re.sub replaces only first 32 matches with re.U flag

[EugeneMorozov]

BPO	11262

^{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 2011-02-20.23:01:03.740>
created_at = <Date 2011-02-20.22:28:09.731>
labels = ['type-security', 'expert-regex', 'invalid']
title = 're.sub replaces only first 32 matches with re.U flag'
updated_at = <Date 2011-02-20.23:01:03.731>
user = 'https://bugs.python.org/EugeneMorozov'

bugs.python.org fields:

activity = <Date 2011-02-20.23:01:03.731>
actor = 'SilentGhost'
assignee = 'none'
closed = True
closed_date = <Date 2011-02-20.23:01:03.740>
closer = 'SilentGhost'
components = ['Regular Expressions']
creation = <Date 2011-02-20.22:28:09.731>
creator = 'Eugene.Morozov'
dependencies = []
files = []
hgrepos = []
issue_num = 11262
keywords = []
message_count = 2.0
messages = ['128923', '128925']
nosy_count = 2.0
nosy_names = ['SilentGhost', 'Eugene.Morozov']
pr_nums = []
priority = 'normal'
resolution = 'not a bug'
stage = None
status = 'closed'
superseder = None
type = 'security'
url = 'https://bugs.python.org/issue11262'
versions = ['Python 2.6']

[EugeneMorozov]

There's a peculiar and difficult to find bug in the re.sub method. Try following example:
>>> text = 'X'*4096
>>> nt = re.sub(u"XX", u".", text, re.U)
>>> nt
u'............XXXXXXXXXXXXXXXXXXX' (only 32 dots, the rest of the string is not changed).

If I first compile regexp, and then perform compiled_regexp.sub, everything seems to work correctly.

[SilentGhost]

If you read docs carefully, you notice that re.sub doesn't accept flags argument. Its 4th argument is count, re.U numerical value is 32.

Closing as invalid. There are some duplicates too, I'm sure.