_ Git - youtube-dl/blob - 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 from .condenast import CondeNastIE
  32
  33
  34 class GenericIE(InfoExtractor):
  35     IE_DESC = 'Generic downloader that works on some sites'
  36     _VALID_URL = r'.*'
  37     IE_NAME = 'generic'
  38     _TESTS = [
  39         {
  40             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  41             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  42             'info_dict': {
  43                 'id': '13601338388002',
  44                 'ext': 'mp4',
  45                 'uploader': 'www.hodiho.fr',
  46                 'title': 'R\u00e9gis plante sa Jeep',
  47             }
  48         },
  49         # bandcamp page with custom domain
  50         {
  51             'add_ie': ['Bandcamp'],
  52             'url': 'http://bronyrock.com/track/the-pony-mash',
  53             'info_dict': {
  54                 'id': '3235767654',
  55                 'ext': 'mp3',
  56                 'title': 'The Pony Mash',
  57                 'uploader': 'M_Pallante',
  58             },
  59             'skip': 'There is a limit of 200 free downloads / month for the test song',
  60         },
  61         # embedded brightcove video
  62         # it also tests brightcove videos that need to set the 'Referer' in the
  63         # http requests
  64         {
  65             'add_ie': ['Brightcove'],
  66             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  67             'info_dict': {
  68                 'id': '2765128793001',
  69                 'ext': 'mp4',
  70                 'title': 'Le cours de bourse : l’analyse technique',
  71                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  72                 'uploader': 'BFM BUSINESS',
  73             },
  74             'params': {
  75                 'skip_download': True,
  76             },
  77         },
  78         {
  79             # https://github.com/rg3/youtube-dl/issues/2253
  80             'url': 'http://bcove.me/i6nfkrc3',
  81             'md5': '0ba9446db037002366bab3b3eb30c88c',
  82             'info_dict': {
  83                 'id': '3101154703001',
  84                 'ext': 'mp4',
  85                 'title': 'Still no power',
  86                 'uploader': 'thestar.com',
  87                 '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.',
  88             },
  89             'add_ie': ['Brightcove'],
  90         },
  91         {
  92             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  93             'md5': 'fb973ecf6e4a78a67453647444222983',
  94             'info_dict': {
  95                 'id': '3414141473001',
  96                 'ext': 'mp4',
  97                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  98                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
  99                 'uploader': 'Championat',
 100             },
 101         },
 102         # Direct link to a video
 103         {
 104             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 105             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 106             'info_dict': {
 107                 'id': 'trailer',
 108                 'ext': 'mp4',
 109                 'title': 'trailer',
 110                 'upload_date': '20100513',
 111             }
 112         },
 113         # ooyala video
 114         {
 115             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 116             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 117             'info_dict': {
 118                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 119                 'ext': 'mp4',
 120                 'title': '2cc213299525360.mov',  # that's what we get
 121             },
 122         },
 123         # google redirect
 124         {
 125             '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',
 126             'info_dict': {
 127                 'id': 'cmQHVoWB5FY',
 128                 'ext': 'mp4',
 129                 'upload_date': '20130224',
 130                 'uploader_id': 'TheVerge',
 131                 '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.',
 132                 'uploader': 'The Verge',
 133                 'title': 'First Firefox OS phones side-by-side',
 134             },
 135             'params': {
 136                 'skip_download': False,
 137             }
 138         },
 139         # embed.ly video
 140         {
 141             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 142             'info_dict': {
 143                 'id': '9ODmcdjQcHQ',
 144                 'ext': 'mp4',
 145                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 146                 'upload_date': '20140225',
 147                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 148                 'uploader': 'Tested',
 149                 'uploader_id': 'testedcom',
 150             },
 151             # No need to test YoutubeIE here
 152             'params': {
 153                 'skip_download': True,
 154             },
 155         },
 156         # funnyordie embed
 157         {
 158             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 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': '65fdff94098e4a607385a60c5177c638',
 184             'info_dict': {
 185                 'id': '1969',
 186                 'ext': 'mp4',
 187                 'title': 'Hidden miracles of the natural world',
 188                 'uploader': 'Louie Schwartzberg',
 189                 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
 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         # Condé Nast embed
 230         {
 231             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 232             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 233             'info_dict': {
 234                 'id': '53501be369702d3275860000',
 235                 'ext': 'mp4',
 236                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 237             }
 238         },
 239         # Dailymotion embed
 240         {
 241             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 242             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 243             'info_dict': {
