+ asset_type = asset.get('asset_type') or asset.get('assetType')
+ if asset_type != 'Video':
+ raise ExtractorError(
+ 'Lecture %s is not a video' % lecture_id, expected=True)
+
+ stream_url = asset.get('stream_url') or asset.get('streamUrl')
+ if stream_url:
+ youtube_url = self._search_regex(
+ r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
+ if youtube_url:
+ return self.url_result(youtube_url, 'Youtube')
+
+ video_id = compat_str(asset['id'])
+ thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')
+ duration = float_or_none(asset.get('data', {}).get('duration'))
+
+ subtitles = {}
+ automatic_captions = {}
+
+ formats = []
+
+ def extract_output_format(src, f_id):
+ return {
+ 'url': src.get('url'),
+ 'format_id': '%sp' % (src.get('height') or f_id),
+ 'width': int_or_none(src.get('width')),
+ 'height': int_or_none(src.get('height')),
+ 'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
+ 'vcodec': src.get('video_codec'),
+ 'fps': int_or_none(src.get('frame_rate')),
+ 'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
+ 'acodec': src.get('audio_codec'),
+ 'asr': int_or_none(src.get('audio_sample_rate')),
+ 'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
+ 'filesize': int_or_none(src.get('file_size_in_bytes')),
+ }
+
+ outputs = asset.get('data', {}).get('outputs')
+ if not isinstance(outputs, dict):
+ outputs = {}
+
+ def add_output_format_meta(f, key):
+ output = outputs.get(key)
+ if isinstance(output, dict):
+ output_format = extract_output_format(output, key)
+ output_format.update(f)
+ return output_format
+ return f
+
+ def extract_formats(source_list):
+ if not isinstance(source_list, list):
+ return
+ for source in source_list:
+ video_url = url_or_none(source.get('file') or source.get('src'))
+ if not video_url:
+ continue
+ if source.get('type') == 'application/x-mpegURL' or determine_ext(video_url) == 'm3u8':
+ formats.extend(self._extract_m3u8_formats(
+ video_url, video_id, 'mp4', entry_protocol='m3u8_native',
+ m3u8_id='hls', fatal=False))
+ continue
+ format_id = source.get('label')
+ f = {
+ 'url': video_url,
+ 'format_id': '%sp' % format_id,
+ 'height': int_or_none(format_id),
+ }
+ if format_id:
+ # Some videos contain additional metadata (e.g.
+ # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
+ f = add_output_format_meta(f, format_id)
+ formats.append(f)
+
+ def extract_subtitles(track_list):
+ if not isinstance(track_list, list):
+ return
+ for track in track_list:
+ if not isinstance(track, dict):
+ continue
+ if track.get('kind') != 'captions':
+ continue
+ src = url_or_none(track.get('src'))
+ if not src:
+ continue
+ lang = track.get('language') or track.get(
+ 'srclang') or track.get('label')
+ sub_dict = automatic_captions if track.get(
+ 'autogenerated') is True else subtitles
+ sub_dict.setdefault(lang, []).append({
+ 'url': src,
+ })
+
+ for url_kind in ('download', 'stream'):
+ urls = asset.get('%s_urls' % url_kind)
+ if isinstance(urls, dict):
+ extract_formats(urls.get('Video'))
+
+ captions = asset.get('captions')
+ if isinstance(captions, list):
+ for cc in captions:
+ if not isinstance(cc, dict):
+ continue
+ cc_url = url_or_none(cc.get('url'))
+ if not cc_url:
+ continue
+ lang = try_get(cc, lambda x: x['locale']['locale'], compat_str)
+ sub_dict = (automatic_captions if cc.get('source') == 'auto'
+ else subtitles)
+ sub_dict.setdefault(lang or 'en', []).append({
+ 'url': cc_url,
+ })
+
+ view_html = lecture.get('view_html')
+ if view_html:
+ view_html_urls = set()
+ for source in re.findall(r'<source[^>]+>', view_html):
+ attributes = extract_attributes(source)
+ src = attributes.get('src')
+ if not src:
+ continue
+ res = attributes.get('data-res')
+ height = int_or_none(res)
+ if src in view_html_urls:
+ continue
+ view_html_urls.add(src)
+ if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
+ m3u8_formats = self._extract_m3u8_formats(
+ src, video_id, 'mp4', entry_protocol='m3u8_native',
+ m3u8_id='hls', fatal=False)
+ for f in m3u8_formats:
+ m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
+ if m:
+ if not f.get('height'):
+ f['height'] = int(m.group('height'))
+ if not f.get('tbr'):
+ f['tbr'] = int(m.group('tbr'))
+ formats.extend(m3u8_formats)
+ else:
+ formats.append(add_output_format_meta({
+ 'url': src,
+ 'format_id': '%dp' % height if height else None,
+ 'height': height,
+ }, res))
+
+ # react rendition since 2017.04.15 (see
+ # https://github.com/ytdl-org/youtube-dl/issues/12744)
+ data = self._parse_json(
+ self._search_regex(
+ r'videojs-setup-data=(["\'])(?P<data>{.+?})\1', view_html,
+ 'setup data', default='{}', group='data'), video_id,
+ transform_source=unescapeHTML, fatal=False)
+ if data and isinstance(data, dict):
+ extract_formats(data.get('sources'))
+ if not duration:
+ duration = int_or_none(data.get('duration'))
+ extract_subtitles(data.get('tracks'))
+
+ if not subtitles and not automatic_captions:
+ text_tracks = self._parse_json(
+ self._search_regex(
+ r'text-tracks=(["\'])(?P<data>\[.+?\])\1', view_html,
+ 'text tracks', default='{}', group='data'), video_id,
+ transform_source=lambda s: js_to_json(unescapeHTML(s)),
+ fatal=False)
+ extract_subtitles(text_tracks)
+
+ if not formats and outputs:
+ for format_id, output in outputs.items():
+ f = extract_output_format(output, format_id)
+ if f.get('url'):
+ formats.append(f)
+
+ self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))
Rapsys Git / youtubedl / blobdiff