youtube_dl/extractor/generic.py

   1 # encoding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5 import os
   6 import re
   7
   8 from .common import InfoExtractor
   9 from .youtube import YoutubeIE
  10 from ..utils import (
  11     compat_urllib_parse,
  12     compat_urlparse,
  13     compat_xml_parse_error,
  14
  15     determine_ext,
  16     ExtractorError,
  17     float_or_none,
  18     HEADRequest,
  19     orderedSet,
  20     parse_xml,
  21     smuggle_url,
  22     unescapeHTML,
  23     unified_strdate,
  24     unsmuggle_url,
  25     url_basename,
  26 )
  27 from .brightcove import BrightcoveIE
  28 from .ooyala import OoyalaIE
  29 from .rutv import RUTVIE
  30 from .smotri import SmotriIE
  31
  32
  33 class GenericIE(InfoExtractor):
  34     IE_DESC = 'Generic downloader that works on some sites'
  35     _VALID_URL = r'.*'
  36     IE_NAME = 'generic'
  37     _TESTS = [
  38         {
  39             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  40             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  41             'info_dict': {
  42                 'id': '13601338388002',
  43                 'ext': 'mp4',
  44                 'uploader': 'www.hodiho.fr',
  45                 'title': 'R\u00e9gis plante sa Jeep',
  46             }
  47         },
  48         # bandcamp page with custom domain
  49         {
  50             'add_ie': ['Bandcamp'],
  51             'url': 'http://bronyrock.com/track/the-pony-mash',
  52             'info_dict': {
  53                 'id': '3235767654',
  54                 'ext': 'mp3',
  55                 'title': 'The Pony Mash',
  56                 'uploader': 'M_Pallante',
  57             },
  58             'skip': 'There is a limit of 200 free downloads / month for the test song',
  59         },
  60         # embedded brightcove video
  61         # it also tests brightcove videos that need to set the 'Referer' in the
  62         # http requests
  63         {
  64             'add_ie': ['Brightcove'],
  65             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  66             'info_dict': {
  67                 'id': '2765128793001',
  68                 'ext': 'mp4',
  69                 'title': 'Le cours de bourse : l’analyse technique',
  70                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  71                 'uploader': 'BFM BUSINESS',
  72             },
  73             'params': {
  74                 'skip_download': True,
  75             },
  76         },
  77         {
  78             # https://github.com/rg3/youtube-dl/issues/2253
  79             'url': 'http://bcove.me/i6nfkrc3',
  80             'md5': '0ba9446db037002366bab3b3eb30c88c',
  81             'info_dict': {
  82                 'id': '3101154703001',
  83                 'ext': 'mp4',
  84                 'title': 'Still no power',
  85                 'uploader': 'thestar.com',
  86                 'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
  87             },
  88             'add_ie': ['Brightcove'],
  89         },
  90         {
  91             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  92             'md5': 'fb973ecf6e4a78a67453647444222983',
  93             'info_dict': {
  94                 'id': '3414141473001',
  95                 'ext': 'mp4',
  96                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  97                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
  98                 'uploader': 'Championat',
  99             },
 100         },
 101         # Direct link to a video
 102         {
 103             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 104             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 105             'info_dict': {
 106                 'id': 'trailer',
 107                 'ext': 'mp4',
 108                 'title': 'trailer',
 109                 'upload_date': '20100513',
 110             }
 111         },
 112         # ooyala video
 113         {
 114             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 115             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 116             'info_dict': {
 117                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 118                 'ext': 'mp4',
 119                 'title': '2cc213299525360.mov',  # that's what we get
 120             },
 121         },
 122         # google redirect
 123         {
 124             'url': 'http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCUQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcmQHVoWB5FY&ei=F-sNU-LLCaXk4QT52ICQBQ&usg=AFQjCNEw4hL29zgOohLXvpJ-Bdh2bils1Q&bvm=bv.61965928,d.bGE',
 125             'info_dict': {
 126                 'id': 'cmQHVoWB5FY',
 127                 'ext': 'mp4',
 128                 'upload_date': '20130224',
 129                 'uploader_id': 'TheVerge',
 130                 'description': 'Chris Ziegler takes a look at the Alcatel OneTouch Fire and the ZTE Open; two of the first Firefox OS handsets to be officially announced.',
 131                 'uploader': 'The Verge',
 132                 'title': 'First Firefox OS phones side-by-side',
 133             },
 134             'params': {
 135                 'skip_download': False,
 136             }