 244                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 245                 'ext': 'mp4',
 246                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 247                 'uploader': 'Spi0n',
 248             },
 249             'add_ie': ['Dailymotion'],
 250         },
 251         # YouTube embed
 252         {
 253             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 254             'info_dict': {
 255                 'id': 'FXRb4ykk4S0',
 256                 'ext': 'mp4',
 257                 'title': 'The NBL Auction 2014',
 258                 'uploader': 'BADMINTON England',
 259                 'uploader_id': 'BADMINTONEvents',
 260                 'upload_date': '20140603',
 261                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 262             },
 263             'add_ie': ['Youtube'],
 264             'params': {
 265                 'skip_download': True,
 266             }
 267         },
 268         # MTVSercices embed
 269         {
 270             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 271             'md5': '35727f82f58c76d996fc188f9755b0d5',
 272             'info_dict': {
 273                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 274                 'ext': 'mp4',
 275                 'title': 'Review',
 276                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 277             },
 278         },
 279         # YouTube embed via <data-embed-url="">
 280         {
 281             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 282             'info_dict': {
 283                 'id': '4vAffPZIT44',
 284                 'ext': 'mp4',
 285                 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
 286                 'uploader': 'Gameloft',
 287                 'uploader_id': 'gameloft',
 288                 'upload_date': '20140828',
 289                 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
 290             },
 291             'params': {
 292                 'skip_download': True,
 293             }
 294         },
 295         # Camtasia studio
 296         {
 297             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 298             'playlist': [{
 299                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 300                 'info_dict': {
 301                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 302                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 303                     'ext': 'flv',
 304                     'duration': 2235.90,
 305                 }
 306             }, {
 307                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 308                 'info_dict': {
 309                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 310                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 311                     'ext': 'flv',
 312                     'duration': 2235.93,
 313                 }
 314             }],
 315             'info_dict': {
 316                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 317             }
 318         },
 319         # Flowplayer
 320         {
 321             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 322             'md5': '9d65602bf31c6e20014319c7d07fba27',
 323             'info_dict': {
 324                 'id': '5123ea6d5e5a7',
 325                 'ext': 'mp4',
 326                 'age_limit': 18,
 327                 'uploader': 'www.handjobhub.com',
 328                 'title': 'Busty Blonde Siri Tit Fuck While Wank at Handjob Hub',
 329             }
 330         },
 331         # RSS feed
 332         {
 333             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 334             'info_dict': {
 335                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 336                 'title': 'Zero Punctuation',
 337                 'description': 're:'
 338             },
 339             'playlist_mincount': 11,
 340         },
 341         # Multiple brightcove videos
 342         # https://github.com/rg3/youtube-dl/issues/2283
 343         {
 344             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 345             'info_dict': {
 346                 'id': 'always-never',
 347                 'title': 'Always / Never - The New Yorker',
 348             },
 349             'playlist_count': 3,
 350             'params': {
 351                 'extract_flat': False,
 352                 'skip_download': True,
 353             }
 354         },
 355         # MLB embed
 356         {
 357             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 358             'md5': '96f09a37e44da40dd083e12d9a683327',
 359             'info_dict': {
 360                 'id': '33322633',
 361                 'ext': 'mp4',
 362                 'title': 'Ump changes call to ball',
 363                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 364                 'duration': 48,
 365                 'timestamp': 1401537900,
 366                 'upload_date': '20140531',
 367                 'thumbnail': 're:^https?://.*\.jpg$',
 368             },
 369         },
 370         # Wistia embed
 371         {
 372             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 373             'md5': '8788b683c777a5cf25621eaf286d0c23',
 374             'info_dict': {
 375                 'id': '1cfaf6b7ea',
 376                 'ext': 'mov',
 377                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 378                 'duration': 643.0,
 379                 'filesize': 182808282,
 380                 'uploader': 'education-portal.com',
 381             },
 382         },
 383         {
 384             'url': 'http://thoughtworks.wistia.com/medias/uxjb0lwrcz',
 385             'md5': 'baf49c2baa8a7de5f3fc145a8506dcd4',
 386             'info_dict': {
 387                 'id': 'uxjb0lwrcz',
 388                 'ext': 'mp4',
 389                 'title': 'Conversation about Hexagonal Rails Part 1 - ThoughtWorks',
 390                 'duration': 1715.0,
 391                 'uploader': 'thoughtworks.wistia.com',
 392             },
 393         },
 394     ]
 395
 396     def report_following_redirect(self, new_url):
 397         """Report information extraction."""
