yt_dlp/extractor/ign.py

   1 import re
   2 import urllib.parse
   3
   4 from .common import InfoExtractor
   5 from ..compat import compat_parse_qs
   6 from ..networking.exceptions import HTTPError
   7 from ..utils import (
   8     ExtractorError,
   9     determine_ext,
  10     error_to_compat_str,
  11     extract_attributes,
  12     int_or_none,
  13     merge_dicts,
  14     parse_iso8601,
  15     strip_or_none,
  16     traverse_obj,
  17     url_or_none,
  18     urljoin,
  19 )
  20
  21
  22 class IGNBaseIE(InfoExtractor):
  23     def _call_api(self, slug):
  24         return self._download_json(
  25             'http://apis.ign.com/{0}/v3/{0}s/slug/{1}'.format(self._PAGE_TYPE, slug), slug)
  26
  27     def _checked_call_api(self, slug):
  28         try:
  29             return self._call_api(slug)
  30         except ExtractorError as e:
  31             if isinstance(e.cause, HTTPError) and e.cause.status == 404:
  32                 e.cause.args = e.cause.args or [
  33                     e.cause.response.url, e.cause.status, e.cause.reason]
  34                 raise ExtractorError(
  35                     'Content not found: expired?', cause=e.cause,
  36                     expected=True)
  37             raise
  38
  39     def _extract_video_info(self, video, fatal=True):
  40         video_id = video['videoId']
  41
  42         formats = []
  43         refs = traverse_obj(video, 'refs', expected_type=dict) or {}
  44
  45         m3u8_url = url_or_none(refs.get('m3uUrl'))
  46         if m3u8_url:
  47             formats.extend(self._extract_m3u8_formats(
  48                 m3u8_url, video_id, 'mp4', 'm3u8_native',
  49                 m3u8_id='hls', fatal=False))
  50
  51         f4m_url = url_or_none(refs.get('f4mUrl'))
  52         if f4m_url:
  53             formats.extend(self._extract_f4m_formats(
  54                 f4m_url, video_id, f4m_id='hds', fatal=False))
  55
  56         for asset in (video.get('assets') or []):
  57             asset_url = url_or_none(asset.get('url'))
  58             if not asset_url:
  59                 continue
  60             formats.append({
  61                 'url': asset_url,
  62                 'tbr': int_or_none(asset.get('bitrate'), 1000),
  63                 'fps': int_or_none(asset.get('frame_rate')),
  64                 'height': int_or_none(asset.get('height')),
  65                 'width': int_or_none(asset.get('width')),
  66             })
  67
  68         mezzanine_url = traverse_obj(
  69             video, ('system', 'mezzanineUrl'), expected_type=url_or_none)
  70         if mezzanine_url:
  71             formats.append({
  72                 'ext': determine_ext(mezzanine_url, 'mp4'),
  73                 'format_id': 'mezzanine',
  74                 'quality': 1,
  75                 'url': mezzanine_url,
  76             })
  77
  78         thumbnails = traverse_obj(
  79             video, ('thumbnails', ..., {'url': 'url'}), expected_type=url_or_none)
  80         tags = traverse_obj(
  81             video, ('tags', ..., 'displayName'),
  82             expected_type=lambda x: x.strip() or None)
  83
  84         metadata = traverse_obj(video, 'metadata', expected_type=dict) or {}
  85         title = traverse_obj(
  86             metadata, 'longTitle', 'title', 'name',
  87             expected_type=lambda x: x.strip() or None)
  88
  89         return {
  90             'id': video_id,
  91             'title': title,
  92             'description': strip_or_none(metadata.get('description')),
  93             'timestamp': parse_iso8601(metadata.get('publishDate')),
  94             'duration': int_or_none(metadata.get('duration')),
  95             'thumbnails': thumbnails,
  96             'formats': formats,
  97             'tags': tags,
  98         }
  99
 100
 101 class IGNIE(IGNBaseIE):
 102     """
 103     Extractor for some of the IGN sites, like www.ign.com, es.ign.com de.ign.com.
 104     Some videos of it.ign.com are also supported
 105     """
 106     _VIDEO_PATH_RE = r'/(?:\d{4}/\d{2}/\d{2}/)?(?P<id>.+?)'
 107     _PLAYLIST_PATH_RE = r'(?:/?\?(?P<filt>[^&#]+))?'