 137         },
 138         # embed.ly video
 139         {
 140             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 141             'info_dict': {
 142                 'id': '9ODmcdjQcHQ',
 143                 'ext': 'mp4',
 144                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 145                 'upload_date': '20140225',
 146                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 147                 'uploader': 'Tested',
 148                 'uploader_id': 'testedcom',
 149             },
 150             # No need to test YoutubeIE here
 151             'params': {
 152                 'skip_download': True,
 153             },
 154         },
 155         # funnyordie embed
 156         {
 157             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 158             'md5': '7cf780be104d40fea7bae52eed4a470e',
 159             'info_dict': {
 160                 'id': '18e820ec3f',
 161                 'ext': 'mp4',
 162                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 163                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 164             },
 165         },
 166         # RUTV embed
 167         {
 168             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 169             'info_dict': {
 170                 'id': '776940',
 171                 'ext': 'mp4',
 172                 'title': 'Охотское море стало целиком российским',
 173                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 174             },
 175             'params': {
 176                 # m3u8 download
 177                 'skip_download': True,
 178             },
 179         },
 180         # Embedded TED video
 181         {
 182             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 183             'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
 184             'info_dict': {
 185                 'id': '981',
 186                 'ext': 'mp4',
 187                 'title': 'My web playroom',
 188                 'uploader': 'Ze Frank',
 189                 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
 190             }
 191         },
 192         # Embeded Ustream video
 193         {
 194             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 195             'md5': '27b99cdb639c9b12a79bca876a073417',
 196             'info_dict': {
 197                 'id': '45734260',
 198                 'ext': 'flv',
 199                 'uploader': 'AU SPA:  The NSA and Privacy',
 200                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 201             }
 202         },
 203         # nowvideo embed hidden behind percent encoding
 204         {
 205             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 206             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 207             'info_dict': {
 208                 'id': '06e53103ca9aa',
 209                 'ext': 'flv',
 210                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 211                 'description': 'No description',
 212             },
 213         },
 214         # arte embed
 215         {
 216             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 217             'md5': '7653032cbb25bf6c80d80f217055fa43',
 218             'info_dict': {
 219                 'id': '048195-004_PLUS7-F',
 220                 'ext': 'flv',
 221                 'title': 'X:enius',
 222                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 223                 'upload_date': '20140320',
 224             },
 225             'params': {
 226                 'skip_download': 'Requires rtmpdump'
 227             }
 228         },
 229         # smotri embed
 230         {
 231             'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
 232             'md5': 'ec40048448e9284c9a1de77bb188108b',
 233             'info_dict': {
 234                 'id': 'v27008541fad',
 235                 'ext': 'mp4',
 236                 'title': 'Крым и Севастополь вошли в состав России',
 237                 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
 238                 'duration': 900,
 239                 'upload_date': '20140318',
 240                 'uploader': 'rbctv_2012_4',
 241                 'uploader_id': 'rbctv_2012_4',
 242             },
 243         },
 244         # Condé Nast embed
 245         {
 246             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 247             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 248             'info_dict': {
 249                 'id': '53501be369702d3275860000',
 250                 'ext': 'mp4',
 251                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 252             }
 253         },
 254         # Dailymotion embed
 255         {
 256             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 257             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 258             'info_dict': {
 259                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 260                 'ext': 'mp4',
 261                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 262                 'uploader': 'Spi0n',
 263             },
 264             'add_ie': ['Dailymotion'],
 265         },
 266         # YouTube embed
