is the responsibility of the Application to keep scheduling the chain of events. m_socket->Send (packet). This is the rate at which the Application produces bits. is probably just what you expected to see.GetSeconds () << ‘‘\t’’ << newCwnd). If you set the data rate of an Application to the same data rate as your underlying Channel you will eventually get a buffer overflow. } } Here. } This should be very familiar to you now. void MyApp::ScheduleTx (void) { if (m_running) { Time tNext (Seconds (m_packetSize * 8 / static_cast<double> (m_dataRate. It has nothing to do with the data rate of an underlying Channel.Chapter 7: The Tracing System 87 void MyApp::SendPacket (void) { Ptr<Packet> packet = Create<Packet> (m_packetSize). This function just logs the current simulation time and the new value of the congestion window every time it is changed. It creates a Packet and then does a Send which. which calls SendPacket again.
we really want to introduce link errors which will drop packets. If we trace a connection that behaves perfectly. Ptr<RateErrorModel> em = CreateObjectWithAttributes<RateErrorModel> ( "RanVar". We are using the RateErrorModel which allows us to introduce errors into a Channel at a given rate.Create (2). so we wanted to demonstrate this working. StringValue ("2ms")). This should tell you that the callback target should be a function that returns void and takes a single parameter which is a Ptr<const Packet> – just what we have above.h for this member variable.3.cc) you will see that this trace source refers to PointToPointNetDevice::m_phyRxDropTrace.4 The Main Program The following code should be very familiar to you by now: int main (int argc. devices = pointToPoint. RandomVariableValue (UniformVariable (0. PointToPointHelper pointToPoint. To see any interesting behavior.SetChannelAttribute ("Delay".Chapter 7: The Tracing System 88 a graphics program (gnuplot or Excel) and immediately see a nice graph of the congestion window behavior over time. pointToPoint. StringValue ("5Mbps")). ns-3 provides ErrorModel objects which can be attached to Channels. The next few lines of code show something new.Install (nodes). This trace source fires when a packet is dropped by the physical layer of a NetDevice. .)). cause duplicate ACKs and trigger the more interesting behaviors of the congestion window. nodes. NetDeviceContainer devices. 1. We are going to add an error model to this code also.4. we will end up with a monotonically increasing congestion window. pointToPoint. If you then look in src/devices/pointto-point/point-to-point-net-device.. "ErrorRate".GetSeconds ()). static void RxDrop (Ptr<const Packet> p) { NS_LOG_UNCOND ("RxDrop at " << Simulator::Now (). DoubleValue (0. just as shown in the illustration at the start of the file. } This trace sink will be connected to the “PhyRxDrop” trace source of the point-to-point NetDevice. If you take a small detour to the source (src/devices/point-topoint/point-to-point-net-device.Get (1)->SetAttribute ("ReceiveErrorModel". devices. This creates two nodes with a point-to-point channel between them. We added a new trace sink to show where packets are dropped. PointerValue (em)). char *argv[]) { NodeContainer nodes.SetDeviceAttribute ("DataRate". 7.00001)). you will find that it is declared as a TracedCallback<Ptr<const Packet> >.
MakeCallback (&CwndChange)). The first statement calls the static member function Socket::CreateSocket and provides a Node and an explicit TypeId for the object factory used to create the socket. sinkPort)).Install (nodes). TcpSocketFactory::GetTypeId ()). Ipv4InterfaceContainer interfaces = address. We also set the “ErrorRate” Attribute. InetSocketAddress (Ipv4Address::GetAny (). uint16_t sinkPort = 8080. Otherwise. to create instances of the Objects created by the factory. Ipv4AddressHelper address. This should all be familiar. This is a slightly lower level call than the PacketSinkHelper call above. This class implements a design pattern called “object factory” which is a commonly used mechanism for specifying a class used to create objects in an abstract way. The PacketSink Application is commonly used in ns-3 for that purpose. PacketSinkHelper packetSinkHelper ("ns3::TcpSocketFactory". we use the convenience function CreateObjectWithAttributes which allows us to do both at the same time. address.Assign (devices). Here. The remaining parameter tells the Application which address and port it should Bind to. InternetStackHelper stack. sinkApps.Chapter 7: The Tracing System 89 The above code instantiates a RateErrorModel Object. it is conceptually the same thing. InetSocketAddress (Ipv4Address::GetAny ().Stop (Seconds (20. with the exception of.GetAddress (1). Since we are using TCP.0’’. ns3TcpSocket->TraceConnectWithoutContext (‘‘CongestionWindow’’. sinkApps. Rather than using the two-step process of instantiating it and then setting Attributes. Ptr<Socket> ns3TcpSocket = Socket::CreateSocket (nodes. It installs internet stacks on our two nodes and creates interfaces and assigns IP addresses for the point-to-point devices.1. We set the “RanVar” Attribute to a random variable that generates a uniform distribution from 0 to 1.1.255. The next two lines of code will create the socket and connect the trace source. in turn. sinkPort)).Get (1)). ‘‘255.252’’).255. stack. The above code should be familiar. . we need something on the destination node to receive TCP connections and data.)). PacketSinkHelper packetSinkHelper ("ns3::TcpSocketFactory".SetBase (‘‘10.Start (Seconds (0. sinkPort)). We then set the resulting instantiated RateErrorModel as the error model used by the point-to-point NetDevice. This will give us some retransmissions and make our plot a little more interesting. ApplicationContainer sinkApps = packetSinkHelper.Get (0).)). This code instantiates a PacketSinkHelper and tells it to create sockets using the class ns3::TcpSocketFactory. and uses an explicit C++ type instead of one referred to by a string. Address sinkAddress (InetSocketAddress(interfaces.Install (nodes. you provide the PacketSinkHelper a string that specifies a TypeId string used to create an object which can then be used. instead of having to create the objects themselves.
3.Get (1)->TraceConnectWithoutContext("PhyRxDrop". what address to connect to. we can use TraceConnectWithoutContext to connect the CongestionWindow trace source to our trace sink. It should now be obvious that we are getting a reference to the receiving Node NetDevice from its container and connecting the trace source defined by the attribute “PhyRxDrop” on that device to the trace sink RxDrop.)). We now have to instantiate that Application. In this case. All of the work we orchestrated by creating the Application and teaching it how to connect and send data actually happens during this function call.5 Running fifth. app->Setup (ns3TcpSocket. 1000. 1040. 7. Simulator::Destroy (). Recall that we coded an Application so we could take that Socket we just made (during configuration time) and use it in simulation time. DataRate ("1Mbps")). how many send events to generate and the rate at which to produce data from those events. how much data to send at each send event.cc for you. } Recall that as soon as Simulator::Run is called. MakeCallback (&RxDrop)).cc Since we have provided the file fifth. . We didn’t go to any trouble to create a helper to manage the Application so we are going to have to create and install it “manually.Chapter 7: The Tracing System 90 Once the TcpSocket is created and attached to the Node. nodes. configuration time ends. Finally.)). Simulator::Run (). if you have built your distribution (in debug mode since it uses NS LOG – recall that optimized builds optimize out NS LOGs) it will be waiting for you to run. Simulator::Destroy takes care of the gory details and we just return a success code after it completes. return 0. Next. sinkAddress. app->Start (Seconds (1. we manually add the MyApp Application to the source node and explicitly call the Start and Stop methods on the Application to tell it when to start and stop doing its thing. The first line creates an Object of type MyApp – our Application. we tell the simulator to override any Applications and just stop processing events at 20 seconds into the simulation. As soon as Simulator::Run returns. the simulation is complete and we enter the teardown phase.Get (0)->AddApplication (app). devices.” This is actually quite easy: Ptr<MyApp> app = CreateObject<MyApp> (). The second line tells the Application what Socket to use. and simulation time begins. app->Stop (Seconds (20. Simulator::Stop (Seconds(20)). We need to actually do the connect from the receiver point-to-point NetDevice to our callback now.
You can probably see immediately a downside of using prints of any kind in your traces.20919 1072 1.480 set output "cwnd.684s) 1. but I’m sure you can’t wait to see the results of all of this work.png" plot "cwnd.dat: .2471 8040 1./waf --run fifth > cwnd.dat 2>&1 Now edit up “cwnd.24895 8576 1.. You can now run gnuplot (if you have it installed) and tell it to generate some pretty pictures: gnuplot> gnuplot> gnuplot> gnuplot> set terminal png size 640..png” in all of its glory. We will remedy that soon.dat” in your favorite editor and remove the waf build status and drop lines. MakeCallback (&RxDrop)).dat" using 1:2 title ’Congestion Window’ with linespoints exit You should now have a graph of the congestion window versus time sitting in the file “cwnd. in the script to get rid of the drop prints just as easily. Let’s redirect that output to a file called cwnd. 1. leaving only the traced data (you could also comment out the TraceConnectWithoutContext("PhyRxDrop". We get those extraneous waf messages printed all over our interesting information along with those RxDrop messages..22103 2144 ./waf --run fifth Waf: Entering directory ‘/home/craigdo/repos/ns-3-allinone-dev/ns-3-dev/build Waf: Leaving directory ‘/home/craigdo/repos/ns-3-allinone-dev/ns-3-dev/build’ ’build’ finished successfully (0. that looks like: .21511 1608 1.25151 ..Chapter 7: The Tracing System 91 .2508 9112 RxDrop at 1.
Chapter 7: The Tracing System 92 .
8. We have really just scratched the surface of ns-3 in this tutorial. We hope to add the following chapters over the next few releases: • The Callback System • The Object System and Memory Management • The Routing System • Adding a New NetDevice and Channel • Adding a New Protocol • Working with Real Networks and Hosts Writing manual and tutorial chapters is not something we all get excited about. please consider contributing to ns-3 by providing one of these chapters. but we hope to have covered enough to get you started doing useful networking research using our favorite simulator. . It is impossible to cover all of the things you will need to know in one small tutorial. or any other chapter you may think is important. – The ns-3 development team.Chapter 8: Closing Remarks 93 8 Closing Remarks 8.2 Closing ns-3 is a large and complicated system. If you are an expert in one of these areas. We hope and expect it to grow over time to cover more and more of the nuts and bolts of ns-3. but it is very important to the project.1 Futures This document is a work in process.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 P parsing ascii traces . . . . . . . . . . . . . . . . . . . . 16 S simulation time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 class Application . . . . . . . . . 10 building with build. . . . . . . . . . . . . 40 ASCII tracing . . . . .tr . . . . . . . . . . . . . . . . . 40 NetDevice . . . . . . . . . . . . . . . . . . . 40 tracing . . . . . . . . . . . . . . . . . . . . . . 15 architecture . . . 3 Mercurial . . . . . . . . . . . . . . . 9 building with Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 system call . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17. . . . . . . . . . . . . . . 15 F first script . . . . 52 topology helper . . . . . . . . . . . 10 configuring Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 release repository . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 C C++ . . . . . . . . . . . . . 3 repository . . . . . . . 41 Python . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29. . . . . . . . . . . . . . . . . . . . . . . 15 node number . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 ascii trace dequeue operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 contributing . . . . . . . . . . 40 ascii trace receive operation . . . . . . . . . . . 4 R regression tests . 4 software configuration management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 T tarball . . 40 sockets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 logging . . . . . . . . . . . . . . . . 4. . . . . . . . . 15 class Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. . . . . . . . 4. . . . . . . 3 E Ethernet . . . . . . . . . . 10 bus network topology . . . . . . . . . . . 28 Logitech . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 pcap tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 toolchain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 topology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 command line arguments . . . . . . . . 33 compiling with Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 B build . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 N net device number . . . . . . . . . 2 Cygwin . . 6 tcpdump . . . . . . . . . . . . . . . . . . . 17 first. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 MinGW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 building debug version with Waf . . . . . . . . . . . . . . . . . . . . . . . 40 ns-3-dev repository . . . . . . . . . . . . . . . . . . . . . . 3 NS LOG . . . . . . . . . . 6 mercurial repository . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Node . . . . 14 D documentation. . . . . . . . . . . . . . . . . . . 17 G GNU. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .py . . 17 L Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 M make . . . . . . . . . . . . . . . . . . . . 40 pcap . 6 . . . . . . . . . . . . . . . . . . . .Index 94 Index A Application . . . . . . . . . . 12 regression tests with Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 smart pointer . . . . . . . . . . . . . . . . . . . . . . . 4 myfirst. . . . . . . . . . . . . .cc. . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 trace event . 40 ascii trace enqueue operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 ascii trace drop operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 running a script with Waf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 H helper . . . . . . . . . . . . 39 tracing packets . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . 12 unit tests with Waf . . . . . . .nsnam. . . 42 www. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . . . . . . . . . . . .Index 95 U unit tests . . . . . 52 Wireshark . . 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 wireless network topology . . . . . . . . . . . . . . . 41. . . . . 10 W Waf . . . . . . . . . . . . . . . .org. . . . . . . . . . . | https://www.scribd.com/document/60189782/Ns3-Tutorial | CC-MAIN-2017-51 | refinedweb | 38,461 | 59.8 |
- wxImage and Group 4 Tiffs?
- isNumber? check
- Prime number module
- Question: Inheritance from a buil-in type
- Google newsgroup "Unable to find thread" message?
- RELEASED Mailman 2.1.3
- PEP-0263 and default encoding
- Threading and Windows.
- data encoding for Python COM interface
- win32traceutil doesn't work in Win32 service when connected viaTerminal Services
- extension call backs and the GIL
- VB6 thin client
- Wavelet package available?
- query
- sendmsg operation on unix socket?
- write a file "copy" program in Python for Unix and Windows...
- Python mascot? How about a Pythoneer?
- Home Environment
- Recursion
- Python parser that records source ranges
- Checking for invalid keyword arguments?
- overriding built-ins
- Distutils Configuration for Visual Studio 6?
- fetching a cvs repository directory from a cgi script
- IDE with a specific feature
- Meta programming question
- Tutor
- FTP access to starship.python.net?
- Module using Numeric array: strange behaviour
- [development doc updates]
- Any tool to translate PHP to Python?
- Microbiology and/or image analysis tools
- unicode 3 digit decimal conversion
- Digest Authentication (RFC 2831)
- DDE Python-C++
- Now that rexec is gone...
- [2.2.1]How To Gracefully Shutdown an XML-RPC Server
- Saving Numeric arrays to file?
- safely closing a socket?
- Slice inconsistency?
- Python Modules for Netegrity's SiteMinder
- py2exe, 2.3 and services (raising flag again)
- Explaining how a (Mind) program works
- COM Dispatch() and multiple instances
- Python SSL Socket Client to Java SSL Server. HELP me PLEASE.
- COM: A bold new adventure in pain
- Access to database other web sites
- PIL: Group4 Tiff Images
- centering a dialog in pygtk
- PEPs link gone from the Python homepage?
- Reduce need of backslash
- Equivalent of std::set
- Is there an extension library?
- Where to import modules?
- server side socket program hangs
- WaitCommEvent error
- Successful installation of mod_python on Apache on Win2k?
- When did Windows start accepting forward slash as a path separator?
- Turning Strings into Functions
- Python for CGI scripting
- Listbox: How do I empty the view in the window
- Another Python rookie trying to port to C/C++ question.
- sockets
- Using Jython with JSPs
- Getting the Directory/Path details
- pythonwin from dos
- Python lexical scanner
- squeeze and package imports
- Packages and PyKyra
- "for" with "else"?
- Stuck building a Python extension
- MySQLdb installed, won't connect on OSX
- getting error in socket connection
- Help with dialog box
- Windows dialog box removal
- adodbapi / string encoding problem
- pilconvert script disappeared ?
- Getting C++ object from PyObject *
- send email
- getting a value from a web-page
- Comment on PEP-0322: Reverse Iteration Methods
- Memory profiling
- Why is ClientCookie/urllib2 using https?
- Others constructors...
- __init__ return value
- Python Glossary
- 64-bit EPIC and some modules
- unknown charset dialog(wxPython problem)
- printing API
- [Tkinter] event problem
- command exit status on windows
- command exit status on windows
- Python <-> C via sockets
- RELEASED Python 2.3.1
- float problem
- CGI Redirect to another page
- Fastest way to get thousands of db records to client in 3 tier?
- Python embedded in C & memory releasing
- why the inconsistency?
- Pre-PEP: reverse iteration methods
- How to tell if a forked process is done?
- Python Database Objects (PDO) 1.0.1 Released
- Passing a String to a Python CGI Script
- Barcode generator for Python?
- Pygtk, widget refreshing
- IE or Netscape com access
- C Extension and multiple namespaces
- eval's local arguement ignored?
- idle import problem
- pymqi sample code
- problem using urllib2: \n
- Bottleneck? More efficient regular expression?
- load changes of subclasses
- Executing a python script from an HTML link?
- TKinter and findertools.launch
- A fistful of proxies.
- command line or idle?
- Factory function to generate a named class
- Anyone who good at PYTHON in Taiwan?
- import pickle succeeds only after two tries??
- How to load the DLL with PyRun_SimpleString?
- compiling python program
- "Protected" property in Python?
- Need arguments for "Python vs. Perl as an OOPL"
- working png output from graphing modules
- How can load dll in C?
- parsing windows event files
- Connect to VBA Objects?
- What is a "method-wrapper" object?
- Thoughts on PEP284
- Thoughts on PEP315
- Compiling informxidb-1.3 on python2.2?
- string.rstrip
- General Password questions
- pyopengl / python2.3 / win32
- Calling functions before that are def'ed
- Trouble with wxListCtrl
- <?standalone yes?>
- Mailbox cleaner on an IMAP basis?
- Playing with dictionaries
- Memory allocation
- how to know the number of menu items in a cascade (sub) menu?
- questions about a C++ COM object accessed in Python via win32com
- Display message while running...
- ISDN / B-channel
- Installing Python and Tkinter on OSX
- Begineer Question : Global string substitution with re
- py2exe: dynamic module does not define init function
- Dr. Dobb's Python-URL! - weekly Python news and links (Sep 22)
- thread question
- Where to publish my code
- Cincinnati PUG
- Where does sys.path get initialized
- Exceptions in threads?
- using poplib to retrieve messages sent from a certain address
- Twisted Matrix and Python Scripts
- how to find a user in a xml document
- math.exp(complex)
- Download to client from a cgi script?
- Creating a new file
- Help me choose a C++ compiler to work with Python
- Has psycopg moved?
- Starship back
- problem about wxPython error
- Trouble sorting lists (unicode/locale related?)
- How to catch socket timeout?
- Tkinter cursor question
- None???
- What is the meaning of the astarisk in Python
- Swen Detection code
- DB2 driver for windows
- 2nd iteration of a character
- XML/XSLT with Python
- Tkinter-PhotoImage question
- starship.python.net DNS problems
- raw string
- I love this language!
- Mutable strings
- Insert into text field?
- python script as an emergency mailbox cleaner
- Visual Basic Procedure Names
- console chat
- Help Compiling 64bit Python
- Problems with string and lists (searching and replaceing)
- urls on securing python on a shared hosting ?
- Sick (yaml library) and Python
- spam killing with poplib
- py2exe web site
- How you chomp in python
- bizarre behavior using .lstrip
- Proxy Authentication using urllib2
- Passing data out of a Sax parser
- Server-side programming
- whitespace
- os.environ and os.path.chdir
- Custom Execution from Python
- Python API
- changing environment variables
- numarray.linear_algebra.eigenvectors bug ?
- Help with parsing and matching a character
- return codes from py2exe python script
- python cgi problem with method = post
- Grail
- mutex? protecting global data/critical section
- How do I learn operator overriding?
- \r for newline in readlines function
- pop3 email header classifier?
- (no subject)
- Webware and memory problem!
- Pmw menubutton enable/disable
- Webware and RAM!
- Anyone has "pyuploader.zip" from Parnassus ?
- utf-8 encoding issue
- convert Unicode to lower/uppercase?
- file position *tell()* works different
- Access to Win API
- PpythonWin: Unix File Mode
- Retrieve class name of caller
- Exception in Threading
- string objects...
- problem with user confirmation
- Unicode 4.0 updates to unicodedata?
- Binding frustration
- pmw MenuBar: delete all menu items of a menu
- print and trailing white space
- Python ISPs
- Join equivalent for tuples of non-strings?
- file object: seek and close?
- Scaling Tk scrollbar handles
- Forcing getopt to process a specific option first??