 108     _VALID_URL = (
 109         r'https?://(?:.+?\.ign|www\.pcmag)\.com/videos(?:%s)'
 110         % '|'.join((_VIDEO_PATH_RE + r'(?:[/?&#]|$)', _PLAYLIST_PATH_RE)))
 111     IE_NAME = 'ign.com'
 112     _PAGE_TYPE = 'video'
 113
 114     _TESTS = [{
 115         'url': 'http://www.ign.com/videos/2013/06/05/the-last-of-us-review',
 116         'md5': 'd2e1586d9987d40fad7867bf96a018ea',
 117         'info_dict': {
 118             'id': '8f862beef863986b2785559b9e1aa599',
 119             'ext': 'mp4',
 120             'title': 'The Last of Us Review',
 121             'description': 'md5:c8946d4260a4d43a00d5ae8ed998870c',
 122             'timestamp': 1370440800,
 123             'upload_date': '20130605',
 124             'tags': 'count:9',
 125             'display_id': 'the-last-of-us-review',
 126             'thumbnail': 'https://assets1.ignimgs.com/vid/thumbnails/user/2014/03/26/lastofusreviewmimig2.jpg',
 127             'duration': 440,
 128         },
 129         'params': {
 130             'nocheckcertificate': True,
 131         },
 132     }, {
 133         'url': 'http://www.pcmag.com/videos/2015/01/06/010615-whats-new-now-is-gogo-snooping-on-your-data',
 134         'md5': 'f1581a6fe8c5121be5b807684aeac3f6',
 135         'info_dict': {
 136             'id': 'ee10d774b508c9b8ec07e763b9125b91',
 137             'ext': 'mp4',
 138             'title': 'What\'s New Now: Is GoGo Snooping on Your Data?',
 139             'description': 'md5:817a20299de610bd56f13175386da6fa',
 140             'timestamp': 1420571160,
 141             'upload_date': '20150106',
 142             'tags': 'count:4',
 143         },
 144         'skip': '404 Not Found',
 145     }, {
 146         'url': 'https://www.ign.com/videos/is-a-resident-evil-4-remake-on-the-way-ign-daily-fix',
 147         'only_matching': True,
 148     }]
 149
 150     @classmethod
 151     def _extract_embed_urls(cls, url, webpage):
 152         grids = re.findall(
 153             r'''(?s)<section\b[^>]+\bclass\s*=\s*['"](?:[\w-]+\s+)*?content-feed-grid(?!\B|-)[^>]+>(.+?)</section[^>]*>''',
 154             webpage)
 155         return filter(None,
 156                       (urljoin(url, m.group('path')) for m in re.finditer(
 157                           r'''<a\b[^>]+\bhref\s*=\s*('|")(?P<path>/videos%s)\1'''
 158                           % cls._VIDEO_PATH_RE, grids[0] if grids else '')))
 159
 160     def _real_extract(self, url):
 161         display_id, filt = self._match_valid_url(url).group('id', 'filt')
 162         if display_id:
 163             return self._extract_video(url, display_id)
 164         return self._extract_playlist(url, filt or 'all')
 165
 166     def _extract_playlist(self, url, display_id):
 167         webpage = self._download_webpage(url, display_id)
 168
 169         return self.playlist_result(
 170             (self.url_result(u, self.ie_key())
 171              for u in self._extract_embed_urls(url, webpage)),
 172             playlist_id=display_id)
 173
 174     def _extract_video(self, url, display_id):
 175         video = self._checked_call_api(display_id)
 176
 177         info = self._extract_video_info(video)
 178
 179         return merge_dicts({
 180             'display_id': display_id,
 181         }, info)
 182
 183
 184 class IGNVideoIE(IGNBaseIE):
 185     _VALID_URL = r'https?://.+?\.ign\.com/(?:[a-z]{2}/)?[^/]+/(?P<id>\d+)/(?:video|trailer)/'
 186     _TESTS = [{
 187         'url': 'http://me.ign.com/en/videos/112203/video/how-hitman-aims-to-be-different-than-every-other-s',
 188         'md5': 'dd9aca7ed2657c4e118d8b261e5e9de1',
 189         'info_dict': {
 190             'id': 'e9be7ea899a9bbfc0674accc22a36cc8',
 191             'ext': 'mp4',
 192             'title': 'How Hitman Aims to Be Different Than Every Other Stealth Game - NYCC 2015',
 193             'description': 'Taking out assassination targets in Hitman has never been more stylish.',
 194             'timestamp': 1444665600,
 195             'upload_date': '20151012',
 196             'display_id': '112203',
 197             'thumbnail': 'https://sm.ign.com/ign_me/video/h/how-hitman/how-hitman-aims-to-be-different-than-every-other-s_8z14.jpg',
 198             'duration': 298,
 199             'tags': 'count:13',
 200         },
 201         'expected_warnings': ['HTTP Error 400: Bad Request'],
 202     }, {
 203         'url': 'http://me.ign.com/ar/angry-birds-2/106533/video/lrd-ldyy-lwl-lfylm-angry-birds',
 204         'only_matching': True,
 205     }, {
 206         # Youtube embed
 207         'url': 'https://me.ign.com/ar/ratchet-clank-rift-apart/144327/trailer/embed',
 208         'only_matching': True,
 209     }, {