 398         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 399
 400     def _extract_rss(self, url, video_id, doc):
 401         playlist_title = doc.find('./channel/title').text
 402         playlist_desc_el = doc.find('./channel/description')
 403         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 404
 405         entries = [{
 406             '_type': 'url',
 407             'url': e.find('link').text,
 408             'title': e.find('title').text,
 409         } for e in doc.findall('./channel/item')]
 410
 411         return {
 412             '_type': 'playlist',
 413             'id': url,
 414             'title': playlist_title,
 415             'description': playlist_desc,
 416             'entries': entries,
 417         }
 418
 419     def _extract_camtasia(self, url, video_id, webpage):
 420         """ Returns None if no camtasia video can be found. """
 421
 422         camtasia_cfg = self._search_regex(
 423             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 424             webpage, 'camtasia configuration file', default=None)
 425         if camtasia_cfg is None:
 426             return None
 427
 428         title = self._html_search_meta('DC.title', webpage, fatal=True)
 429
 430         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 431         camtasia_cfg = self._download_xml(
 432             camtasia_url, video_id,
 433             note='Downloading camtasia configuration',
 434             errnote='Failed to download camtasia configuration')
 435         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 436
 437         entries = []
 438         for n in fileset_node.getchildren():
 439             url_n = n.find('./uri')
 440             if url_n is None:
 441                 continue
 442
 443             entries.append({
 444                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 445                 'title': '%s - %s' % (title, n.tag),
 446                 'url': compat_urlparse.urljoin(url, url_n.text),
 447                 'duration': float_or_none(n.find('./duration').text),
 448             })
 449
 450         return {
 451             '_type': 'playlist',
 452             'entries': entries,
 453             'title': title,
 454         }
 455
 456     def _real_extract(self, url):
 457         if url.startswith('//'):
 458             return {
 459                 '_type': 'url',
 460                 'url': self.http_scheme() + url,
 461             }
 462
 463         parsed_url = compat_urlparse.urlparse(url)
 464         if not parsed_url.scheme:
 465             default_search = self._downloader.params.get('default_search')
 466             if default_search is None:
 467                 default_search = 'fixup_error'
 468
 469             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 470                 if '/' in url:
 471                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 472                     return self.url_result('http://' + url)
 473                 elif default_search != 'fixup_error':
 474                     if default_search == 'auto_warning':
 475                         if re.match(r'^(?:url|URL)$', url):
 476                             raise ExtractorError(
 477                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 478                                 expected=True)
 479                         else:
 480                             self._downloader.report_warning(
 481                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 482                     return self.url_result('ytsearch:' + url)
 483
 484             if default_search in ('error', 'fixup_error'):
 485                 raise ExtractorError(
 486                     ('%r is not a valid URL. '
 487                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 488                     ) % (url, url), expected=True)
 489             else:
 490                 if ':' not in default_search:
 491                     default_search += ':'
 492                 return self.url_result(default_search + url)
 493
 494         url, smuggled_data = unsmuggle_url(url)
 495         force_videoid = None
 496         is_intentional = smuggled_data and smuggled_data.get('to_generic')
 497         if smuggled_data and 'force_videoid' in smuggled_data:
 498             force_videoid = smuggled_data['force_videoid']
 499             video_id = force_videoid
 500         else:
 501             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 502
 503         self.to_screen('%s: Requesting header' % video_id)
 504
 505         head_req = HEADRequest(url)
 506         head_response = self._request_webpage(
 507             head_req, video_id,
 508             note=False, errnote='Could not send HEAD request to %s' % url,
 509             fatal=False)
 510
 511         if head_response is not False:
 512             # Check for redirect
 513             new_url = head_response.geturl()
 514             if url != new_url:
 515                 self.report_following_redirect(new_url)
 516                 if force_videoid:
 517                     new_url = smuggle_url(
 518                         new_url, {'force_videoid': force_videoid})
 519                 return self.url_result(new_url)
 520
 521         full_response = None
 522         if head_response is False:
 523             full_response = self._request_webpage(url, video_id)
 524             head_response = full_response
 525
 526         # Check for direct link to a video
 527         content_type = head_response.headers.get('Content-Type', '')
 528         m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 529         if m:
 530             upload_date = unified_strdate(
 531                 head_response.headers.get('Last-Modified'))
 532             return {
 533                 'id': video_id,
 534                 'title': os.path.splitext(url_basename(url))[0],
 535                 'formats': [{
 536                     'format_id': m.group('format_id'),
 537                     'url': url,
 538                     'vcodec': 'none' if m.group('type') == 'audio' else None
 539                 }],
 540                 'upload_date': upload_date,
 541             }
 542
 543         if not self._downloader.params.get('test', False) and not is_intentional:
 544             self._downloader.report_warning('Falling back on generic information extractor.')