- SciTE
- win32comm
- How to inheritance overwrite non virtual?
- "tuple index out of range" side effect of C extension??
- ASN.1 source reader
- semantic operation in simpleparse?
- How to pass parameter to a module?
- Closures in python
- Multithreading and locking
- MATLAB & Python?
- DLL info
- variable to be transmitted from a CGI script to a module
- sendf EOF by socket
- Help with regular expression
- Checking if the computer is online
- PySQLite 0.4.3 install problem on Mandrake9.1
- Setting environment variables
- Where's UserList.ListMixin?
- About Python's C API
- Module to generate OpenOffice Writer documents
- win32: 'Tailing' the NT event log using pipes?
- convert ints in a range to strings
- database
- Clearing file object buffers
- Did my SMTPlib question go through ?
- smtplib question
- how to write-protect names
- New style classes
- Money data type
- GTK problems on gnome-python install
- build python using mingw
- "New" style classes
- Sound, time and platform issues
- Mandrake Linux 9: Why isn't bsddb working?
- mx.DateTime bogus warning: "float where int expected"
- SMTPlib Emailing Attachments
- fast way to filter a set?
- Python 2.1 Compilation error - can't find krb5.h
- import on modules/files that don't have .py extension
- Fsrit Propttoye: SCLABMRE
- Simple python cgi question
- scp in python
- Wrapper for libevent?
- htmllib.HTMLParser and unicode
- Graphic python debugger
- Re[2]: GTK help
- replace %(word) in a string
- Python exercises
- Can Python Module Locate Itself?
- GTK help
- Hurricane Isabel
- testing the data-type of a stream
- How can I rewrite the JPython to Python?
- audio module
- Streams causing heap crash using Python C-Extensions
- MORE volunteers for voting project needed
- Extending the python language
- Help with Python
- When passing functions as args,how to pass extra args for passed function?
- texi2html in a cgi ?
- scoping with lambda in loops
- how to build a list of mx.DateTime objects from a start and enddate?
- xmlrpclib/timeoutsocket not happy together in 2.3
- Tkinter: tooltips (Windows)
- Missing python23_d.lib in Windows Python 2.3
- is importing redundantly harmful/wasteful?
- why pass statement?
- None, False, True
- Indexing list of lists
- App development for Sharp Zaurus PDA using QTopia?
- distutils on win32 with link.exe -- use dlls?
- Variable lookup in calling context
- tuples vs lists
- Tkinter.Text can never be empty?
- Python Audio (Alpy, Fastaudio, Etc Etc)
- py2exe & pyxml
- Tkinter and Movie Player
- Tkinter and Pygame
- (no subject)
- getch() in Python
- Duck Typing
- how can I use Berkeley DB XML 1.1.0 from Python 2.3 (on Windows) ?
- PyQt: Can't show QFileDialog.getExistingDirectory in Python Thread, bug or not?
- Anyone? Anyone?
- Re[2]: from clipper move to python
- object.attribute vs. object.getAttribute()
- wanted($): mac and linux python gui coders for porting
- compound conditional statements
- Ways to improve this clock algorithm?
- _ssl.so build problems on Solaris 8 for 2.3
- calling objects parents' __init__?
- Calling Python from PHP
- unreferenced (???) variable; prob. simple
- Dr. Dobb's Python-URL! - weekly Python news and links (Sep 15)
- irritating problem
- strptime() backported to Jython 2.1
- Powerset
- unique unions of several dict keys
- Multiple instances of modules
- Processes.
- Datetime utility functions
- list of lists
- py2exe endcodings/SAXReader Issues
- Win32 Com + ADO: How to compare the result of a recordset to 'nothing'
- distributing a standalone python app. under Linux
- I: caching the sql queries
- R: caching the sql queries
- caching the sql queries
- configuration files and 'distutils'
- Could somebody please explain what is happening ....
- pyc2py
- from clipper move to python
- datetime: How to get diff between 2 dates in month units?
- Simples examples for standard libraries
- stripping a string
- Is there "let binding" in Python?
- Anybody know the Etymology of the word 'Sprint' as in coding Sprint?
- embedding python code in html, like PHP
- gdmodule true type font size
- Finally found a use for lambda!
- Expat version in Python 2.3: 1.95.6?
- package _mysql with McMillan
- problems with threaded socket app
- singleton or Borg?
- No "side effect" assignment!
- Huh?!?!? What is it that I'm seeing here?
- Print always puts newline (or adds a space)
- Encoding problems with DCOracle2
- checking type of my own objects
- Building patterns
- Importing modules from within other modules
- better use of os.system()
- Java Checked Exceptions
- referencing in cmd string
- Postgres BYTEA support python module
- thread focus question.
- String functions
- Sending SMS...
- Boost Python and MS Visual Studio 7 - hit compiler limit
- Pipe problem
- Algorithmic complexity of StringIO?
- Problems with pyPgSQL
- Directory names from untrusted data
- commands.getstatusoutput("FOO") -- won't work for me
- Python / Chinese Encodings
- How to make URLLIB .urlopen(url1) .urlopen(url2) session related ?
- Dislin compilation
- Changing Mouse Cursor
- GUI versions using PythonCard are available
- authenticating against shadow passwords
- (long message)Request advie on an app I'm doing
- pygtk/gnome warning 989
- Printing a Tk canvas under windows
- date formatting
- Copying latest version of the file
- read from file
- Internal Debugger Question
- Toronto-area Python user group September meeting
- RAD with Python
- Redhat 9.0 & tkinter
- Problems with Installer V-5b5_5 on linux
- Is there a web templating system like ruby's amrita ?
- python program not running from different folders
- asyncore question
- using weave with Windows
- formatting number in the CSV module
- Usenet posting wiht python
- Trouble with source-code encoding
- Can This Code Be Made Faster?
- looking for MOP documentation
- Overlaying transparent images with PIL
- ascii codec missing under py2exe
- CUTE 0.2 released
- Python 2.3 and Mac OS X
- Recursive import within packages
- Parsing XML streams
- python on mac (os x): application automation
- list to string
- Web server with Python
- jython registry file lookup in J2EE environment
- Flash Remoting in Python?
- dumb q: repeated inheritance in python?
- why is this failing?
- mixing for x in file: and file.readline
- getting ttf font/family name; fontTools?
- gdmodule clipping
- Canvas - Rectangle. Is there an easy way to detect if you're inside?
- JPEG comments and PIL
- tkinter, sockets and threads together
- ftplib question: how to upload files?
- Problems with Python/Pycrust on SUSE 8.2
- mod_python installation
- Import question
- PyQt: Can't show PY variable in QT filedialog as initially parameter
- Problem with custom extension: help needed
- GET and POST
- Fetching address-string from web-browser
- How to compile 3rd party python into single large binary
- Building Imaging-1.1.4 on Solaris 9
- Cool & Useful EmPy project, prior art?
- how to use .wsdl files?
- socket error 10061
- Fatal Python error: PyEval_RestoreThread: NULL tstate
- interscript:developers needed to take over project
- Arctec newsletter mentions Guido's move
- a quick program to download tv listings
- FIFO problems
- Learning Python, 2'nd Edition, O'Reilly
- Tkinter Canvas/Grid question
- Pickling/unpickling extensions types
- metaclasses for type comparison
- Embeddor Woes...
- Trouble Learning Zope
- getrusage
- Python C extension: Value different if passed as list than if passed as number
- ANN hashtar 0.1: archival encryption to corruptible media
- TTF fonts rendered in Python
- Heisenberg strikes again!
- Background image in toplevel
- Opposite of yield?
- Running user scripts in a Tkinter app?
- sort by file/directory
- Problems with setting up mod_python and Apache
- SWIG, MinGW, and Python 2.3 problem
- python module for MS SQL Server 7 or 2000?
- negative indices for sequence types
- Testing
- Question on Standard Types
- Ternery operator
- win32com: create email message in windows default app?
- Accessing a file with contents
- Package importing problem
- Anyone ever overridden a builtin by accident?
- Redesign of Python site
- Python glossary - contribute your definitions
- Need help on UNICODE conversion
- Great Book on Web Services
- Any tools to print source code call hierarchy
- [2.3] object does not appear to be a reserved word
- Content violation
- cygwin python with Tkinter and pexpect module
- in CGI, how to include html pages?
- Unreachable
- Standard Lib To Use New-Style Classes?
- XRC, wxMenuBar's, and wxMenuBarPtr's
- just curious
- list to string
- Packaging?
- dbm under Windows
- _imp__* linking errors when compiling extension using Mingw
- Replacing python: in syslog log messages
- I need a little help with Tkinter Entry widget
- Help with a utility class and having a method call another method from within
- ezmlm response
- Why python???
- ezmlm response
- Coding Style: Defining Functions within Methods?
- Automated code generation
- How to override distutils install?
- Why the 'self' argument?
- Saving IDLE output
- Creating a PyFunction
- Executing a Jython function from Java
- Database over 2gb
- compiling python
- Coloring markup simply...
- spyce hangs on 'file not found'
- Starting point for unicode conversion
- multiline repr (code generation)
- in Ram database?
- Embedded Perl or Python (XPost)
- Python port to Swiss Ephemeris work with 2.3?
- ezmlm response
- Anybody else having problem *sending* mail to this list?
- this is my own python-setuid.c
- Selecting elements from a list
- Content violation
- [development doc updates]
- Output in color
- Massive unit test vs MySQL
- Strange problem with xml.dom.minidom Text object (Python 2.3)
- Pmw ScrolledText Widget
- Comparing objects - is there a maximum object?
- Report to Sender
- removing terminal control characters
- unicode memory usage
- pyOpenSSL + windows
- xml.dom.minidom - bug ? future ?
- PYTHONPATH x *.pth???
- Search and Replace in Zope
- Tracking Sessions in a Python CGI app.
- Zope locking
- How do I match this with re
- basic language question
- Circle Hell
- Optionmenu.
- Where to post useful code
- Zope mailing lists dead ?
- question: How can I put empty directories into a zip-archive?
- built with --enable-shared but get error: libpython2.3.so.1.0: can't open shared object file
- Using python RPM on SuSE 8.1 : No libpython.a or libpython.so present. Why?
- python plugin for excel (ala gnumeric)
- Importing/reloading modules
- py2exe copies to much dlls
- Active Scripting in Python
- strings
- questions
- Drag and drop in Tkinter. How difficult is it?
- python 2.3, cvs module specific question
- quick and smart way for parsing a CSV file?
- installing PyQt
- Lisp with Kenny! <g>
- Invisible function attributes
- Tkinter Text widget getting too slow
- Extracting TIFF from emails
- wxGrid?
- How to accelerate python application GUI speed
- Trouble with script fetching site | https://bytes.com/sitemap/f-292-p-88.html | CC-MAIN-2020-45 | refinedweb | 2,765 | 58.48 |
The Yorick Programming Language
The Yorick Programming Language, written by David Munro, "...is an interpreted programming language for scientific simulations or calculations, postprocessing or steering large simulation codes, interactive scientific graphics, and reading, writing, or translating large files of numbers..."
The size of the Yorick 2.2.03 binaries and supporting files, i.e., the relocate directory with doc and g removed for my ARM9 cross compilation was ~5MB. The interpreter binary was ~1.2MB with ~1.5MB for the Yorick include files (demo and test files deleted) but you can trim this further depending on what you use. Things like the C include files can be removed (you only need these if you're writing your own Yorick extensions). The lib can also be trimmed if you were not using the hex or drat libraries and didn't ever need to run codger on target. You could also trim out Gist, which is a Computer Graphics Metafile (CGM) file viewer.
Page Contents
Useful Links
Easy Yorick Install On Ubuntu
Really very very easy. Just do sudo apt-get install yorick! If you want to develop your own Yorick plugins then your will also need to do a sudo apt-get install yorick-dev.
apt-get install yorick yorick-dbg yorick-dev
Get Yorick Home Path, Y_HOME
For some of my little make jobs I want to be able to compile Yorick on two targets. On one target it is built from scratch so Yorick lives in one directory but on my Linux box it lives under /usr/bin and /usr/lib. Really the /usr/bin/yorick binary just points to /usr/lib/yorick/bin/yorick as Yorick tends to live under the one installation directory.
Usually to generate a plugin Makefile you'd just type...
yorick -batch make.i
... and get an auto generated Makefile with variables pointing to your Yorick installation written into the Makefile for you. For example, on my Ubuntu atp-gett'ed installation I get the following.
# these values filled in by "yorick -batch make.i" Y_MAKEDIR=/usr/lib/yorick Y_EXE=/usr/lib/yorick/bin/yorick Y_EXE_PKGS= Y_EXE_HOME=/usr/lib/yorick Y_EXE_SITE=/usr/lib/yorick Y_HOME_PKG=
One way to get the Yorick installation directory for a system that is independent of the particular Yorick installation is as follows.
YORICK_HOME="$(echo "print, Y_HOME" | yorick | grep '^".*yorick' | sed -e 's/"//g' | tr -d '\r' | sed -r 's/\/ *$//g')" echo $YORICK_HOME
You can then substitute YORICK_HOME into the auto-generated Makefile so that it should work regardless who has the plugin and where their particular Yorick is installed (it might not be /usr/lib/yorick if they have installed in manually, for example).
Yorick Arrays
Major Differences VS C Arrays
- Yorick arrays start at index 1 but C arrays start at index 0.
- Yorick arrays are column-major whereas C arrays are row-major.
- This means that array[1][2] in C addresses row 1, column 2 (remember because C indices start at 0 this means the second row, third column). I.e., the C-syntax is
array[ROW][COL]
- But in Yorick, array(1,2), addresses column 1, row 2 (in C column 0, row 1 taking into account the start index of 1 instead of 0. I.e, the Yorick-syntax is
array(COL, ROW).
Yorick arrays are column major and indices start at 1.
What Is Row/Column Major?
For all languages, arrays are are contiguous blocks of items. Therefore, in memory we can store a 1D arrays as follows:
| ITEM[0] | ITEM[1] | ITEM[2] | ... | ITEM[N] | | | | | 0 *sz | | | | 1 *sz | | | 2 *sz | | 3 * sz : | N * sz
Here we store the first array element at the lowest memory address and increment upwards. You can see that each element is contiguous in memory. It is worth noting that the array is contiguous in terms of the language's model of the memory. For systems with an MMU this means virtual memory (which if the array crosses a page boundary it is only contiguous in virtual memory, not necessarily physical memory)
What happens when we try to store a 2D array? Consider the following array:
int someArray[2][3] = { { 1, 2, 3}, { 4, 5, 6} }
This is something we're very used to in C. We read this as an array with 2 rows and 3 columns. But how is this laid out in memory? It still has to be a contiguous chunk of memory but now there are two layout options. The C way is to lay out the array as follows:
| ITEM[0][0] | ITEM[0][1] | ITEM[0][2] | ITEM[1][0] | ITEM[1][1] | ... | | | | ^^^ | 0 *sz | | | Note: new | 1 *sz | | row starts | 2 *sz | | 3* sz | | 4 * sz
The array items are represented as a contiguous flow of values. The first row is laid out as it was for the 1D array, then the second row is appended to this block as if it were a 1D array and so on. Each row, therefore, starts at address row_index*3.
The C syntax specifies the fastest increasing index in memory from right to left. What does "fastest" mean? I'm using Munro speak here, and what he means is that, if we looked at the above memory layout, as we move one position from left to right the first dimension to increase is the "fastest", so in the C sense this is column.
Another way to think about this is considering an int p* = someArray pointer. As we increment this pointer by one (p++) it will traverse the columns in a row, which is what Munro calls the fastest increasing dimension, and then once all the columns in one row are iterated over it will move to the columns in the next row (so therefore the row is the second fastest increasing dimension) and so on.
We can see this in the above definition, int [2][3], where the right most dimension is [3] and represents the fastest increasing dimension in memory. The next dimension to the left is [2] and is the second fastest increasing dimension: rows.
This method of thinking about an array can be used for any number of dimensions. Consider int anotherArray[2][3][4][5]. We know that the fastest increasing dimension is 5, so if int *p = anotherArray then...
- *p == anotherArray[0][0][0][0]
- *(p+1) == anotherArray[0][0][0][1]
- ...
- *(p+4) == anotherArray[0][0][0][4]
We also know that:
- *(p+a*5) == anotherArray[0][0][a][0]
- *(p+b*4*5) == anotherArray[0][b][0][0]
- *(p+c*3*4*5) == anotherArray[c][0][0][0]
So when we say row-major, we mean that, in terms of the syntax, the right most index is the fastest incrementing dimension and going left the speed of increment decreases.
Column-major therefore means, that, again in terms of the syntax, the right most index is the slowest incrementing dimensions and going left the speed of increment increases.
I keep saying "in terms of syntax" because as Yorick is implemented in C its physical array storage is C-like. It is only in the scripting language itself that column-major notation is used.
Row-major means that, syntactically, the right most index is the
fastest incrementing dimension and going left the speed of increment decreases.
Column-major means that, syntactically, the right most index is the slowest incrementing dimensions and going left the speed of increment increases.
Adding More Dimensions: Row/Column Major Syntax in C VS Yorick
So, now moving onto ad 3D example... I'd normally visualise a 2x2x2 array as is shown to the right.The 2x2x2 array is an array of two 2x2 arrays. Each 2x2 array is what one would normally expect: a set of rows and a set of columns. So, the front 2x2 block has 2 rows and 2 columns, as does the back 2x2 block. The new dimension is the depth (or what I call "depth" anyway), the third dimension.
Given the description so far I would intuitively expect C to access
the array using this syntax:
array[DEPTH][ROW][COL]. I would
expect Yorick to access the array using this sytax:
array(COL, ROW, DEPTH).
In C:
array[DEPTH][ROW][COL]
To be double sure, look at the noddy little program below:
#include
int main(int argc, char *argv[]) { int array[2][2][2] = { { { 1, 2}, /* Depth, 0, row 0 */ { 3, 4} /* Depth, 0, row 1 */ }, { { 5, 6}, /* Depth, 1, row 0 */ { 7, 8} /* Depth, 0, row 1 */ } }; int r,c,d; for(d = 0; d < 2; ++d) for(r = 0; r < 2; ++r) for(c = 0; c < 2; ++c) printf("[d%i][r%i][c%i] == %i\n", d, r, c, array[d][r][c]); return 0; }
It outputs the following...
[d0][r0][c0] == 1 [d0][r0][c1] == 2 [d0][r1][c0] == 3 [d0][r1][c1] == 4 [d1][r0][c0] == 5 [d1][r0][c1] == 6 [d1][r1][c0] == 7 [d1][r1][c1] == 8
So, in C, one have to visualise the inner most brace as being your slowest-incrementing dimension.
In Yorick:
array(COL, ROW, DEPTH)
So, we can test how Yorick dimensions work. I would assume, given the
Yorick syntax, Yorick would represent the array above using the
syntax
array(COL, ROW, DEPTH).
And again, to be sure, we make a little noddy program:
myarray = [ [ [ 1, 2], /* Depth, 0, row 0 */ [ 3, 4] /* Depth, 0, row 1 */ ], [ [ 5, 6], /* Depth, 1, row 0 */ [ 7, 8] /* Depth, 0, row 1 */ ] ]; for (d=1; d<3; ++d) for (r=1; r<3; ++r) for (c=1; c<3; ++c) print, swrite(format="[d%i][r%i][c%i] == %i", d-1, r-1, c-1, myarray(c, r, d))
And, this outputs exactly the same as our little C program (the
d/r/c-1 in the
swrite() function is so that
we print out the indexing from zero, as the C program would have done).
This shows that our intuitive understanding of the syntax works for
adding extra dimensions to arrays in Yorick is correct...
