[cnet] improve extraction

- relex data json regex
- extract the platform metadata once
- extract hds formats
- extract duration
- extract thumbnail

youtube_dl/extractor/cnet.py

@@ -1,33 +1,33 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
-
-from .common import InfoExtractor
 from .theplatform import ThePlatformIE
+from ..utils import int_or_none
 
 
-class CNETIE(InfoExtractor):
+class CNETIE(ThePlatformIE):
     _VALID_URL = r'https?://(?:www\.)?cnet\.com/videos/(?P<id>[^/]+)/'
     _TESTS = [{
         'url': 'http://www.cnet.com/videos/hands-on-with-microsofts-windows-8-1-update/',
         'info_dict': {
             'id': '56f4ea68-bd21-4852-b08c-4de5b8354c60',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 'Hands-on with Microsoft Windows 8.1 Update',
             'description': 'The new update to the Windows 8 OS brings improved performance for mouse and keyboard users.',
             'uploader_id': '6085384d-619e-11e3-b231-14feb5ca9861',
             'uploader': 'Sarah Mitroff',
+            'duration': 70,
         },
     }, {
         'url': 'http://www.cnet.com/videos/whiny-pothole-tweets-at-local-government-when-hit-by-cars-tomorrow-daily-187/',
         'info_dict': {
             'id': '56527b93-d25d-44e3-b738-f989ce2e49ba',
-            'ext': 'mp4',
+            'ext': 'flv',
+            'title': 'Whiny potholes tweet at local government when hit by cars (Tomorrow Daily 187)',
             'description': 'Khail and Ashley wonder what other civic woes can be solved by self-tweeting objects, investigate a new kind of VR camera and watch an origami robot self-assemble, walk, climb, dig and dissolve. #TDPothole',
             'uploader_id': 'b163284d-6b73-44fc-b3e6-3da66c392d40',
             'uploader': 'Ashley Esqueda',
-            'title': 'Whiny potholes tweet at local government when hit by cars (Tomorrow Daily 187)',
+            'duration': 1482,
         },
     }]
 
@@ -36,10 +36,10 @@ def _real_extract(self, url):
         webpage = self._download_webpage(url, display_id)
 
         data_json = self._html_search_regex(
-            r"<div class=\"videoPlayer\"\s+.*?data-cnet-video-uvp-options='([^']+)'",
+            r"data-cnet-video(?:-uvp)?-options='([^']+)'",
             webpage, 'data json')
-        data = json.loads(data_json)
+        data = self._parse_json(data_json, display_id)
-        vdata = data['videos'][0]
+        vdata = data.get('video') or data['videos'][0]
 
         video_id = vdata['id']
         title = vdata['title']
@@ -52,19 +52,22 @@ def _real_extract(self, url):
             uploader_id = None
 
         mpx_account = data['config']['uvpConfig']['default']['mpx_account']
-        tp = ThePlatformIE(self._downloader)
+
+        metadata = self.get_metadata('%s/%s' % (mpx_account, list(vdata['files'].values())[0]), video_id)
+        description = vdata.get('description') or metadata.get('description')
+        duration = int_or_none(vdata.get('duration')) or metadata.get('duration')
+
         formats = []
         subtitles = {}
-        description = vdata.get('description')
-
         for (fkey, vid) in vdata['files'].items():
             if fkey == 'hls_phone' and 'hls_tablet' in vdata['files']:
                 continue
-            result = tp.extract(('http://link.theplatform.com/s/%s/%s' % (mpx_account, vid)))
+            release_url = 'http://link.theplatform.com/s/%s/%s?format=SMIL&mbr=true' % (mpx_account, vid)
-            formats.extend(result['formats'])
+            if fkey == 'hds':
-            subtitles = self._merge_subtitles(subtitles, result['subtitles'])
+                release_url += '&manifest=f4m'
-            description = description or result.get('description')
+            tp_formats, tp_subtitles = self._extract_theplatform_smil(release_url, video_id, 'Downloading %s SMIL data' % fkey)
-
+            formats.extend(tp_formats)
+            subtitles = self._merge_subtitles(subtitles, tp_subtitles)
         self._sort_formats(formats)
 
         return {
@@ -72,6 +75,8 @@ def _real_extract(self, url):
             'display_id': display_id,
             'title': title,
             'description': description,
+            'thumbnail': metadata.get('thumbnail'),
+            'duration': duration,
             'uploader': uploader,
             'uploader_id': uploader_id,
             'subtitles': subtitles,