fix calc_expected_status_length for urls not at beginning on string

Author: jeremylow

tests/test_twitter_utils.py

@@ -5,6 +5,7 @@
 import twitter
 
 from twitter.twitter_utils import (
+    calc_expected_status_length,
     parse_media_file
 )
 
@@ -58,3 +59,18 @@ def test_utils_error_checking(self):
         self.assertRaises(
             twitter.TwitterError,
             lambda: twitter.twitter_utils.enf_type('test', int, 'hi'))
+
+    def test_calc_expected_status_length(self):
+        status = 'hi a tweet there'
+        len_status = calc_expected_status_length(status)
+        self.assertEqual(len_status, 16)
+
+    def test_calc_expected_status_length_with_url(self):
+        status = 'hi a tweet there example.com'
+        len_status = calc_expected_status_length(status)
+        self.assertEqual(len_status, 40)
+
+    def test_calc_expected_status_length_with_url_and_extra_spaces(self):
+        status = 'hi a tweet          there               example.com'
+        len_status = calc_expected_status_length(status)
+        self.assertEqual(len_status, 63)

twitter/twitter_utils.py

@@ -161,12 +161,13 @@ def calc_expected_status_length(status, short_url_length=23):
         Expected length of the status message as an integer.
 
     """
-    replaced_chars = 0
-    status_length = len(status)
-    match = re.findall(URL_REGEXP, status)
-    if len(match) >= 1:
-        replaced_chars = len(''.join(match))
-    status_length = status_length - replaced_chars + (short_url_length * len(match))
+    status_length = 0
+    for word in re.split(r'\s', status):
+        if is_url(word):
+            status_length += short_url_length
+        else:
+            status_length += len(word)
+    status_length += len(re.findall(r'\s', status))
     return status_length