Array Indexing
The Yorick docs on this pretty much explain everything but the Yorick equivalent of "fancy" indexing on arrays with 2 or more dimensions puzzled me for a second... had to think it through so here's a graphic :)
Lets say we define the following array
a = [ [1,2,3], [4,5,6] ]
This is an array with 3 columns and 2 rows, which we can see by using dimsof():
> dimsof(a) [2,3,2] ^ ^ ^ ^ ^ # rows ^ ^ ^ # columns ^ # dimensions
The array can be indexed as if it were a flat object:
> a(1) 1 > a(2) 2 > a(3) 3 > a(4) 4 > a(5) 5 > a(6) 6
And you can slice it as you might expect:
> a(2:3, 1:2) [[2,3],[5,6]]
The "fancy" indexing took me a second though:
> a([2,1],[1,2]) [[2,1],[5,4]]
The Yorick docs says
dimensions from the dimensions of the index list; values from the array
being indexed... which is very true, but I still needed a pretty picture :)
The dimensions used in the fancy index will be the dimensions of the resulting array. So in the above example the array in the column index has 2 elements to the resulting array will have 2 columns. The array in the row index as 2 elements so the resulting array will have 2 rows. Therefore we know the result is a 2-by-2 array.
Now for the values. TODO
Slices Are Views Onto Arrays But Beware Of Array Copy
In Python's NumPy, slices are like views into the array, and this is also the case in Yorick. BUT... if we transliterate the NumPy example into Yorick script, we will see a different result! See below:
> a = [1,2,3,4,5,6] > b = a(2:4) > b [2,3,4] > b(:) = 111 > b [111,111,111] > a /*<-- LOOK: Unlike the NumPy example, a has not been affected! */ [1,2,3,4,5,6]
In the Python NumPy example a would have been effected. Now, it's not quite that the slice isn't a view. It still is. Observe the following.
> a(2:4) = 111 > a [1,111,111,111,5,6] /*< Aha! the slice is a view into the array */
So, the slice is in fact a view into the array. The caveat is that in Yorick when we did b = a(...whatever...), b will be a copy of and not a reference to the array (slice). This can be quite an expensive operation so beware!
In Python's NumPy assigning one array (or non primative) variable to another copies a reference and not the value. In Yorick however, in the specific case of array = array, the entire array is copied... Beware of this as for large arrays this can get expensive!
The Yorick manual entry for
eq_nocopy() says that
having multiple variables
reference the same data can be confusing, which is why the default
"=" operation copies the array.
To copy an an array by reference use
eq_nocopy(). Note, however, that you can only do this for the
entire array, not slices. Munro
explains this here:
...Unlike NumPy, Yorick does not have any way to refer to
a slice of an object. All slicing operations happen immediately and result in a
temporary array.... Munro discusses it further in the
All slicing operations happen immediately and result in a temporary array (i.e., a copy-by-value).
When Arrays Are Not Copied By Value
Arrays Passed To Functions By Reference
Arrays are passed to functions by reference in the sense that if you modify the array in the function, the caller's array is modified...
> y = [1,2,3,4] > func cheeseit(x) { x(1) = 9999 /*<-- NOTE: Will change array in caller's scope */ } > cheeseit(y) > y /*<-- NOTE: y has been changed by function */ [9999,2,3,4]
For a pass-by-copy kind of semantic, do the following, but note copying an array will be expensive if this is a large array!
> y = [1,2,3,4] > func cheeseit(x) { x_local = x; /*<-- x is copied to x_local, but copy is EXPENSIVE! */ x_local(1) = 9999; /*<-- Changes to x_local will NOT affect array in caller's scope */ } > cheeseit(y) > y /*<-- NOTE: y is NOT changed by the function */ [1,2,3,4]
OXY Objects Store References
Above we just saw that when copying an entire array, or a slice of an array, the copy is done by value and not by reference. There is a caveat however... when assigning an array belonging to an OXY object to another OXY object, a reference to the array is copied. The array is not copied by value in this instance!
> a = save(q=[1,2,3,4]) > b = save(q=a.q) /* b's copy of a.q is a reference to a.q. unlike bare > a.q * array-to-array copy, which is by value, this copy is [1,2,3,4] * by reference! */ > b.q [1,2,3,4] > a.q(1)=999 > a.q [999,2,3,4] > b.q [999,2,3,4] /* LOOK! b.q must be a reference to a.q! */
This is not just true of copying OXY members between objects. If a.q is replaced by a vanilla array (just normal variable and not an OXY object), the result is the same! Note, however, that assigning from an OXY object array member to a normal variable is a copy-by-value.
Broadcasting
Yorick, being a nicely vectorised language, lets you do mathematical operations between scalars and arrays, and arrays and arrays. The only condition is that the arrays are what is called "conformable".Two operands are conformable if the dimensions which they share in common have the same length. This means that if we have two operands, A and B, where A has the smaller rank, that dimsof(A)(2:) == dimsof(B)(-dimsof(A)(1)+1:). Eek! That's a little horrid right?! It just expresses that the dimensions shared in common have the same length.
Try thinking of it this way: if you can create the array B by joining together a load of A arrays then they will be conformable because the dimensions they share in common must be of equal length as B is just made up of many A's. The image below is meant to make that a little more clear...
The image is trying to show that, as the manual says, the shorter operand repeats its values for every dimension that is does not have. This is what is called broadcasting.
As a little side note, it appears that Yorick actually created the term broadcasting, to the extent that it inspired Python's NumPy broadcasting... wowzers!
We can see that because, for example, the 2D array is made up of two 1D arrays, the 1D array can be multiplied, added, etc with the 2D array by "expanding" into a 2D array by repeating itself in the second dimension so that it has the same shape. The same goes for 1D op 3D or 2D op 3D in the above example.
Let's see this in practice...
Copyright (c) 2005. The Regents of the University of California. All rights reserved. Yorick 2.2.01 ready. For help type 'help' > a = [1,2] > b = [a,a] > c = [b,b,b] > a [1,2] > b [[1,2],[1,2]] > c [[[1,2],[1,2]],[[1,2],[1,2]],[[1,2],[1,2]]] > > a+b [[2,4],[2,4]] > > a+c [[[2,4],[2,4]],[[2,4],[2,4]],[[2,4],[2,4]]] > c+a [[[2,4],[2,4]],[[2,4],[2,4]],[[2,4],[2,4]]] > > b+c [[[2,4],[2,4]],[[2,4],[2,4]],[[2,4],[2,4]]] > c+b [[[2,4],[2,4]],[[2,4],[2,4]],[[2,4],[2,4]]]
As per our little image above... the arrays are conformable and so can be operands to the normal mathematical operations. BUT... the above does not give the whole story. "In common" can be more flexible. For example, we can do the following...
> aa = [1] > > dimsof(aa) [1,1] > dimsof(b) [2,2,2] > > aa + b [[2,3],[2,3]] > aa + c [[[2,3],[2,3]],[[2,3],[2,3]],[[2,3],[2,3]]] > bb = [[1], [1]] > bb+c [[[2,3],[2,3]],[[2,3],[2,3]],[[2,3],[2,3]]]
Now, given the above definition, one might thing that the common dimension between aa and b does not have the same length, yet they are clearly conformable as far as the Yorick interpreter is concerned. What's going on?
The reason that this works is that
Yorick will broadcast any
unit-length dimension in addition to a missing final dimension.
Therefore, pictorially, we can see the following...
So far we seem to have the following definition...
Yorick operands are conformable if the dimensions they share in common have exactly the same length or the shared dimensions in the "smaller" operand that are not the same length as in the other are unit length.
One does have to take a little care in what is understood by "dimensions in common", however. This might appear obvious to you, but I had to do a double take...
> a = [1,2] > dimsof(a) [1,2] > b = [[7,8,9,0],[6,5,4,3]] > dimsof(b) [2,4,2] > a+b ERROR (*main*) operands not conformable in binary + WARNING source code unavailable (try dbdis function)
If we look at the dimensions with the following formatting we might be fooled into thinking that a and b share a common dimension of exactly the same length, and therefore should be conformable.
/* Caution: This is an INcorrect grouping of the dimsof result */ +++ dim_a = [ 1, |2| ] dim_b = [ 2, 4, |2| ] +++
Remember that because Yorick is column major the dimsof array has the format [2, cols, rows]! Writing out the dimensions as I have above is quite misleading, and the mistake I at first made, because a has 2 columns and b has 2 rows and 4 columns. They are therefore NOT conformable. Remember to group your dimensions correctly:
/* This is the right way! */ +++ dim_a = [ 1, |2| ] dim_b = [ 2, |4|, 2 ] +++
Pseudo Index
The Yorick manual example starts with the outer product of two vectors. The following is taken from the linked-to Wikipedia article...
Assume that there are two column vectors $u$ and $v$.
$u = \begin{pmatrix} u1 \\ u2 \\ u3 \\ u4 \end{pmatrix}$ and $v = \begin{pmatrix} v1 \\ v2 \\ v3 \end{pmatrix}$
The outer product of the two column vectors is defined as follows...
$u \otimes v = uv^T = $ $\begin{pmatrix} u_1 \\ u_2 \\ u_3 \\ u_4 \\ \end{pmatrix}$ $\begin{pmatrix} v_1 & v_2 & v_3 \\ \end{pmatrix} = $ $\begin{pmatrix} u_1v_1 & u_1v_2 & u_1v_3 \\ u_2v_1 & u_2v_2 & u_2v_3 \\ u_3v_1 & u_3v_2 & u_3v_3 \\ u_4v_1 & u_4v_2 & u_4v_3 \\ \end{pmatrix}$
The matrix multiplication of those two vectors puzzled me for a little bit (I've only ever been used to multiplying matrices with dimension sizes greater than 1). It may be easier to think of it as follows...
$\begin{pmatrix} u_1 \\ u_2 \\ u_3 \\ u_4 \\ \end{pmatrix}$ $\begin{pmatrix} v_1 & v_2 & v_3 \\ \end{pmatrix} \equiv $ $\begin{pmatrix} u_1 & 0\\ u_2 & 0\\ u_3 & 0\\ u_4 & 0\\ \end{pmatrix}$ $\begin{pmatrix} v_1 & v_2 & v_3 \\ 0 & 0 & 0 \end{pmatrix}$
How do we do this using Yorick? All one dimension vectors in Yorick are column vectors (even if they are displayed textually as-if they are a row!). So can we just transpose() the $v$ vector?
> a * transpose(b) ERROR (*main*) operands not conformable in binary * WARNING source code unavailable (try dbdis function) > transpose(b) [100,200,300] > dimsof(transpose(b)) [1,3]
Apparently not! transpose(b) didn't do what it might have... I thought it would turn the column vector into a row vector, but it doesn't. The reason for this is that a row vector is actually a 2D array, i.e., it has an extra dimension.
And this makes sense... as we saw in our section on adding extra
dimensions to arrays and how to syntactically express this:
we know a 1D vector is
array(col) and a 2D vector is
array(col, row). Thus, is makes sense that a row vector
is a 2D array with one column!
So, to convert $v$ into a row vector we need to add an extra dimension...
To convert $v$ into a row vector we need to make each column element it's own row in a 2D vector. If $v$ has 3 columns, $v^T$ has 3 rows, where each row has only one element. To do this in Yorick, as said, we need an extra dimension... enter the pseudo index...
> b [100,200,300] > b(-,) [[100],[200],[300]] > b(,-) [[100,200,300]]
The syntax (-,), really the symbol "-", is the pseudo index. "-" means "add an extra dimension here".
Munro calls
the - sign, when used as an index, a
pseudo-index because it actually inserts an additional dimension into
the result[ing] array which was not present in the array being indexed.
And, as we noted above, to go from a column vector to a row vector (i.e, transpose
the column vector) we need to increase the dimensionality of our array.
So, based on this, we can say that b(-,) is saying add a row dimension to the column vector, or more generally add another most-slowly-increasing dimension to the existing vector. In the example above we get a 2D vector with 3 rows and 1 column.
How the extra dimension is added is both interesting and important. If b(-,) is saying add a row dimension, why does b(,-) == [[100],[200],[300]] and not [[100,200,300]]? They both have a row dimension added!
I think of it in the following way. Remembering that Yorick is column major, anything to the left of the "," in (,) indicates a faster dimension and anything to the right a slower dimension. Therefore, (-,) is saying add a faster dimension and (,-) is saying add a slower dimension... and this kinda works...
This generalises... suppose we can c a two dimensional array and we write c(-,).
> c = [[1,2,3],[4,5,6]] > c [[1,2,3],[4,5,6]] > c(-,) [[[1],[2],[3]],[[4],[5],[6]]] /* * c hand-formatted is... * [ * [ 1, 2, 3], * [ 4, 5, 6] * ] | | | * | | | * c(-,) hand-formatted is... * [ | | | * [[1],[2],[3]], * [[4],[5],[6]] * ] */
To add the extra dimension, each column element becomes a row with one column, and therefore the rows become depths. More generally it is like we shift the dimensions up one... The fastest increasing dimension is changed so that each individual element inside it is now the fastest increasing dimension and all the containing dimensions become one slower! Ie., we've added a new fastest dimension.
Now we can look at b(,-) . Given the discussion so far it looks like we are saying add a column dimension. We have seen that b(,-) == [[100,200,300]]. And this also makes sense... we've added a new slowest dimension.
So back to the original example of outer product. To accomplish this in Yorick we need to do transpose(u*v(-,)) I.e., the transpose of v is v(-,):
> u = [1,2,3,4] > v = [1, 10, 100] > v(-,) [[1],[10],[100]] > u * v(-,) /* This won't be quite right! */ [ [ 1, 2, 3, 4], [ 10, 20, 30, 40], [100, 200, 300, 400] ] > transpose(u*v(-,)) /* YAY, now we're good! */ [ [1, 10, 100], [2, 20, 200], [3, 30, 300], [4, 40, 400] ]
Matrix Multiplication & Inner Product
Remember to keep in mind that Yorick is column-major so the indices are a(col, row)!
The concept of matrix multiplication, the type we learnt in school, is shown in the little animation below. For each row vector of the LHS matrix we take the dot product of that vector with each column vector from the RHS matrix to produce the result:
In Yorick the '+' sign labels an array index for use in the dot product (a.k.a. inner product or scalar product) and can be used to calculate vector inner products or matrix multiplications. Munro has generalised the "school-learnt" matrix multiplication concept so that we are not constrained as to the choice of row/column vectors we take from the LHS and RHS. For example, we might want to take the dot product of column vectors from the LHS and column vectors from the RHS.
For two 1D arrays, a and b, the operation is fairly straight forward. Writing a(+) and b(+) marks the only dimension in each array as being used for the inner product. Therefore, the calculation becomes $\Sigma a(i)b(i)$. Let's take a look at this pictorially, otherwise the 2D example melts my brain a little...
Now we move on to the more complex examples... as always, remember to keep in mind that Yorick is column-major, so the indices are M(row, col).
LHS column vectors and RHS row vectors
I feel like the above needs a little explanation before continuing with the others.
Munro describes the plus sybol as "marking an index for use in an inner product". But what exactly do we mean? It means that the index used is the one that we iterate over by summing the multiplication of each element with it's respective "partner" in the dot product. Thus a(, +) means iterate over the row indices in the dot product summation. If we are iterating over the row index, then the column index must remain constant, so we will be taking column vectors.
To understand what the + sign means we could think of it as meaning "selecting the dimension to be iterated over in the summation of the inner product", because what the above does is to select all dimensions not marked by a + sign, and then iterate over the marked dimension marked for the inner product.
So, if the + sign marks rows, we are NOT selecting row vectors and then using the row vector for the inner product. We ARE selecting column
vectors and using the rows from each column vector in the inner product. The same is also happening with respect
to
b(+,): we are selecting rows and iterating over the columns.
I.e., for the + sign we select the vectors in the other dimensions and iterate over the marked dimension in those selected vectors.
Munro describes the plus sybol as "marking an index for use in an inner product". I prefer
to think of it as "selecting the dimension to be iterated over in the summation of the inner product".
I.e., for the + sign we select the vectors in the other dimensions and iterate over the marked dimension in those selected vectors.
So, now we can continue looking at the other possible ways to multiply out these square matrices...
LHS column vectors and RHS column vectors
LHS row vectors and RHS row vectors
LHS row vectors and RHS column vectors
We can see from the above that the index marked with the + sign is used for the inner product. Note that all these combinations are only possible because the matrices are square (for rectangular matrices, normal matrix multiplication shapes apply).
For example, in the pictures above, a(,+) marks the row index as being used for the inner product. I.e by selecting all rows in each column we create the vectors being used on the LHS of the dot product. Therefore we select a(1,:), a(2,:) all the way through to a(n,:).
Therefore, the indices marked with the + sign must have the same length.
Indicies marked with the + sign must have the same length.
Lets say dimsof(a) = (2, n, m) and dimsof(b) = (2, m, l). The process for c = a(,+)*b(+,) becomes:
for i in 1..n v1 = a(i, :) // 1d vector, length m for j in 1..l v2 = b(:, j) // 1d vector, length m r(i,j) = v1(+) * v2(+)
Normal or "School-Learnt" Matrix Multiplication vs Yorick
The only matrix multiplication I was ever taught was "normal" matrix multiplication:
$$\begin{pmatrix} a(1,1) & a(2,1)\\ a(1,2) & a(2,2)\\ \end{pmatrix} \begin{pmatrix} b(1,1) & b(2,1)\\ b(1,2) & b(2,2)\\ \end{pmatrix} = $$ $$\begin{pmatrix} a(1,1)b(1,1) + a(2,1)b(1,2) & a(1,1)b(2,1) + a(2,1)b(2,2)\\ a(1,2)b(1,1) + a(2,2)b(1,2) & a(1,2)b(2,1) + a(2,2)b(2,2)\\ \end{pmatrix} = $$ $$\begin{pmatrix} a(:,1)b(1,:) & a(:,1)b(2,:)\\ a(:,2)b(1,:) & a(:,2)b(2,:)\\ \end{pmatrix}$$
What you might notice is that the way the matrix multiplication works in Yorick, as described above, the index of the RHS changes most quickly. In "normal" matrix multiplication, it is the LHS index that is changing most quickly.
The figure below compares the closest operation we've seen so far with what I'd think of as a "normal" matrix multiplication...
Visually we can see that transpose(a(+,)*b(,+)) is equivalent to the "normal" matrix multiplication.
Another way to do this is to note that... $$\begin{pmatrix} a(:,1)b(1,:) & a(:,1)b(2,:)\\ a(:,2)b(1,:) & a(:,2)b(2,:)\\ \end{pmatrix} \equiv \begin{pmatrix} b(1,:)a(:,1) & b(2,:)a(:,1)\\ b(1,:)a(:,2) & b(2,:)a(:,2)\\ \end{pmatrix}$$
Therefore, to accomplish this "normal" matrix multiplication between a and b, in Yorick we would write b(,+) * a(+,) to do the "normal" matrix multiplication AB.
transpose(a(+,)*b(,+)) is equivalent to the "normal" matrix multiplication.
As we've noted, all these combinations above are only possible because the matrices are square. Remember that the indices marked for inner product must share the same length. If, for example a has 3 rows and 2 columns and b had 8 rows and 2 columns, the only possible operation would be a(+,)*b(,+).
An Example With SVsolve
The function SVsolve can be used to solve sets of simultaneous linear equations. Let's say that we have two arrays. Array a will represent the coefficients of the unknowns in our set of equations. Lets say that that we have n unknowns and m equations. The array a is an n x m matrix. Our linear equations, as we would normally write them (i.e., as I was taught in school) therefore look like this:
So if I were to pass the coefficient matrix (2D array) and the result vector (1D array) to SVsolve everything will work fine right? Err... no! Doh! Why not? It comes back to the issue above... the matrix multiplication is (at least for me) slightly unintuitive (probably just me!).
SVsolve() solves for A(,+)*x(+) = B. As we've seen, A(,+) will multiply the columns of A with x(+). Oops: SVsolve() is trying to do this:
So, the solution is to transpose() the array you are using before passing it to SVsolve()!
