[cbs] improve extraction(closes #6321)

youtube_dl/extractor/cbs.py

@@ -1,21 +1,24 @@
 from __future__ import unicode_literals
 
-from .common import InfoExtractor
+from .theplatform import ThePlatformIE
 from ..utils import (
-    sanitized_Request,
+    xpath_text,
-    smuggle_url,
+    xpath_element,
+    int_or_none,
+    ExtractorError,
+    find_xpath_attr,
 )
 
 
-class CBSIE(InfoExtractor):
+class CBSIE(ThePlatformIE):
     _VALID_URL = r'https?://(?:www\.)?(?:cbs\.com/shows/[^/]+/(?:video|artist)|colbertlateshow\.com/(?:video|podcasts))/[^/]+/(?P<id>[^/]+)'
 
     _TESTS = [{
         'url': 'http://www.cbs.com/shows/garth-brooks/video/_u7W953k6la293J7EPTd9oHkSPs6Xn6_/connect-chat-feat-garth-brooks/',
         'info_dict': {
-            'id': '4JUVEwq3wUT7',
+            'id': '_u7W953k6la293J7EPTd9oHkSPs6Xn6_',
             'display_id': 'connect-chat-feat-garth-brooks',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Connect Chat feat. Garth Brooks',
             'description': 'Connect with country music singer Garth Brooks, as he chats with fans on Wednesday November 27, 2013. Be sure to tune in to Garth Brooks: Live from Las Vegas, Friday November 29, at 9/8c on CBS!',
             'duration': 1495,
@@ -47,22 +50,55 @@ class CBSIE(InfoExtractor):
         'url': 'http://www.colbertlateshow.com/podcasts/dYSwjqPs_X1tvbV_P2FcPWRa_qT6akTC/in-the-bad-room-with-stephen/',
         'only_matching': True,
     }]
+    TP_RELEASE_URL_TEMPLATE = 'http://link.theplatform.com/s/dJ5BDC/%s?manifest=m3u&mbr=true'
+
+    def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'):
+        closed_caption_e = find_xpath_attr(smil, self._xpath_ns('.//param', namespace), 'name', 'ClosedCaptionURL')
+        return {
+            'en': [{
+                'ext': 'ttml',
+                'url': closed_caption_e.attrib['value'],
+            }]
+        } if closed_caption_e is not None and closed_caption_e.attrib.get('value') else []
 
     def _real_extract(self, url):
         display_id = self._match_id(url)
-        request = sanitized_Request(url)
+        webpage = self._download_webpage(url, display_id)
-        # Android UA is served with higher quality (720p) streams (see
+        content_id = self._search_regex(
-        # https://github.com/rg3/youtube-dl/issues/7490)
+            [r"video\.settings\.content_id\s*=\s*'([^']+)';", r"cbsplayer\.contentId\s*=\s*'([^']+)';"],
-        request.add_header('User-Agent', 'Mozilla/5.0 (Linux; Android 4.4; Nexus 5)')
+            webpage, 'content id')
-        webpage = self._download_webpage(request, display_id)
+        items_data = self._download_xml(
-        real_id = self._search_regex(
+            'http://can.cbs.com/thunder/player/videoPlayerService.php',
-            [r"video\.settings\.pid\s*=\s*'([^']+)';", r"cbsplayer\.pid\s*=\s*'([^']+)';"],
+            content_id, query={'partner': 'cbs', 'contentId': content_id})
-            webpage, 'real video ID')
+        video_data = xpath_element(items_data, './/item')
-        return {
+        title = xpath_text(video_data, 'videoTitle', 'title', True)
-            '_type': 'url_transparent',
+
-            'ie_key': 'ThePlatform',
+        subtitles = {}
-            'url': smuggle_url(
+        formats = []
-                'http://link.theplatform.com/s/dJ5BDC/%s?mbr=true&manifest=m3u' % real_id,
+        for item in items_data.findall('.//item'):
-                {'force_smil_url': True}),
+            pid = xpath_text(item, 'pid')
+            if not pid:
+                continue
+            try:
+                tp_formats, tp_subtitles = self._extract_theplatform_smil(
+                    self.TP_RELEASE_URL_TEMPLATE % pid, content_id, 'Downloading %s SMIL data' % pid)
+            except ExtractorError:
+                continue
+            formats.extend(tp_formats)
+            subtitles = self._merge_subtitles(subtitles, tp_subtitles)
+        self._sort_formats(formats)
+
+        info = self.get_metadata('dJ5BDC/media/guid/2198311517/%s' % content_id, content_id)
+        info.update({
+            'id': content_id,
             'display_id': display_id,
-        }
+            'title': title,
+            'series': xpath_text(video_data, 'seriesTitle'),
+            'season_number': int_or_none(xpath_text(video_data, 'seasonNumber')),
+            'episode_number': int_or_none(xpath_text(video_data, 'episodeNumber')),
+            'duration': int_or_none(xpath_text(video_data, 'videoLength'), 1000),
+            'thumbnail': xpath_text(video_data, 'previewImageURL'),
+            'formats': formats,
+            'subtitles': subtitles,
+        })
+        return info