Merge pull request #488 from mouse-reeve/url-regex

Better matching for links in statuses

bookwyrm/outgoing.py

@@ -293,13 +293,16 @@ def find_mentions(content):
         yield (match.group(), mention_user)
 
 
-def to_markdown(content):
+def format_links(content):
-    ''' catch links and convert to markdown '''
+    ''' detect and format links '''
-    content = re.sub(
+    return re.sub(
-        r'([^(href=")])(https?:\/\/([A-Za-z\.\-_\/]+' \
+        r'([^(href=")]|^)(https?:\/\/(%s([\w\.\-_\/])*))' % regex.domain,
-            r'\.[A-Za-z]{2,}[A-Za-z\.\-_\/]+))',
         r'\g<1><a href="\g<2>">\g<3></a>',
         content)
+
+def to_markdown(content):
+    ''' catch links and convert to markdown '''
+    content = format_links(content)
     content = markdown(content)
     # sanitize resulting html
     sanitizer = InputHtmlParser()

bookwyrm/tests/test_outgoing.py

@@ -447,3 +447,59 @@ class Outgoing(TestCase):
         self.assertEqual(reply.user, user)
         self.assertTrue(self.remote_user in reply.mention_users.all())
         self.assertTrue(self.local_user in reply.mention_users.all())
+
+    def test_find_mentions(self):
+        ''' detect and look up @ mentions of users '''
+        user = models.User.objects.create_user(
+            'nutria@%s' % DOMAIN, 'nutria@nutria.com', 'password',
+            local=True, localname='nutria')
+        self.assertEqual(user.username, 'nutria@%s' % DOMAIN)
+
+        self.assertEqual(
+            list(outgoing.find_mentions('@nutria'))[0],
+            ('@nutria', user)
+        )
+        self.assertEqual(
+            list(outgoing.find_mentions('leading text @nutria'))[0],
+            ('@nutria', user)
+        )
+        self.assertEqual(
+            list(outgoing.find_mentions('leading @nutria trailing text'))[0],
+            ('@nutria', user)
+        )
+        self.assertEqual(
+            list(outgoing.find_mentions('@rat@example.com'))[0],
+            ('@rat@example.com', self.remote_user)
+        )
+
+        multiple = list(outgoing.find_mentions('@nutria and @rat@example.com'))
+        self.assertEqual(multiple[0], ('@nutria', user))
+        self.assertEqual(multiple[1], ('@rat@example.com', self.remote_user))
+
+        with patch('bookwyrm.outgoing.handle_remote_webfinger') as rw:
+            rw.return_value = self.local_user
+            self.assertEqual(
+                list(outgoing.find_mentions('@beep@beep.com'))[0],
+                ('@beep@beep.com', self.local_user)
+            )
+        with patch('bookwyrm.outgoing.handle_remote_webfinger') as rw:
+            rw.return_value = None
+            self.assertEqual(list(outgoing.find_mentions('@beep@beep.com')), [])
+
+        self.assertEqual(
+            list(outgoing.find_mentions('@nutria@%s' % DOMAIN))[0],
+            ('@nutria@%s' % DOMAIN, user)
+        )
+
+    def test_format_links(self):
+        ''' find and format urls into a tags '''
+        url = 'http://www.fish.com/'
+        self.assertEqual(
+            outgoing.format_links(url),
+            '<a href="%s">www.fish.com/</a>' % url)
+        url = 'https://archive.org/details/dli.granth.72113/page/n25/mode/2up'
+        self.assertEqual(
+            outgoing.format_links(url),
+            '<a href="%s">' \
+                'archive.org/details/dli.granth.72113/page/n25/mode/2up</a>' \
+                % url)

bookwyrm/utils/regex.py

@@ -1,10 +1,10 @@
 ''' defining regexes for regularly used concepts '''
 
-domain = r'[a-z-A-Z0-9_\-]+\.[a-z]+'
+domain = r'[\w_\-\.]+\.[a-z]{2,}'
-localname = r'@?[a-zA-Z_\-\.0-9]+'
+localname = r'@?[a-zA-Z_\-\.0-9]+\b'
 strict_localname = r'@[a-zA-Z_\-\.0-9]+'
 username = r'%s(@%s)?' % (localname, domain)
-strict_username = r'%s(@%s)?' % (strict_localname, domain)
+strict_username = r'\B%s(@%s)?\b' % (strict_localname, domain)
-full_username = r'%s@%s' % (localname, domain)
+full_username = r'\B%s@%s\b' % (localname, domain)
 # should match (BookWyrm/1.0.0; or (BookWyrm/99.1.2;
 bookwyrm_user_agent = r'\(BookWyrm/[0-9]+\.[0-9]+\.[0-9]+;'