Rubber Indicies In Yorick
Yorick has one other indexing syntax which has proven useful, which I
call rubber indices. They address the problem of writing
interpreted code which extracts slices of arrays when you don't know
beforehand how many dimensions the array has. An example is an
opacity array for which you know that the most slowly varying index
represents photon wavelength, but there might be zero, one, two, or
three faster varying dimensions representing position in space. --
PYTHON MATRIX-SIG
> a = [ [ [1,2,3], [4,5,6] ],[ [11, 22, 33], [44,55,66] ] ] > a [[[1,2,3],[4,5,6]],[[11,22,33],[44,55,66]]] > a(1) 1 > a(1,1,1) ## Give me the first element in the inner array, from 1st array 1 ## in the middle array, from the first array (of arrays) from the ## outer array > a(1,:) ## Give me the first element in each inner array, contained in the [1,4] ## same parent array. This is a little incorrect because we don’t ## specify all array dimensions in the indices > a(1,:,:) ## Give me the first element in each inner array, from each [[1,4],[11,44]] ## outer array, contained in the overall array > a(1,..) ## The RUBBER INDEX says the same thing: Give me the first [[1,4],[11,44]] ## element from the innermost array and do the same recursing ## outwards > a(1,*) ## Another RUBBER INDEX that collapses all arrays into one. [1,4,11,44] > a(:,1,1) ## Give me everything from the first array, i.e., the array [1,2,3] ## itself, from the element in the middle array from the first ## element in the outer array. > a(:,1,..) ## Give me everything from the first array, ie. the [[1,2,3],[11,22,33]] ## array itself, from the first element in the middle ## array from all elements in the outermost array.
Oxy Objects
Copying Oxy Objects
When you assigned an OXY object from one variable to another is is copied by reference. Thus changing an object member will be reflected in both variables as assignment is not copy for OXY objects.
> a = save(q = 123) > a.q 123 > b = a > save, a, q=321 /*<-- Changing a's member variable will change b's! */ > a.q 321 > b.q 321 /*<-- LOOK, b has been affected by the change to a! */
Given the rational behind Yorick as a number cruncher, this makes a lot of sense. It's a little like passing an array into a function. If we think as an OXY object as a potential "bag" of large data, copying it would be hugely expensive, so that is why an object assignment is NOT a copy!
To actually make a real copy of the object you have to manually copy each member of one object to the other.
> a = save(q=101) > b = save() /*<-- Must manually re-create a new... */ > save, b, q=a.q /* ...object if we are to copy a */ > a object [1]: group > a.q 101 > b.q 101 > save, a, q=999 > a.q 999 > b.q 101 /*<-- Changing a.q has not affected b.q. We did a proper copy!
BUT, this is further complicated by arrays as member variables and OXY objects as member variables. For example, examine the following.
> a = save(q=[1,2,3]) > b = save() > save, b, q=a.q > b.q [1,2,3] > a.q [1,2,3] > a.q(1) = 111 > a.q [111,2,3] > b.q [111,2,3]
In the above example we used the same recipe as above, but because q is now an array the reference to the array is copied and not the value. Again, make sense for Yorick as a number cruncher of large data arrays, but if you wanted a copy, this could throw you. The same copy-of-reference problem would occur if q has been another OXY object.
You make have done a double-take here because previosly, when describing arrays, we said array assignment copies the array values, i.e. a complete copy of the array is made. This is clearly not the case when saving an OXY object member as an array. OXY object members appear always to be references unless the member stores a non-array primative.
OXY object members appear always to be references unless the member stores a non-array primative. Be cautious, when setting an OXY object member to equal an array: remember, unlike vanilla array to array copy, the OXY object member stores a reference to the asignee!
If you want to copy an array "into" an OXY object (actually create a nameless array and assign the reference to the OXY object member), use array slicing and take, if you want the entire array, a slice of the entire array:
> a = [1,2,3] > b = save(a=a(:)) # By taking a slice of the entire array we force a copy. > b.a # Slicing operations happen immediately and result in a [1,2,3] # temporary array, the reference to which is then stored > a # in the OXY object member. [1,2,3] > a(1)=99 > a [99,2,3] > b.a [1,2,3]
To fully, deep copy, an OXY object your therefore have to build a new object from scratch, and copy in all primative types and then recursively copy in all member OXY objects and also take care to copy the arrays correctly.
Yorick "Namespaces"
If you are writing a lot of Yorick scripts that get included in other scripts etc etc, you can end up with a lot of variable and function names in the global scope... this can lead to global namespace pollution.
You can create a kind-of namespace in Yorick. There is no "namespace" command but you do have access to save() and restore() which can be used to save a set of variables in an object and restore those variables from that object respectively. The general idea is this:
/* a and blah are already in the global namespace */ a = 1 func blah(void) { ... } /* Our script starts here... * Now we defined our own namespace that will be accessed through the * object my_new_namespace. The save() function saves the variable a and * function blah in this object. */ my_new_namespace = save(a, blah) //{ /* Here we overwrite the variable a and function blah in the global namespace * with these new values */ a = 101 func blah(does_something_else) { ... } //} /* We restore() the original namespace that existed before we modified it by * swapping the data from my_new_namespace.a and my_new_namespace.blah with * their the current values in the global namespace - the ones we just defined. * Thus the global namespace appears untouched by our script, and the variables * and functions in our script can be accessed using my_new_namespace.xxx. */ my_new_namespace = restore(my_new_namespace)
The are a couple of subtleties to watch out for here and they have to do with when statements are evaluated. For example, if you assign from one variable in your module's "namespace" to another variable in your module's "namespace", you would not use the namespace prefix my_new_namespace.xxx. For example:
//------------ // my_module.i //------------ my_new_namespace = save(a, b, blah, blah_oops) a = 222 b = a func blah(void) { print, my_new_namespace.a; } func blah_oops(void) { print, a; } my_new_namespace = restore(my_new_namespace) //------------ // main_program.i //------------ a = 111 b = 999 func blah(void) { print, a; } require, "my_module.i" print, "global a", a print, "global blah" blah, [] print, "my_module a", my_new_namespace.a print, "my_module blah" my_new_namespace, blah, [] print, "my_module blah_oops" my_new_namespace, blah_oops, []
The program will output the following:
"global a" 111 "global blah" 111 "my_module a" 222 "my_module blah" 222 "my_module blah_oops" 111
We can see that although my_module.i has it's own "global" symbols a, b, blah and blah_oops, when it is included from the main module, it does not effect that module's global scope. Instead to access these symbols from my_module.i the prefix my_new_namespace.xxx has to be used.
Note the difference between blah() and blah_oops() in my_module.i: when the function is called, it is evaluated after the save()/restore() commands, so to access the module's "globals" it now has to used the prefix my_new_namespace.xxx. The function blah_oops() represents an error! It does not access the module's variable, it accesses the variable from the global scope (if it exists - otherwise Yorick will error out!).
Also note the that my_module.i asigns a to b but in this case it is accessing my_module.i's symbol a. This is because the evaluation of this bit of code is immediate and occurs before the restore().
Watch out for these subtle little caveats!!
Writing Yorick Plug-Ins
Here I cover "manually" writing a plugin, i.e, not using Codger to auto-generate the C code using the PROTOTYPE Yorick comment method. Most of the stuff you need to know in order to do this can be found in yapi.h and the source code comments are pretty comprehensive.
One point to note about terminology. Whenever I talk about "stack" I will, unless made explicitly clear, be talking about the Yorick stack. This is a stack maintained by the Yorick interpreter and has nothing to do with your process/thread stack!
An Intro: Create The Plugin Shell And The Basics
Create The Shell
To create the Yorick makefile, there must be at least one .i file available. You will most likely also need the equivalent .c file. In this example I will call the plugin "jehtech", so I have created the files "jehtech.i" and "jehtech.c".
/* * FILE: jehtech.i */ plug_in, "jehtech" extern jehtech_Version; /* DOCUMENT void jehtech_Version() * * Return string describing version of the library */
/* * FILE: jehtech.c */ #include "yapi.h" #include "ydata.h" #ifndef NULL #define NULL '\0' #endif void Y_jehtech_Version(int argc) { ystring_t *verStr = NULL; /* ystring_t is char*, so this is char** */ /* Push a ystring_t* onto the stack. This is a char** */ verStr = ypush_q(0); if( !verStr ) y_error("Could not push return string onto stack"); /* p_strcpy is Yorick's mem-managed of strcpy(). Returns char* */ *verStr = (ystring_t)p_strcpy("v1.0"); }
Once these are created we can create the Makefile for this plugin. From the linux command line:
$ yorick -batch make.i created Makefile automatically generated make macros and dependencies: PKG_NAME=jehtech PKG_I=jehtech.i OBJS=jehtech.o edit Makefile by hand to provide PKG_DEPLIBS or other changes
In your working directory you will now have a ready made Makefile. Run it by typeing make all.
In the directory of compilation you will now have the object and library files jehtech.o and jehtech.so. There will also be two new files ywrap.c and ywrap.o, the object file having been added into the library.
Now, from this directory, fire up Yorick to test our little plugin...
$ rlwrap yorick Copyright (c) 2005. The Regents of the University of California. All rights reserved. Yorick 2.2.01 ready. For help type 'help' > #include "jehtech.i" > jehtech_Version() "v1.0"
The function jehtech_Version() has correctly returned the version string!
So, we kinda just dived into the deep end of the pool here so now to explain what we've done...
The example above shows us how to extend Yorick by binding a Yorick function name to a C function. There are two component files, the Yorick include file and the C file that implements the functionality. This C-implemented functionality will often call upon other libraries that you are in some way wrapping.
The Yorick include file so far has two lines of note:
- plug_in, "jehtech"
This line declares this script file as one that defines a plugin or library. When this file is included Yorick will go off and try to find the dynamic library jehtech.so and attempt to load it.
- extern jehtech_Version; Declares a symbol that will be added to the Yorick interpreters list of recognised symbols. The magic Yorick binding will associate this symbol with the C function Y_jehtech_version. Whenever a Yorick script calls the function jehtech_Version(), under the hood Y_jehtech_version() will be called.
Quick Debug Tip
If Yorick is complaining that it cannot find a library you can always try setting export LD_DEBUG=libs before running Yorick =.
Passing Values To C Functions From Yorick
Parameters Passed On Yorick Stack
Yorick maintains a parameter stack that it uses to pass values to an interface function (for example Y_jehtech_version()) and receive return values. Function parameters are pushed in-order onto the stack and whatever is at the top of the stack when the interface function exists, is it's return value.
So, if we have a Yorick function with parameters a, b, c, they will be passed to the C interface function on the Yorick stack. At stack position 0 will be c, at position 1, b, and at position 2, a. If the function wanted to return the multiple a*b*c it would put the result on the top of the stack, at position 0, to return this value to the Yorick interpreter.
All interface functions will have the same return value, void, and the same parameter list, consisting of one integer, usually called argc, which gives the number of parameters pushed onto the stack (note I mean the Yorick stack and *not* the C stack!).
void Y_functionName(int argc) { /* argc gives the number of elements pushed onto the **Yorick** stack. * Note this has nothing to do with the C stack! */ ... /* Anything left or placed on the top of the **Yorick** stack is the * function's return value */ }
Lets see this in action by creating a new function jehtech_ParameterOrder(a,b,c) and the associated interface function Y_jehtech_ParameterOrder():
/* * FILE: jehtech.i */ ... snip ... extern jehtech_ParameterOrder;
/* * FILE: jehtech.c */ ... snip ... void Y_jehtech_ParameterOrder(int argc) { long hundreds = ygets_l(0), tens = ygets_l(1), units = ygets_l(2); ypush_long(hundreds*100 + tens*10 + units); }
With the above modifications added and the plugin recompiled (just use make all, no need to regenerate the Makefile!) we can run the Yorick interpreter and see the results.
$ rlwrap yorick Copyright (c) 2005. The Regents of the University of California. All rights reserved. Yorick 2.2.01 ready. For help type 'help' > #include "jehtech.i" > jehtech_ParameterOrder(1,2,3) 321
The signature, written in C parlance, for jehtech_ParameterOrder() is long jehtech_Parameters(long units, long tens, long hundreds).
We can clearly see that the last parameter is at the top of the stack and the first parameter is at the bottom.
Function parameters, passed on the Yorick stack, are pushed on in parameter order. This means that the last parameter will be at the top of the stack and the first at the bottom.
As explained in yapi.h, the stack has space for at least 8 new elements when your plug-in's interface C function is called. It notes that if you are going to push more than 8 things onto the stack however, that you must reserve that space to avoid stack overflow.
If you push more than 8 items onto the Yorick stack you must reserve enough space using ypush_check() to avoid stack overflow!
Getting Scalar Parameters
The previous example showed how a function could retrieve 3 long parameters. The function ygets_l(stack_index) reads the long value at position stack_index in the stack. Note that it does not pop the value, it only peeks at the value. Position zero is the top of the stack.
All of the functions that peek at scalars on the stack are called ygets_X, where X is a single character representing the type of value, l for long, for example. Here is a list of the scalar-peaking functions. Each function has one intargument, the stack position to peak.
Getting Array Parameters
All yorick array dimensions are represented by the list [rank, ndim1, ndim2, ..., ndimY]. The first parameter of the list, rank, gives the number of dimensions that the array possesses. The size of each dimension is given by ndimX. For example, a 2D array with 10 columns and 90 rows will be described by the list [2, 10, 90]. The size of the array dimension list is limited in Yorick. It will be a maximum of Y_DIMSIZE long's. This means that Yorick arrays cannot have more than Y_DIMSIZE-1 dimensions.
The first thing to remember is that Yorick arrays are column major. Therefore a 2D array is described by the dimensions list [2, #columns, #rows]. Very easy to see this in action. We'll make the following additions to our project (note: generally you should not use printf like I'm doing here).
/* * FILE: jehtech.i */ ... extern jehtech_ArrayTest;
/* * FILE: jehtech.c */ ... #include <stdio.h> ... void Y_jehtech_ArrayTest(int argc) { long i; long dimInf[Y_DIMSIZE]; long ntot; long *ary = ygeta_l(0, &ntot, dimInf); printf("jehtech_2DArray: ntot == %ld\n", ntot); printf("jehtech_2DArray: dimInf = ["); for(i = 0; i <= dimInf[0]; ++i) { printf("%ld%s", dimInf[i], i == dimInf[0] ? "]" : ", "); } printf("\n"); }
Recompile the plugin and running Yorick, we can get the following:
$ make all ... $ rlwrap yorick Copyright (c) 2005. The Regents of the University of California. All rights reserved. Yorick 2.2.01 ready. For help type 'help' > #include "jehtech.i" > jehtech_ArrayTest, [1,2,3] jehtech_2DArray: ntot == 3 jehtech_2DArray: dimInf = [1, 3] > jehtech_ArrayTest, [[1,2,3], [1,2,3]] jehtech_2DArray: ntot == 6 jehtech_2DArray: dimInf = [2, 3, 2] > jehtech_ArrayTest, [[[1,2,3,4], [1,2,3,4]], [[1,2,3,4], [1,2,3,4]]] jehtech_2DArray: ntot == 16 jehtech_2DArray: dimInf = [3, 4, 2, 2]
Great so we can, from our C code, get an array of long's, or at least so far, it's dimensions so that we can determine what shape the array is. But what about accessing the array values? Yorick specifies everything column major, so in the array [[1,2,3,4], [5,6,7,8]] to access the third column, second row, we would write a(3,2). However, under the hood Yorick stores arrays C-style... i.e., from C, when passed an array, we would access it as a[2][3] or a[2*num_cols + 3].
Lets see this in action to see that this is in fact the case. We'll make the following additions to out project files...
/* * FILE: jehtech.i */ ... extern jehtech_Dump2D;
/* * FILE: jehtech.c */ ... void Y_jehtech_Dump2D(int argc) { /* PRE: array is 2D! */ long row, col; long dimInf[Y_DIMSIZE]; /* [rank, #cols, #rows] */ long *ary = ygeta_l(0, NULL, dimInf); for(row = 0; row < dimInf[2]; ++row) { /* dimInf[2] is #rows */ for(col = 0; col < dimInf[1]; ++col) { /* dimInf[1] is #cols */ printf("%ld\t", ary[row*dimInf[1] + col]); } printf("\n"); } }
Recompiling and running Yorick we then see the following...
$ rlwrap yorick Copyright (c) 2005. The Regents of the University of California. All rights reserved. Yorick 2.2.01 ready. For help type 'help' > require, "jehtech.i" > jehtech_Dump2D, [[1,2,3,4], [5,6,7,8]] 1 2 3 4 5 6 7 8
Run Yorick From Python
Cool... this looks promising as to test some Yorick I have been using Robot Framework so being able to do this would make tests easier! Haven't done anything with it yet but Munro has written a Python module interact with Yorick. | https://jehtech.com/yorick-programming-language.html | CC-MAIN-2021-25 | refinedweb | 8,679 | 63.19 |
JQuery :: Validate Multiple Checkbox In Each Row Of Gridview?Jan 31, 2011
i want to validate 3 checkbox in each row of gridview.
i want one or two checkbox be checked,but no 3 checkbox.
if 3checkbox is checked showe error msg.
i want to validate 3 checkbox in each row of gridview.
i want one or two checkbox be checked,but no 3 checkbox.
if 3checkbox is checked showe error msg.]....
I have an ASP.NET gridview where I allow the user to edit a record. I need to validate multiple fields as one. I need to require that the user either enter a first and last name OR a company name. All three cannot be blank. Most of the sample code I am finding does not address the text boxes only being visible while the gridview is in edit mode. When not in edit mode, the text boxes do not exist so
document.getElementById('<%= editFirstName.ClientID %>') throws an error upon page load..View 10 Replies
I have a Dynamic GridView on a page as below. The user can add or delete rows. When inserting record, the code is executed one row at a time. I have 3 rows to be inserted to the database. Here is the issue:
-if the first row is not completed, then no record is added to the database
-if the first row is completed and the second row is not, then only the first row is inserted in the database
-if the first row is completed, the second row is not complete, and the third row is completed, then only the first row is inserted in the database.
below is the code:
protected void GridView1_AddNewRecord(object sender, EventArgs e)
{
string constr = ConfigurationManager.ConnectionStrings["DatabaseVestConnectionString1"].ConnectionString;
SqlConnection con = new SqlConnection(constr);
SqlCommand cmd;
string query1 = "INSERT INTO kart_Bestilling(SakBehandlingCode, KystverketRegionID, KystverketAvdelingID, BestillerReferanse, BestillerLeveringAdresse, BestillerPostKodeBy, BestillerNavn, BestillerStilling, BestillerEpost, BestillerTelefon, BestillingBeskrivelse, BestillingDato, BehandlingSakNummer, BehandlingBeskrivelse, BehandlingStatus)" +
[code]...
How to arrange this piece of code to validate all rows before any insert? or avoid inserting a row if there is one row not complete?the conditions starts with int rowIndex = 0;
if (txtBestillerReferanse.Text != "" && txtBestillerEpost.Text != "" && txtBestillerNavn.Text != "" && txtBestillerPostKodeBy.Text != "" && ddKartTypeName.SelectedValue != "Velg KartType" && ddKartNummerName.SelectedValue != "Velg KartNummer" && tBestillingAntallKart1.Text != "")
I have a gridview which displays following columns:
TaskID, TaskDescription, IsComplete, DateOfCompletion
Now what I want to achieve is, when user clicks IsComplete checkbox, she must enter the date of completion. So, I want a validation (client side) on DateOfCompletion if the checkbox is selected (checked). And, I also want to use Validator Callout Extender
if possible.
How can I achieve this?
I had a gridview with checkbox to select records for generate report.
There are column "From Date", "To Date", "Applicant"...
How can I validate the selected row (checked checkbox) are on same date (From Date) ?
I have a nested gridview with 45 checkboxes. I want the user to be able to click a checkbox and have that value held in a separate gridview until they wish to act on it (similar to a shopping cart).
The checkboxes are nested in a 2nd level gridview behind a repeater.
<repeater>
<gridview>
<gridview>
checkbox
<gridview/>
<girdview />
<repeater />
I was having a heck of a time trying to get the value of the checkbox that deep and want to learn jQuery and thought this is a good time. What I was thinking was the user would click the checkbox, jQuery would get the id of the control (and value), I could then pass that to a ajax postback trigger and fill the 'shopping cart' gridview. The reason why I need to go through ajax is I need to get more values from a database based on the user selected checkbox. I think I could take it from there. My biggest problem right now is figuring out how to get the id and value from the checkbox.
I created a gridview with a checkbox in front of some columns. I need to grab the data the user is delecting and building an xml file.
Here is my code so far.
[code]....
On click of checkbox, I need to retrieve immediate parent span's class value:
The checkbox column is defined in an ItemTemplate as:
<asp:CheckBox
The JS function is defined as:
function CartItemCheckClicked() {
alert($(this).parent().attr('class')); //Undefined
//alert($(this).attr('id')); //Undefined [code]....
But the result is always 'undefined'. How do I access the checkbox or parent span?
