[3.13] Revert "[3.13] gh-140797: Forbid capturing groups in re.Scanner lexicon patterns (GH-140944) (GH-140983)" (GH-142231)

Revert "[3.13] gh-140797: Forbid capturing groups in re.Scanner lexicon patterns (GH-140944) (GH-140983)"

This reverts commit ee894d2abb.
This commit is contained in:
Serhiy Storchaka 2025-12-05 15:31:34 +02:00 committed by GitHub
parent 29100c8e87
commit 6f47a134bc
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 4 additions and 22 deletions

View file

@ -399,12 +399,9 @@ class Scanner:
s = _parser.State()
s.flags = flags
for phrase, action in lexicon:
sub_pattern = _parser.parse(phrase, flags)
if sub_pattern.state.groups != 1:
raise ValueError("Cannot use capturing groups in re.Scanner")
gid = s.opengroup()
p.append(_parser.SubPattern(s, [
(SUBPATTERN, (gid, 0, 0, sub_pattern)),
(SUBPATTERN, (gid, 0, 0, _parser.parse(phrase, flags))),
]))
s.closegroup(gid, p[-1])
p = _parser.SubPattern(s, [(BRANCH, (None, p))])

View file

@ -1638,24 +1638,6 @@ class ReTests(unittest.TestCase):
(['sum', 'op=', 3, 'op*', 'foo', 'op+', 312.5,
'op+', 'bar'], ''))
def test_bug_gh140797(self):
# gh140797: Capturing groups are not allowed in re.Scanner
msg = r"Cannot use capturing groups in re\.Scanner"
# Capturing group throws an error
with self.assertRaisesRegex(ValueError, msg):
Scanner([("(a)b", None)])
# Named Group
with self.assertRaisesRegex(ValueError, msg):
Scanner([("(?P<name>a)", None)])
# Non-capturing groups should pass normally
s = Scanner([("(?:a)b", lambda scanner, token: token)])
result, rem = s.scan("ab")
self.assertEqual(result,['ab'])
self.assertEqual(rem,'')
def test_bug_448951(self):
# bug 448951 (similar to 429357, but with single char match)
# (Also test greedy matches.)

View file

@ -0,0 +1,3 @@
Revert changes to the undocumented :class:`!re.Scanner` class. Capturing
groups are still allowed for backward compatibility, although using them can
lead to incorrect result. They will be forbidden in future Python versions.