re.sub does NOT substitute all the matching patterns when re.IGNORECASE is used

[anitrajpurohit28]

BPO	41664
Nosy	@ezio-melotti

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2020-08-29.22:32:24.365>
created_at = <Date 2020-08-29.21:47:33.268>
labels = ['expert-regex', 'invalid', 'type-bug', '3.8']
title = 're.sub does NOT substitute all the matching patterns when re.IGNORECASE is used'
updated_at = <Date 2020-08-29.22:32:24.364>
user = 'https://bugs.python.org/anitrajpurohit28'

bugs.python.org fields:

activity = <Date 2020-08-29.22:32:24.364>
actor = 'mrabarnett'
assignee = 'none'
closed = True
closed_date = <Date 2020-08-29.22:32:24.365>
closer = 'mrabarnett'
components = ['Regular Expressions']
creation = <Date 2020-08-29.21:47:33.268>
creator = 'anitrajpurohit28'
dependencies = []
files = []
hgrepos = []
issue_num = 41664
keywords = []
message_count = 2.0
messages = ['376083', '376086']
nosy_count = 3.0
nosy_names = ['ezio.melotti', 'mrabarnett', 'anitrajpurohit28']
pr_nums = []
priority = 'normal'
resolution = 'not a bug'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue41664'
versions = ['Python 3.8']

[anitrajpurohit28]

Usage of re flags leads to inconsistent results when

1. The pattern directly used in re.sub
2. The pattern is re.compile'd and used

Note 1: Input string is all in the lowercase 'all is fair in love and war'
Note 2: Results are always consistent in case of re.compile'd pattern
=======================================

1. The pattern directly used in re.sub
   =======================================

>>> import re
>>> re.sub(r'[aeiou]', '#', 'all is fair in love and war')
'#ll #s f##r #n l#v# #nd w#r'
>>> 
>>> re.sub(r'[aeiou]', '#', 'all is fair in love and war', re.IGNORECASE)
'#ll #s fair in love and war'
>>> 
>>> re.sub(r'[aeiou]', '#', 'all is fair in love and war', re.IGNORECASE|re.DOTALL)
'#ll #s f##r #n l#v# #nd w#r'
>>> 
>>> 

=======================================
2. The pattern is re.compile'd and used
=======================================

>>> pattern = re.compile(r'[aeiou]', re.IGNORECASE)
>>> re.sub(pattern, '#', 'all is fair in love and war')
'#ll #s f##r #n l#v# #nd w#r'
>>> 
>>> pattern = re.compile(r'[aeiou]')
>>> re.sub(pattern, '#', 'all is fair in love and war')
'#ll #s f##r #n l#v# #nd w#r'
>>> 
>>> pattern = re.compile(r'[aeiou]', re.IGNORECASE | re.DOTALL)
>>> re.sub(pattern, '#', 'all is fair in love and war')
'#ll #s f##r #n l#v# #nd w#r'

[mrabarnett]

The 4th argument of re.sub is 'count', not 'flags'.

re.IGNORECASE has the numeric value of 2, so:

 re.sub(r'[aeiou]', '#', 'all is fair in love and war', re.IGNORECASE)

is equivalent to:

 re.sub(r'[aeiou]', '#', 'all is fair in love and war', count=2)