I have a form with three Gridview's. ie. GridView1,GridView2,GridView3.Each gridview contains two templatefield and a boundfield.The first templatefield is a checkbox field and the second one is a labeltemplate with databinding to field "Usr_Id" and the third field is for Name.I have enabled paging for all the three girdview's and different data's are populated on form_load.how can i maintain the checkbox status while paging for all the three gridview's.ie, when i check checkboxes in two pages in gridview1 and checked three pages in gridview2 and so on,i should get the check state of all the gridview checkboxes maintained.I know to maintain the checkstate for a single gridview but when i am checking the pages in second gridview ,the checkbox values of some pages of gridview1 is lostView 5 Replies
i am stucked in Checkbox Filteration issue... My problem is If first checkbox list is checked and user clicks on second checkbox list then i want to compare both checkboxes and populate result based on both checkboxes. using System;
using System.Collections.Generic;
using System.Linq;
using System.Web;
using System.Web.UI;
using System.Web.UI.WebControls;
using System.Configuration;
using System.Data.SqlClient;
[code]...
I have a checkbox column in a Gridview that I would like to use to insert the value of one cell in the row into an access database. There will be a maximun of five cells allowed to be selected so there can be one value written to one field in the database up to the maximum of the five fields. The access database has five columns: Selection1 ,Selection2, Selection3 ,Selection4 and Selection5
When the checkbox is checked, the value of one cell (PliD) should be written to each field in the database. Ie: checkbox in row 1 should write PliD value to Selection1, checkbox in row 2 should write PliD value to Selection2, checkbox in row 3 should write PliD value to Selection3 and on depending on which checkbox has been checked. I am not sure how to get the value of the checkbox and write it into the corresponding fields in the database.
[Code]....
I have gridview that contains a check box, messageId and firstName. The gridview is getting information from the database and showing them in the gridview and in a checkbox. The gridview shows the fields that are selected and that are not selected. I want to send an email to all people that the check box is selected.
I have already created the email system and it is working. What I don't to know is how to get the information from the gridview and place in the mail.To.Add(?); so I can send the information to all people into the database.
---.aspx showing the gridview---
[Code]....
--CODE BEHIND IN C#---
[Code]....
I am trying to delete multiple rows using a checkbox in a Gridview. When someone checks the checkbox and then click on button Delete that the rows is chosen will be deleting.
GUI:
[Code]....
when i click on button Delete : Error: Index was out of range. Must be non-negative and less than the size of the collection. Parameter name: index
I have a GridView which is databound and I added two columns which are checkboxes. There is no link between the two columns. On a check box click , I need to check to see any other check boxes on the same column checked, if so, show a javascript confirmation . If click ok, need to do postback to update the database.if cancel, no change. The goal is ,I need to make sure that only one check box is checked in a check box column.View 2 Replies
how to download selected files from grid view. Initial I will display list of files with file paths in grid view along with check boxes. when user select some files, I should be able to download selected files in a zip file.View 1 Replies
I have 5 checkbox columns in my grid .. like
Id Chk1 Chk2 Chk3 Chk4 Chk5
I want select only one checkbox among 5 checkboxes if user selects one checkbox another which are checked are need to uncheck. How can i do this in client side .....
I want to send mail tot multiple user using checkbox within gridbiew , but problem is it is sending mail to only first column email ID , and suppose i have four email id in databse then it will send 4 times mail to first email id,
OleDbCommand cmd = new OleDbCommand("select * from sendmail",con);
OleDbDataReader red = cmd.ExecuteReader();
if (red.Read())
{
StringBuilder str = new StringBuilder();
for (int i = 0; i < GridView1.Rows.Count; i++)
{
GridViewRow row = GridView1.Rows[i];
bool isChecked = ((CheckBox)row.FindControl("chkSelect")).Checked;
if (isChecked)
{
str.Append(GridView1.Rows[i].Cells[3].Text);
SmtpClient sc = new SmtpClient();
MailMessage mm = new MailMessage("dnk247@gmail.com", red["nemail"].ToString());
mm.Subject = "hello";
mm.Body = "just say hi";
sc.EnableSsl = true;
sc.Send(mm);
}
}
}
I want to send gridview multiple selected values (using Checkbox) to another page gridview..on click of a button.View 1 Replies
I have a gridview with 5 columns , i am bind gridview on pageload .... each row of gridview has a checkbox aslo ..... and i have a hyperlink outside the gridview..
I want that when i check checkbox of particular row the 'Firstname' field of that row should display as link , i am using jquery to avoid postback .. but not able to get that field value as link.
here is my gridview code:
<asp:GridView
<Columns>
[Code]..... had problems with jQuery fancy checkbox plugin [URL]
However after that I want to "combine" 2 plugins, the one mentioned in first post:
[URL]
and this one ("Safari"):
[URL]
So I've changed the picture of checkbox and wanted to add hover effects as it is shown in second link (according to state of checkbox).
However since I'm quite noob with jQuery (and JS) the thing only works fine on "default" mode, when checkbox is not selected or checked on page load.
JS:
[Code]....
CSS:
[Code].... | http://asp.net.bigresource.com/JQuery-Validate-multiple-checkbox-in-each-row-of-gridview--USqVeq1r8.html | CC-MAIN-2018-39 | refinedweb | 1,761 | 65.22 |
Opened 10 years ago
Closed 10 years ago
#3519 closed (invalid)
IndexError when creating related objects in admin
Description
I have a model with
ForeignKey(Entry, edit_inline=models.TABULAR) field. When i try to edit the Entry object in admin and ask to create Topic object too, django throws and IndexError:
POST data: ... topic.0.id: '' ...
Attachments (2)
Change History (17)
Changed 10 years ago by
comment:1 Changed 10 years ago by
Could you craft and post a minimal
models.py file that reproduces the problem in order to check if this is a new one or something already reported?
comment:2 Changed 10 years ago by
Will mark accepted once someone produces the code that reproduces this error.
comment:3 Changed 10 years ago by
comment:4 Changed 10 years ago by
huh... always popping out on original machine, but can't reproduce it on new one with the same apps. Looks like misconfiguration.
comment:5 Changed 10 years ago by
steps to try:
failed to attach it directly, so test project is at data.cod.ru/569408970
syncdb, launch server and try to add a news entry. entry will create but entry's topic can fail here.
comment:6 Changed 10 years ago by
The URL you posted shows what seems to be an error message in russian.
Anyways, no need to attach the full project, just paste the
models.py here as a comment wraping it with {{{ }}} to preserve the python formatting.
comment:7 Changed 10 years ago by
Another bug in Admin interface. Again on submitting inline objects.
There are error page with submitted data and full error traceback and a screenshot just before submitting: media.twogre.aenor.ru/html/admin_bug/
The model page are here: twogre.googlecode.com/svn/trunk/blog/models.py
comment:8 Changed 10 years ago by
Posting model code here for convenience...
from django.db import models from django.contrib.auth.models import User class Entry(models.Model): title = models.CharField(maxlength=255) slug = models.CharField(maxlength=40, prepopulate_from=("title",)) published = models.DateTimeField(blank=True, null=True) author = models.ForeignKey(User) def get_uncut(self): return self.part_set.filter(cut=False) def __str__(self): return self.title def get_absolute_url(self): return '/blog/%d.%02d.%02d/%s/' % (self.published.year, self.published.month, self.published.day, self.slug) class Admin: pass class Meta: ordering = ['-published'] class Part(models.Model): entry = models.ForeignKey(Entry, edit_inline=models.TABULAR, num_in_admin=5) header = models.CharField(maxlength=255) order = models.SmallIntegerField(default=0) text = models.TextField(core=True) cut = models.BooleanField(default=False) def __str__(self): return self.header class Meta: ordering = ['order', 'id'] class Reply(models.Model): object = models.ForeignKey(Entry) author = models.ForeignKey(User) text = models.TextField() posted = models.DateTimeField(auto_now_add=True) def __str__(self): return '%s @ %s on "%s"' % (self.author.first_name, self.posted.strftime("%H:%M, %d %b %y"), self.object.title) class Admin: pass class Meta: ordering = ['posted']
comment:9 Changed 10 years ago by
I've gotten the same error using python2.5 but it works fine when using python2.4?!
Changed 10 years ago by
Patch to fix the problem
comment:10 follow-up: 12 Changed 10 years ago by
So, line 133 of django/db/models/manipulators.py reads:
expanded_data = DotExpandedDict(dict(new_data))
This works fine in python 2.4 (not sure why). But in python 2.5, the new_data is converted into a new dictionary by calling getitem, which for MultiValueDict returns the last value stored in the list rather than the list itself. In python2.4 the lists were preserved by this call. The subsequent code depended on having lists stored at each key value and used the [0] index to grab the first item.
I've just changed line 133 to read:
expanded_data = DotExpandedDict(dict([(k,new_data.getlist(k)) for k in new_data.keys()]))
which explicitly describes the conversion into a normal dict and therefore preserves the lists at each key which allows the subsequent code to work in the same way for python 2.5 and python 2.4.
I haven't had time to write unit tests for this yet, and I'm also not certain how to write a test for the admin interface. If someone can enlighten me I'd be more than happy to do it.
See patch.
comment:11 Changed 10 years ago by
Sorry, that previous comment was by me. (Forgot to add my email address)
comment:12 Changed 10 years ago by
It works now with the patch. Thank you!
comment:13 Changed 10 years ago by
This problem was caused by a regression in Python 2.5.1 pre-releases which has been fixed in 2.5.1-final. Anybody using 2.5.1c1 or anything like that should upgrade to 2.5.1-final or downgrade to 2.5.0-final. It's not a Django bug.
comment:14 Changed 10 years ago by
I still receive this error with Python 2.5.1-final on Windows XP SP2.
comment:15 Changed 10 years ago by
Nevermind. My mistake.
workaround | https://code.djangoproject.com/ticket/3519 | CC-MAIN-2017-17 | refinedweb | 844 | 61.22 |
Red Hat Bugzilla – Bug 530541
Free space check on /boot not thorough enough
Last modified: 2014-01-21 18:11:51 EST
Description of problem:
The free space check on /boot is not thorough enough, causing the user to get stuck with too little disk space on /boot to complete the upgrade.
Version-Release number of selected component (if applicable):
preupgrade-1.1.0-1.fc11.noarch
I try to preupgrade to (pre-)F12. After the preupgrade assistant finishes, I have:
/dev/sda1 190M 177M 3.7M 98% /boot
After a reboot, Anaconda tells me it can't finish the upgrade because there's not enough space on /boot. I would have expected preupgrade to fail instead of allowing the reboot to Anaconda.
I suppose it's hard to tell beforehand how much space the rpm transaction will need, but perhaps a hard-coded minimum size check would be in order, then. (About twice the size of the current kernel and initrd together?)
So from what I just looked at - we need for anaconda to put a kernel an initrd and a systemmap file in there. So files are roughly: 3M, 4M and 1.5M respectively.
So, maybe add 10M to the freespace check as a margin of error?
Will, does that sound reasonable?
Sure, that's fine, although we may need to doublecheck since dracut initrds can be larger than our usual initrds sometimes.
This was supposed to be handled by the rpm transaction test, but the rpm transaction test takes several minutes to run and is (in the testing we performed) no more accurate than our rough estimates.
*** Bug 534052 has been marked as a duplicate of this bug. ***
and when rpm runs it makes some tmp files which suck up some space, too and are not always added to the calculation - I've learned that one the hard way.
sigh - fun.
*** Bug 534055 has been marked as a duplicate of this bug. ***
Since changing the size of a previously installed F-11 /boot is not an option, Wwoods outlined several approaches to mitigate this problem for users upgrading to F-12 from F-11.
[Nov 10 15:06:51] < wwoods> | - write an anaconda patch to make /boot bigger and format it without reserved space
[Nov 10 15:07:23] < wwoods> | - add instructions for doing tune2fs and removing old kernels to a wiki page (common bugs etc)
[Nov 10 15:07:30] < wwoods> | - maybe link to that wiki page from the error dialog
[Nov 10 15:08:46] < wwoods> | - figure out why the heck RPM / mkinitrd / whatever need so much extra overhead on /boot
I'm going to add the CommonBugs keyword to this issue which will add it to our list of bugs to document for the Common_F12_Bugs wiki.
I produced this on ppc,when upgrade from f11->f12rc4,will output error of insufficient disk space ,need 8M space in /mnt/sysimage/boot/
I ran into this as well with a preupgrade, telling me I needed 1M more space. I deleted an old kernel and tried again and got the same error. The initrd file in /boot/upgrade was quite large. I ended up using yum upgrade to upgrade since I could not resize /boot since my other partitions where LVMs and I could not figure out how to resize and move them.
/boot is typically partitioned with 5% (10MB) reserved for the root user, which anaconda/RPM seems to ignore. So 'tune2fs -r 0 /dev/your-boot-partition' would have freed up 10MB and let the upgrade complete.
We may add a dialog to offer to do this for you, but I'm wary of twiddling the filesystem bits this way. You gotta be careful with this kind of thing.
(In reply to comment #6)
> Since changing the size of a previously installed F-11 /boot is not an option,
> Wwoods outlined several approaches to mitigate this problem for users upgrading
> to F-12 from F-11.
<>
> [Nov 10 15:08:46] < wwoods> | - figure out why the heck RPM / mkinitrd /
> whatever need so much extra overhead on /boot
Think what you need to have happen is that the /upgrade/install.img in /boot should be moved off of /boot and into /mnt/sysimage/tmp just before yum/rpm is called. That should free up the needed space. I've have a untested patch.. Just looking for feedback at this point... I'll whip up an updates.img for testing and post it here later if there is interest.
cat freeram1246.diff
diff -up ./backend.py.orig ./backend.py
--- ./backend.py.orig 2009-11-12 10:28:33.000000000 -0600
+++ ./backend.py 2009-11-12 12:46:09.000000000 -0600
@@ -154,28 +154,51 @@ class AnacondaBackend:
return
if self._loopbackFile and os.path.exists(self._loopbackFile):
+ log.info("and exists %s" % self._loopbackFile ) they booted with a boot.iso, just continue using that install.img.
+ if os.path.exists("/mnt/stage2/images/install.img"):
+ log.info("Don't need to transfer stage2 image")):
+
+ log.info("mediaDevice is %s" % anaconda.mediaDevice )
+ log.info("Using %s as stage2 image" % installimg)
- free = anaconda.id.storage.fsFreeSpace
self._loopbackFile = "%s%s/rhinstall-install.img" % (anaconda.rootPath,
- free[0][0])
+ "/tmp")
+ log.info("New image name %s" % self._loopbackFile )
+
+ if os.path.exists("/mnt/sysimage/boot/upgrade/install.img"):
+ log.info("OVERRIDE using stage2 image from /mnt/sysimage/boot/upgrade")
+ stage2img = "/mnt/sysimage/boot/upgrade/install.img"
+ stage2boot = 1
+
+ # +218,16 @@ class AnacondaBackend:
return 1
isys.lochangefd("/dev/loop0", self._loopbackFile)
- isys.umount("/mnt/stage2")
+
+ if stage2ram or stage2boot
+ try:
+ os.unlink(stage2img)
+ except:
+ pass
+ try:
+ isys.umount("/mnt/stage2")
+ except:
+ pass
def removeInstallImage(self):
if self._loopbackFile:
diff -up ./yuminstall.py.orig ./yuminstall.py
--- ./yuminstall.py.orig 2009-11-12 10:28:33.000000000 -0600
+++ ./yuminstall.py 2009-11-12 12:04:54.000000000 -0600
@@ -631,6 +63158,12 +861:
Not a great idea, since AFAIK anaconda is being run from *inside* install.img.
Moving the file that contains your root filesystem to a different device while the system is running might cause some problems. Not sure how/if the loop driver handles that.
(In reply to comment #11)
> Not a great idea, since AFAIK anaconda is being run from *inside* install.img.
>
isys.lochangefd("/dev/loop0", self._loopbackFile) anaconda does that now after copying install.img from a cdrom/dvd, so it does work.
> Moving the file that contains your root filesystem to a different device while
> the system is running might cause some problems. Not sure how/if the loop
> driver handles that.
That is what lochangefd is for, problem there is that install.img becomes
/mnt/sysimage/boot/rhinstall-install.img.
free = anaconda.id.storage.fsFreeSpace
self._loopbackFile = "%s%s/rhinstall-install.img" %(anaconda.rootPath,
free[0][0])
I'm just trying to move rhinstall-install.img from being placed in /boot by anaconda to /tmp and then freeing up the space in /boot if needed. As a bonus anaconda's /tmp wouldn't be holding the install.img if it was present, freeing up the ram before calling yum.
just my 2 cents
Created attachment 369306 [details]
patched updates.img
updates.img containing the above patched files. place in /boot/upgrade to be picked up automatically by anaconda.
Created attachment 369310 [details]
revised patch
a bit of a thinko... with preupgrade, the install method would be using hdinstall.c, install.img would already be in ram. revised patch
Created attachment 369311 [details]
revised updates.img
Created attachment 369316 [details]
typo in other patch
Created attachment 369317 [details]
updates.img with typo fixed
Hi. Suggestions on how to make preupgrade better able to handle a small /boot are welcome, but I think they should go in separate bugs. This is about how to make the calculations on free and required space more accurate or at least more conservative. Thanks. :)
Well if the 100+ meg of install.img is no longer in /boot when yum starts to install the rpms, this problem goes away, and the initial calculation of space needed by preupgrade is then valid IMHO :)
Sorry for being a bug nazi. I made bug 537243.
(In reply to comment #19)
> is then valid IMHO :)
Well... :)
preupgrade-1.1.3-1.fc12 has been submitted as an update for Fedora 12.
preupgrade-1.1.3-1.fc10 has been submitted as an update for Fedora 10.
preupgrade-1.1.3-1.fc11 has been submitted as an update for Fedora 11.
preupgrade-1.1.3-1.fc11 has been pushed to the Fedora 11 testing repository. If problems still persist, please make note of it in this bug report.
If you want to test the update, you can install it with
su -c 'yum --enablerepo=updates-testing update preupgrade'. You can provide feedback for this update here:
The Common_F12_Bugs page has been updated with guidance on this issue. See and. Feedback on the content encouraged.
*** Bug 538584 has been marked as a duplicate of this bug. ***
preupgrade-1.1.3-1.fc11 has been pushed to the Fedora 11 stable repository. If problems still persist, please make note of it in this bug report.
preupgrade-1.1.3-1.fc12 has been pushed to the Fedora 12 stable repository. If problems still persist, please make note of it in this bug report.
preupgrade-1.1.3-1.fc10 has been pushed to the Fedora 10 stable repository. If problems still persist, please make note of it in this bug report.
this bug has persisted after installing preupgrade-1.1.3-1.fc11.
I did a clean install for fc11 - so have the default /boot size.
I also ran into Bug 538118 and had to wipe my preupgrade cache files to start again. But can't see any traces in /boot of the earlier run that could be clogging it up.
Preupgrade warned me to clean out old kernels - which I did and then allowed me to finish the process. But after rebooting it came up with the error of not enough space in the /boot partition.
df -h /boot shows the following:
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 194M 158M 27M 86% /boot
(In reply to comment #32)
> df -h /boot shows the following:
> Filesystem Size Used Avail Use% Mounted on
> /dev/sda1 194M 158M 27M 86% /boot
Greetings David! The bugfix for this issue also came with several recommended manual steps to alleviate the disk space shortage in /boot. Please take a look at for additional suggestions on addressing the low disk space on /boot when using preupgrade. Thanks!
Thanks - I did find that resource and removing the reserve faced fixed the problem.
My point, though, preupgrade still allowed me to go through the whole process without warning me I had insufficient space. So it is not allowing enough margin for larger install requirements.
Btw, this was only the second of many problems with the install using preupgrade, that eventually led me to abandon it altogether and do a fresh install from media.
Hi!
I'm using the latest preupgrade for F11 (preupgrade-1.1.3-1.fc11.noarch), it gave me the warning I didn't have enough disk space, I cleaned it up until preupgrade was happy, and I started the installation. Anaconda then failed on me with not enough disk space, it claimed I needed to free up 0M. I'll remove some more stuff, but it seems the amount of free space required might need an extra buffer or something..
Hi, people!
Last week I tried to upgrade my two Fedora 10 servers to Fedora 12. And on the two servers I have problems with /boot free space.
