sabirbagwan commited on
Commit
2e6e05f
·
1 Parent(s): cc82d58

Update preprocess.py

Browse files
Files changed (1) hide show
  1. preprocess.py +6 -0
preprocess.py CHANGED
@@ -10,6 +10,7 @@ def preprocess(data):
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
 
@@ -43,6 +44,11 @@ def preprocess(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
 
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
+ pattern6 = r'\d{1,2}/\d{1,2}/\d{2}, \d{1,2}:\d{2} [APM]{2} -'
14
 
15
  # data = "30/06/21, 1:46pm - Gauri Mami turned off disappearing messages."
16
 
 
44
  result.append(data)
45
  messages = re.split(pattern5, data)[1:]
46
  dates = re.findall(pattern5, data)
47
+
48
+ if re.search(pattern6, data):
49
+ result.append(data)
50
+ messages = re.split(pattern6, data)[1:]
51
+ dates = re.findall(pattern6, data)
52
 
53
  if not result:
54
  result = None