[utils] sanitize_path: Fix some incorrect behavior (#11923)

Authored by: Grub4K
This commit is contained in:
Simon Sawicki 2025-01-25 22:32:00 +01:00 committed by GitHub
parent 61ae5dc34a
commit fc12e724a3
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
2 changed files with 34 additions and 12 deletions

View File

@ -249,17 +249,36 @@ def _test_sanitize_path(self):
self.assertEqual(sanitize_path('abc/def...'), 'abc\\def..#')
self.assertEqual(sanitize_path('abc.../def'), 'abc..#\\def')
self.assertEqual(sanitize_path('abc.../def...'), 'abc..#\\def..#')
self.assertEqual(sanitize_path('../abc'), '..\\abc')
self.assertEqual(sanitize_path('../../abc'), '..\\..\\abc')
self.assertEqual(sanitize_path('./abc'), 'abc')
self.assertEqual(sanitize_path('./../abc'), '..\\abc')
self.assertEqual(sanitize_path('\\abc'), '\\abc')
self.assertEqual(sanitize_path('C:abc'), 'C:abc')
self.assertEqual(sanitize_path('C:abc\\..\\'), 'C:..')
self.assertEqual(sanitize_path('C:\\abc:%(title)s.%(ext)s'), 'C:\\abc#%(title)s.%(ext)s')
# Check with nt._path_normpath if available
try:
import nt
nt_path_normpath = getattr(nt, '_path_normpath', None)
except Exception:
nt_path_normpath = None
for test, expected in [
('C:\\', 'C:\\'),
('../abc', '..\\abc'),
('../../abc', '..\\..\\abc'),
('./abc', 'abc'),
('./../abc', '..\\abc'),
('\\abc', '\\abc'),
('C:abc', 'C:abc'),
('C:abc\\..\\', 'C:'),
('C:abc\\..\\def\\..\\..\\', 'C:..'),
('C:\\abc\\xyz///..\\def\\', 'C:\\abc\\def'),
('abc/../', '.'),
('./abc/../', '.'),
]:
result = sanitize_path(test)
assert result == expected, f'{test} was incorrectly resolved'
assert result == sanitize_path(result), f'{test} changed after sanitizing again'
if nt_path_normpath:
assert result == nt_path_normpath(test), f'{test} does not match nt._path_normpath'
def test_sanitize_url(self):
self.assertEqual(sanitize_url('//foo.bar'), 'http://foo.bar')
self.assertEqual(sanitize_url('httpss://foo.bar'), 'https://foo.bar')

View File

@ -685,7 +685,8 @@ def _sanitize_path_parts(parts):
elif part == '..':
if sanitized_parts and sanitized_parts[-1] != '..':
sanitized_parts.pop()
sanitized_parts.append('..')
else:
sanitized_parts.append('..')
continue
# Replace invalid segments with `#`
# - trailing dots and spaces (`asdf...` => `asdf..#`)
@ -702,7 +703,8 @@ def sanitize_path(s, force=False):
if not force:
return s
root = '/' if s.startswith('/') else ''
return root + '/'.join(_sanitize_path_parts(s.split('/')))
path = '/'.join(_sanitize_path_parts(s.split('/')))
return root + path if root or path else '.'
normed = s.replace('/', '\\')
@ -721,7 +723,8 @@ def sanitize_path(s, force=False):
root = '\\' if normed[:1] == '\\' else ''
parts = normed.split('\\')
return root + '\\'.join(_sanitize_path_parts(parts))
path = '\\'.join(_sanitize_path_parts(parts))
return root + path if root or path else '.'
def sanitize_url(url, *, scheme='http'):