The most bizarre problem was that message telling me to free up 0M. After follow the steps on , I finally could upgrade without problems.
I think the 'free up 0M' message is given because preupgrade (or anaconda) does not consider the root reserved space on this counts, assuming that it will have this space in case it will not.
The 'tune2fs' step solved the problem for me. Maybe it could be integrated to the preupgrade install scripts...
I was hit by this problem when (trying to) update from F12 to F13.
The /boot partition is the (F12 default) 194MB (which is not enough for the F13 install.img).
Preupgrade is version 1.1.7 - far never than version 1.1.3 that was supposed to fix the problem.
Prerequisite:
* preupgrade is not running.
Steps:
1) run preupgrade in a terminal
2) watch output in the window.
At some point there is a lot of chatter in the terminal window :
"urlgrabber" keeps casting IOErrors (no space on device) but preupgrade continues regardless.
The last thing in the log is "raise KeyboardInterrupt" - does that make sense?!
I had a 100MB filler file to force preupgrade to download "stage2" from the network. That didn't work - instead i was left with a (truncated) 60MB install.img in /boot/upgrade - and with no changes to grub.conf. (I guess because /boot is full).
Nevertheless - preupgrade proudly proclaims that it is ready to upgrade as soon as the "Reboot" button is pressed!
Result:
1) Rebooting just start F12 again.
2) Rebooting and setting "root/kernel/initrd" hangs (i assume that the truncated install.img is blindly loaded and executed).
3) removing install.img, and doing step 2) leaves you with a not very helpful installer. I can select that the install.img is on a "URL" but what does the input field expect?
I'm out of ideas :-( | https://bugzilla.redhat.com/show_bug.cgi?id=530541 | CC-MAIN-2016-26 | refinedweb | 2,261 | 66.23 |
We recently had a customer who needed to create TIFF Class F images from a PDF. TIFF class F is a subset of TIFF used for Faxing, and it has some very specific requirements, namely Group 3 compression and an image resolution of 204 x 196 dpi.
When creating black and white TIFF images, our PDF library can only create the more modern (and smaller) Group 4, so some other approach was needed. The JAI library can create TIFF images in a number of configurations, so we thought we'd take a look.
Integrating with a third party library requires supplying the image as a
BufferedImage,
which you can retrieve very easily from the
PagePainter class. Here was our first cut of the code:
import java.io.*; import java.util.*; import java.awt.image.*; import javax.media.jai.*; import com.sun.media.jai.codec.*; import org.faceless.pdf2.*; import java.awt.Graphics2D; import java.awt.geom.AffineTransform; public class G3Test2 { public static void main(String[] args) throws Exception { PDF pdf = new PDF(new PDFReader(new File(args[0]))); PDFParser parser = new PDFParser(pdf); // 1. Create a bi-level ColorModel - 0=white, 1=black // JAI requires it round this way for no good reason, // which is a shame as PDFParser.BLACKANDWHITE is this // but the other way around! byte[] b = new byte[] { -1, 0 }; ColorModel model = new IndexColorModel(1, 2, b, b, b); int num = pdf.getNumberOfPages(); BufferedImage[] images = new BufferedImage[num]; for (int i=0;i < num;i++) { PagePainter painter = parser.getPagePainter(i); images[i] = getImage(painter, model); } FileOutputStream out = new FileOutputStream("out.tif"); TIFFEncodeParam param = new TIFFEncodeParam(); param.setCompression(TIFFEncodeParam.COMPRESSION_GROUP3_2D); ImageEncoder enc = ImageCodec.createImageEncoder("TIFF", out, param); if (num > 1) { // Add subsequent pages using the odd approach // required by JAI. List therest = Arrays.asList(images).subList(1, num); param.setExtraImages(therest.iterator()); } enc.encode(images[0]); out.close(); } static BufferedImage getImage(PagePainter painter, ColorModel model, int dpi) { return painter.getImage(dpi, model); } }
This worked well, creating a Group 3 compressed image. However the resolution was 200x200dpi, not the required 204x196dpi.
The solution isn't necessarily obvious, but the approach is very flexible and shows
the kind of tricks that can be done by supplying your own image to
Graphics2D object to the
PagePainter class instead of relying on it to create one for you. Essentially we create an appropriate
sized bitmap, scale its Graphics object appropriately and then paint the PDF to it.
Here's the new
getImage method.
static BufferedImage getImage(PagePainter painter, ColorModel model, int dpix, int dpiy) { PDFPage page = painter.getPage(); // Find size of image in pixels int w = Math.round(page.getWidth() * dpix / 72); int h = Math.round(page.getHeight() * dpiy / 72); WritableRaster r = Raster.createPackedRaster( DataBuffer.TYPE_BYTE, w, h, 1, 1, null); BufferedImage img = new BufferedImage(model, r, false, null); Graphics2D g = (Graphics2D)img.createGraphics(); // Apply the anamorphic transform float dpi = (float)Math.max(dpix, dpiy); if (dpix!=dpiy) { g.transform(AffineTransform.getScaleInstance(dpix/dpi, dpiy/dpi)); } // Paint the area of the page we want to see. float[] box = page.getBox("ViewBox"); painter.drawSubImage(g, box[0], box[1], box[2], box[3], dpi); g.dispose(); return img; }
Passing 204 and 196 as the resolution into that method will give an image of the correct size - the last thing we need to to is tell JAI to write the correct resolution units to the image. We added the lines in bold to the code:
param.setCompression(TIFFEncodeParam.COMPRESSION_GROUP3_2D); TIFFField xres = new TIFFField(0x11A, TIFFField.TIFF_RATIONAL, 1, new long[][] { { 204, 1 } }); TIFFField yres = new TIFFField(0x11B, TIFFField.TIFF_RATIONAL, 1, new long[][] { { (196, 1 } }); param.setExtraFields(new TIFFField[] { xres, yres }); ImageEncoder enc = ImageCodec.createImageEncoder("TIFF", out, param);
Success! The resulting TIFF image is a Class F TIFF. You can download a complete, quick and dirty version of this code here. | http://bfo.com/blog/2009/07/17/creating_tiff_class_f_images/ | CC-MAIN-2017-04 | refinedweb | 635 | 50.53 |
Your Raspberry Pi works great on your local network and you can even use a Raspberry Pi as a web server there. But what if you want to write a web application that will allow someone to access your Pi’s hardware, its camera for example, from a web browser anywhere on the Internet? It’s pretty easy using Anvil, a Python web service that you can run locally.
Anvil provides a powerful, yet simple way to write full stack web apps in pure Python and it has been with us for a few years as a service that has been online only, requiring the use of Anvil’s servers to create projects. Your project is hosted on their servers and it can talk to a remote machine, for example a Raspberry Pi, using an uplink script on the Pi. However, there’s no need to rely on Anvil’s own servers anymore as the Anvil team recently open-sourced their app server enabling anyone to make their own projects using the Anvil web service and then download and run the same project on their own server..
What You Need
- Raspberry Pi (preferably a Raspberry Pi 4, but any other model should work also)
- Raspberry Pi Camera (Pi Quality Camera preferred)
- The latest version of Raspbian installed on your microSD card
Connecting the Camera Module
1. Insert the camera connector to the Camera port (CSI) which for model B boards is between the HDMI and composite port. Gently lift the plastic cover and insert the cable with the blue tab facing the Ethernet / USB ports. Push the plastic cover back in place to lock the cable
2. Enable the camera interface on your Raspberry Pi. Just launch the Preferences app on y our Pi, go to the Interfaces tab, enable the Camera interface and then reboot the Pi.
Getting Started with Anvil
3. Install the Anvil App Server by typing the following at the command prompt on your Pi.
$ sudo pip3 install anvil-app-server
4. Create a new blank app at anvil.works. Open a web browser to and click on “Start Building” to sign up for a free account. Create a new blank app and use the Material Design theme.
The Anvil editor now opens and in the center of the screen is a form which is where we create our user interface.
Building Your Photo App in Anvil
5. Drag a button from the toolbox into Form1 and place it at the top of the form.
6. Change the text for this button by altering the text field in Properties.
7. Drag an image from the Toolbox and place it under the button. Drag the image area to resize as you require. The image captured by the camera will be shown here and scaled to fit the image area.
8. Double clicking on the button will open the code editor and highlight the section of code for the button.
9. Add the following code to line 3 of your app.
import anvil.server
Then on line 15 we will add the following code which will change the image displayed on the app by running code on our app server. This code is a function that we shall create later in the project. Note that this code is automatically indented to show that it belongs inside the function “button_1_click()”
self.image_1.source = anvil.server.call('takepic')
10. Navigate to SERVER CODE on the left side of your screen and click Camera_Controller, then click “Add a new Server Module” This will create an area where we can write the code that will trigger the camera to take a picture.
11. Import two modules of Python code on lines 2 and 3. The first will enable Anvil to handle images in our app. The second enables access to the Raspberry Pi camera.
import anvil.media from picamera import PiCamera
12. Add the camera object. At line 4 we create an object which we can use to control the camera. Line 5 configures the camera to use the maximum resolution for the HQ camera. If you are using an older camera, change this to reflect your needs.
camera = PiCamera() camera.resolution = (4056, 3040)
13. Create the takepic() function. Line 20 sees a Python decorator used to identify an area of code which can be called by the Anvil server. This code is our “takepic” function, called when the button is clicked. This function will capture an image and save it as ‘foo.jpg’, then on line 23 we return this file as a media object which Anvil can then show in the app.
@anvil.server.callable def takepic(): camera.capture('foo.jpg') return anvil.media.from_file('foo.jpg')
Making Anvil Code Run on Your Raspberry Pi
14. Click on the Cog icon, and then select “Share app…” then click on “Clone with Git”. This will download all of the code to our Raspberry Pi.
15. Copy the URL starting “git clone” and paste this into a terminal, then press Enter. You will need your Anvil account email and password to authenticate.
16. Navigate to the directory containing the Anvil code, called Camera_Controller, in your terminal window.
$ cd Camera_Controller
17. Type the following to run the Anvil App Server, and load the configuration file for the app. Note that this will take some time, especially at first as it needs to download ~200MB of files
$ anvil-app-server --config-file anvil.yaml
Once we see “Downlink authenticated OK” we know that the Anvil code has been run successfully.
18. Open a browser to localhost:3030 and you will see the app. Click on the TakePic button and after a few seconds the image will appear in the app
Another device on the same network can also control the app we just need to replace “localhost” with the IP address of our Pi. Which can be found by hovering the mouse over the WiFi icon for a pop up to appear with the details.
Adding a Secure Tunnel for Internet Access
Now it’s time to create a secure tunnel to our Raspberry Pi server so we can access it from outside of our network. If you don’t wish to do this, you can skip these steps.
19. Download and install the Linux ARM archive from.
$ unzip ngrok-stable-linux-arm.zip
20. Launch ngrok to start a tunnel that will create a URL directly to our app. Look for the forwarding URL and make a note of it.
$ ./ngrok http 3030
21. Type the URL into a browser on another device (phone, laptop, tablet) and the camera interface is ready for use. From anywhere in the world!
Failed to start built-in Postgres database: java.lang.IllegalStateException: Process failed
More logs are available in .anvil-data/postgres.log.
Some common causes of this problem:
- Are you launching this server as 'root' on a UNIX system?
Postgres will not run as root; try launching the server as an ordinary user.
- Are you running this server on an unusual architecture or OS? (Linux/arm)
tail .anvil-data/postgres.log.
initdb: cannot be run as root
Please log in (using, e.g., "su") as the (unprivileged) user that will
own the server process.
What could be the solution to this?
1. Check ownership of the files
ls -la /home/pi/cam
If you see root as the owner, move on to the next step
2. Change ownership. I don't think anything needs to be owned by root so you should be safe recursively changing:
chown -R pi: /home/pi/cam
That should do it
camera_ controller does not appear on server code, is there another way | https://www.tomshardware.com/how-to/raspberry-pi-remote-control-camera-from-web | CC-MAIN-2021-43 | refinedweb | 1,283 | 72.56 |
Writing JavaScript code in the form of ES6 modules has become a common industry practice. Browsers today already have the support for ES6 modules, but just like with most other problems in web development, not all browsers support it.
So, instead of serving the exact ES modules to the client, we have to rely on a build system to transpile and bundle the code into something that all browsers can process, not just for production but also for development.
This seems like an awful lot of unnecessary bundling during the development cycle, since the developers can just use a modern browser while they’re developing the app.
Would it be nice if we had a way to serve the ES modules directly only for development, and we can let the build system bundle the code for production as usual? That way we get all the speed benefits of running the modules directly, without dropping support for IE.
Vite can help us to do just that.
Vite is a Vue.js application build tool, authored by the creator of Vue, Evan You. In the rest of this article, we’ll take a closer look at what it is and what it does, with a brief hands-on code demonstration.
Modules in a Nutshell
To get a feel of what a modern browser is capable of, let’s start with a simple experiment to run ES modules directly in the browser, all without the help of any build system.
First, a module like this:
📄hello.js
function hello() { alert("Hello"); } export default hello;
And then, a second module importing the above module:
📄main.js
import hello from './hello' hello();
Finally, a script tag in your HTML connecting to the second module:
<script type="module" src="main.js"></script>
This code setup will work on most modern browsers, except IE. (But, we still need to serve the files using a static server, because importing a module using the
<script> tag is subject to CORS restriction.)
The Problem
Since IE still has a sizable share of the market, it’s not practical to just serve
the “future” code and ignore IE. That’s why the current standard workflow is to convert the “future” code into something more “traditional” that all browsers can understand. That’s usually taken care of by tools like Webpack and Babel.
During the development cycle, we have to change and save the code a few hundred times on a daily basis. The hot reloading process involves putting a module through the bundling pipeline every time we change the code in the module, and this is as slow as it sounds.
Now Enter Vite
Vite is an alternative to the standard Vue CLI that intends to fix this particular speed problem.
So during development, you can continue to write your code in ES modules, and Vite will serve the ES modules directly to your browser. All you need to do is to avoid using a browser that doesn’t support ES modules during development.
Since the build system will skip the bundling process and serve the modules directly to the browser, it’s fast when refreshing the page on code changes.
And for production, it will be the same old way, similar to how Vue CLI handles it. Since not all your clients will be browsing your website with a modern browser, Vite will convert the modules into a build that all browsers can understand.
As you can see, Vite focuses on speed during development, not production (because the production build is usually optimized already).
Trying It Out
Getting started with Vite is simple. We just need to run the vite-app initializer with the npm init command:
npm init vite-app my-app
And then, install the dependencies:
cd my-app
npm install
Finally, run the app:
npm run dev
Go to localhost:3000, and you’ll see the default welcome page.
Inspecting the Code
If you check out the rendered HTML using the Code Inspector in your browser, you will find a script tag importing an ES module.
As you can see, it fetches a module from src/main.js.
If you follow the trail and inspect what’s inside src/main.js, you’ll find that the code your browser has is more or less the same as the source.
Browser:
Source:
📄src/main.js
import { createApp } from 'vue' import App from './App.vue' import './index.css' createApp(App).mount('#app')
The import paths are slightly different between the two versions, but nonetheless, this proves that Vite serves modules instead of a bundle.
If we dig deeper and inspect what’s inside src/App.vue, we’ll see another module, but this time the code is very different from the source module:
Browser:
Source:
📄
<template> <img alt="Vue logo" src="./assets/logo.png" /> <HelloWorld msg="Hello Vue 3.0 + Vite" /> </template> <script> import HelloWorld from './components/HelloWorld.vue' export default { name: 'App', components: { HelloWorld } } </script>
Now it should be clear that Vite doesn’t just serve the source code directly to the browser, it still compiles the source whenever there’s something that doesn’t make sense to the browser. For example, the
<template> tags and
<script> tags inside a .vue file aren’t valid JavaScript, so Vite will transpile them into actual JS code that the browser can understand.
So basically, Vite still runs the modules through a transpilation process, just not a bundling process.
It’s still just Vue
With Vite, the way you would develop your Vue app is still the same.
For demonstration, let’s create a new module called Counter:
📄src/components/Counter.vue
<template> <p>{{ count }}</p> </template> <script> export default { name: 'Counter', data() { return { count: 0 } } } </script>
And then, add a click event:
📄src/components/Counter.vue
<template> <p v-on:{{ count }}</p> </template> <script> export default { name: 'Counter', data() { return { count: 0 } } } </script>
Finally swap it with the default HelloWorld component:
📄src/App.vue
<template> <Counter> </template> <script> import Counter from './components/Counter.vue' export default { name: 'App', components: { Counter } } </script>
Now you should see a number in the browser, clicking the number will increase its count.
Change of Era
Vite is a special tool in the sense that it highlights a gradual change of era in web development, and in turn, foreshadows some future JavaScript development practices.
Choosing to create a JavaScript app in an NPM setup over a pure front-end setup is no longer an exotic choice. Instead, it’s considered the default standard choice. Since the modules can be served to the browser directly without bundling, debugging these modules can be as straight-forward as debugging traditional JavaScript code.
A natural next phase could be a new class of tools that can serve the native modules to supported browsers conditionally, not just in development but also in production. Then, JavaScript bundling will no longer be the norm. Instead, it will just be an exception only for browsers that don’t support modules.
One of these days, web developers will rejoice in the realization that the future is practically now. | https://www.vuemastery.com/blog/faster-hot-reloading-for-vue-development-with-vite/ | CC-MAIN-2020-29 | refinedweb | 1,183 | 61.06 |
102209/how-do-i-use-raw-input-in-python-3
import sys
print(sys.platform)
print(2**100)
raw_input()
I am using Python 3.1 and can't get the raw_input to "freeze" the dos pop-up. The book I'm reading is for Python 2.5 and I'm using Python 3.1
What should I do to fix this?
Starting with Python 3, raw_input() was renamed to input().
The raw_input() function reads a line from input (i.e. the user) and returns a string by stripping a trailing newline. This page shows some common and useful raw_input() examples for new users. Please note that raw_input() was renamed to input() in Python version 3
raw_input() was renamed to input() in Python 3. Another example method, to mix the prompt using print, if you need to make your code simpler. x= int(raw_input()) -- Gets the input number as a string from raw_input() and then converts it to an integer using int()
For Python 3, try doing this:
import urllib.request, ...READ MORE
can u give an example? READ MORE
The reduce() function in Python takes in ...READ MORE
copy a file in python
from shutil ...READ MORE
suppose you have a string with a ...READ MORE
You can also use the random library's ...READ MORE
Syntax :
list. count(value)
Code:
colors = ['red', 'green', ...READ MORE
can you give an example using a ...READ MORE
The first part starts with grep , followed by ...READ MORE
1886
Use del and specify the index of the element ...READ MORE
OR
Already have an account? Sign in. | https://www.edureka.co/community/102209/how-do-i-use-raw-input-in-python-3 | CC-MAIN-2021-21 | refinedweb | 264 | 86.71 |
SemanticWiki
What hybrid vigor can we get from cross-breeding Wiki and the SemanticWeb?
What Wiki Offers the SemanticWeb
Wiki provides: Easy Names, Easy Editing, Communal Ownership, ...
InterWiki names almost perfectly match XML or N3 namespaces. The namespace list is fixed for the wiki, but maybe that could be just the default, and per-page overrides could be allowed. @prefix on a wiki page! See how "RFC':2396" turns into a link: RFC:2396 ? This does raise the usual spectre of DansCar.
Easy Editing -- well, sure, it's fine as it is. For formal languages saving should offer or require syntax checking.
Communal Ownership -- fine as is. (If it gets to be a problem we can require logins... There's an intermediate position where (perhaps on a per-page basis) changes can be offered by anyone, and anyone can view the changes offered, but they don't get committed and shown on the official page until someone with write access commits them. For simplicity, all unofficial changes happen in the same series, and no official changes can be made without accepted or rejecting the current chain of unnofficial changes. This must be an old Wiki topic... LinkMe?)
This is what is called staging in a CMS, and I don't remember much discussion about it in wikis. If you find such things, then either in MeatBall (MeatBall:FrontPage) or the C2 wiki. -- JürgenHermann DateTime(2003-03-30T03:39:44)
So how do you provide (RDF) data on a Wiki page?
then they still look like normal pages. maybe that's good/okay. Needs system-wide mime.types config file. can't change the format without changing the name.
