https://github.com/python/cpython/commit/6f47a134bc76d32d2c357249cbafa451453c254f
commit: 6f47a134bc76d32d2c357249cbafa451453c254f
branch: 3.13
author: Serhiy Storchaka <[email protected]>
committer: serhiy-storchaka <[email protected]>
date: 2025-12-05T15:31:34+02:00
summary:
[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 ee894d2abbdb9678f4cecdc15707496a0eea7f85.
files:
A Misc/NEWS.d/next/Library/2025-12-03-19-33-17.gh-issue-140797.YxB27u.rst
M Lib/re/__init__.py
M Lib/test/test_re.py
diff --git a/Lib/re/__init__.py b/Lib/re/__init__.py
index 5a821411ad6e50..7e8abbf6ffe155 100644
--- a/Lib/re/__init__.py
+++ b/Lib/re/__init__.py
@@ -399,12 +399,9 @@ def __init__(self, lexicon, flags=0):
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))])
diff --git a/Lib/test/test_re.py b/Lib/test/test_re.py
index 8d39e7f2086666..813cb4a3f54f23 100644
--- a/Lib/test/test_re.py
+++ b/Lib/test/test_re.py
@@ -1638,24 +1638,6 @@ def s_int(scanner, token): return int(token)
(['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.)
diff --git
a/Misc/NEWS.d/next/Library/2025-12-03-19-33-17.gh-issue-140797.YxB27u.rst
b/Misc/NEWS.d/next/Library/2025-12-03-19-33-17.gh-issue-140797.YxB27u.rst
new file mode 100644
index 00000000000000..ebbe06fddfb372
--- /dev/null
+++ b/Misc/NEWS.d/next/Library/2025-12-03-19-33-17.gh-issue-140797.YxB27u.rst
@@ -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.
_______________________________________________
Python-checkins mailing list -- [email protected]
To unsubscribe send an email to [email protected]
https://mail.python.org/mailman3//lists/python-checkins.python.org
Member address: [email protected]
