Broadcast mentions and retweets

README.rst

@@ -36,6 +36,17 @@ Usage::
 
 If you need different HTML output just subclass and override the ``format_*`` methods.
 
+"broadcast" mentions and old-style retweets are now available as well:
+
+    >>> from ttp import ttp
+    >>> p = ttp.Parser()
+    >>> result = p.parse(".@eadmundo has added broadcast mentions!")
+    >>> result.broadcast
+    'eadmundo'
+    >>> result = p.parse("RT @eadmundo, also old-style retweets")
+    >>> result.retweet
+    'eadmundo'
+
 You can also ask for the span tags to be returned for each entity::
 
     >>> p = ttp.Parser(include_spans=True)

ttp/tests.py

@@ -509,6 +509,50 @@ def test_username_non_reply(self):
         self.assertEqual(result.users, [u'username'])
         self.assertEqual(result.reply, None)
 
+    # Broadcast mentions
+    def test_username_broadcast_mention_at_start(self):
+        result = self.parser.parse(u'.@username')
+        self.assertEqual(result.html, u'.<a href="http://twitter.com/username">@username</a>')
+        self.assertEqual(result.users, [u'username'])
+        self.assertEqual(result.broadcast, u'username')
+
+    def test_username_broadcast_mention_in_middle(self):
+        result = self.parser.parse(u'something .@username')
+        self.assertEqual(result.html, u'something .<a href="http://twitter.com/username">@username</a>')
+        self.assertEqual(result.users, [u'username'])
+        self.assertEqual(result.broadcast, u'username')
+
+    # Retweets
+    def test_username_old_style_retweet(self):
+        result = self.parser.parse(u'retweet RT @username something')
+        self.assertEqual(result.html, u'retweet RT <a href="http://twitter.com/username">@username</a> something')
+        self.assertEqual(result.retweet, u'username')
+
+    def test_username_old_style_retweet_at_beginning(self):
+        result = self.parser.parse(u'RT @username something')
+        self.assertEqual(result.html, u'RT <a href="http://twitter.com/username">@username</a> something')
+        self.assertEqual(result.retweet, u'username')
+
+    def test_username_quoted_retweet(self):
+        result = self.parser.parse(u'retweet "@username something"')
+        self.assertEqual(result.html, u'retweet "<a href="http://twitter.com/username">@username</a> something"')
+        self.assertEqual(result.retweet, u'username')
+
+    def test_username_curly_quoted_retweet(self):
+        result = self.parser.parse(u'retweet “@username something”')
+        self.assertEqual(result.html, u'retweet “<a href="http://twitter.com/username">@username</a> something”')
+        self.assertEqual(result.retweet, u'username')
+
+    def test_username_quoted_retweet_at_beginning(self):
+        result = self.parser.parse(u'"@username something"')
+        self.assertEqual(result.html, u'"<a href="http://twitter.com/username">@username</a> something"')
+        self.assertEqual(result.retweet, u'username')
+
+    def test_username_curly_quoted_retweet_at_beginning(self):
+        result = self.parser.parse(u'“@username something”')
+        self.assertEqual(result.html, u'“<a href="http://twitter.com/username">@username</a> something”')
+        self.assertEqual(result.retweet, u'username')
+
     # List tests ---------------------------------------------------------------
     # --------------------------------------------------------------------------
     def test_list_preceeded(self):

ttp/ttp.py

@@ -30,6 +30,7 @@
 AT_SIGNS = ur'[@\uff20]'
 UTF_CHARS = ur'a-z0-9_\u00c0-\u00d6\u00d8-\u00f6\u00f8-\u00ff'
 SPACES = ur'[\u0020\u00A0\u1680\u180E\u2002-\u202F\u205F\u2060\u3000]'
+QUOTES = ur'[\u0022\u201C]'
 
 # Lists
 LIST_PRE_CHARS = ur'([^a-z0-9_]|^)'
@@ -41,6 +42,8 @@
 USERNAME_REGEX = re.compile(ur'\B' + AT_SIGNS + LIST_END_CHARS, re.IGNORECASE)
 REPLY_REGEX = re.compile(ur'^(?:' + SPACES + ur')*' + AT_SIGNS
                          + ur'([a-z0-9_]{1,20}).*', re.IGNORECASE)
+BROADCAST_REGEX = re.compile('.' + AT_SIGNS + LIST_END_CHARS, re.IGNORECASE)
+RETWEET_REGEX = re.compile('(?:RT' + SPACES + '|' + QUOTES + ')' + AT_SIGNS + LIST_END_CHARS, re.IGNORECASE)
 
 # Hashtags
 HASHTAG_EXP = ur'(^|[^0-9A-Z&/]+)(#|\uff03)([0-9A-Z_]*[A-Z_]+[%s]*)' % UTF_CHARS
@@ -86,6 +89,15 @@ class ParseResult(object):
         Note: It's generally better to rely on the Tweet JSON/XML in order to
         find out if it's a reply or not.
 
+    - broadcast
+        A string containing the username this tweet was a broadcast mention to
+        (e.g. preceeded by a dot to ensure that mention is broadcast even to
+            non-followers of that username)
+
+    - retweet
+        A string containing the username this tweet was a retweet of (as indicated by
+            either old-style RT or encased in quotes)
+
     - lists
         A list containing all the valid lists in the Tweet.
         Each list item is a tuple in the format (username, listname).
@@ -100,11 +112,13 @@ class ParseResult(object):
 
     '''
 
-    def __init__(self, urls, users, reply, lists, tags, html):
+    def __init__(self, urls, users, reply, broadcast, retweet, lists, tags, html):
         self.urls = urls if urls else []
         self.users = users if users else []
         self.lists = lists if lists else []
         self.reply = reply if reply else None
+        self.broadcast = broadcast if broadcast else None
+        self.retweet = retweet if retweet else None
         self.tags = tags if tags else []
         self.html = html
 
@@ -127,8 +141,14 @@ def parse(self, text, html=True):
         reply = REPLY_REGEX.match(text)
         reply = reply.groups(0)[0] if reply is not None else None
 
+        broadcast = BROADCAST_REGEX.search(text)
+        broadcast = broadcast.groups(0)[0] if broadcast is not None else None
+
+        retweet = RETWEET_REGEX.search(text)
+        retweet = retweet.groups(0)[0] if retweet is not None else None
+
         parsed_html = self._html(text) if html else self._text(text)
-        return ParseResult(self._urls, self._users, reply,
+        return ParseResult(self._urls, self._users, reply, broadcast, retweet,
                            self._lists, self._tags, parsed_html)
 
     def _text(self, text):