yt_dlp/extractor/ted.py

   1 import itertools
   2 import re
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     int_or_none,
   7     parse_duration,
   8     str_to_int,
   9     try_get,
  10     unified_strdate,
  11     url_or_none,
  12 )
  13
  14
  15 class TedBaseIE(InfoExtractor):
  16     _VALID_URL_BASE = r'https?://www\.ted\.com/(?:{type})(?:/lang/[^/#?]+)?/(?P<id>[\w-]+)'
  17
  18     def _parse_playlist(self, playlist):
  19         for entry in try_get(playlist, lambda x: x['videos']['nodes'], list):
  20             if entry.get('__typename') == 'Video' and entry.get('canonicalUrl'):
  21                 yield self.url_result(entry['canonicalUrl'], TedTalkIE.ie_key())
  22
  23
  24 class TedTalkIE(TedBaseIE):
  25     _VALID_URL = TedBaseIE._VALID_URL_BASE.format(type='talks')
  26     _TESTS = [{
  27         'url': 'https://www.ted.com/talks/candace_parker_how_to_break_down_barriers_and_not_accept_limits',
  28         'md5': '47e82c666d9c3261d4fe74748a90aada',
  29         'info_dict': {
  30             'id': '86532',
  31             'ext': 'mp4',
  32             'title': 'How to break down barriers and not accept limits',
  33             'description': 'md5:000707cece219d1e165b11550d612331',
  34             'view_count': int,
  35             'tags': ['personal growth', 'equality', 'activism', 'motivation', 'social change', 'sports'],
  36             'uploader': 'Candace Parker',
  37             'duration': 676.0,
  38             'upload_date': '20220114',
  39             'release_date': '20211201',
  40             'thumbnail': r're:http.*\.jpg',
  41         },
  42     }]
  43
  44     def _real_extract(self, url):
  45         display_id = self._match_id(url)
  46         webpage = self._download_webpage(url, display_id)
  47         talk_info = self._search_nextjs_data(webpage, display_id)['props']['pageProps']['videoData']
  48         video_id = talk_info['id']
  49         player_data = self._parse_json(talk_info.get('playerData'), video_id)
  50
  51         http_url = None
  52         formats, subtitles = [], {}
  53         for format_id, resources in (player_data.get('resources') or {}).items():
  54             if format_id == 'hls':
  55                 stream_url = url_or_none(try_get(resources, lambda x: x['stream']))
  56                 if not stream_url:
  57                     continue
  58                 m3u8_formats, m3u8_subs = self._extract_m3u8_formats_and_subtitles(
  59                     stream_url, video_id, 'mp4', m3u8_id=format_id, fatal=False)
  60                 formats.extend(m3u8_formats)
  61                 subtitles = self._merge_subtitles(subtitles, m3u8_subs)
  62                 continue
  63
  64             if not isinstance(resources, list):
  65                 continue
  66             if format_id == 'h264':
  67                 for resource in resources:
  68                     h264_url = resource.get('file')
  69                     if not h264_url:
  70                         continue
  71                     bitrate = int_or_none(resource.get('bitrate'))
  72                     formats.append({
  73                         'url': h264_url,
  74                         'format_id': f'{format_id}-{bitrate}k',
  75                         'tbr': bitrate,
  76                     })
  77                     if re.search(r'\d+k', h264_url):
  78                         http_url = h264_url
  79             elif format_id == 'rtmp':
  80                 streamer = talk_info.get('streamer')
  81                 if not streamer:
  82                     continue
  83                 formats.extend({
  84                     'format_id': '{}-{}'.format(format_id, resource.get('name')),
  85                     'url': streamer,
  86                     'play_path': resource['file'],
  87                     'ext': 'flv',
  88                     'width': int_or_none(resource.get('width')),
  89                     'height': int_or_none(resource.get('height')),
  90                     'tbr': int_or_none(resource.get('bitrate')),
  91                 } for resource in resources if resource.get('file'))
  92
  93         if http_url:
  94             m3u8_formats = [f for f in formats if f.get('protocol') == 'm3u8' and f.get('vcodec') != 'none']
  95             for m3u8_format in m3u8_formats:
  96                 bitrate = self._search_regex(r'(\d+k)', m3u8_format['url'], 'bitrate', default=None)
  97                 if not bitrate:
  98                     continue
  99                 bitrate_url = re.sub(r'\d+k', bitrate, http_url)
 100                 if not self._is_valid_url(
 101                         bitrate_url, video_id, f'{bitrate} bitrate'):
 102                     continue
 103                 f = m3u8_format.copy()
 104                 f.update({
 105                     'url': bitrate_url,
 106                     'format_id': m3u8_format['format_id'].replace('hls', 'http'),
 107                     'protocol': 'http',
 108                 })
 109                 if f.get('acodec') == 'none':
 110                     del f['acodec']
 111                 formats.append(f)
 112
 113         audio_download = talk_info.get('audioDownload')
 114         if audio_download:
 115             formats.append({
 116                 'url': audio_download,
 117                 'format_id': 'audio',
 118                 'vcodec': 'none',
 119             })
 120
 121         if not formats:
 122             external = player_data.get('external') or {}
 123             service = external.get('service') or ''
 124             ext_url = external.get('code') if service.lower() == 'youtube' else None
 125             return self.url_result(ext_url or external['uri'])
 126
 127         thumbnail = player_data.get('thumb') or self._og_search_property('image', webpage)
 128         if thumbnail:
 129             # trim thumbnail resize parameters
 130             thumbnail = thumbnail.split('?')[0]
 131
 132         return {
 133             'id': video_id,
 134             'title': talk_info.get('title') or self._og_search_title(webpage),
 135             'uploader': talk_info.get('presenterDisplayName'),
 136             'thumbnail': thumbnail,
 137             'description': talk_info.get('description') or self._og_search_description(webpage),
 138             'subtitles': subtitles,
 139             'formats': formats,
 140             'duration': talk_info.get('duration') or parse_duration(self._og_search_property('video:duration', webpage)),
 141             'view_count': str_to_int(talk_info.get('viewedCount')),
 142             'upload_date': unified_strdate(talk_info.get('publishedAt')),
 143             'release_date': unified_strdate(talk_info.get('recordedOn')),
 144             'tags': try_get(player_data, lambda x: x['targeting']['tag'].split(',')),
 145         }
 146
 147
 148 class TedSeriesIE(TedBaseIE):
 149     _VALID_URL = fr'{TedBaseIE._VALID_URL_BASE.format(type=r"series")}(?:#season_(?P<season>\d+))?'
