sabirbagwan commited on
Commit
0d1237a
·
1 Parent(s): dffb86a

Update preprocess.py

Browse files
Files changed (1) hide show
  1. preprocess.py +6 -0
preprocess.py CHANGED
@@ -9,6 +9,7 @@ def preprocess(data):
9
  pattern2 = r'\[\d{2}/\d{2}/\d{2}, \d{2}:\d{2}:\d{2}\]'
10
  pattern3 = r'\d{2}/\d{2}/\d{2}, \d{1,2}:\d{2} -'
11
  pattern4 = r'\d{1,2}/\d{1,2}/\d{2}, \d{2}:\d{2} -'
 
12
 
13
  # data = "30/06/21, 1:46pm - Gauri Mami turned off disappearing messages."
14
 
@@ -38,6 +39,11 @@ def preprocess(data):
38
  messages = re.split(pattern4, data)[1:]
39
  dates = re.findall(pattern4, data)
40
 
 
 
 
 
 
41
  if not result:
42
  result = None
43
 
 
9
  pattern2 = r'\[\d{2}/\d{2}/\d{2}, \d{2}:\d{2}:\d{2}\]'
10
  pattern3 = r'\d{2}/\d{2}/\d{2}, \d{1,2}:\d{2} -'
11
  pattern4 = r'\d{1,2}/\d{1,2}/\d{2}, \d{2}:\d{2} -'
12
+ pattern5 = r'\d{2}/\d{2}/\d{2}, \d{1,2}:\d{2}[AP]M -'
13
 
14
  # data = "30/06/21, 1:46pm - Gauri Mami turned off disappearing messages."
15
 
 
39
  messages = re.split(pattern4, data)[1:]
40
  dates = re.findall(pattern4, data)
41
 
42
+ if re.search(pattern5, data):
43
+ result.append(data)
44
+ messages = re.split(pattern5, data)[1:]
45
+ dates = re.findall(pattern5, data)
46
+
47
  if not result:
48
  result = None
49