commit 5cb2d36c82abf3b753910afe3013b274e31a247a
parent fcca0d53a8fa47614a39a433a3da7d1ab1d88ed9
Author: Sergey M․ <dstftw@gmail.com>
Date: Tue, 14 Feb 2017 22:56:39 +0700
[ceskatelevize] Extract DASH formats (closes #12119, closes #12133)
Diffstat:
1 file changed, 88 insertions(+), 64 deletions(-)
diff --git a/youtube_dl/extractor/ceskatelevize.py b/youtube_dl/extractor/ceskatelevize.py
@@ -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)
