yt_dlp/extractor/condenast.py

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