 150     _TESTS = [{
 151         'url': 'https://www.ted.com/series/small_thing_big_idea',
 152         'info_dict': {
 153             'id': '3',
 154             'title': 'Small Thing Big Idea',
 155             'series': 'Small Thing Big Idea',
 156             'description': 'md5:6869ca52cec661aef72b3e9f7441c55c',
 157         },
 158         'playlist_mincount': 16,
 159     }, {
 160         'url': 'https://www.ted.com/series/the_way_we_work#season_2',
 161         'info_dict': {
 162             'id': '8_2',
 163             'title': 'The Way We Work Season 2',
 164             'series': 'The Way We Work',
 165             'description': 'md5:59469256e533e1a48c4aa926a382234c',
 166             'season_number': 2,
 167         },
 168         'playlist_mincount': 8,
 169     }]
 170
 171     def _real_extract(self, url):
 172         display_id, season = self._match_valid_url(url).group('id', 'season')
 173         webpage = self._download_webpage(url, display_id, 'Downloading series webpage')
 174         info = self._search_nextjs_data(webpage, display_id)['props']['pageProps']
 175
 176         entries = itertools.chain.from_iterable(
 177             self._parse_playlist(s) for s in info['seasons'] if season in [None, s.get('seasonNumber')])
 178
 179         series_id = try_get(info, lambda x: x['series']['id'])
 180         series_name = try_get(info, lambda x: x['series']['name']) or self._og_search_title(webpage, fatal=False)
 181
 182         return self.playlist_result(
 183             entries,
 184             f'{series_id}_{season}' if season and series_id else series_id,
 185             f'{series_name} Season {season}' if season else series_name,
 186             self._og_search_description(webpage),
 187             series=series_name, season_number=int_or_none(season))
 188
 189
 190 class TedPlaylistIE(TedBaseIE):
 191     _VALID_URL = TedBaseIE._VALID_URL_BASE.format(type=r'playlists(?:/\d+)?')
 192     _TESTS = [{
 193         'url': 'https://www.ted.com/playlists/171/the_most_popular_talks_of_all',
 194         'info_dict': {
 195             'id': '171',
 196             'title': 'The most popular talks of all time',
 197             'description': 'md5:d2f22831dc86c7040e733a3cb3993d78',
 198         },
 199         'playlist_mincount': 25,
 200     }]
 201
 202     def _real_extract(self, url):
 203         display_id = self._match_id(url)
 204         webpage = self._download_webpage(url, display_id)
 205         playlist = self._search_nextjs_data(webpage, display_id)['props']['pageProps']['playlist']
 206
 207         return self.playlist_result(
 208             self._parse_playlist(playlist), playlist.get('id'),
 209             playlist.get('title') or self._og_search_title(webpage, default='').replace(' | TED Talks', '') or None,
 210             self._og_search_description(webpage))
 211
 212
 213 class TedEmbedIE(InfoExtractor):
 214     _VALID_URL = r'https?://embed(?:-ssl)?\.ted\.com/'
 215     _EMBED_REGEX = [rf'<iframe[^>]+?src=(["\'])(?P<url>{_VALID_URL}.+?)\1']
 216
 217     _TESTS = [{
 218         'url': 'https://embed.ted.com/talks/janet_stovall_how_to_get_serious_about_diversity_and_inclusion_in_the_workplace',
 219         'info_dict': {
 220             'id': '21802',
 221             'ext': 'mp4',
 222             'title': 'How to get serious about diversity and inclusion in the workplace',
 223             'description': 'md5:0978aafe396e05341f8ecc795d22189d',
 224             'view_count': int,
 225             'tags': list,
 226             'uploader': 'Janet Stovall',
 227             'duration': 664.0,
 228             'upload_date': '20180822',
 229             'release_date': '20180719',
 230             'thumbnail': r're:http.*\.jpg',
 231         },
 232     }]
 233
 234     def _real_extract(self, url):
 235         return self.url_result(re.sub(r'://embed(-ssl)?', '://www', url), TedTalkIE.ie_key())