 545
 546         if full_response:
 547             webpage = _webpage_read_content(url, video_id)
 548         else:
 549             webpage = self._download_webpage(url, video_id)
 550         self.report_extraction(video_id)
 551
 552         # Is it an RSS feed?
 553         try:
 554             doc = parse_xml(webpage)
 555             if doc.tag == 'rss':
 556                 return self._extract_rss(url, video_id, doc)
 557         except compat_xml_parse_error:
 558             pass
 559
 560         # Is it a Camtasia project?
 561         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 562         if camtasia_res is not None:
 563             return camtasia_res
 564
 565         # Sometimes embedded video player is hidden behind percent encoding
 566         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 567         # Unescaping the whole page allows to handle those cases in a generic way
 568         webpage = compat_urllib_parse.unquote(webpage)
 569
 570         # it's tempting to parse this further, but you would
 571         # have to take into account all the variations like
 572         #   Video Title - Site Name
 573         #   Site Name | Video Title
 574         #   Video Title - Tagline | Site Name
 575         # and so on and so forth; it's just not practical
 576         video_title = self._html_search_regex(
 577             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 578             default='video')
 579
 580         # Try to detect age limit automatically
 581         age_limit = self._rta_search(webpage)
 582         # And then there are the jokers who advertise that they use RTA,
 583         # but actually don't.
 584         AGE_LIMIT_MARKERS = [
 585             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 586         ]
 587         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 588             age_limit = 18
 589
 590         # video uploader is domain name
 591         video_uploader = self._search_regex(
 592             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 593
 594         # Helper method
 595         def _playlist_from_matches(matches, getter, ie=None):
 596             urlrs = orderedSet(
 597                 self.url_result(self._proto_relative_url(getter(m)), ie)
 598                 for m in matches)
 599             return self.playlist_result(
 600                 urlrs, playlist_id=video_id, playlist_title=video_title)
 601
 602         # Look for BrightCove:
 603         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 604         if bc_urls:
 605             self.to_screen('Brightcove video detected.')
 606             entries = [{
 607                 '_type': 'url',
 608                 'url': smuggle_url(bc_url, {'Referer': url}),
 609                 'ie_key': 'Brightcove'
 610             } for bc_url in bc_urls]
 611
 612             return {
 613                 '_type': 'playlist',
 614                 'title': video_title,
 615                 'id': video_id,
 616                 'entries': entries,
 617             }
 618
 619         # Look for embedded (iframe) Vimeo player
 620         mobj = re.search(
 621             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 622         if mobj:
 623             player_url = unescapeHTML(mobj.group('url'))
 624             surl = smuggle_url(player_url, {'Referer': url})
 625             return self.url_result(surl)
 626
 627         # Look for embedded (swf embed) Vimeo player
 628         mobj = re.search(
 629             r'<embed[^>]+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 630         if mobj:
 631             return self.url_result(mobj.group(1))
 632
 633         # Look for embedded YouTube player
 634         matches = re.findall(r'''(?x)
 635             (?:
 636                 <iframe[^>]+?src=|
 637                 data-video-url=|
 638                 <embed[^>]+?src=|
 639                 embedSWF\(?:\s*|
 640                 new\s+SWFObject\(
 641             )
 642             (["\'])
 643                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 644                 (?:embed|v|p)/.+?)