- a) Have wiki store some metadata elsewhere. kinda messy -- or is it already doing
that?
- a) Have wiki store metadata in the file, but not show it (like Twiki did with %TOPIC).
- b) Put some metadata in a standard format in the
file, like -*- Content-Type: "foo/bar" -*-
- Have a section of a page be data
Is there any implementation of a SemanticWiki (or Wiki:SemanticWikiWikiWeb)? I'm very interested on it. -- PaoloCastagna
I made a prototype. It is very young. See (oh, I'm sorry... it's in french.)
There is also a very different approch with RDFWiki : see -- CharlesNepote
gnowsis rdf / wiki
the research prototype project has a wiki approach that goes a few step beyond. I, Leo Sauermann am building the gnowsis / gnogno system as a personal RDF Server. I use a wiki as user interface to access complex RDF data. Some concepts It contains:
- wiki permalinks and the wiki urls will definitely be used as rdf:about identifiers, they are correct uris and work fine.
- the content of a wiki page IS DansCar. I have the feeling (and will prove it scientifically) that this works fine in personal, subjective way. So the wiki page = the idea = the concept behind. how this works exactly you will find out in November.
- You will be able to author RDF data using the wiki.
- you will be able to "link" a wiki page to other resources.
The system will be published in November 2003, I am open for personal discussion by email: leo(at)gnowsis.com
Also PeriPeri, rx4rdf, PlatypusWiki and UseMod:RdfForWikis
I've got some basic Java Semantic Wiki code ("Stiki"), will publish asap. -- DannyAyers | http://www.w3.org/wiki/SemanticWiki | CC-MAIN-2015-22 | refinedweb | 555 | 67.55 |
@Chris, I do hope you find that all Microsoft bloggers are accessible, this forum provides individuals at Microsoft with an opportunity to communicate at a much more grounded level.
Rob declined two events, only one of which was preceded by the ODF interop event. As for the specific reasons why Microsoft did not attend, you’ll have to post that question to Doug, he (or folks in his team) would typically be the ones to attend. ()..)
@"Smion"
I will criticize Rob’s unprofessional manner for pointing it out, thank you. As for the part about asking for input, we did () Rob was invited more than once, but elected not to attend. You are welcome to attend as well.
@Jan, "Instead of bitching around, let’s all move to productive mode and work on the technical problems by solving them to create true interoperability based on open standards."
We’re in total agreement here. Let’s make sure that for Open XML implementers and ODF implementers, we all have the same expectations for how that works. And please attend the DII to share your feedback with us in the future.
@Ian
"What I fail to understand is how this interoperability forum ended up with no spreadsheet formulae from Microsoft Office being readable by any other application?"
Perhaps because you were not there? — (really, I’m not trying to wind you up here.) Please do attend DII to share your feedback and discuss the implementation. You will find people ready and willing to engage on the topics. This is what the forum is intended to surface.
@Matthew Flaschen
If it isn’t Rob’s duty to baby-sit, then perhaps it is also not his duty (as a person responsible for conducting a vendor-neutral forum for ODF) to evaluate it either.
You are correct that I do not sit in the ODF TC or any other stanadrds body. As you are likely aware, there are thousands of people who are affected by document format standards who do not sit in the ODF TC or any other standards body. Am I to take your comment to mean that nobody but ODF TC members should take an interest in the development of the standard? Woudl you close the process off so that nobody but ODF TC members are entitled to opinions?
As a consumer of the ODF TC process, I think I’m entitled to my opinion, and I think I’m not out of line in asking for professional and neutral conduct from its leader.
If you are interested in learning about Patrick Durusa’s definition of a "supporter" is, read here:.
Let me know if you think Rob Weir fits the definition of an ODF Supporter: "PS: Every keystroke for a negative message about some other standard, corporation or process is a
keystroke taken away from promotion of OpenDocument. If enough such keystrokes fall, so will
OpenDocument. It’s your choice."
@Andrew Dar
"What seems painfully obvious to me however is that a large number of competing implementations appear to have achieved interoperability *despite the absence* of such definition. One of these implementations is even an existing plugin for MS Office. So it is demonstrably achievable."
— yes, based on the implementation of a single product (OpenOffice). If it is not appropriate for Open XML implementers to be forced to achieve interoperability with Microsoft Office (As was claimed by Rob in the past on many occasions), it is not appropriate to expect ODF implementers be required to implement OpenOffice features either.
Thank you Dave James for the support:
"let’s face it – to the outside world you’re all just representatives of the same monolith),"
Let’s face it. Rob Weir is an engineer from IBM with a long track record of bashing Microsoft. And while Rob is certainly entitled to his opinion, I find it hard to separate those opinions from his role of being the ODF TC co-chair.
Please explain how it is different for him than it is for me? I might be mistaking your point of view.
@Jan, I’d say this is the reality we live in. There are implementations of ODF 1.2 (which is not a standard), IS26300, ODF 1.1. Standards evolve. Developers work to keep in touch with those evolutions.
@Chris R.
He’s a Microsoft employee, too.
@Mad Hatter… if you read the posts, you’ll see the problem. Because ODF does not define a syntax for spreadsheet formulas, implementers are left to choose how to represent them when writing the format. I suggest you also read for a little more detail.
@Dirk
.
"
OpenOffice is not the most widely used program for Spreadsheets, and the "common convention" for ODF implementers is not based on the program that is most widely used.
Are you recommending that ODF implementers switch to using the Open XML implementation for formulas? — (you said it, not me.) I don’t think I would be THAT bold or foolish.
Why does ODF require implementers to mimic the behavior of a single application?
@.
@Dave Pawson
Thank you Dave, you’re not the only one noticing!.
@Allin
I’m all for people speaking the truth. And believe me, we welcome the feedback on ODF. We WANT a good implementation. We would not have shown up to the TC if we didn’t.
.
As for the "truth" — make sure you read this as well. ::
From Doug’s Post:
?"
@Jose_X, for all the points that you raise, none of them really account for the fact that ODF has no definition for formulas. Dragging out the ‘reuses exisiting standards’ talking points, for example, has no meaning when you are comparing it to something ODF doesn’t support.
Furhtermore, I’ll take your insistence that ODF comes with an open source reference implementation as a confirmation that ODF cannot be implemented from the specification alone, and I’ll note again the degree to which this was used as a reason to not standardize Open XML..
Thanks for the pointer James, but I’d rather just read it on the ISO site. Besides, the Groklaw camp are doing a good job at finding their way here (thanks for the traffic PJ).
You can view a list of Open XML applications here:. Note that this (long) list has not been updated for some time.
If you want to learn more about the nature of Open XML Development, visit here:. You’ll find a large and healthy community ready to assist you with any questions you may have.
@Jan
"Anyway, ultimately it is up to OASIS to announce their POV on the compliance of SP2’s ODF implementation. The practical tests I have seen are however not really promising."
I think this is exactly what we’re talking about in my post. The co-Chair of the ODF TC did exactly that.
@Bob Ross..
@Jan, I agree. It would be good to have neutral leadership as well.
@Andre, @Fiery… YES, I completely agree. Thank you! This is exactly what I am saying. (with apologies for the American reference) Would you see Barak Obama going home at night to post his personal opinions on the way US Auto companies are managed? While
he may certainly do so in an official capacity, I doubt that he’d be reflecting his offline thoughts in a blog.
While significanly different in scale and importance, Rob’s position is the same. It is my (personal) opinion that he should act similarly as the designated co-chair of the committee.
And let’s be clear about the history. It was MICROSOFT who first published detailed notes about the implementation of formulas & ODF in SP2, welll in advance of the SP2 release. Those notes are here if you have not seen them:.
All Rob did was act as a publicity engine… and you are exactly right. This has nothing to do with ODF TC work.
And frankly, had the "malice or incompetence" comment (or similarly negative and uwarranted criticism) not been written, my post would have never existed. Had he played it down the line (especially given that we’re in an area where the ODF Spec has no answer),
the conversation would be much more constructive.
I don’t think I’m being unreasonable in asking Rob to choose between being the leader of the Anti-Microsoft sentiment and the chair of the ODF TC. It seems counter productive for him to continue to be both.
Read this:
— and before you deem him a "shill" make sure you read everything he’s written on Office in the past. ().
@S P Arif Sahari Wibowo
Interesting point about the use of Symphony 1.3. It it is beta software, and was not available to Doug. Rob has said in the past that he only likes to test shipping products, so that part was a little strange, to be sure.
@Jose_X
“It surprises me that the Openoffice developers can figure out many details of Microsoft’s closed formats [this requires a lot of hard work and desire for interoperability], “
And it surprises me even more that Rob’s & Doug’s tests are likely to have an “OK” mark in every table cell if binary formats were used instead of ODF.
And please remember back to the Open XML standardization process, where Microsoft and Open XML were so sharply criticized with this accusation: “one [supposedly] cannot implement Open XML using the specification alone.”
So many people on my post (and Rob himself) are claiming that this is fine for ODF, but a reason to vote no on Open XML. It’s too bad that this conversation seems to have a different set of rules based on the standard that is up for discussion.
Apologies to the folks who have been waiting for comments to pass through the moderation queue for a while. It’s Mother’s day weekend as you know, and I’m taking a bit of time for my family. I’ll get to the responses to these when I have a minute.
Thank you for the comments and the discussion. Keep them coming.
.
@Rick
Thank you Rick, I couldn’t agree with you or Patrick Dursau more.
@Rick, thank you for your comment. I appreciate the balance given the stream of Groklaw traffic to the post (currently about 80% of its hits).
For me this is all about conduct. The technical debate is one worth having. The DII is a great forum for sharing that feedback directly with those who are responsible for writing the code in Office.
James, try this:
this,
and this.
and this.
All available today.?)
@HWIN
I can state for the record a few things:
1. This blog is my opinion, not the opinion of Microsoft.
2. Thus, I can state for the record that there is not coup or conspiracy to "Fracture ODF" or to take over the TC..
A super-majority of countries voted to approve Open XML. I think there is an axiom out there somewhere about what to do when it seems like everybody else is the problem :).
Med anledning av Computer Swedens artikel “ Uselt stöd för öppet filformat i nya Office ” tänkte vi tipsa
@Martin, that’s one way of looking at it. Or you could look at it in the way that we do. Microsoft is actively soliciting the participation of the community that is most vocal about their desire to see ODF implemented in Microsoft Office. — wise for any software product to do… face to face discussion on feature requirements is a very good way to solicit feedback. ‘advising’.
As to your comments about ‘influencing’.
Dave
I don’t get this…how could Rob Weir’s position as chair of the ODF TC make him responsible for quality checking of Microsofts ODF implementation? Wouldn’t it be Microsoft’s resposibility to perform tests with loading ODF files from other vendors? It is bizare to blame Weir because he did not participate in your internal beta testing and quality verification.
The problem that Microsoft has slowed down development of OpenFormula by not providing basic information about what different Office versions does is not a reason to ignore interoperability. What the other Office suits really do with spreadsheets is known information and in the absence of specification of formulas any vendor that care a tiny bit about interoperability should at least make an effort to provide interoperability.
As far as I can see Rob criticised an implementation which is useless for consumers. This has nothing to do with ODF TC work.
Hi Gray,
Can you state for the record that your attack on Rob Weir is based on the desire to see a proper standards process, rather than part of an organized coup to take over chairmanship of ODF, thus better to destroy this fundamental threat to Microsoft’s hegemony?
Following the hysterically evil antics of your company in the ISO vote, do you think you have any credibility left at all?
Just curious.
Hwin
“Attack the messenger not the message”
You Idiots are not saying anything about $Microsoft not complying with the ODF world standard.
Your just attacking Rob Weir for telling the Truth, and that is not what $Microsoft wants.
The ODF is bigger than $Microsoft and they will not be able to kill it.
@Chris Auld
You point out the situation precisely:
."
I have been arguing the same point over on Doug Mahugh’s blog, albeit in a more emotive way, which is the only way the anti-OOXML crowd seem to be capable of operating.
@Dave Lane
As I mentioned over on Doug’s blog:
"Perhaps if the people that feel they are qualified to be judge and jury about how Microsoft should implement standards in software could produce a guidance document with advice on when and where they should not be rigid, then that would probably help."
"Perhaps a formal request and a pass for using "conventions" instead of standards from the ODF / OOO good and the great might have enabled them to do so without fear being accused of any wrongdoing."
It really is starting to sound that the definition of interoperability is a fluid concept, changed to suit any anti-Microsoft arguments du jour.
Bearing grudges about reputations is all well and good, but there are plenty of other companies out there with a unpleasant track record, I hope you apply the same principles to them too.
Gareth
.”
Well said.
Microsoft has no ethics.
Their software is just bad and why should I bother buying when I can install ubuntu linux and do everything I need to do and be ODF compliant.
Just an fyi my forces windows on me because of microsoft’s so call “ethics”. we had option to chose vista or xp and I stuck with xp.
I won’t touch vista with my enemies ten foot pole.
.”
I invite you to read groklaw.net and see what happened with the voting of ooxml. There were documented high amount of irregularities that went that did not go on when ODF passed with much greater approval.
Let’s see how many products out there use ooxml? I believe there are far more greater products that use odf than ooxml.
"@."
That’s it – that is all the applications? Why should Rob or anybody help a company of microsoft’s size debug their software?
I help the open source community because they need and they appreciate my help and I can see the fruits of labor in the products I use everyday. If we don’t use your products why should we help make them better? It is your company that gets people fired or resign their jobs. Now you are trying to get Rob to resign when all he does is state facts. Peter Quinn is another example.
"Thanks for the pointer James, but I’d rather just read it on the ISO site. Besides, the Groklaw camp are doing a good job at finding their way here (thanks for the traffic PJ)."
well I didn’t get here through groklaw but I read sites from both sides of the argument and keep a neutral balance. You really need to read from both sides of the argument.
Every time microsoft comes up short when I look at subjects from a objective point of view.
Microsoft does what haagan daas tried to do to ben and jerrys. they punish distributors for trying to sell another software product with their hardware.
This is illegal and to this day microsoft still does this illegal business practice and is why the european union and doj are watching and will continue watching. Your odf in sp2 was nothing but a ploy to get the european union off your backs. But I am confident they can’t be bought and will be objective about microsoft and if they are I am confident they will see through your actions.
Here is something that would solve the whole problem:
.”
When you even meet half of these I might start listening but debug your problems are way off into the future unless you start changing your behavior. ‘should’ 😉 ‘practical.
"I’d like everyone reading the post to know that Rob was invited to participate in the DII events"
Hey people ( Microsof ), why don’t you try to work yourselves and stop asking help to people to do *basic* engineering tasks ?? this is not rocket science, just put square brackets in formulas ( you know, the "[" and "]" … do you need the ascii/unicode codes? let me know ;-).
The same happened with the OOXML rush to standardization, at the end, lot of people around the world had to do your homework and ended generating lot of fixes/erratas/corrigenda/addenda ?
Leave the "workshops" for more advanced things , ok?
Just my humble take on this issue. Greatings from Argentina
Martin Elizondo ( SA )
@Rick, I’ll be explaining our approach to formulas in a blog post shortly, and how it interrelates to our guiding principles.
@Jan Wildeboer, the DII site currently documents the ECMA-376 implementation in Office 2007; before Office 2010 is available, the same site will document the IS29500 implementation in Office 2010. You’re correct, the ECMA-376 spec on the DII site doesn’t mention IS29500, because ECMA-376 was published two years before IS29500.
Why do you say "ultimately it is up to OASIS to announce their POV on the compliance of SP2’s ODF implementation?" In actual fact, OASIS can not — by their own rules — do such a thing, for SP2 or any other implementation. Historically, you’ll notice that has never happened.
Hmnn, let me see if I understand this:
* MS does a bad ODF implementation that will not allow
interoperability.
* Rob Weir points it out.
* Therefore, Weir must leave the committee.
Is it a necessity to have people in the committee that
silently accept MS’ attempts at doing what – accept it –
we alredy know they have at least tried to do with things
like … in example, Java? I disagree.
@Jose_X:
"It surprises me that the Openoffice developers can figure out many details of Microsoft’s closed formats [this requires a lot of hard work and desire for interoperability], yet Microsoft can’t be bothered to attempt interoperability in such an important area…"
Microsoft HAS been capable of figuring out such things… They managed to deal with Lotus 123 compatibility.
No, this ‘feature’ is no accident. It’s classic Microsoft… I wonder why anybody would be surprised about it.
Microsoft should move its headquarter to tenaha, TX. Looks like they share at lot of values about how to conduct ‘business’..
Micro-Soft, for which your are obviously speaking, more and more behaves like Scientology or other cults.
The desire of individuals to make money to have a good live is one thing. What Micro-Soft is doing is beyond anything related to THAT, now that you as Micro-Soft’s tool again are trying to destroy an individual’s reputation and life just to remove another vocal opponent.
I hope you have at least enough respect not to touch human live for your desire to continue to rule your software empire.
Reading Ballmer’s quotes in this context, I am not really sure about this anymore.
"Our people, our shareholders, me, Bill Gates, we expect to change the world in every way, to succeed wildly at everything we touch, to have the broadest impact of any company in the world. "
Steve Ballmer, CEO Micro-Soft
This is getting scary. Glad to be a European.
"As for the "truth" — make sure you read this as well. ::"
@Gray: fair point, Mahugh’s observations are mostly reasonable. But he doesn’t mention the fact that Weir was strongly critical of non-Microsoft products that had introduced a non-interoperable default for formulas.
And this (from Mahugh) is a bit rich: "The nearly 400 pages of formula syntax documentation in ISO/IEC 29500 (Part 1, section 18.17) enables reliable formula interoperability in the Open XML community, and soon the ODF community will have a similar level
of formula interoperability."
The "reliable interoperability" of the currently non-existent "community" which implements ISO/IEC 29500 is entirely hypothetical, while the not-bad de facto interoperability of the greater part of the currently existing ODF community (minus Excel 2007
SP2) is a demonstrable fact.
Hi Gray,
As I’m sure you know, the purpose of standards is to provide an independent description of a complex vocabulary and grammar <em>with the goal of practical interoperability</em>. All languages suffer from problems of ambiguity, and people can speak the same language and still not understand one another… If, on the other hand, they truly <em>want</em> to understand one another, people can overcome all sorts of language barriers.
The problem here is that Microsoft doesn’t really want to understand anyone else. So you, Gray, argue about insignificant minutiae of the specification so that you don’t have admit that, in fact, you’d really rather skill keep your monopoly profits, thank you very much. I need only point to the fact that OpenOffice.org provides at least as good support for legacy MS Office documents as MS Office does – there was <em>NO common standard</em> to aid OpenOffice.org developers. They achieved practical interoperability through will alone.
It seems to me that <em>a sincere will</em> to interoperate would smooth over all of the troubles with the standard you’ve been going on about. In fact, the standard should only act as a convenient way for you to cost effectively achieve interoperability. Implementing a standard without interoperability is pointless for the market.
Standards compliance <em>without interoprability</em> only has a value to Microsoft, and that’s only because many national bodies legislating open standards compliance seem shortsighted enough to tick the "ok for procurement" box based on MS’s <em>statement</em> of "standards compliance" rather than confirmed real-world interoperability.
Now it’s time for MS to show goodwill and meet the market half way (like the many OpenOffice, KOffice, Google Docs, etc. users out there) for a change – even if <em>it doesn’t help MS maximise its profit this quarter</em>… because it’s better to lose some profit in the short term to avoid being despised by the market and being dropped like a bad habit as soon as a viable alternative presents itself.
Goodwill is money in the bank for the longer term, and Gray, from my perspective your corporate employer has got pretty much none of it.
Sincerely,
Dave
@Gary – Thanks for taking the time to read all these comments it’s quite nice that I, a basic no-name, am able to have a conversation with the Group Product Manager for the Microsoft Office System. I guess the internet is a great equalizer
Anyways, on to more comments.
"So then we’re really not talking about interoperability and standards, are we?" Actually we are, why should I trust MS to implement standards to be truly interoperable when then have such a history of implementing standard in a way that benefit ONLY Microsoft? Sure they aren’t breaking the standard (yet) but sure are breaking the intent of a standard.
Also I noticed several times you invited people to attend DII and pointed to their website, which it appears to be owned by Microsoft. You also enjoy pointing out that how Rob Weir did not attend.