 267         {
 268             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 269             'info_dict': {
 270                 'id': 'FXRb4ykk4S0',
 271                 'ext': 'mp4',
 272                 'title': 'The NBL Auction 2014',
 273                 'uploader': 'BADMINTON England',
 274                 'uploader_id': 'BADMINTONEvents',
 275                 'upload_date': '20140603',
 276                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 277             },
 278             'add_ie': ['Youtube'],
 279             'params': {
 280                 'skip_download': True,
 281             }
 282         },
 283         # MTVSercices embed
 284         {
 285             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 286             'md5': '35727f82f58c76d996fc188f9755b0d5',
 287             'info_dict': {
 288                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 289                 'ext': 'mp4',
 290                 'title': 'Review',
 291                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 292             },
 293         },
 294         # YouTube embed via <data-embed-url="">
 295         {
 296             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 297             'info_dict': {
 298                 'id': 'jpSGZsgga_I',
 299                 'ext': 'mp4',
 300                 'title': 'Asphalt 8: Airborne - Launch Trailer',
 301                 'uploader': 'Gameloft',
 302                 'uploader_id': 'gameloft',
 303                 'upload_date': '20130821',
 304                 'description': 'md5:87bd95f13d8be3e7da87a5f2c443106a',
 305             },
 306             'params': {
 307                 'skip_download': True,
 308             }
 309         },
 310         # Camtasia studio
 311         {
 312             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 313             'playlist': [{
 314                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 315                 'info_dict': {
 316                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 317                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 318                     'ext': 'flv',
 319                     'duration': 2235.90,
 320                 }
 321             }, {
 322                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 323                 'info_dict': {
 324                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 325                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 326                     'ext': 'flv',
 327                     'duration': 2235.93,
 328                 }
 329             }],
 330             'info_dict': {
 331                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 332             }
 333         },
 334         # Flowplayer
 335         {
 336             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 337             'md5': '9d65602bf31c6e20014319c7d07fba27',
 338             'info_dict': {
 339                 'id': '5123ea6d5e5a7',
 340                 'ext': 'mp4',
 341                 'age_limit': 18,
 342                 'uploader': 'www.handjobhub.com',
 343                 'title': 'Busty Blonde Siri Tit Fuck While Wank at Handjob Hub',
 344             }
 345         },
 346         # RSS feed
 347         {
 348             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 349             'info_dict': {
 350                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 351                 'title': 'Zero Punctuation',
 352                 'description': 're:'
 353             },
 354             'playlist_mincount': 11,
 355         },
 356         # Multiple brightcove videos
 357         # https://github.com/rg3/youtube-dl/issues/2283
 358         {
 359             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 360             'info_dict': {
 361                 'id': 'always-never',
 362                 'title': 'Always / Never - The New Yorker',
 363             },
 364             'playlist_count': 3,
 365             'params': {
 366                 'extract_flat': False,
 367                 'skip_download': True,
 368             }
 369         },
 370         # MLB embed
 371         {
 372             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 373             'md5': '96f09a37e44da40dd083e12d9a683327',
 374             'info_dict': {
 375                 'id': '33322633',
 376                 'ext': 'mp4',
 377                 'title': 'Ump changes call to ball',
 378                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 379                 'duration': 48,
 380                 'timestamp': 1401537900,
 381                 'upload_date': '20140531',
 382                 'thumbnail': 're:^https?://.*\.jpg$',
 383             },
 384         },
 385     ]
 386
 387     def report_download_webpage(self, video_id):
 388         """Report webpage download."""
 389         if not self._downloader.params.get('test', False):
 390             self._downloader.report_warning('Falling back on generic information extractor.')
 391         super(GenericIE, self).report_download_webpage(video_id)
 392
 393     def report_following_redirect(self, new_url):
 394         """Report information extraction."""