 645             \1''', webpage)
 646         if matches:
 647             return _playlist_from_matches(
 648                 matches, lambda m: unescapeHTML(m[1]))
 649
 650         # Look for embedded Dailymotion player
 651         matches = re.findall(
 652             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 653         if matches:
 654             return _playlist_from_matches(
 655                 matches, lambda m: unescapeHTML(m[1]))
 656
 657         # Look for embedded Dailymotion playlist player (#3822)
 658         m = re.search(
 659             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
 660         if m:
 661             playlists = re.findall(
 662                 r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
 663             if playlists:
 664                 return _playlist_from_matches(
 665                     playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
 666
 667         # Look for embedded Wistia player
 668         match = re.search(
 669             r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 670         if match:
 671             embed_url = self._proto_relative_url(
 672                 unescapeHTML(match.group('url')))
 673             return {
 674                 '_type': 'url_transparent',
 675                 'url': embed_url,
 676                 'ie_key': 'Wistia',
 677                 'uploader': video_uploader,
 678                 'title': video_title,
 679                 'id': video_id,
 680             }
 681
 682         match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 683         if match:
 684             return {
 685                 '_type': 'url_transparent',
 686                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 687                 'ie_key': 'Wistia',
 688                 'uploader': video_uploader,
 689                 'title': video_title,
 690                 'id': match.group('id')
 691             }
 692
 693         # Look for embedded blip.tv player
 694         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 695         if mobj:
 696             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 697         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 698         if mobj:
 699             return self.url_result(mobj.group(1), 'BlipTV')
 700
 701         # Look for embedded condenast player
 702         matches = re.findall(
 703             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 704             webpage)
 705         if matches:
 706             return {
 707                 '_type': 'playlist',
 708                 'entries': [{
 709                     '_type': 'url',
 710                     'ie_key': 'CondeNast',
 711                     'url': ma,
 712                 } for ma in matches],
 713                 'title': video_title,
 714                 'id': video_id,
 715             }
 716
 717         # Look for Bandcamp pages with custom domain
 718         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 719         if mobj is not None:
 720             burl = unescapeHTML(mobj.group(1))
 721             # Don't set the extractor because it can be a track url or an album
 722             return self.url_result(burl)
 723
 724         # Look for embedded Vevo player
 725         mobj = re.search(
 726             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 727         if mobj is not None:
 728             return self.url_result(mobj.group('url'))
 729
 730         # Look for Ooyala videos
 731         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 732              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 733         if mobj is not None:
 734             return OoyalaIE._build_url_result(mobj.group('ec'))
 735
 736         # Look for Aparat videos
 737         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 738         if mobj is not None:
 739             return self.url_result(mobj.group(1), 'Aparat')
 740
 741         # Look for MPORA videos
 742         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 743         if mobj is not None:
 744             return self.url_result(mobj.group(1), 'Mpora')
 745
 746         # Look for embedded NovaMov-based player
 747         mobj = re.search(
 748             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 749                     (?P<url>http://(?:(?:embed|www)\.)?
 750                         (?:novamov\.com|
 751                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 752                            videoweed\.(?:es|com)|
 753                            movshare\.(?:net|sx|ag)|
 754                            divxstage\.(?:eu|net|ch|co|at|ag))
 755                         /embed\.php.+?)\1''', webpage)
 756         if mobj is not None:
 757             return self.url_result(mobj.group('url'))
 758
 759         # Look for embedded Facebook player
 760         mobj = re.search(
 761             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 762         if mobj is not None:
 763             return self.url_result(mobj.group('url'), 'Facebook')
 764
 765         # Look for embedded VK player
 766         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 767         if mobj is not None:
 768             return self.url_result(mobj.group('url'), 'VK')
 769
 770         # Look for embedded ivi player
 771         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 772         if mobj is not None:
 773             return self.url_result(mobj.group('url'), 'Ivi')
 774
 775         # Look for embedded Huffington Post player
 776         mobj = re.search(
 777             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 778         if mobj is not None:
 779             return self.url_result(mobj.group('url'), 'HuffPost')
 780
 781         # Look for embed.ly
 782         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 783         if mobj is not None:
 784             return self.url_result(mobj.group('url'))
 785         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 786         if mobj is not None:
 787             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 788
 789         # Look for funnyordie embed
 790         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 791         if matches:
 792             return _playlist_from_matches(
 793                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 794
 795         # Look for embedded RUTV player
 796         rutv_url = RUTVIE._extract_url(webpage)
 797         if rutv_url:
 798             return self.url_result(rutv_url, 'RUTV')
 799
 800         # Look for embedded TED player
 801         mobj = re.search(
 802             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 803         if mobj is not None:
 804             return self.url_result(mobj.group('url'), 'TED')
 805
 806         # Look for embedded Ustream videos
