[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / tmz.py

diff --git a/youtube_dl/extractor/tmz.py b/youtube_dl/extractor/tmz.py
index 7dbe68b5c228ea6d3fa20d835a60ecf38b820816..419f9d92eea375e4aeabc0c39f736955b818c921 100644 (file)
--- a/youtube_dl/extractor/tmz.py
+++ b/youtube_dl/extractor/tmz.py
@@ -5,43 +5,42 @@ from .common import InfoExtractor
 
 
 class TMZIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tmz\.com/videos/(?P<id>[^/]+)/?'
-    _TEST = {
+    _VALID_URL = r'https?://(?:www\.)?tmz\.com/videos/(?P<id>[^/?#]+)'
+    _TESTS = [{
         'url': 'http://www.tmz.com/videos/0_okj015ty/',
-        'md5': '791204e3bf790b1426cb2db0706184c0',
+        'md5': '4d22a51ef205b6c06395d8394f72d560',
         'info_dict': {
             'id': '0_okj015ty',
-            'url': 'http://tmz.vo.llnwd.net/o28/2014-03/13/0_okj015ty_0_rt8ro3si_2.mp4',
             'ext': 'mp4',
             'title': 'Kim Kardashian\'s Boobs Unlock a Mystery!',
             'description': 'Did Kim Kardasain try to one-up Khloe by one-upping Kylie???  Or is she just showing off her amazing boobs?',
-            'thumbnail': r're:http://cdnbakmi\.kaltura\.com/.*thumbnail.*',
+            'timestamp': 1394747163,
+            'uploader_id': 'batchUser',
+            'upload_date': '20140313',
         }
-    }
+    }, {
+        'url': 'http://www.tmz.com/videos/0-cegprt2p/',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
-        webpage = self._download_webpage(url, video_id)
-
-        return {
-            'id': video_id,
-            'url': self._html_search_meta('VideoURL', webpage, fatal=True),
-            'title': self._og_search_title(webpage),
-            'description': self._og_search_description(webpage),
-            'thumbnail': self._html_search_meta('ThumbURL', webpage),
-        }
+        video_id = self._match_id(url).replace('-', '_')
+        return self.url_result('kaltura:591531:%s' % video_id, 'Kaltura', video_id)
 
 
 class TMZArticleIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?tmz\.com/\d{4}/\d{2}/\d{2}/(?P<id>[^/]+)/?'
     _TEST = {
         'url': 'http://www.tmz.com/2015/04/19/bobby-brown-bobbi-kristina-awake-video-concert',
-        'md5': 'e482a414a38db73087450e3a6ce69d00',
+        'md5': '3316ff838ae5bb7f642537825e1e90d2',
         'info_dict': {
             'id': '0_6snoelag',
-            'ext': 'mp4',
+            'ext': 'mov',
             'title': 'Bobby Brown Tells Crowd ... Bobbi Kristina is Awake',
             'description': 'Bobby Brown stunned his audience during a concert Saturday night, when he told the crowd, "Bobbi is awake.  She\'s watching me."',
+            'timestamp': 1429467813,
+            'upload_date': '20150419',
+            'uploader_id': 'batchUser',
         }
     }
 
@@ -49,12 +48,9 @@ class TMZArticleIE(InfoExtractor):
         video_id = self._match_id(url)
 
         webpage = self._download_webpage(url, video_id)
-        embedded_video_info_str = self._html_search_regex(
-            r'tmzVideoEmbedV2\("([^)]+)"\);', webpage, 'embedded video info')
-
-        embedded_video_info = self._parse_json(
-            embedded_video_info_str, video_id,
-            transform_source=lambda s: s.replace('\\', ''))
+        embedded_video_info = self._parse_json(self._html_search_regex(
+            r'tmzVideoEmbed\(({.+?})\);', webpage, 'embedded video info'),
+            video_id)
 
         return self.url_result(
             'http://www.tmz.com/videos/%s/' % embedded_video_info['id'])