 395         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 396
 397     def _extract_rss(self, url, video_id, doc):
 398         playlist_title = doc.find('./channel/title').text
 399         playlist_desc_el = doc.find('./channel/description')
 400         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 401
 402         entries = [{
 403             '_type': 'url',
 404             'url': e.find('link').text,
 405             'title': e.find('title').text,
 406         } for e in doc.findall('./channel/item')]
 407
 408         return {
 409             '_type': 'playlist',
 410             'id': url,
 411             'title': playlist_title,
 412             'description': playlist_desc,
 413             'entries': entries,
 414         }
 415
 416     def _extract_camtasia(self, url, video_id, webpage):
 417         """ Returns None if no camtasia video can be found. """
 418
 419         camtasia_cfg = self._search_regex(
 420             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 421             webpage, 'camtasia configuration file', default=None)
 422         if camtasia_cfg is None:
 423             return None
 424
 425         title = self._html_search_meta('DC.title', webpage, fatal=True)
 426
 427         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 428         camtasia_cfg = self._download_xml(
 429             camtasia_url, video_id,
 430             note='Downloading camtasia configuration',
 431             errnote='Failed to download camtasia configuration')
 432         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 433
 434         entries = []
 435         for n in fileset_node.getchildren():
 436             url_n = n.find('./uri')
 437             if url_n is None:
 438                 continue
 439
 440             entries.append({
 441                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 442                 'title': '%s - %s' % (title, n.tag),
 443                 'url': compat_urlparse.urljoin(url, url_n.text),
 444                 'duration': float_or_none(n.find('./duration').text),
 445             })
 446
 447         return {
 448             '_type': 'playlist',
 449             'entries': entries,
 450             'title': title,
 451         }
 452
 453     def _real_extract(self, url):
 454         if url.startswith('//'):
 455             return {
 456                 '_type': 'url',
 457                 'url': self.http_scheme() + url,
 458             }
 459
 460         parsed_url = compat_urlparse.urlparse(url)
 461         if not parsed_url.scheme:
 462             default_search = self._downloader.params.get('default_search')
 463             if default_search is None:
 464                 default_search = 'fixup_error'
 465
 466             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 467                 if '/' in url:
 468                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 469                     return self.url_result('http://' + url)
 470                 elif default_search != 'fixup_error':
 471                     if default_search == 'auto_warning':
 472                         if re.match(r'^(?:url|URL)$', url):
 473                             raise ExtractorError(
 474                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 475                                 expected=True)
 476                         else:
 477                             self._downloader.report_warning(
 478                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 479                     return self.url_result('ytsearch:' + url)
 480
 481             if default_search in ('error', 'fixup_error'):
 482                 raise ExtractorError(
 483                     ('%r is not a valid URL. '
 484                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 485                     ) % (url, url), expected=True)
 486             else:
 487                 assert ':' in default_search
 488                 return self.url_result(default_search + url)
 489
 490         url, smuggled_data = unsmuggle_url(url)
 491         force_videoid = None
 492         if smuggled_data and 'force_videoid' in smuggled_data:
 493             force_videoid = smuggled_data['force_videoid']
 494             video_id = force_videoid
 495         else:
 496             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 497
 498         self.to_screen('%s: Requesting header' % video_id)
 499
 500         head_req = HEADRequest(url)
 501         response = self._request_webpage(
 502             head_req, video_id,
 503             note=False, errnote='Could not send HEAD request to %s' % url,
 504             fatal=False)
 505
 506         if response is not False:
 507             # Check for redirect
 508             new_url = response.geturl()
 509             if url != new_url:
 510                 self.report_following_redirect(new_url)
 511                 if force_videoid:
 512                     new_url = smuggle_url(
 513                         new_url, {'force_videoid': force_videoid})
 514                 return self.url_result(new_url)
 515
 516             # Check for direct link to a video
 517             content_type = response.headers.get('Content-Type', '')
 518             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 519             if m:
 520                 upload_date = response.headers.get('Last-Modified')
 521                 if upload_date:
 522                     upload_date = unified_strdate(upload_date)
 523                 return {
 524                     'id': video_id,
 525                     'title': os.path.splitext(url_basename(url))[0],
 526                     'formats': [{
 527                         'format_id': m.group('format_id'),
 528                         'url': url,
 529                         'vcodec': 'none' if m.group('type') == 'audio' else None
 530                     }],
 531                     'upload_date': upload_date,
 532                 }
 533
 534         try:
 535             webpage = self._download_webpage(url, video_id)
 536         except ValueError:
 537             # since this is the last-resort InfoExtractor, if
 538             # this error is thrown, it'll be thrown here
 539             raise ExtractorError('Failed to download URL: %s' % url)
 540
 541         self.report_extraction(video_id)
 542
 543         # Is it an RSS feed?
 544         try:
 545             doc = parse_xml(webpage)
 546             if doc.tag == 'rss':
 547                 return self._extract_rss(url, video_id, doc)
 548         except compat_xml_parse_error:
 549             pass
 550
 551         # Is it a Camtasia project?
 552         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 553         if camtasia_res is not None:
 554             return camtasia_res
 555
 556         # Sometimes embedded video player is hidden behind percent encoding
 557         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 558         # Unescaping the whole page allows to handle those cases in a generic way
 559         webpage = compat_urllib_parse.unquote(webpage)
 560
 561         # it's tempting to parse this further, but you would
 562         # have to take into account all the variations like
 563         #   Video Title - Site Name
 564         #   Site Name | Video Title
 565         #   Video Title - Tagline | Site Name
 566         # and so on and so forth; it's just not practical
 567         video_title = self._html_search_regex(
 568             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 569             default='video')
 570
 571         # Try to detect age limit automatically
 572         age_limit = self._rta_search(webpage)
 573         # And then there are the jokers who advertise that they use RTA,
 574         # but actually don't.
 575         AGE_LIMIT_MARKERS = [
 576             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 577         ]
 578         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 579             age_limit = 18
 580
 581         # video uploader is domain name
 582         video_uploader = self._search_regex(
 583             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 584
 585         # Helper method
 586         def _playlist_from_matches(matches, getter, ie=None):
 587             urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
 588             return self.playlist_result(
 589                 urlrs, playlist_id=video_id, playlist_title=video_title)
 590
 591         # Look for BrightCove:
 592         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 593         if bc_urls:
 594             self.to_screen('Brightcove video detected.')
 595             entries = [{
 596                 '_type': 'url',
 597                 'url': smuggle_url(bc_url, {'Referer': url}),
 598                 'ie_key': 'Brightcove'
 599             } for bc_url in bc_urls]
 600
 601             return {
 602                 '_type': 'playlist',
 603                 'title': video_title,
 604                 'id': video_id,
 605                 'entries': entries,
 606             }
 607
 608         # Look for embedded (iframe) Vimeo player
 609         mobj = re.search(
 610             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 611         if mobj:
 612             player_url = unescapeHTML(mobj.group('url'))
 613             surl = smuggle_url(player_url, {'Referer': url})
 614             return self.url_result(surl, 'Vimeo')
 615
 616         # Look for embedded (swf embed) Vimeo player
 617         mobj = re.search(
 618             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 619         if mobj:
 620             return self.url_result(mobj.group(1), 'Vimeo')
 621
 622         # Look for embedded YouTube player
 623         matches = re.findall(r'''(?x)
 624             (?:
 625                 <iframe[^>]+?src=|
 626                 data-video-url=|
 627                 <embed[^>]+?src=|
 628                 embedSWF\(?:\s*
 629             )
 630             (["\'])
 631                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 632                 (?:embed|v)/.+?)
 633             \1''', webpage)
 634         if matches:
 635             return _playlist_from_matches(
 636                 matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
 637
 638         # Look for embedded Dailymotion player
 639         matches = re.findall(
 640             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 641         if matches:
 642             return _playlist_from_matches(
 643                 matches, lambda m: unescapeHTML(m[1]))
 644
 645         # Look for embedded Wistia player
 646         match = re.search(
 647             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 648         if match:
 649             return {
 650                 '_type': 'url_transparent',
 651                 'url': unescapeHTML(match.group('url')),
 652                 'ie_key': 'Wistia',
 653                 'uploader': video_uploader,
 654                 'title': video_title,
 655                 'id': video_id,
 656             }
 657
 658         # Look for embedded blip.tv player
 659         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 660         if mobj:
 661             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 662         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 663         if mobj:
 664             return self.url_result(mobj.group(1), 'BlipTV')
 665
 666         # Look for embedded condenast player
 667         matches = re.findall(
 668             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 669             webpage)
 670         if matches:
 671             return {
 672                 '_type': 'playlist',
 673                 'entries': [{
 674                     '_type': 'url',
 675                     'ie_key': 'CondeNast',
 676                     'url': ma,
 677                 } for ma in matches],
 678                 'title': video_title,
 679                 'id': video_id,
 680             }
 681
 682         # Look for Bandcamp pages with custom domain
 683         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 684         if mobj is not None:
 685             burl = unescapeHTML(mobj.group(1))
 686             # Don't set the extractor because it can be a track url or an album
 687             return self.url_result(burl)
 688
 689         # Look for embedded Vevo player
 690         mobj = re.search(
 691             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 692         if mobj is not None:
 693             return self.url_result(mobj.group('url'))
 694
 695         # Look for Ooyala videos
 696         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 697              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 698         if mobj is not None:
 699             return OoyalaIE._build_url_result(mobj.group('ec'))
 700
 701         # Look for Aparat videos
 702         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 703         if mobj is not None:
 704             return self.url_result(mobj.group(1), 'Aparat')
 705
 706         # Look for MPORA videos
 707         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 708         if mobj is not None:
 709             return self.url_result(mobj.group(1), 'Mpora')
 710
 711         # Look for embedded NovaMov-based player
 712         mobj = re.search(
 713             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 714                     (?P<url>http://(?:(?:embed|www)\.)?
 715                         (?:novamov\.com|
 716                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 717                            videoweed\.(?:es|com)|
 718                            movshare\.(?:net|sx|ag)|
 719                            divxstage\.(?:eu|net|ch|co|at|ag))
 720                         /embed\.php.+?)\1''', webpage)
 721         if mobj is not None:
 722             return self.url_result(mobj.group('url'))
 723
 724         # Look for embedded Facebook player
 725         mobj = re.search(
 726             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 727         if mobj is not None:
 728             return self.url_result(mobj.group('url'), 'Facebook')
 729
 730         # Look for embedded VK player
 731         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 732         if mobj is not None:
 733             return self.url_result(mobj.group('url'), 'VK')
 734
 735         # Look for embedded ivi player
 736         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 737         if mobj is not None:
 738             return self.url_result(mobj.group('url'), 'Ivi')
 739
 740         # Look for embedded Huffington Post player
 741         mobj = re.search(
 742             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 743         if mobj is not None:
 744             return self.url_result(mobj.group('url'), 'HuffPost')
 745
 746         # Look for embed.ly
 747         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 748         if mobj is not None:
 749             return self.url_result(mobj.group('url'))
 750         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 751         if mobj is not None:
 752             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 753
 754         # Look for funnyordie embed
 755         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 756         if matches:
 757             return _playlist_from_matches(
 758                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 759
 760         # Look for embedded RUTV player
 761         rutv_url = RUTVIE._extract_url(webpage)
 762         if rutv_url:
 763             return self.url_result(rutv_url, 'RUTV')
 764
 765         # Look for embedded TED player
 766         mobj = re.search(
 767             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 768         if mobj is not None:
 769             return self.url_result(mobj.group('url'), 'TED')
 770
 771         # Look for embedded Ustream videos
 772         mobj = re.search(
 773             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 774         if mobj is not None:
 775             return self.url_result(mobj.group('url'), 'Ustream')
 776
 777         # Look for embedded arte.tv player
 778         mobj = re.search(
 779             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 780             webpage)
 781         if mobj is not None:
 782             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 783
 784         # Look for embedded smotri.com player
 785         smotri_url = SmotriIE._extract_url(webpage)
 786         if smotri_url:
 787             return self.url_result(smotri_url, 'Smotri')
 788
 789         # Look for embeded soundcloud player
 790         mobj = re.search(
 791             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 792             webpage)
 793         if mobj is not None:
 794             url = unescapeHTML(mobj.group('url'))
 795             return self.url_result(url)
 796
 797         # Look for embedded vulture.com player
 798         mobj = re.search(
 799             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 800             webpage)
 801         if mobj is not None:
 802             url = unescapeHTML(mobj.group('url'))
 803             return self.url_result(url, ie='Vulture')
 804
 805         # Look for embedded mtvservices player
 806         mobj = re.search(
 807             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 808             webpage)
 809         if mobj is not None:
 810             url = unescapeHTML(mobj.group('url'))
 811             return self.url_result(url, ie='MTVServicesEmbedded')
 812
 813         # Look for embedded yahoo player
 814         mobj = re.search(
 815             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 816             webpage)
 817         if mobj is not None:
 818             return self.url_result(mobj.group('url'), 'Yahoo')
 819
 820         # Look for embedded sbs.com.au player
 821         mobj = re.search(
 822             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 823             webpage)
 824         if mobj is not None:
 825             return self.url_result(mobj.group('url'), 'SBS')
 826
 827         mobj = re.search(
 828             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 829             webpage)
 830         if mobj is not None:
 831             return self.url_result(mobj.group('url'), 'MLB')
 832
 833         # Start with something easy: JW Player in SWFObject
 834         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 835         if not found:
 836             # Look for gorilla-vid style embedding
 837             found = re.findall(r'''(?sx)
 838                 (?:
 839                     jw_plugins|
 840                     JWPlayerOptions|
 841                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 842                 )
 843                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 844         if not found:
 845             # Broaden the search a little bit
 846             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 847         if not found:
 848             # Broaden the findall a little bit: JWPlayer JS loader
 849             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 850         if not found:
 851             # Flow player
 852             found = re.findall(r'''(?xs)
 853                 flowplayer\("[^"]+",\s*
 854                     \{[^}]+?\}\s*,
 855                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 856                         ["']?url["']?\s*:\s*["']([^"']+)["']
 857             ''', webpage)
 858         if not found:
 859             # Try to find twitter cards info
 860             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 861         if not found:
 862             # We look for Open Graph info:
 863             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 864             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 865             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 866             if m_video_type is not None:
 867                 def check_video(vurl):
 868                     vpath = compat_urlparse.urlparse(vurl).path
 869                     vext = determine_ext(vpath)
 870                     return '.' in vpath and vext not in ('swf', 'png', 'jpg')
 871                 found = list(filter(
 872                     check_video,
 873                     re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
 874         if not found:
 875             # HTML5 video
 876             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)
 877         if not found:
 878             found = re.search(
 879                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 880                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
 881                 webpage)
 882             if found:
 883                 new_url = found.group(1)
 884                 self.report_following_redirect(new_url)
 885                 return {
 886                     '_type': 'url',
 887                     'url': new_url,
 888                 }
 889         if not found:
 890             raise ExtractorError('Unsupported URL: %s' % url)
 891
 892         entries = []
 893         for video_url in found:
 894             video_url = compat_urlparse.urljoin(url, video_url)
 895             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 896
 897             # Sometimes, jwplayer extraction will result in a YouTube URL
 898             if YoutubeIE.suitable(video_url):
 899                 entries.append(self.url_result(video_url, 'Youtube'))
 900                 continue
 901
 902             # here's a fun little line of code for you:
 903             video_id = os.path.splitext(video_id)[0]
 904
 905             entries.append({
 906                 'id': video_id,
 907                 'url': video_url,
 908                 'uploader': video_uploader,
 909                 'title': video_title,
 910                 'age_limit': age_limit,
 911             })
 912
 913         if len(entries) == 1:
 914             return entries[0]
 915         else:
 916             for num, e in enumerate(entries, start=1):
 917                 e['title'] = '%s (%d)' % (e['title'], num)
 918             return {
 919                 '_type': 'playlist',
 920                 'entries': entries,
 921             }
 922