[ceskatelevize] Extract DASH formats (closes #12119, closes #12133)
authorSergey M․ <dstftw@gmail.com>
Tue, 14 Feb 2017 15:56:39 +0000 (22:56 +0700)
committerSergey M․ <dstftw@gmail.com>
Tue, 14 Feb 2017 15:57:38 +0000 (22:57 +0700)
youtube_dl/extractor/ceskatelevize.py

index 0f1453b994a4897c4c86ab9ff1ad5dd16292deb5..e08bf264cb4b576cae3e5cdf5c172d67bd28a252 100644 (file)
@@ -13,6 +13,7 @@ from ..utils import (
     float_or_none,
     sanitized_Request,
     urlencode_postdata,
+    USER_AGENTS,
 )
 
 
@@ -114,71 +115,94 @@ class CeskaTelevizeIE(InfoExtractor):
             'requestSource': 'iVysilani',
         }
 
-        req = sanitized_Request(
-            'http://www.ceskatelevize.cz/ivysilani/ajax/get-client-playlist',
-            data=urlencode_postdata(data))
-
-        req.add_header('Content-type', 'application/x-www-form-urlencoded')
-        req.add_header('x-addr', '127.0.0.1')
-        req.add_header('X-Requested-With', 'XMLHttpRequest')
-        req.add_header('User-agent', 'Mozilla/5.0')
-        req.add_header('Referer', url)
-
-        playlistpage = self._download_json(req, playlist_id)
-
-        playlist_url = playlistpage['url']
-        if playlist_url == 'error_region':
-            raise ExtractorError(NOT_AVAILABLE_STRING, expected=True)
-
-        req = sanitized_Request(compat_urllib_parse_unquote(playlist_url))
-        req.add_header('Referer', url)
-
-        playlist_title = self._og_search_title(webpage, default=None)
-        playlist_description = self._og_search_description(webpage, default=None)
-
-        playlist = self._download_json(req, playlist_id)['playlist']
-        playlist_len = len(playlist)
-
         entries = []
-        for item in playlist:
-            is_live = item.get('type') == 'LIVE'
-            formats = []
-            for format_id, stream_url in item['streamUrls'].items():
-                formats.extend(self._extract_m3u8_formats(
-                    stream_url, playlist_id, 'mp4',
-                    entry_protocol='m3u8' if is_live else 'm3u8_native',
-                    fatal=False))
-            self._sort_formats(formats)
-
-            item_id = item.get('id') or item['assetId']
-            title = item['title']
-
-            duration = float_or_none(item.get('duration'))
-            thumbnail = item.get('previewImageUrl')
-
-            subtitles = {}
-            if item.get('type') == 'VOD':
-                subs = item.get('subtitles')
-                if subs:
-                    subtitles = self.extract_subtitles(episode_id, subs)
-
-            if playlist_len == 1:
-                final_title = playlist_title or title
-                if is_live:
-                    final_title = self._live_title(final_title)
-            else:
-                final_title = '%s (%s)' % (playlist_title, title)
-
-            entries.append({
-                'id': item_id,
-                'title': final_title,
-                'description': playlist_description if playlist_len == 1 else None,
-                'thumbnail': thumbnail,
-                'duration': duration,
-                'formats': formats,
-                'subtitles': subtitles,
-                'is_live': is_live,
-            })
+
+        for user_agent in (None, USER_AGENTS['Safari']):
+            req = sanitized_Request(
+                'http://www.ceskatelevize.cz/ivysilani/ajax/get-client-playlist',
+                data=urlencode_postdata(data))
+
+            req.add_header('Content-type', 'application/x-www-form-urlencoded')
+            req.add_header('x-addr', '127.0.0.1')
+            req.add_header('X-Requested-With', 'XMLHttpRequest')
+            if user_agent:
+                req.add_header('User-Agent', user_agent)
+            req.add_header('Referer', url)
+
+            playlistpage = self._download_json(req, playlist_id, fatal=False)
+
+            if not playlistpage:
+                continue
+
+            playlist_url = playlistpage['url']
+            if playlist_url == 'error_region':
+                raise ExtractorError(NOT_AVAILABLE_STRING, expected=True)
+
+            req = sanitized_Request(compat_urllib_parse_unquote(playlist_url))
+            req.add_header('Referer', url)
+
+            playlist_title = self._og_search_title(webpage, default=None)
+            playlist_description = self._og_search_description(webpage, default=None)
+
+            playlist = self._download_json(req, playlist_id, fatal=False)
+            if not playlist:
+                continue
+
+            playlist = playlist.get('playlist')
+            if not isinstance(playlist, list):
+                continue
+
+            playlist_len = len(playlist)
+
+            for num, item in enumerate(playlist):
+                is_live = item.get('type') == 'LIVE'
+                formats = []
+                for format_id, stream_url in item.get('streamUrls', {}).items():
+                    if 'playerType=flash' in stream_url:
+                        formats.extend(self._extract_m3u8_formats(
+                            stream_url, playlist_id, 'mp4',
+                            entry_protocol='m3u8' if is_live else 'm3u8_native',
+                            fatal=False))
+                    else:
+                        formats.extend(self._extract_mpd_formats(
+                            stream_url, playlist_id, fatal=False))
+
+                if user_agent and len(entries) == playlist_len:
+                    entries[num]['formats'].extend(formats)
+                    continue
+
+                item_id = item.get('id') or item['assetId']
+                title = item['title']
+
+                duration = float_or_none(item.get('duration'))
+                thumbnail = item.get('previewImageUrl')
+
+                subtitles = {}
+                if item.get('type') == 'VOD':
+                    subs = item.get('subtitles')
+                    if subs:
+                        subtitles = self.extract_subtitles(episode_id, subs)
+
+                if playlist_len == 1:
+                    final_title = playlist_title or title
+                    if is_live:
+                        final_title = self._live_title(final_title)
+                else:
+                    final_title = '%s (%s)' % (playlist_title, title)
+
+                entries.append({
+                    'id': item_id,
+                    'title': final_title,
+                    'description': playlist_description if playlist_len == 1 else None,
+                    'thumbnail': thumbnail,
+                    'duration': duration,
+                    'formats': formats,
+                    'subtitles': subtitles,
+                    'is_live': is_live,
+                })
+
+        for e in entries:
+            self._sort_formats(e['formats'])
 
         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)