 807         mobj = re.search(
 808             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 809         if mobj is not None:
 810             return self.url_result(mobj.group('url'), 'Ustream')
 811
 812         # Look for embedded arte.tv player
 813         mobj = re.search(
 814             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 815             webpage)
 816         if mobj is not None:
 817             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 818
 819         # Look for embedded smotri.com player
 820         smotri_url = SmotriIE._extract_url(webpage)
 821         if smotri_url:
 822             return self.url_result(smotri_url, 'Smotri')
 823
 824         # Look for embeded soundcloud player
 825         mobj = re.search(
 826             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 827             webpage)
 828         if mobj is not None:
 829             url = unescapeHTML(mobj.group('url'))
 830             return self.url_result(url)
 831
 832         # Look for embedded vulture.com player
 833         mobj = re.search(
 834             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 835             webpage)
 836         if mobj is not None:
 837             url = unescapeHTML(mobj.group('url'))
 838             return self.url_result(url, ie='Vulture')
 839
 840         # Look for embedded mtvservices player
 841         mobj = re.search(
 842             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 843             webpage)
 844         if mobj is not None:
 845             url = unescapeHTML(mobj.group('url'))
 846             return self.url_result(url, ie='MTVServicesEmbedded')
 847
 848         # Look for embedded yahoo player
 849         mobj = re.search(
 850             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 851             webpage)
 852         if mobj is not None:
 853             return self.url_result(mobj.group('url'), 'Yahoo')
 854
 855         # Look for embedded sbs.com.au player
 856         mobj = re.search(
 857             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 858             webpage)
 859         if mobj is not None:
 860             return self.url_result(mobj.group('url'), 'SBS')
 861
 862         mobj = re.search(
 863             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 864             webpage)
 865         if mobj is not None:
 866             return self.url_result(mobj.group('url'), 'MLB')
 867
 868         mobj = re.search(
 869             r'<iframe[^>]+?src=(["\'])(?P<url>%s)\1' % CondeNastIE.EMBED_URL,
 870             webpage)
 871         if mobj is not None:
 872             return self.url_result(self._proto_relative_url(mobj.group('url'), scheme='http:'), 'CondeNast')
 873
 874         def check_video(vurl):
 875             vpath = compat_urlparse.urlparse(vurl).path
 876             vext = determine_ext(vpath)
 877             return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
 878
 879         def filter_video(urls):
 880             return list(filter(check_video, urls))
 881
 882         # Start with something easy: JW Player in SWFObject
 883         found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
 884         if not found:
 885             # Look for gorilla-vid style embedding
 886             found = filter_video(re.findall(r'''(?sx)
 887                 (?:
 888                     jw_plugins|
 889                     JWPlayerOptions|
 890                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 891                 )
 892                 .*?file\s*:\s*["\'](.*?)["\']''', webpage))
 893         if not found:
 894             # Broaden the search a little bit
 895             found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
 896         if not found:
 897             # Broaden the findall a little bit: JWPlayer JS loader
 898             found = filter_video(re.findall(
 899                 r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
 900         if not found:
 901             # Flow player
 902             found = filter_video(re.findall(r'''(?xs)
 903                 flowplayer\("[^"]+",\s*
 904                     \{[^}]+?\}\s*,
 905                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 906                         ["']?url["']?\s*:\s*["']([^"']+)["']
 907             ''', webpage))
 908         if not found:
 909             # Try to find twitter cards info
 910             found = filter_video(re.findall(
 911                 r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
 912         if not found:
 913             # We look for Open Graph info:
 914             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 915             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 916             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 917             if m_video_type is not None:
 918                 found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
 919         if not found:
 920             # HTML5 video
 921             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)
 922         if not found:
 923             found = re.search(
 924                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 925                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
 926                 webpage)
 927             if found:
 928                 new_url = found.group(1)
 929                 self.report_following_redirect(new_url)
 930                 return {
 931                     '_type': 'url',
 932                     'url': new_url,
 933                 }
 934         if not found:
 935             raise ExtractorError('Unsupported URL: %s' % url)
 936
 937         entries = []
 938         for video_url in found:
 939             video_url = compat_urlparse.urljoin(url, video_url)
 940             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 941
 942             # Sometimes, jwplayer extraction will result in a YouTube URL
 943             if YoutubeIE.suitable(video_url):
 944                 entries.append(self.url_result(video_url, 'Youtube'))
 945                 continue
 946
 947             # here's a fun little line of code for you:
 948             video_id = os.path.splitext(video_id)[0]
 949
 950             entries.append({
 951                 'id': video_id,
 952                 'url': video_url,
 953                 'uploader': video_uploader,
 954                 'title': video_title,
 955                 'age_limit': age_limit,
 956             })
 957
 958         if len(entries) == 1:
 959             return entries[0]
 960         else:
 961             for num, e in enumerate(entries, start=1):
 962                 e['title'] = '%s (%d)' % (e['title'], num)
 963             return {
 964                 '_type': 'playlist',
 965                 'entries': entries,
 966             }
 967