From c637b9cf44cf267e7883dd4e3ed529ea7b17a004 Mon Sep 17 00:00:00 2001
From: Leo Franchi <leo@humbughq.com>
Date: Wed, 6 Feb 2013 09:33:24 -0500
Subject: [PATCH] Handle chars at end of link better

(imported from commit 53842bc17bbb9cfb555738ee6b72291d7ce61d67)
---
 zephyr/lib/bugdown/__init__.py |  4 ++--
 zephyr/tests.py                | 12 ++++++++++--
 2 files changed, 12 insertions(+), 4 deletions(-)

diff --git a/zephyr/lib/bugdown/__init__.py b/zephyr/lib/bugdown/__init__.py
index 5539e03a5f..cd333a3b33 100644
--- a/zephyr/lib/bugdown/__init__.py
+++ b/zephyr/lib/bugdown/__init__.py
@@ -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',
diff --git a/zephyr/tests.py b/zephyr/tests.py
index 72b4a3d0e3..30c2e8c25a 100644
--- a/zephyr/tests.py
+++ b/zephyr/tests.py
@@ -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: