Skip to content

Commit d89a5f6

Browse files
sethmlarsonZeroIntensity
sethmlarson
and
ZeroIntensity
authored
gh-105704: Disallow square brackets ([ and ]) in domain names for parsed URLs (#129418)
* gh-105704: Disallow square brackets ( and ) in domain names for parsed URLs * Use Sphinx references Co-authored-by: Peter Bierma <zintensitydev@gmail.com> * Add mismatched bracket test cases, fix news format * Add more test coverage for ports --------- Co-authored-by: Peter Bierma <zintensitydev@gmail.com>
1 parent 54f74b8 commit d89a5f6

File tree

3 files changed

+58
-3
lines changed

3 files changed

+58
-3
lines changed

‎Lib/test/test_urlparse.py‎

Lines changed: 36 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1412,16 +1412,51 @@ def test_invalid_bracketed_hosts(self):
14121412
self.assertRaises(ValueError, urllib.parse.urlsplit, 'Scheme://user@[0439:23af::2309::fae7:1234]/Path?Query')
14131413
self.assertRaises(ValueError, urllib.parse.urlsplit, 'Scheme://user@[0439:23af:2309::fae7:1234:2342:438e:192.0.2.146]/Path?Query')
14141414
self.assertRaises(ValueError, urllib.parse.urlsplit, 'Scheme://user@]v6a.ip[/Path')
1415+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[v6a.ip]')
1416+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[v6a.ip].suffix')
1417+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[v6a.ip]/')
1418+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[v6a.ip].suffix/')
1419+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[v6a.ip]?')
1420+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[v6a.ip].suffix?')
1421+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]')
1422+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix')
1423+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]/')
1424+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix/')
1425+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]?')
1426+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix?')
1427+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]:a')
1428+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix:a')
1429+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]:a1')
1430+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix:a1')
1431+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]:1a')
1432+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix:1a')
1433+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]:')
1434+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[::1].suffix:/')
1435+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[::1]:?')
1436+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://user@prefix.[v6a.ip]')
1437+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://user@[v6a.ip].suffix')
1438+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://[v6a.ip')
1439+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://v6a.ip]')
1440+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://]v6a.ip[')
1441+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://]v6a.ip')
1442+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://v6a.ip[')
1443+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix.[v6a.ip')
1444+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://v6a.ip].suffix')
1445+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix]v6a.ip[suffix')
1446+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://prefix]v6a.ip')
1447+
self.assertRaises(ValueError, urllib.parse.urlsplit, 'scheme://v6a.ip[suffix')
14151448

14161449
def test_splitting_bracketed_hosts(self):
1417-
p1 = urllib.parse.urlsplit('scheme://user@[v6a.ip]/path?query')
1450+
p1 = urllib.parse.urlsplit('scheme://user@[v6a.ip]:1234/path?query')
14181451
self.assertEqual(p1.hostname, 'v6a.ip')
14191452
self.assertEqual(p1.username, 'user')
14201453
self.assertEqual(p1.path, '/path')
1454+
self.assertEqual(p1.port, 1234)
14211455
p2 = urllib.parse.urlsplit('scheme://user@[0439:23af:2309::fae7%test]/path?query')
14221456
self.assertEqual(p2.hostname, '0439:23af:2309::fae7%test')
14231457
self.assertEqual(p2.username, 'user')
14241458
self.assertEqual(p2.path, '/path')
1459+
self.assertIs(p2.port, None)
14251460
p3 = urllib.parse.urlsplit('scheme://user@[0439:23af:2309::fae7:1234:192.0.2.146%test]/path?query')
14261461
self.assertEqual(p3.hostname, '0439:23af:2309::fae7:1234:192.0.2.146%test')
14271462
self.assertEqual(p3.username, 'user')

‎Lib/urllib/parse.py‎

Lines changed: 18 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -439,6 +439,23 @@ def _checknetloc(netloc):
439439
raise ValueError("netloc '" + netloc + "' contains invalid " +
440440
"characters under NFKC normalization")
441441

442+
def _check_bracketed_netloc(netloc):
443+
# Note that this function must mirror the splitting
444+
# done in NetlocResultMixins._hostinfo().
445+
hostname_and_port = netloc.rpartition('@')[2]
446+
before_bracket, have_open_br, bracketed = hostname_and_port.partition('[')
447+
if have_open_br:
448+
# No data is allowed before a bracket.
449+
if before_bracket:
450+
raise ValueError("Invalid IPv6 URL")
451+
hostname, _, port = bracketed.partition(']')
452+
# No data is allowed after the bracket but before the port delimiter.
453+
if port and not port.startswith(":"):
454+
raise ValueError("Invalid IPv6 URL")
455+
else:
456+
hostname, _, port = hostname_and_port.partition(':')
457+
_check_bracketed_host(hostname)
458+
442459
# Valid bracketed hosts are defined in
443460
# https://www.rfc-editor.org/rfc/rfc3986#page-49 and https://url.spec.whatwg.org/
444461
def _check_bracketed_host(hostname):
@@ -505,8 +522,7 @@ def _urlsplit(url, scheme=None, allow_fragments=True):
505522
(']' in netloc and '[' not in netloc)):
506523
raise ValueError("Invalid IPv6 URL")
507524
if '[' in netloc and ']' in netloc:
508-
bracketed_host = netloc.partition('[')[2].partition(']')[0]
509-
_check_bracketed_host(bracketed_host)
525+
_check_bracketed_netloc(netloc)
510526
if allow_fragments and '#' in url:
511527
url, fragment = url.split('#', 1)
512528
if '?' in url:

‎Misc/NEWS.d/next/Security/2025-01-28-14-08-03.gh-issue-105704.EnhHxu.rst‎

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,4 @@
1+
When using :func:`urllib.parse.urlsplit` and :func:`urllib.parse.urlparse` host
2+
parsing would not reject domain names containing square brackets (``[`` and
3+
``]``). Square brackets are only valid for IPv6 and IPvFuture hosts according to
4+
`RFC 3986 Section 3.2.2 <https://www.rfc-editor.org/rfc/rfc3986#section-3.2.2>`__.

0 commit comments

Comments
 (0)