Be careful with trailing punctuation (fixes mozilla#30)

bleach/__init__.py

@@ -65,6 +65,8 @@
 
 proto_re = re.compile(r'^[\w-]+:/{0,3}')
 
+punct_re = re.compile(r'([\.,]+)$')
+
 email_re = re.compile(
     r"(?<!//)"
     r"(([-!#$%&'*+/=?^_`{}|~0-9A-Z]+(\.[-!#$%&'*+/=?^_`{}|~0-9A-Z]+)*"  # dot-atom
@@ -163,14 +165,19 @@ def email_repl(match):
 
     def link_repl(match):
         url = match.group(0)
+        end = u''
+        m = re.search(punct_re, url)
+        if m:
+            end = m.group(0)
+            url = url[0:m.start()]
         if re.search(proto_re, url):
             href = url
         else:
-            href = u''.join(['http://', url])
+            href = u''.join([u'http://', url])
 
-        repl = u'<a href="%s"%s>%s</a>'
+        repl = u'<a href="%s"%s>%s</a>%s'
 
-        return repl % (filter_url(href), rel, filter_text(url))
+        return repl % (filter_url(href), rel, filter_text(url), end)
 
     linkify_nodes(forest)
 

bleach/tests/test_links.py

@@ -211,5 +211,24 @@ def test_libgl():
 
 def test_end_of_sentence():
     """example.com. should match."""
-    eq_('<a href="http://example.com" rel="nofollow">example.com</a>.',
-        linkify('example.com.'))
+    out = u'<a href="http://%s" rel="nofollow">%s</a>%s'
+    in_ = u'%s%s'
+
+    def check(u, p):
+        eq_(out % (u, u, p), linkify(in_ % (u, p)))
+
+    tests = (
+        ('example.com', '.'),
+        ('example.com', '...'),
+        ('ex.com/foo', '.'),
+        ('ex.com/foo', '....'),
+    )
+
+    for u, p in tests:
+        yield check, u, p
+
+
+def test_end_of_clause():
+    """example.com/foo, shouldn't include the ,"""
+    eq_('<a href="http://ex.com/foo" rel="nofollow">ex.com/foo</a>, bar',
+        linkify('ex.com/foo, bar'))