mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-08-01 07:33:08 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 4

[3.13] gh-126505: Fix bugs in compiling case-insensitive character classes (GH-126557) (GH-126689)

Browse source 
* upper-case non-BMP character was ignored
* the ASCII flag was ignored when matching a character range whose
  upper bound is beyond the BMP region
(cherry picked from commit 819830f34a)

Co-authored-by: Serhiy Storchaka <storchaka@gmail.com>
This commit is contained in:
Miss Islington (bot) 2024-11-11 17:54:57 +01:00 • committed by GitHub
parent fc10908f7d
commit 7db6d4282f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 73 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

23
Lib/re/_compiler.py
View file

@ -248,11 +248,11 @@ def _optimize_charset(charset, iscased=None, fixup=None, fixes=None):
while True: while True:
try: try:
if op is LITERAL: if op is LITERAL:
if fixup: if fixup: # IGNORECASE and not LOCALE
lo = fixup(av) av = fixup(av)
charmap[lo] = 1 charmap[av] = 1
if fixes and lo in fixes: if fixes and av in fixes:
for k in fixes[lo]: for k in fixes[av]:
charmap[k] = 1 charmap[k] = 1
if not hascased and iscased(av): if not hascased and iscased(av):
hascased = True hascased = True
@ -260,7 +260,7 @@ def _optimize_charset(charset, iscased=None, fixup=None, fixes=None):
charmap[av] = 1 charmap[av] = 1
elif op is RANGE: elif op is RANGE:
r = range(av[0], av[1]+1) r = range(av[0], av[1]+1)
if fixup: if fixup: # IGNORECASE and not LOCALE
if fixes: if fixes:
for i in map(fixup, r): for i in map(fixup, r):
charmap[i] = 1 charmap[i] = 1
@ -287,8 +287,7 @@ def _optimize_charset(charset, iscased=None, fixup=None, fixes=None):
# Character set contains non-BMP character codes. # Character set contains non-BMP character codes.
# For range, all BMP characters in the range are already # For range, all BMP characters in the range are already
# proceeded. # proceeded.
if fixup: if fixup: # IGNORECASE and not LOCALE
hascased = True
# For now, IN_UNI_IGNORE+LITERAL and # For now, IN_UNI_IGNORE+LITERAL and
# IN_UNI_IGNORE+RANGE_UNI_IGNORE work for all non-BMP # IN_UNI_IGNORE+RANGE_UNI_IGNORE work for all non-BMP
# characters, because two characters (at least one of # characters, because two characters (at least one of
@ -299,7 +298,13 @@ def _optimize_charset(charset, iscased=None, fixup=None, fixes=None):
# Also, both c.lower() and c.lower().upper() are single # Also, both c.lower() and c.lower().upper() are single
# characters for every non-BMP character. # characters for every non-BMP character.
if op is RANGE: if op is RANGE:
op = RANGE_UNI_IGNORE if fixes: # not ASCII
op = RANGE_UNI_IGNORE
hascased = True
else:
assert op is LITERAL
if not hascased and iscased(av):
hascased = True
tail.append((op, av)) tail.append((op, av))
break break

55
Lib/test/test_re.py
View file

@ -1137,6 +1137,39 @@ class ReTests(unittest.TestCase):
self.assertTrue(re.match(br'[19a]', b'a', re.I)) self.assertTrue(re.match(br'[19a]', b'a', re.I))
self.assertTrue(re.match(br'[19a]', b'A', re.I)) self.assertTrue(re.match(br'[19a]', b'A', re.I))
self.assertTrue(re.match(br'[19A]', b'a', re.I)) self.assertTrue(re.match(br'[19A]', b'a', re.I))
self.assertTrue(re.match(r'[19\xc7]', '\xc7', re.I))
self.assertTrue(re.match(r'[19\xc7]', '\xe7', re.I))
self.assertTrue(re.match(r'[19\xe7]', '\xc7', re.I))
self.assertTrue(re.match(r'[19\xe7]', '\xe7', re.I))
self.assertTrue(re.match(r'[19\u0400]', '\u0400', re.I))
self.assertTrue(re.match(r'[19\u0400]', '\u0450', re.I))
self.assertTrue(re.match(r'[19\u0450]', '\u0400', re.I))
self.assertTrue(re.match(r'[19\u0450]', '\u0450', re.I))
self.assertTrue(re.match(r'[19\U00010400]', '\U00010400', re.I))
self.assertTrue(re.match(r'[19\U00010400]', '\U00010428', re.I))
self.assertTrue(re.match(r'[19\U00010428]', '\U00010400', re.I))
self.assertTrue(re.match(r'[19\U00010428]', '\U00010428', re.I))
self.assertTrue(re.match(br'[19A]', b'A', re.I))
self.assertTrue(re.match(br'[19a]', b'a', re.I))
self.assertTrue(re.match(br'[19a]', b'A', re.I))
self.assertTrue(re.match(br'[19A]', b'a', re.I))
self.assertTrue(re.match(r'[19A]', 'A', re.I|re.A))
self.assertTrue(re.match(r'[19a]', 'a', re.I|re.A))
self.assertTrue(re.match(r'[19a]', 'A', re.I|re.A))
self.assertTrue(re.match(r'[19A]', 'a', re.I|re.A))
self.assertTrue(re.match(r'[19\xc7]', '\xc7', re.I|re.A))
self.assertIsNone(re.match(r'[19\xc7]', '\xe7', re.I|re.A))
self.assertIsNone(re.match(r'[19\xe7]', '\xc7', re.I|re.A))
self.assertTrue(re.match(r'[19\xe7]', '\xe7', re.I|re.A))
self.assertTrue(re.match(r'[19\u0400]', '\u0400', re.I|re.A))
self.assertIsNone(re.match(r'[19\u0400]', '\u0450', re.I|re.A))
self.assertIsNone(re.match(r'[19\u0450]', '\u0400', re.I|re.A))
self.assertTrue(re.match(r'[19\u0450]', '\u0450', re.I|re.A))
self.assertTrue(re.match(r'[19\U00010400]', '\U00010400', re.I|re.A))
self.assertIsNone(re.match(r'[19\U00010400]', '\U00010428', re.I|re.A))
self.assertIsNone(re.match(r'[19\U00010428]', '\U00010400', re.I|re.A))
self.assertTrue(re.match(r'[19\U00010428]', '\U00010428', re.I|re.A))
# Two different characters have the same lowercase. # Two different characters have the same lowercase.
assert 'K'.lower() == '\u212a'.lower() == 'k' # 'K' assert 'K'.lower() == '\u212a'.lower() == 'k' # 'K'
@ -1173,8 +1206,10 @@ class ReTests(unittest.TestCase):
self.assertTrue(re.match(br'[9-a]', b'_', re.I)) self.assertTrue(re.match(br'[9-a]', b'_', re.I))
self.assertIsNone(re.match(br'[9-A]', b'_', re.I)) self.assertIsNone(re.match(br'[9-A]', b'_', re.I))
self.assertTrue(re.match(r'[\xc0-\xde]', '\xd7', re.I)) self.assertTrue(re.match(r'[\xc0-\xde]', '\xd7', re.I))
self.assertTrue(re.match(r'[\xc0-\xde]', '\xe7', re.I))
self.assertIsNone(re.match(r'[\xc0-\xde]', '\xf7', re.I)) self.assertIsNone(re.match(r'[\xc0-\xde]', '\xf7', re.I))
self.assertTrue(re.match(r'[\xe0-\xfe]', '\xf7', re.I)) self.assertTrue(re.match(r'[\xe0-\xfe]', '\xf7', re.I))
self.assertTrue(re.match(r'[\xe0-\xfe]', '\xc7', re.I))
self.assertIsNone(re.match(r'[\xe0-\xfe]', '\xd7', re.I)) self.assertIsNone(re.match(r'[\xe0-\xfe]', '\xd7', re.I))
self.assertTrue(re.match(r'[\u0430-\u045f]', '\u0450', re.I)) self.assertTrue(re.match(r'[\u0430-\u045f]', '\u0450', re.I))
self.assertTrue(re.match(r'[\u0430-\u045f]', '\u0400', re.I)) self.assertTrue(re.match(r'[\u0430-\u045f]', '\u0400', re.I))
@ -1185,6 +1220,26 @@ class ReTests(unittest.TestCase):
self.assertTrue(re.match(r'[\U00010400-\U00010427]', '\U00010428', re.I)) self.assertTrue(re.match(r'[\U00010400-\U00010427]', '\U00010428', re.I))
self.assertTrue(re.match(r'[\U00010400-\U00010427]', '\U00010400', re.I)) self.assertTrue(re.match(r'[\U00010400-\U00010427]', '\U00010400', re.I))
self.assertTrue(re.match(r'[\xc0-\xde]', '\xd7', re.I|re.A))
self.assertIsNone(re.match(r'[\xc0-\xde]', '\xe7', re.I|re.A))
self.assertTrue(re.match(r'[\xe0-\xfe]', '\xf7', re.I|re.A))
self.assertIsNone(re.match(r'[\xe0-\xfe]', '\xc7', re.I|re.A))
self.assertTrue(re.match(r'[\u0430-\u045f]', '\u0450', re.I|re.A))
self.assertIsNone(re.match(r'[\u0430-\u045f]', '\u0400', re.I|re.A))
self.assertIsNone(re.match(r'[\u0400-\u042f]', '\u0450', re.I|re.A))
self.assertTrue(re.match(r'[\u0400-\u042f]', '\u0400', re.I|re.A))
self.assertTrue(re.match(r'[\U00010428-\U0001044f]', '\U00010428', re.I|re.A))
self.assertIsNone(re.match(r'[\U00010428-\U0001044f]', '\U00010400', re.I|re.A))
self.assertIsNone(re.match(r'[\U00010400-\U00010427]', '\U00010428', re.I|re.A))
self.assertTrue(re.match(r'[\U00010400-\U00010427]', '\U00010400', re.I|re.A))
self.assertTrue(re.match(r'[N-\x7f]', 'A', re.I|re.A))
self.assertTrue(re.match(r'[n-\x7f]', 'Z', re.I|re.A))
self.assertTrue(re.match(r'[N-\uffff]', 'A', re.I|re.A))
self.assertTrue(re.match(r'[n-\uffff]', 'Z', re.I|re.A))
self.assertTrue(re.match(r'[N-\U00010000]', 'A', re.I|re.A))
self.assertTrue(re.match(r'[n-\U00010000]', 'Z', re.I|re.A))
# Two different characters have the same lowercase. # Two different characters have the same lowercase.
assert 'K'.lower() == '\u212a'.lower() == 'k' # 'K' assert 'K'.lower() == '\u212a'.lower() == 'k' # 'K'
self.assertTrue(re.match(r'[J-M]', '\u212a', re.I)) self.assertTrue(re.match(r'[J-M]', '\u212a', re.I))

4
Misc/NEWS.d/next/Library/2024-11-07-22-41-47.gh-issue-126505.iztYE1.rst Normal file
View file

@ -0,0 +1,4 @@
Fix bugs in compiling case-insensitive :mod:`regular expressions <re>` with
character classes containing non-BMP characters: upper-case non-BMP
character did was ignored and the ASCII flag was ignored when
matching a character range whose upper bound is beyond the BMP region.