yt_dlp/extractor/condenast.py

   1 import re
   2 import urllib.parse
   3
   4 from .common import InfoExtractor
   5 from ..utils import (
   6     determine_ext,
   7     extract_attributes,
   8     int_or_none,
   9     js_to_json,
  10     mimetype2ext,
  11     orderedSet,
  12     parse_iso8601,
  13     strip_or_none,
  14     try_get,
  15     urljoin,
  16 )
  17
  18
  19 class CondeNastIE(InfoExtractor):
  20     """
  21     Condé Nast is a media group, some of its sites use a custom HTML5 player
  22     that works the same in all of them.
  23     """
  24
  25     # The keys are the supported sites and the values are the name to be shown
  26     # to the user and in the extractor description.
  27     _SITES = {
  28         'allure': 'Allure',
  29         'architecturaldigest': 'Architectural Digest',
  30         'arstechnica': 'Ars Technica',
  31         'bonappetit': 'Bon Appétit',
  32         'brides': 'Brides',
  33         'cnevids': 'Condé Nast',
  34         'cntraveler': 'Condé Nast Traveler',
  35         'details': 'Details',
  36         'epicurious': 'Epicurious',
  37         'glamour': 'Glamour',
  38         'golfdigest': 'Golf Digest',
  39         'gq': 'GQ',
  40         'newyorker': 'The New Yorker',
  41         'self': 'SELF',
  42         'teenvogue': 'Teen Vogue',
  43         'vanityfair': 'Vanity Fair',
  44         'vogue': 'Vogue',
  45         'wired': 'WIRED',
  46         'wmagazine': 'W Magazine',
  47     }
  48
  49     _VALID_URL = r'''(?x)https?://(?:video|www|player(?:-backend)?)\.(?:{})\.com/
  50         (?:
  51             (?:
  52                 embed(?:js)?|
  53                 (?:script|inline)/video
  54             )/(?P<id>[0-9a-f]{{24}})(?:/(?P<player_id>[0-9a-f]{{24}}))?(?:.+?\btarget=(?P<target>[^&]+))?|
  55             (?P<type>watch|series|video)/(?P<display_id>[^/?#]+)
  56         )'''.format('|'.join(_SITES.keys()))
  57     IE_DESC = 'Condé Nast media group: {}'.format(', '.join(sorted(_SITES.values())))
  58
  59     _EMBED_REGEX = [r'''(?x)
  60         <(?:iframe|script)[^>]+?src=(["\'])(?P<url>
  61             (?:https?:)?//player(?:-backend)?\.(?:{})\.com/(?:embed(?:js)?|(?:script|inline)/video)/.+?
  62         )\1'''.format('|'.join(_SITES.keys()))]
  63
  64     _TESTS = [{
  65         'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
  66         'md5': '1921f713ed48aabd715691f774c451f7',
  67         'info_dict': {
  68             'id': '5171b343c2b4c00dd0c1ccb3',
  69             'ext': 'mp4',
  70             'title': '3D Printed Speakers Lit With LED',
  71             'description': 'Check out these beautiful 3D printed LED speakers.  You can\'t actually buy them, but LumiGeek is working on a board that will let you make you\'re own.',
  72             'uploader': 'wired',
  73             'upload_date': '20130314',
  74             'timestamp': 1363219200,
  75         },
  76     }, {
  77         'url': 'http://video.gq.com/watch/the-closer-with-keith-olbermann-the-only-true-surprise-trump-s-an-idiot?c=series',
  78         'info_dict': {
  79             'id': '58d1865bfd2e6126e2000015',
  80             'ext': 'mp4',
  81             'title': 'The Only True Surprise? Trump’s an Idiot',
  82             'uploader': 'gq',
  83             'upload_date': '20170321',
  84             'timestamp': 1490126427,
  85             'description': 'How much grimmer would things be if these people were competent?',
  86         },
  87     }, {
  88         # JS embed
  89         'url': 'http://player.cnevids.com/embedjs/55f9cf8b61646d1acf00000c/5511d76261646d5566020000.js',
  90         'md5': 'f1a6f9cafb7083bab74a710f65d08999',
  91         'info_dict': {
  92             'id': '55f9cf8b61646d1acf00000c',
  93             'ext': 'mp4',
  94             'title': '3D printed TSA Travel Sentry keys really do open TSA locks',
  95             'uploader': 'arstechnica',
  96             'upload_date': '20150916',
  97             'timestamp': 1442434920,
  98         },
  99     }, {
 100         'url': 'https://player.cnevids.com/inline/video/59138decb57ac36b83000005.js?target=js-cne-player',
 101         'only_matching': True,
 102     }, {
 103         'url': 'http://player-backend.cnevids.com/script/video/59138decb57ac36b83000005.js',
 104         'only_matching': True,
 105     }]
 106
 107     def _extract_series(self, url, webpage):
 108         title = self._html_search_regex(
 109             r'(?s)<div class="cne-series-info">.*?<h1>(.+?)</h1>',
 110             webpage, 'series title')
 111         url_object = urllib.parse.urlparse(url)
 112         base_url = f'{url_object.scheme}://{url_object.netloc}'
 113         m_paths = re.finditer(
 114             r'(?s)<p class="cne-thumb-title">.*?<a href="(/watch/.+?)["\?]', webpage)
 115         paths = orderedSet(m.group(1) for m in m_paths)
 116         entries = [self.url_result(urljoin(base_url, path), 'CondeNast') for path in paths]
 117         return self.playlist_result(entries, playlist_title=title)
 118
 119     def _extract_video_params(self, webpage, display_id):
 120         query = self._parse_json(
 121             self._search_regex(
 122                 r'(?s)var\s+params\s*=\s*({.+?})[;,]', webpage, 'player params',
 123                 default='{}'),
 124             display_id, transform_source=js_to_json, fatal=False)
 125         if query:
 126             query['videoId'] = self._search_regex(
 127                 r'(?:data-video-id=|currentVideoId\s*=\s*)["\']([\da-f]+)',
 128                 webpage, 'video id', default=None)
 129         else:
 130             params = extract_attributes(self._search_regex(
 131                 r'(<[^>]+data-js="video-player"[^>]+>)',
 132                 webpage, 'player params element'))
 133             query.update({
 134                 'videoId': params['data-video'],
 135                 'playerId': params['data-player'],
 136                 'target': params['id'],
 137             })
 138         return query
 139
 140     def _extract_video(self, params):
 141         video_id = params['videoId']
 142
 143         video_info = None
 144
 145         # New API path
 146         query = params.copy()
 147         query['embedType'] = 'inline'
 148         info_page = self._download_json(
 149             'http://player.cnevids.com/embed-api.json', video_id,
 150             'Downloading embed info', fatal=False, query=query)
 151
 152         # Old fallbacks
 153         if not info_page:
 154             if params.get('playerId'):
 155                 info_page = self._download_json(
 156                     'http://player.cnevids.com/player/video.js', video_id,
 157                     'Downloading video info', fatal=False, query=params)
 158         if info_page:
 159             video_info = info_page.get('video')
 160         if not video_info:
 161             info_page = self._download_webpage(
 162                 'http://player.cnevids.com/player/loader.js',
 163                 video_id, 'Downloading loader info', query=params)
 164         if not video_info:
 165             info_page = self._download_webpage(
 166                 f'https://player.cnevids.com/inline/video/{video_id}.js',
 167                 video_id, 'Downloading inline info', query={
 168                     'target': params.get('target', 'embedplayer'),
 169                 })
 170
 171         if not video_info:
 172             video_info = self._parse_json(
 173                 self._search_regex(
 174                     r'(?s)var\s+config\s*=\s*({.+?});', info_page, 'config'),
 175                 video_id, transform_source=js_to_json)['video']
 176
 177         title = video_info['title']
 178
 179         formats = []
 180         for fdata in video_info['sources']:
 181             src = fdata.get('src')
 182             if not src:
 183                 continue
 184             ext = mimetype2ext(fdata.get('type')) or determine_ext(src)
 185             if ext == 'm3u8':
 186                 formats.extend(self._extract_m3u8_formats(
 187                     src, video_id, 'mp4', entry_protocol='m3u8_native',
 188                     m3u8_id='hls', fatal=False))
 189                 continue
 190             quality = fdata.get('quality')
 191             formats.append({
 192                 'format_id': ext + (f'-{quality}' if quality else ''),
 193                 'url': src,
 194                 'ext': ext,
 195                 'quality': 1 if quality == 'high' else 0,
 196             })
 197
 198         subtitles = {}
 199         for t, caption in video_info.get('captions', {}).items():
 200             caption_url = caption.get('src')
 201             if not (t in ('vtt', 'srt', 'tml') and caption_url):
 202                 continue
 203             subtitles.setdefault('en', []).append({'url': caption_url})
 204
 205         return {
 206             'id': video_id,
 207             'formats': formats,
 208             'title': title,
 209             'thumbnail': video_info.get('poster_frame'),
 210             'uploader': video_info.get('brand'),
 211             'duration': int_or_none(video_info.get('duration')),
 212             'tags': video_info.get('tags'),
 213             'series': video_info.get('series_title'),
 214             'season': video_info.get('season_title'),
 215             'timestamp': parse_iso8601(video_info.get('premiere_date')),
 216             'categories': video_info.get('categories'),
 217             'subtitles': subtitles,
 218         }
 219
 220     def _real_extract(self, url):
 221         video_id, player_id, target, url_type, display_id = self._match_valid_url(url).groups()
 222
 223         if video_id:
 224             return self._extract_video({
 225                 'videoId': video_id,
 226                 'playerId': player_id,
 227                 'target': target,
 228             })
 229
 230         webpage = self._download_webpage(url, display_id)
 231
 232         if url_type == 'series':
 233             return self._extract_series(url, webpage)
 234         else:
 235             video = try_get(self._parse_json(self._search_regex(
 236                 r'__PRELOADED_STATE__\s*=\s*({.+?});', webpage,
 237                 'preload state', '{}'), display_id),
 238                 lambda x: x['transformed']['video'])
 239             if video:
 240                 params = {'videoId': video['id']}
 241                 info = {'description': strip_or_none(video.get('description'))}
 242             else:
 243                 params = self._extract_video_params(webpage, display_id)
 244                 info = self._search_json_ld(
 245                     webpage, display_id, fatal=False)
 246             info.update(self._extract_video(params))
 247             return info