[ceskatelevize] Extract DASH formats (closes #12119, closes #12133)

youtube_dl/extractor/ceskatelevize.py

@@ -13,6 +13,7 @@ from ..utils import (
     float_or_none,
     sanitized_Request,
     urlencode_postdata,
+    USER_AGENTS,
 )
 
 
@@ -114,6 +115,9 @@ class CeskaTelevizeIE(InfoExtractor):
             'requestSource': 'iVysilani',
         }
 
+        entries = []
+
+        for user_agent in (None, USER_AGENTS['Safari']):
             req = sanitized_Request(
                 'http://www.ceskatelevize.cz/ivysilani/ajax/get-client-playlist',
                 data=urlencode_postdata(data))
@@ -121,10 +125,14 @@ class CeskaTelevizeIE(InfoExtractor):
             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')
+            if user_agent:
+                req.add_header('User-Agent', user_agent)
             req.add_header('Referer', url)
 
-        playlistpage = self._download_json(req, playlist_id)
+            playlistpage = self._download_json(req, playlist_id, fatal=False)
+
+            if not playlistpage:
+                continue
 
             playlist_url = playlistpage['url']
             if playlist_url == 'error_region':
@@ -136,19 +144,32 @@ class CeskaTelevizeIE(InfoExtractor):
             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 = 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)
 
-        entries = []
-        for item in playlist:
+            for num, item in enumerate(playlist):
                 is_live = item.get('type') == 'LIVE'
                 formats = []
-            for format_id, stream_url in item['streamUrls'].items():
+                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))
-            self._sort_formats(formats)
+                    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']
@@ -180,6 +201,9 @@ class CeskaTelevizeIE(InfoExtractor):
                     'is_live': is_live,
                 })
 
+        for e in entries:
+            self._sort_formats(e['formats'])
+
         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 
     def _get_subtitles(self, episode_id, subs):