[tmz] Fix and improve extraction (closes #24603, closes #24687, closes #28211)

1 files changed, 76 insertions, 21 deletions

diff --git a/youtube_dl/extractor/tmz.py b/youtube_dl/extractor/tmz.py
index 419f9d92e..3d1bf75ff 100644
--- a/youtube_dl/extractor/tmz.py
+++ b/youtube_dl/extractor/tmz.py
@@ -2,55 +2,110 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from .jwplatform import JWPlatformIE
+from .kaltura import KalturaIE
+from ..utils import (
+    int_or_none,
+    unified_timestamp,
+)
 
 
 class TMZIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tmz\.com/videos/(?P<id>[^/?#]+)'
+    _VALID_URL = r'https?://(?:www\.)?tmz\.com/videos/(?P<id>[^/?#&]+)'
     _TESTS = [{
-        'url': 'http://www.tmz.com/videos/0_okj015ty/',
-        'md5': '4d22a51ef205b6c06395d8394f72d560',
+        'url': 'http://www.tmz.com/videos/0-cegprt2p/',
+        'md5': '31f9223e20eef55954973359afa61a20',
         'info_dict': {
-            'id': '0_okj015ty',
+            'id': 'P6YjLBLk',
             '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?',
-            'timestamp': 1394747163,
-            'uploader_id': 'batchUser',
-            'upload_date': '20140313',
-        }
+            'title': "No Charges Against Hillary Clinton? Harvey Says It Ain't Over Yet",
+            'description': 'md5:b714359fc18607715ebccbd2da8ff488',
+            'timestamp': 1467831837,
+            'upload_date': '20160706',
+        },
+        'add_ie': [JWPlatformIE.ie_key()],
     }, {
-        'url': 'http://www.tmz.com/videos/0-cegprt2p/',
+        'url': 'http://www.tmz.com/videos/0_okj015ty/',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.tmz.com/videos/071119-chris-morgan-women-4590005-0-zcsejvcr/',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.tmz.com/videos/2021-02-19-021921-floyd-mayweather-1043872/',
         'only_matching': True,
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url).replace('-', '_')
-        return self.url_result('kaltura:591531:%s' % video_id, 'Kaltura', video_id)
+
+        webpage = self._download_webpage(url, video_id, fatal=False)
+        if webpage:
+            tmz_video_id = self._search_regex(
+                r'nodeRef\s*:\s*["\']tmz:video:([\da-fA-F]{8}-[\da-fA-F]{4}-[\da-fA-F]{4}-[\da-fA-F]{4}-[\da-fA-F]{12})',
+                webpage, 'video id', default=None)
+            video = self._download_json(
+                'https://www.tmz.com/_/video/%s' % tmz_video_id, video_id,
+                fatal=False)
+            if video:
+                message = video['message']
+                info = {
+                    '_type': 'url_transparent',
+                    'title': message.get('title'),
+                    'description': message.get('description'),
+                    'timestamp': unified_timestamp(message.get('published_at')),
+                    'duration': int_or_none(message.get('duration')),
+                }
+                jwplatform_id = message.get('jwplayer_media_id')
+                if jwplatform_id:
+                    info.update({
+                        'url': 'jwplatform:%s' % jwplatform_id,
+                        'ie_key': JWPlatformIE.ie_key(),
+                    })
+                else:
+                    kaltura_entry_id = message.get('kaltura_entry_id') or video_id
+                    kaltura_partner_id = message.get('kaltura_partner_id') or '591531'
+                    info.update({
+                        'url': 'kaltura:%s:%s' % (kaltura_partner_id, kaltura_entry_id),
+                        'ie_key': KalturaIE.ie_key(),
+                    })
+                return info
+
+        return self.url_result(
+            'kaltura:591531:%s' % video_id, KalturaIE.ie_key(), video_id)
 
 
 class TMZArticleIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?tmz\.com/\d{4}/\d{2}/\d{2}/(?P<id>[^/]+)/?'
+    _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': '3316ff838ae5bb7f642537825e1e90d2',
         'info_dict': {
-            'id': '0_6snoelag',
-            'ext': 'mov',
+            'id': 'PAKZa97W',
+            'ext': 'mp4',
             '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,
+            'timestamp': 1429466400,
             'upload_date': '20150419',
-            'uploader_id': 'batchUser',
-        }
+        },
+        'params': {
+            'skip_download': True,
+        },
+        'add_ie': [JWPlatformIE.ie_key()],
     }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
         webpage = self._download_webpage(url, video_id)
+
+        tmz_url = self._search_regex(
+            r'clickLink\s*\(\s*["\'](?P<url>%s)' % TMZIE._VALID_URL, webpage,
+            'video id', default=None, group='url')
+        if tmz_url:
+            return self.url_result(tmz_url, ie=TMZIE.ie_key())
+
         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'])
+            'http://www.tmz.com/videos/%s/' % embedded_video_info['id'],
+            ie=TMZIE.ie_key())