Skip to content

Commit 5b76eaf

Browse files
[3.11] gh-106052: Fix bug in the matching of possessive quantifiers (GH-106515) (GH-107795)
It did not work in the case of a subpattern containing backtracking. Temporary implement possessive quantifiers as equivalent greedy qualifiers in atomic groups. (cherry picked from commit 7b6e34e)
1 parent edaa0db commit 5b76eaf

File tree

3 files changed

+21
-0
lines changed

3 files changed

+21
-0
lines changed

Lib/re/_compiler.py

Lines changed: 7 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -100,6 +100,13 @@ def _compile(code, pattern, flags):
100100
emit(ANY_ALL)
101101
else:
102102
emit(ANY)
103+
elif op is POSSESSIVE_REPEAT:
104+
# gh-106052: Possessive quantifiers do not work when the
105+
# subpattern contains backtracking, i.e. "(?:ab?c)*+".
106+
# Implement it as equivalent greedy qualifier in atomic group.
107+
p = [(MAX_REPEAT, av)]
108+
p = [(ATOMIC_GROUP, p)]
109+
_compile(code, p, flags)
103110
elif op in REPEATING_CODES:
104111
if flags & SRE_FLAG_TEMPLATE:
105112
raise error("internal: unsupported template operator %r" % (op,))

Lib/test/test_re.py

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2396,6 +2396,16 @@ def test_template_function_and_flag_is_deprecated(self):
23962396
self.assertTrue(template_re1.match('ahoy'))
23972397
self.assertFalse(template_re1.match('nope'))
23982398

2399+
def test_bug_gh106052(self):
2400+
self.assertEqual(re.match("(?>(?:ab?c)+)", "aca").span(), (0, 2))
2401+
self.assertEqual(re.match("(?:ab?c)++", "aca").span(), (0, 2))
2402+
self.assertEqual(re.match("(?>(?:ab?c)*)", "aca").span(), (0, 2))
2403+
self.assertEqual(re.match("(?:ab?c)*+", "aca").span(), (0, 2))
2404+
self.assertEqual(re.match("(?>(?:ab?c)?)", "a").span(), (0, 0))
2405+
self.assertEqual(re.match("(?:ab?c)?+", "a").span(), (0, 0))
2406+
self.assertEqual(re.match("(?>(?:ab?c){1,3})", "aca").span(), (0, 2))
2407+
self.assertEqual(re.match("(?:ab?c){1,3}+", "aca").span(), (0, 2))
2408+
23992409
@unittest.skipIf(multiprocessing is None, 'test requires multiprocessing')
24002410
def test_regression_gh94675(self):
24012411
pattern = re.compile(r'(?<=[({}])(((//[^\n]*)?[\n])([\000-\040])*)*'
@@ -2492,6 +2502,7 @@ def test_atomic_group(self):
24922502
17: SUCCESS
24932503
''')
24942504

2505+
@unittest.expectedFailure # gh-106052
24952506
def test_possesive_repeat_one(self):
24962507
self.assertEqual(get_debug_out(r'a?+'), '''\
24972508
POSSESSIVE_REPEAT 0 1
@@ -2504,6 +2515,7 @@ def test_possesive_repeat_one(self):
25042515
12: SUCCESS
25052516
''')
25062517

2518+
@unittest.expectedFailure # gh-106052
25072519
def test_possesive_repeat(self):
25082520
self.assertEqual(get_debug_out(r'(?:ab)?+'), '''\
25092521
POSSESSIVE_REPEAT 0 1
Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,2 @@
1+
:mod:`re` module: fix the matching of possessive quantifiers in the case of
2+
a subpattern containing backtracking.

0 commit comments

Comments
 (0)