 210         # Twitter embed
 211         'url': 'http://adria.ign.com/sherlock-season-4/9687/trailer/embed',
 212         'only_matching': True,
 213     }, {
 214         # Vimeo embed
 215         'url': 'https://kr.ign.com/bic-2018/3307/trailer/embed',
 216         'only_matching': True,
 217     }]
 218
 219     def _real_extract(self, url):
 220         video_id = self._match_id(url)
 221         parsed_url = urllib.parse.urlparse(url)
 222         embed_url = urllib.parse.urlunparse(
 223             parsed_url._replace(path=parsed_url.path.rsplit('/', 1)[0] + '/embed'))
 224
 225         webpage, urlh = self._download_webpage_handle(embed_url, video_id)
 226         new_url = urlh.url
 227         ign_url = compat_parse_qs(
 228             urllib.parse.urlparse(new_url).query).get('url', [None])[-1]
 229         if ign_url:
 230             return self.url_result(ign_url, IGNIE.ie_key())
 231         video = self._search_regex(r'(<div\b[^>]+\bdata-video-id\s*=\s*[^>]+>)', webpage, 'video element', fatal=False)
 232         if not video:
 233             if new_url == url:
 234                 raise ExtractorError('Redirect loop: ' + url)
 235             return self.url_result(new_url)
 236         video = extract_attributes(video)
 237         video_data = video.get('data-settings') or '{}'
 238         video_data = self._parse_json(video_data, video_id)['video']
 239         info = self._extract_video_info(video_data)
 240
 241         return merge_dicts({
 242             'display_id': video_id,
 243         }, info)
 244
 245
 246 class IGNArticleIE(IGNBaseIE):
 247     _VALID_URL = r'https?://.+?\.ign\.com/(?:articles(?:/\d{4}/\d{2}/\d{2})?|(?:[a-z]{2}/)?(?:[\w-]+/)*?feature/\d+)/(?P<id>[^/?&#]+)'
 248     _PAGE_TYPE = 'article'
 249     _TESTS = [{
 250         'url': 'http://me.ign.com/en/feature/15775/100-little-things-in-gta-5-that-will-blow-your-mind',
 251         'info_dict': {
 252             'id': '72113',
 253             'title': '100 Little Things in GTA 5 That Will Blow Your Mind',
 254         },
 255         'playlist': [
 256             {
 257                 'info_dict': {
 258                     'id': '5ebbd138523268b93c9141af17bec937',
 259                     'ext': 'mp4',
 260                     'title': 'Grand Theft Auto V Video Review',
 261                     'description': 'Rockstar drops the mic on this generation of games. Watch our review of the masterly Grand Theft Auto V.',
 262                     'timestamp': 1379339880,
 263                     'upload_date': '20130916',
 264                     'tags': 'count:12',
 265                     'thumbnail': 'https://assets1.ignimgs.com/thumbs/userUploaded/2021/8/16/gta-v-heistsjpg-e94705-1629138553533.jpeg',
 266                     'display_id': 'grand-theft-auto-v-video-review',
 267                     'duration': 501,
 268                 },
 269             },
 270             {
 271                 'info_dict': {
 272                     'id': '638672ee848ae4ff108df2a296418ee2',
 273                     'ext': 'mp4',
 274                     'title': 'GTA 5 In Slow Motion',
 275                     'description': 'The twisted beauty of GTA 5 in stunning slow motion.',
 276                     'timestamp': 1386878820,
 277                     'upload_date': '20131212',
 278                     'duration': 202,
 279                     'tags': 'count:25',
 280                     'display_id': 'gta-5-in-slow-motion',
 281                     'thumbnail': 'https://assets1.ignimgs.com/vid/thumbnails/user/2013/11/03/GTA-SLO-MO-1.jpg',
 282                 },
 283             },
 284         ],
 285         'params': {
 286             'skip_download': True,
 287         },
 288         'expected_warnings': ['Backend fetch failed'],
 289     }, {
 290         'url': 'http://www.ign.com/articles/2014/08/15/rewind-theater-wild-trailer-gamescom-2014?watch',
 291         'info_dict': {
 292             'id': '53ee806780a81ec46e0790f8',
 293             'title': 'Rewind Theater - Wild Trailer Gamescom 2014',
 294         },
 295         'playlist_count': 1,
 296         'expected_warnings': ['Backend fetch failed'],
 297     }, {
 298         # videoId pattern
 299         'url': 'http://www.ign.com/articles/2017/06/08/new-ducktales-short-donalds-birthday-doesnt-go-as-planned',
 300         'only_matching': True,
 301     }, {
 302         # Youtube embed
 303         'url': 'https://www.ign.com/articles/2021-mvp-named-in-puppy-bowl-xvii',
 304         'only_matching': True,
 305     }, {
 306         # IMDB embed
 307         'url': 'https://www.ign.com/articles/2014/08/07/sons-of-anarchy-final-season-trailer',
 308         'only_matching': True,
