improve is_hashtag

tests/test_models.py

@@ -156,10 +156,13 @@ def test_is_username(self):
  self.failIf(is_username(user))
 
  def test_is_hashtag(self):
- valid = ['#turses', '#cúrcuma', '#4n_4Wfu1_US3RN4M3']
+ valid = ['#turses', '#cúrcuma', '#4n_4Wfu1_H45hT46']
  for hashtag in valid:
  self.failUnless(is_hashtag(hashtag))
- # TODO: test invalid hashtags
+
+ invalid = ['s#turses', '#']
+ for hashtag in invalid:
+ self.failIf(is_hashtag(hashtag))
 
  def test_sanitize_username(self):
  dirty_and_clean = [

turses/models.py

@@ -20,6 +20,8 @@
 
 username_regex = re.compile(r'[A-Za-z0-9_]+')
 
+hashtag_regex = re.compile(r'#.+')
+
 prepend_at = lambda username: '@%s' % username
 
 def is_DM(status):
@@ -92,8 +94,15 @@ def is_username(username):
  return match.start() == 0 and match.end() == len(username)
  return False 
 
-def is_hashtag(string):
- return len(string) > 1 and string.startswith('#')
+def is_hashtag(hashtag):
+ """
+ Return `True` if `hashtag` is a valid Twitter hashtag, `False`
+ otherwise.
+ """
+ match = hashtag_regex.match(hashtag)
+ if match:
+ return match.start() == 0 and match.end() == len(hashtag)
+ return False 
 
 def sanitize_username(username):
  """