]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/cnet.py
Merge tag 'upstream/2016.02.22'
[youtubedl] / youtube_dl / extractor / cnet.py
index a94f42571746f21cc70ba2527e661952caca1c6f..5c3908f72b2f94c5557aaeaaa2e19ec78716c0f0 100644 (file)
@@ -1,75 +1,84 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
-import re
+from .theplatform import ThePlatformIE
+from ..utils import int_or_none
 
-from .common import InfoExtractor
-from ..utils import (
-    ExtractorError,
-    int_or_none,
-)
 
-
-class CNETIE(InfoExtractor):
+class CNETIE(ThePlatformIE):
     _VALID_URL = r'https?://(?:www\.)?cnet\.com/videos/(?P<id>[^/]+)/'
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.cnet.com/videos/hands-on-with-microsofts-windows-8-1-update/',
-        'md5': '041233212a0d06b179c87cbcca1577b8',
         '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.',
-            'thumbnail': 're:^http://.*/flmswindows8.jpg$',
-            'uploader_id': 'sarah.mitroff@cbsinteractive.com',
+            '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': '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',
+            'duration': 1482,
+        },
+    }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        display_id = mobj.group('id')
-
+        display_id = self._match_id(url)
         webpage = self._download_webpage(url, display_id)
+
         data_json = self._html_search_regex(
-            r"<div class=\"cnetVideoPlayer\"\s+.*?data-cnet-video-options='([^']+)'",
+            r"data-cnet-video(?:-uvp)?-options='([^']+)'",
             webpage, 'data json')
-        data = json.loads(data_json)
-        vdata = data['video']
-        if not vdata:
-            vdata = data['videos'][0]
-        if not vdata:
-            raise ExtractorError('Cannot find video data')
+        data = self._parse_json(data_json, display_id)
+        vdata = data.get('video') or data['videos'][0]
 
         video_id = vdata['id']
-        title = vdata['headline']
-        description = vdata.get('dek')
-        thumbnail = vdata.get('image', {}).get('path')
+        title = vdata['title']
         author = vdata.get('author')
         if author:
             uploader = '%s %s' % (author['firstName'], author['lastName'])
-            uploader_id = author.get('email')
+            uploader_id = author.get('id')
         else:
             uploader = None
             uploader_id = None
 
-        formats = [{
-            'format_id': '%s-%s-%s' % (
-                f['type'], f['format'],
-                int_or_none(f.get('bitrate'), 1000, default='')),
-            'url': f['uri'],
-            'tbr': int_or_none(f.get('bitrate'), 1000),
-        } for f in vdata['files']['data']]
+        mpx_account = data['config']['uvpConfig']['default']['mpx_account']
+
+        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 = {}
+        for (fkey, vid) in vdata['files'].items():
+            if fkey == 'hls_phone' and 'hls_tablet' in vdata['files']:
+                continue
+            release_url = 'http://link.theplatform.com/s/%s/%s?format=SMIL&mbr=true' % (mpx_account, vid)
+            if fkey == 'hds':
+                release_url += '&manifest=f4m'
+            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 {
             'id': video_id,
             'display_id': display_id,
             'title': title,
-            'formats': formats,
             'description': description,
+            'thumbnail': metadata.get('thumbnail'),
+            'duration': duration,
             'uploader': uploader,
             'uploader_id': uploader_id,
-            'thumbnail': thumbnail,
+            'subtitles': subtitles,
+            'formats': formats,
         }