 309     }, {
 310         # Facebook embed
 311         'url': 'https://www.ign.com/articles/2017/09/20/marvels-the-punisher-watch-the-new-trailer-for-the-netflix-series',
 312         'only_matching': True,
 313     }, {
 314         # Brightcove embed
 315         'url': 'https://www.ign.com/articles/2016/01/16/supergirl-goes-flying-with-martian-manhunter-in-new-clip',
 316         'only_matching': True,
 317     }]
 318
 319     def _checked_call_api(self, slug):
 320         try:
 321             return self._call_api(slug)
 322         except ExtractorError as e:
 323             if isinstance(e.cause, HTTPError):
 324                 e.cause.args = e.cause.args or [
 325                     e.cause.response.url, e.cause.status, e.cause.reason]
 326                 if e.cause.status == 404:
 327                     raise ExtractorError(
 328                         'Content not found: expired?', cause=e.cause,
 329                         expected=True)
 330                 elif e.cause.status == 503:
 331                     self.report_warning(error_to_compat_str(e.cause))
 332                     return
 333             raise
 334
 335     def _real_extract(self, url):
 336         display_id = self._match_id(url)
 337         article = self._checked_call_api(display_id)
 338
 339         if article:
 340             # obsolete ?
 341             def entries():
 342                 media_url = traverse_obj(
 343                     article, ('mediaRelations', 0, 'media', 'metadata', 'url'),
 344                     expected_type=url_or_none)
 345                 if media_url:
 346                     yield self.url_result(media_url, IGNIE.ie_key())
 347                 for content in (article.get('content') or []):
 348                     for video_url in re.findall(r'(?:\[(?:ignvideo\s+url|youtube\s+clip_id)|<iframe[^>]+src)="([^"]+)"', content):
 349                         if url_or_none(video_url):
 350                             yield self.url_result(video_url)
 351
 352             return self.playlist_result(
 353                 entries(), article.get('articleId'),
 354                 traverse_obj(
 355                     article, ('metadata', 'headline'),
 356                     expected_type=lambda x: x.strip() or None))
 357
 358         webpage = self._download_webpage(url, display_id)
 359
 360         playlist_id = self._html_search_meta('dable:item_id', webpage, default=None)
 361         if playlist_id:
 362
 363             def entries():
 364                 for m in re.finditer(
 365                         r'''(?s)<object\b[^>]+\bclass\s*=\s*("|')ign-videoplayer\1[^>]*>(?P<params>.+?)</object''',
 366                         webpage):
 367                     flashvars = self._search_regex(
 368                         r'''(<param\b[^>]+\bname\s*=\s*("|')flashvars\2[^>]*>)''',
 369                         m.group('params'), 'flashvars', default='')
 370                     flashvars = compat_parse_qs(extract_attributes(flashvars).get('value') or '')
 371                     v_url = url_or_none((flashvars.get('url') or [None])[-1])
 372                     if v_url:
 373                         yield self.url_result(v_url)
 374         else:
 375             playlist_id = self._search_regex(
 376                 r'''\bdata-post-id\s*=\s*("|')(?P<id>[\da-f]+)\1''',
 377                 webpage, 'id', group='id', default=None)
 378
 379             nextjs_data = self._search_nextjs_data(webpage, display_id)
 380
 381             def entries():
 382                 for player in traverse_obj(
 383                         nextjs_data,
 384                         ('props', 'apolloState', 'ROOT_QUERY', lambda k, _: k.startswith('videoPlayerProps('), '__ref')):
 385                     # skip promo links (which may not always be served, eg GH CI servers)
 386                     if traverse_obj(nextjs_data,
 387                                     ('props', 'apolloState', player.replace('PlayerProps', 'ModernContent')),
 388                                     expected_type=dict):
 389                         continue
 390                     video = traverse_obj(nextjs_data, ('props', 'apolloState', player), expected_type=dict) or {}
 391                     info = self._extract_video_info(video, fatal=False)
 392                     if info:
 393                         yield merge_dicts({
 394                             'display_id': display_id,
 395                         }, info)
 396
 397         return self.playlist_result(
 398             entries(), playlist_id or display_id,
 399             re.sub(r'\s+-\s+IGN\s*$', '', self._og_search_title(webpage, default='')) or None)