Rob We."
Is what Rob Weir stated true? Did Microsoft not attend, if not, why not? From my perspective this does not look like Microsoft is sincere in their desire for interoperability, which of course makes me not trust them.
"I’d like everyone reading the post to know that Rob was invited to participate in the DII events leading up to the SP2 release, and offered the opportunity to test the beta software specifically for the purpose of providing feedback on the implementation."
It’s your responsibility to /comply with the standard/, and support interoperability, to the best of your ability. It’s not his responsibility to personally baby-sit as you do so.
"If departments within 18 various governments really do use ODF as their standard, should we be comfortable with an ODF TC chair that is trying very hard to discredit and divide its supporters?"
To count yourself among the supporters of ODF 1.1 is laughable. You didn’t even vote /on/ (let alone /for/) the ODF 1.1 standard!
"Is it time for Rob to step down as chair? I think so."
Even if Rob Weir were completely wrong about Office’s interoperability failures (which you know he is not), you have no right to ask him to step down when you remain so aloof from the actual standards process.
I’m terribly sorry I got your name wrong, Gray. I feel rather foolish now that I see the pun in the title of your blog.
I visited the website you linked, it is clear that this issue was discussed in advance; I look forward to your blog explaining/justifying the approach to formulas.
You don’t have to show or respond to this post if you would like, I understand you must be inundated with posts.
Again, terribly sorry I got your name wrong,
Simon.
@Gray Knowlton
"."
I may be missing something here but there _were_ 7 different implementations used in the test. Your fixation on only a few is fascinating though. It’s not Robs fault that the outcome was the way it was.
Also, I’m not finding anywhere where Rob actually calls anyone names or assigns the labels you’ve mentioned. He uses "incompetence" as a descriptor to explain his thought processes but does stop short of actually calling anyone anything.
A few people here have mentioned the difference between the messenger and the message. Reading your post, it does come across as an ad hominem attack against Rob. This does nothing towards reinforcing your reputation as someone knowledgeable in any technical field.
."
yup, agreed. You are definitely not the ones trashing ISO. You were the one who made others trashed ISO.
The form of these conflicts is really ages old, from way before there were even computers,
It used to be called "eminent domain" back in the day – and which side one was on nearly completely guided the tenor of one’s response in the struggle du jour.
Without pushing the analogy too far, it is probably safe to say that ODF developed as a result of real needs which were not being met, and which existed largely because the *structure* of Microsoft’s dominance effectively precluded such needs fror being satisfied by them.
It is also safe to conclude that Gray’s responses are unsurprisingly oblivious to this context, since from his point of view there is no discernible reason for any struggle. Since Microsoft is dominant, its ineluctability is a foregone conclusion to such adherents – hence any serious conflict ends up representing a direct attack on the joint efforts of the world he inhabits, regardless of any other possible interpretations.
Sadly, many "open" advocates get pulled into this false representation as well, and much "bashing" and "we can never trust you" moments abound as a result.
Please people try to be clear about this – it is not a war, it is fundamentally different paradigms in confrontation. Microsoft cannot help operating from the "hindbrain", it’s just the kind of beast it is.
All you nice li’l mammals can take your warm blood & communal sharing into a future where you eventually leave all this behind – we all learned something about trails of tears, so we don’t need to repeat that any more …
Dear Mr Knowlton, unfortunately, I have to point out a false statement in one of your comments: i.e., that Microsoft is actively soliciting community involvement in the process of ODF interoperability. I am the leader of the OOo community in Italy, and I happen to know your colleague Sam Ramij.
When I discovered the DII labs I wrote Sam and I suggested to invite the OOo community to these labs. I still remember that Sam forwarded my message to a number of your colleagues (unfortunately, I only remember the name of Mr Paoli among them) pointing out that – according to his experience – my ideas deserved some attention.
I never got any answer, I never got any contact from anyone at Microsoft (Sam told me that he couldn’t help on this subject).
So, your statement about Microsoft pushing for the involvement of the community is blatantly false. Microsoft doesn’t want the community around because it wants the discussion to be a Microsoft/IBM battle. And what has happened after the launch of Office 2007 SP2 is a proof of this situation. Unfortunately, IBM has never been as smart as Microsoft in terms of Marketing, and has brought a technical discussion into a marketing arena.
Although I am the leader of the OOo community in Italy, I am not a developer nor a technical guy. I am a marketer by education and experience, and during the process it was perfectly clear to me that Microsoft sees the ODF/OOXML debate not as a technical issues but as just another marketing campaign.
But lies are lies also in the marketing field, and your statement about the involvement of the community is a lie (at least to my eyes).
@Doug, @Gray
In my opinion it doesn’t make much sense if implementors and vendors are left alone in interpreting a standard. It would be far better if there is a neutral instance that helps in ironing out interoperability issues.
And exactly this is already done for example with CSS and (X)HTML. So why not have a similar validator for ODF? Maintained by OASIS, just as the W3C is doing?
Make it open source, so that anyone can run his own instance.
Jan
Gray, I think it is admirable for a chair of a committee actually spent time to dirty his hand, did hands-on sanity check to see if everything works well. Off course his check probably not so thorough than a dedicated team so all the "Ok"s should be taken lightly, but all the "Fail"s are definitely a concern.
I generally agreed with Rob that interoperability should not be limited on the scope of a standard alone. It is normal that a standard have several things it does not cover, and people then make the most reasonable implementation. It is also very reasonable for newcomers to the standard to learn from others how things were done and try follow it the best they could.
For the poor interoperability you may use excuse that Microsoft does not pay enough people to "learn from others" and CleverAge agreement does not allow reuse in Microsoft, but whining "the standard does not cover it" is a really bad excuse.
You compare Rob’s result on Symphony with Doug Mahugh’s result and use that to claim that Rob was bias. Unfortunately you failed to point out that they were testing different software: Doug Mahugh tested Symphony 1.2 while Rob tested Symphony 1.3 beta. Nothing bad can be inferred when the newer version works better.
I am sure if you have another office beta in the work with better interoperability many will be happy to test it for you. I guess even Rob will welcome it, assuming it come with clear statement of its license.
You aren’t the first to note this. Alex pointed me to James Clarks comments in 2005.
Compliance means what Rob wants it to mean. *to* the spec
seems to mean nothing.
regards DaveP
Everyone else can do ODF properly, but not Microsoft. Why am I not in the least bit surprised by this ? Perhaps it could be Microsoft’s long history of anticompetitive practices and the harm it has done to the industry and consumers alike. Absolutely pathetic.
And attempting to smear and bring into question Rob Weir’s integrity because he dares to point out that Microsoft, as always, fails to comply with a standard and remains as mendaciously non-interoperable as ever, is simply disgusting..
As I posted in Rob’s blog, there are very few possible reasons for Microsoft’s failure, when other organizations have managed to implement the standard properly, and none of those possibilities reflect very well on Microsoft.
mad hatter asked:
."
gray knowlton answered
"@Mad Hatter… if you read the posts, you’ll see the problem. Because ODF does not define a syntax for spreadsheet formulas, implementers are left to choose how to represent them when writing the format. I suggest you also read for a little more detail. "
This is not an answer. He is asking why are you the *only* different and no interoperable implementation. Do you thought in ODF users when you took this decision or was a pure strategical/marketing drived one?
Marc, the ODF 1.1 implementation in SP2 is the *only* ODF 1.1 implementation that uses a namespace prefix (as specified in the spec) to define the syntax and semantics of its formula markup through reference to an approved, published standard.
OpenOffice and KOffice use an undocumented non-standardized extension to achieve interoperability, which will no longer be allowed in ODF 1.2. We’ve heard loud and clear for a long time that the ODF community does not want Microsoft to extend the standard, and we’ve followed that advice in SP2’s implementation.’
Doug, in ODF 1.2 conformant documents, the used formula *prefix* will have to be "of:", and none of the current OOo or SP2 prefix will remain valid.
OOo and Koffice DON’T use a non-standardised *extension*: they simply use an allowed prefix, exactly in the same way as SP2 !
What you explain above as being the Microsoft excuse is simply a lie !
Doug, after reading again your comment, I now see what you meant by "undocumented non-standardized extension to achieve interoperability". Please replace in my comment above the word "lie" by "far fetched explanation". Apologises for an inappropriate word.
I was confused by the fact that, while standards are normally made to improve interoperability, Microsoft manages to use standards as an excuse to break interoperability !
Doug, one thing puzzles me. You wrote that SP2 "uses a namespace prefix (as specified in the spec) to define the syntax and semantics of its formula markup through reference to an approved, published standard.""…
Luc,
"…"
I never actually thought about that. One thing that might speak against using "ooxml" or "is29500" or similar would be, that the namespace used for spreadsheet formula in SP2 is not defined in OOXML, so it might cause a bit of confusion.
So rob weir actually refused a beta version of the specification ?
Could he explain what possible real reason he could have for doing that.
The license issue seems moot as millions of people use MS beta products every day without any problem and have done so for ages. and that included a ton of IBM people as well as I have seen very often as well. (I actually recieved an Office SP2 beta copy from an IBM guy).
As ODF TC chairman getting the chance to preview the ODF implementation of the leading Office product suite should be quite normal. And even in a role of IBM office product representative involved in producing new IBM Office prodcuts he and/or his IBM collegues should have been evaluating MS Office SP2 for at least 6 months or so.
There is just no way in hell that Rob Weir only knew of interoperability issues untill the official release of MS Office SP2.
Most likely his comments were timed to a moment that the Symphony 1.3 beta was able to show compatibility with OpenOffice.
It is a total disgrace that Rob Weir apperantly refuses to test a beta product of the leading Office suite he was offered to do so and then critisises this product in a post where he actually compared it with non public IBM ‘s beta products that he never prior to his post offered to scrutiny or commenting by others.
Can anybody seriously persieve IBM development team fully ignoring devlopment of the leading spreadsheet product in producing their own new product and not testing interoperability themselves using the MS Office SP2 beta ?
If so I cannot see IBM as a serious Office developer but in reality we could all easily see that IBM (and Rob) were already aware of the interoperability with MS Office SP2 for months if not even already last year.
The currently available Symphony 1.2 versions actually mutilates OOo 3.x files in a way that is horrible to any users making the information in cetain formulas unavailable and this makes interoperability a total joke.
Compared to IBM’s current Symphony implementation the Microsoft approach which micht not preserve formula’s but only values represent a much more valued approach as the reciever will at least have a spreadsheet with reliable data preserved for the user, something that IBM’s product is unable to accomplish.
@Jan wildeboer
[quote]And exactly this is already done for example with CSS and (X)HTML. So why not have a similar validator for ODF? Maintained by OASIS, just as the W3C is doing?[/quote]
Actually the MS Offfice SP2 files seem to validate fine against the OASIS ODF schema’s.
A validator would not be of much use on ODF on such issues as it has several crucial parts not validating. Both formula’s and MathML 2.0 is not in the schema’s.
Formula’s in ODF are just arbitrary strings in the XML schema’s and MathML even is arbitrary markup in the ODF schema’s.
The latter being helpfull to OOo as including MathML 2.0 schema’s or use the official schema’s as a normative reference would make all OpenOffice files using math fail.
hAl,
“If so I cannot see IBM as a serious Office developer but in reality we could all easily see that IBM (and Rob) were already aware of the interoperability with MS Office SP2 for months if not even already last year.”
Actually, I am sure they have been aware of how Microsoft Office 2007 SP2 handles formulas in ODF since at least August 2008. I documented the approach by SP2 after the first DII-workshop when I wrote the article “DII ODF workshop – the good stuff”. This was August 18th 2008. At that time we were playing with a pre-alpha-release of Microsoft Office 2007 SP2 and two of my conclusions were these:
1. Microsoft Office 2007 SP2 uses their own formulas in ODF spreadsheets
2. Microsoft Office 2007 SP2 strips away formulas from “unknown” namespaces
(the last test file in the article, Testfile_20.ods)
@Jesper: "One thing that might speak against using "ooxml" or "is29500" or similar would be, that the namespace used for spreadsheet formula in SP2 is not defined in OOXML, so it might cause a bit of confusion."
There is clearly a serious contradiction here: Microsoft insists that they use the "only standardised formula syntax", but then it appears that the namespace is not defined in the standard, and Microsoft has to use a proprietary namespace to implement the "standard" syntax.
The impact is that if somebody else (e.g. OpenOffice.org) wants to be interoperable with the ISO OOXML formula syntax, it has to use the Microsoft proprietary namespace ! This is a nonsense.
Isn’t it an additional proof that OOXML is much more a proprietary Microsoft standard than an open ISO standard ?
At least, it is a proof that more time was needed to correctly review OOXML before making it an ISO standard. (Another proof of this is that there is no mechanism in OOXML to identify the version, but this is another story)
Luc,
"There is clearly a serious contradiction here: Microsoft insists that they use the "only standardised formula syntax", but then it appears that the namespace is not defined in the standard, and Microsoft has to use a proprietary namespace to implement the "standard" syntax."
Excuse me – but you seem to be completely missing crucial technical points here.
1.
Microsoft uses the formula syntax in OOXML – the only standardised formula syntax. The format is perfectly (well, ahem) defined in ISO/IEC 29500. The formulas are part of the specification of SpreadsheetML in Part 1. They are defined in BNF-notation in section 18.7 (Formulas) from page 2268.
2.
About the namespace: it’s a pointer to where information about how to interpret it can be found. As Doug wrote somewhere, ODF requires the following for formulas:
"a namespace prefix specifying the syntax and semantics used within the formula."
And this is exactly that Microsoft does – they supply a namespace prefix (msoxl) while explaining in their implementer’s notes that this refers to Excel’s formula syntax. Excel’s formula syntax is defined in … tadaa … ISO/IEC 29500.
(I do think that they should elaborate a bit more in their notes – explaining exactly where the syntax and specification of Excel’s formulas can be found – I’ll log a comment on this in a bit)
"Isn’t it an additional proof that OOXML is much more a proprietary Microsoft standard than an open ISO standard ?"
No – it’s additional proof that you don’t always know what you are talking about. | https://blogs.technet.microsoft.com/gray_knowlton/2009/05/06/rethinking-odf-leadership/ | CC-MAIN-2016-30 | refinedweb | 7,104 | 61.26 |
Linux initial RAM disk (initrd) overview
Learn about its anatomy, creation, and use in the Linux boot process)
# mkdir temp ; cd temp # cp /boot/initrd.img.gz . # gunzip initrd.img.gz # mount -t ext -o loop initrd.img /mnt/initrd # ls -la /mnt/initrd #)
# mkdir temp ; cd temp # cp /boot/initrd-2.6.14.2.img initrd-2.6.14.2.img.gz # gunzip initrd-2.6.14.2.img.gz # cpio -i --make-directories < initrd-2.6.14.2.img #
#. The resulting file (ramdisk.img.gz) is
copied to the /boot subdirectory so it can be loaded via GNU GRUB.
To build the initial RAM disk, you simply invoke
mkird, and
the image is automatically created and copied to /boot.
Testing the custom initial RAM disk
Your new initrd image is in /boot, so the next step is to test it with your
default kernel. You can now restart your Linux system. When GRUB appears,
press the C key to enable the command-line utility within GRUB. You can
now interact with GRUB to define the specific kernel and initrd image to
load. The
kernel command allows you to define the kernel
file, and the
initrd command allows you to specify the
particular initrd image file. When these are defined, use the
boot.
After the kernel starts, it checks to see if an initrd image is available
(more on this later), and then loads and mounts it as the root file
system. You can see the end of this particular Linux startup in Listing 6.
When started, the ash shell is available to enter commands. In this
example, I explore the root file system and interrogate a virtual proc
file system entry. I also demonstrate that you can write to the file
system by touching a file (thus creating it). Note here that the first
process created is
linuxrc (commonly
init).
Booting with an initial RAM disk
Now that you've seen how to build and use a custom initial RAM disk, this section explores how the kernel identifies and mounts the initrd as its root file system. I walk through some of the major functions in the boot chain and explain what's happening.
The boot loader, such as GRUB, identifies the kernel that is to be loaded and copies this kernel image and any associated initrd into memory. You can find much of this functionality in the ./init subdirectory under your Linux kernel source directory.
After the kernel and initrd images are decompressed and copied into memory,
the kernel is invoked. Various initialization is performed and,
eventually, you find yourself in
init/main.c:init()
(subdir/file:function). This function performs a large amount of subsystem
initialization. A call is made here to
init/do_mounts.c:prepare_namespace(), which is used to
prepare the namespace (mount the dev file system, RAID, or md, devices,
and, finally, the initrd). Loading the initrd is done through a call to
init/do_mounts_initrd.c:initrd_load().
The
initrd_load() function calls
init/do_mounts_rd.c:rd_load_image(), which determines the RAM
disk image to load through a call to
init/do_mounts_rd.c:identify_ramdisk_image(). This function
checks the magic number of the image to determine if it's a minux, etc2,
romfs, cramfs, or gzip format. Upon return to
initrd_load_image, a call is made to
init/do_mounts_rd:crd_load(). This function allocates space
for the RAM disk, calculates the cyclic redundancy check (CRC), and then
uncompresses and loads the RAM disk image into memory. At this point, you
have the initrd image in a block device suitable for mounting.
Mounting the block device now as root begins with a call to
init/do_mounts.c:mount_root(). The root device is created,
and then a call is made to
init/do_mounts.c:mount_block_root(). From here,
init/do_mounts.c:do_mount_root() is called, which calls
fs/namespace.c:sys_mount() to actually mount the root file
system and then
chdir to it. This is where you see the
familiar message shown in Listing 6:
VFS: Mounted root (ext2 file system).
Finally, you return to the
init function and call
init/main.c:run_init_process. This results in a call to
execve to start the init process (in this case
/linuxrc). The linuxrc can be an executable or a script (as
long as a script interpreter is available for it).
The hierarchy of functions called is shown in Listing 7. Not all functions that are involved in copying and mounting the initial RAM disk are shown here, but this gives you a rough overview of the overall flow.
Much like embedded booting scenarios, a local disk (floppy or CD-ROM) isn't necessary to boot a kernel and ramdisk root filesystem. The Dynamic Host Configuration Protocol (or DHCP) can be used to identify network parameters such as IP address and subnet mask. The Trivial File Transfer Protocol (or TFTP) can then be used to transfer the kernel image and the initial ramdisk image to the local device. Once transferred, the Linux kernel can be booted and initrd mounted, as is done in a local image boot.
Shrinking your initrd
When you're building an embedded system and want the smallest initrd image possible, there are a few tips to consider. The first is to use BusyBox (demonstrated in this article). BusyBox takes several megabytes of utilities and shrinks them down to several hundred kilobytes.
In this example, the BusyBox image is statically linked so that no libraries are required. However, if you need the standard C library (for your custom binaries), there are other options beyond the massive glibc. The first small library is uClibc, which is a minimized version of the standard C library for space-constrained systems. Another library that's ideal for space-constrained environments is dietlib. Keep in mind that you'll need to recompile the binaries that you want in your embedded system using these libraries, so some additional work is required (but worth it).
Summary
The initial RAM disk was originally created to support bridging the kernel to the ultimate root file system through a transient root file system. The initrd is also useful as a non-persistent root file system mounted in a RAM disk for embedded Linux systems. | https://www.ibm.com/developerworks/linux/library/l-initrd/ | CC-MAIN-2018-30 | refinedweb | 1,031 | 64.41 |
Code Indirection
stereobooster
Originally published at
stereobooster.com
on
・2 min read
Code indirection:
import a from "./a.ts"; const b = a(); function c() { b; }
-:
let b = a(); // ... b;:
function c(b) { b; }
to find a source of
b you will need to find where
c is called and where it gets arguments.
The more code indirections you have the harder it is navigating through the code base, the less readable code.
Direct code:
e ← d ← c ← b ← a
Indirect code:
e → d ← c ← b ← a
PS
More about readability:
🎩 JavaScript Enhanced Scss mixins! 🎩 concepts explained
In the next post we are going to explore CSS @apply to supercharge what we talk about here....
| https://practicaldev-herokuapp-com.global.ssl.fastly.net/stereobooster/code-indirection-42hp | CC-MAIN-2020-24 | refinedweb | 114 | 72.26 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.