CodeCow/zulip
1
0
Fork 0
mirror of https://github.com/zulip/zulip.git synced 2026-07-12 21:04:41 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Handle chars at end of link better

Browse Source 
(imported from commit 53842bc17bbb9cfb555738ee6b72291d7ce61d67)
This commit is contained in:
Leo Franchi 2013-02-06 09:33:24 -05:00
parent b0b304e58b
commit c637b9cf44
2 changed files with 12 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
zephyr/lib/bugdown/__init__.py
View File

@ -161,8 +161,8 @@ class Bugdown(markdown.Extension):
# we match at maximum one set of matching parens in a url. We could extend this
# to match two parenthetical groups, at the cost of more regex complexity.
tlds = '|'.join(['co.uk', 'com', 'co', 'biz', 'gd', 'org', 'net', 'ly', 'edu', 'mil',
'gov', 'info', 'me', 'it', '.ca', 'tv', 'fm',])
link_regex = r"\b(?P<url>[^\s]+\.(%s)(?:/[^\s()\"]*|([^\s()\"]*\([^\s()\"]+\)[^\s()\"]*))?)(?=[\s:;\?\),\.\'\"]|\Z)" % (tlds,)
'gov', 'info', 'me', 'it', '.ca', 'tv', 'fm', 'io', 'gl'])
link_regex = r"\b(?P<url>[^\s]+\.(%s)(?:/[^\s()\":]*?|([^\s()\":]*\([^\s()\":]*\)[^\s()\":]*))?)(?=([:;\?\),\.\'\"]\Z|[:;\?\),\.\'\"]\s|\Z|\s))" % (tlds,)
md.inlinePatterns.add('autolink', AutoLink(link_regex), '>link')
md.preprocessors.add('hanging_ulists',

12
zephyr/tests.py
View File

@ -1754,6 +1754,11 @@ int x = 3
('(Something like http://foo.com/blah_blah)', "<p>(Something like %s)</p>", 'http://foo.com/blah_blah'),
('"is.gd/foo/"', "<p>\"%s\"</p>", 'is.gd/foo/'),
('end with a quote www.google.com"', "<p>end with a quote %s\"</p>", 'www.google.com'),
('http://www.guardian.co.uk/foo/bar', "<p>%s</p>", 'http://www.guardian.co.uk/foo/bar'),
('from http://supervisord.org/running.html:', "<p>from %s:</p>", 'http://supervisord.org/running.html'),
('http://raven.io', "<p>%s</p>", 'http://raven.io'),
('at https://humbughq.com/api. Check it!', "<p>at %s. Check it!</p>", 'https://humbughq.com/api'),
('goo.gl/abc', "<p>%s</p>", 'goo.gl/abc'),
# XSS Sanitization
('javascript:alert(\'hi\');.com', "<p>%s</p>", None), # None marks xss
@ -1765,11 +1770,14 @@ int x = 3
('http://leo@foo.com/my/file', "<p>%s</p>", None),
('http://example.com/something?with,commas,in,url, but not at end',
"<p>%s but not at end</p>", 'http://example.com/something?with,commas,in,url,'),
"<p>%s, but not at end</p>", 'http://example.com/something?with,commas,in,url'),
(' some text https://www.google.com/baz_(match)?with=foo&bar=baz with extras',
"<p>some text %s with extras</p>", 'https://www.google.com/baz_(match)?with=foo&amp;bar=baz'),
('hash it http://foo.com/blah_(wikipedia)_blah#cite-1',
"<p>hash it %s</p>", 'http://foo.com/blah_(wikipedia)_blah#cite-1'),]
"<p>hash it %s</p>", 'http://foo.com/blah_(wikipedia)_blah#cite-1'),
('http://technet.microsoft.com/en-us/library/Cc751099.rk20_25_big(l=en-us).gif',
"<p>%s</p>",
'http://technet.microsoft.com/en-us/library/Cc751099.rk20_25_big(l=en-us).gif')]
for inline_url, reference, url in conversions:
try: