_ 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
  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         # Wistia embed
 386         {
 387             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 388             'md5': '8788b683c777a5cf25621eaf286d0c23',
 389             'info_dict': {
 390                 'id': '1cfaf6b7ea',
 391                 'ext': 'mov',
 392                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 393                 'duration': 643.0,
 394                 'filesize': 182808282,
 395                 'uploader': 'education-portal.com',
 396             },
 397         },
 398     ]
 399
 400     def report_download_webpage(self, video_id):
 401         """Report webpage download."""
 402         if not self._downloader.params.get('test', False):
 403             self._downloader.report_warning('Falling back on generic information extractor.')
 404         super(GenericIE, self).report_download_webpage(video_id)
 405
 406     def report_following_redirect(self, new_url):
 407         """Report information extraction."""
 408         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 409
 410     def _extract_rss(self, url, video_id, doc):
 411         playlist_title = doc.find('./channel/title').text
 412         playlist_desc_el = doc.find('./channel/description')
 413         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 414
 415         entries = [{
 416             '_type': 'url',
 417             'url': e.find('link').text,
 418             'title': e.find('title').text,
 419         } for e in doc.findall('./channel/item')]
 420
 421         return {
 422             '_type': 'playlist',
 423             'id': url,
 424             'title': playlist_title,
 425             'description': playlist_desc,
 426             'entries': entries,
 427         }
 428
 429     def _extract_camtasia(self, url, video_id, webpage):
 430         """ Returns None if no camtasia video can be found. """
 431
 432         camtasia_cfg = self._search_regex(
 433             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 434             webpage, 'camtasia configuration file', default=None)
 435         if camtasia_cfg is None:
 436             return None
 437
 438         title = self._html_search_meta('DC.title', webpage, fatal=True)
 439
 440         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 441         camtasia_cfg = self._download_xml(
 442             camtasia_url, video_id,
 443             note='Downloading camtasia configuration',
 444             errnote='Failed to download camtasia configuration')
 445         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 446
 447         entries = []
 448         for n in fileset_node.getchildren():
 449             url_n = n.find('./uri')
 450             if url_n is None:
 451                 continue
 452
 453             entries.append({
 454                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 455                 'title': '%s - %s' % (title, n.tag),
 456                 'url': compat_urlparse.urljoin(url, url_n.text),
 457                 'duration': float_or_none(n.find('./duration').text),
 458             })
 459
 460         return {
 461             '_type': 'playlist',
 462             'entries': entries,
 463             'title': title,
 464         }
 465
 466     def _real_extract(self, url):
 467         if url.startswith('//'):
 468             return {
 469                 '_type': 'url',
 470                 'url': self.http_scheme() + url,
 471             }
 472
 473         parsed_url = compat_urlparse.urlparse(url)
 474         if not parsed_url.scheme:
 475             default_search = self._downloader.params.get('default_search')
 476             if default_search is None:
 477                 default_search = 'fixup_error'
 478
 479             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 480                 if '/' in url:
 481                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 482                     return self.url_result('http://' + url)
 483                 elif default_search != 'fixup_error':
 484                     if default_search == 'auto_warning':
 485                         if re.match(r'^(?:url|URL)$', url):
 486                             raise ExtractorError(
 487                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 488                                 expected=True)
 489                         else:
 490                             self._downloader.report_warning(
 491                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 492                     return self.url_result('ytsearch:' + url)
 493
 494             if default_search in ('error', 'fixup_error'):
 495                 raise ExtractorError(
 496                     ('%r is not a valid URL. '
 497                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 498                     ) % (url, url), expected=True)
 499             else:
 500                 assert ':' in default_search
 501                 return self.url_result(default_search + url)
 502
 503         url, smuggled_data = unsmuggle_url(url)
 504         force_videoid = None
 505         if smuggled_data and 'force_videoid' in smuggled_data:
 506             force_videoid = smuggled_data['force_videoid']
 507             video_id = force_videoid
 508         else:
 509             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 510
 511         self.to_screen('%s: Requesting header' % video_id)
 512
 513         head_req = HEADRequest(url)
 514         response = self._request_webpage(
 515             head_req, video_id,
 516             note=False, errnote='Could not send HEAD request to %s' % url,
 517             fatal=False)
 518
 519         if response is not False:
 520             # Check for redirect
 521             new_url = response.geturl()
 522             if url != new_url:
 523                 self.report_following_redirect(new_url)
 524                 if force_videoid:
 525                     new_url = smuggle_url(
 526                         new_url, {'force_videoid': force_videoid})
 527                 return self.url_result(new_url)
 528
 529             # Check for direct link to a video
 530             content_type = response.headers.get('Content-Type', '')
 531             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 532             if m:
 533                 upload_date = response.headers.get('Last-Modified')
 534                 if upload_date:
 535                     upload_date = unified_strdate(upload_date)
 536                 return {
 537                     'id': video_id,
 538                     'title': os.path.splitext(url_basename(url))[0],
 539                     'formats': [{
 540                         'format_id': m.group('format_id'),
 541                         'url': url,
 542                         'vcodec': 'none' if m.group('type') == 'audio' else None
 543                     }],
 544                     'upload_date': upload_date,
 545                 }
 546
 547         try:
 548             webpage = self._download_webpage(url, video_id)
 549         except ValueError:
 550             # since this is the last-resort InfoExtractor, if
 551             # this error is thrown, it'll be thrown here
 552             raise ExtractorError('Failed to download URL: %s' % url)
 553
 554         self.report_extraction(video_id)
 555
 556         # Is it an RSS feed?
 557         try:
 558             doc = parse_xml(webpage)
 559             if doc.tag == 'rss':
 560                 return self._extract_rss(url, video_id, doc)
 561         except compat_xml_parse_error:
 562             pass
 563
 564         # Is it a Camtasia project?
 565         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 566         if camtasia_res is not None:
 567             return camtasia_res
 568
 569         # Sometimes embedded video player is hidden behind percent encoding
 570         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 571         # Unescaping the whole page allows to handle those cases in a generic way
 572         webpage = compat_urllib_parse.unquote(webpage)
 573
 574         # it's tempting to parse this further, but you would
 575         # have to take into account all the variations like
 576         #   Video Title - Site Name
 577         #   Site Name | Video Title
 578         #   Video Title - Tagline | Site Name
 579         # and so on and so forth; it's just not practical
 580         video_title = self._html_search_regex(
 581             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 582             default='video')
 583
 584         # Try to detect age limit automatically
 585         age_limit = self._rta_search(webpage)
 586         # And then there are the jokers who advertise that they use RTA,
 587         # but actually don't.
 588         AGE_LIMIT_MARKERS = [
 589             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 590         ]
 591         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 592             age_limit = 18
 593
 594         # video uploader is domain name
 595         video_uploader = self._search_regex(
 596             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 597
 598         # Helper method
 599         def _playlist_from_matches(matches, getter, ie=None):
 600             urlrs = orderedSet(
 601                 self.url_result(self._proto_relative_url(getter(m)), ie)
 602                 for m in matches)
 603             return self.playlist_result(
 604                 urlrs, playlist_id=video_id, playlist_title=video_title)
 605
 606         # Look for BrightCove:
 607         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 608         if bc_urls:
 609             self.to_screen('Brightcove video detected.')
 610             entries = [{
 611                 '_type': 'url',
 612                 'url': smuggle_url(bc_url, {'Referer': url}),
 613                 'ie_key': 'Brightcove'
 614             } for bc_url in bc_urls]
 615
 616             return {
 617                 '_type': 'playlist',
 618                 'title': video_title,
 619                 'id': video_id,
 620                 'entries': entries,
 621             }
 622
 623         # Look for embedded (iframe) Vimeo player
 624         mobj = re.search(
 625             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 626         if mobj:
 627             player_url = unescapeHTML(mobj.group('url'))
 628             surl = smuggle_url(player_url, {'Referer': url})
 629             return self.url_result(surl, 'Vimeo')
 630
 631         # Look for embedded (swf embed) Vimeo player
 632         mobj = re.search(
 633             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 634         if mobj:
 635             return self.url_result(mobj.group(1), 'Vimeo')
 636
 637         # Look for embedded YouTube player
 638         matches = re.findall(r'''(?x)
 639             (?:
 640                 <iframe[^>]+?src=|
 641                 data-video-url=|
 642                 <embed[^>]+?src=|
 643                 embedSWF\(?:\s*
 644             )
 645             (["\'])
 646                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 647                 (?:embed|v|p)/.+?)
 648             \1''', webpage)
 649         if matches:
 650             return _playlist_from_matches(
 651                 matches, lambda m: unescapeHTML(m[1]))
 652
 653         # Look for embedded Dailymotion player
 654         matches = re.findall(
 655             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 656         if matches:
 657             return _playlist_from_matches(
 658                 matches, lambda m: unescapeHTML(m[1]))
 659
 660         # Look for embedded Wistia player
 661         match = re.search(
 662             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 663         if match:
 664             return {
 665                 '_type': 'url_transparent',
 666                 'url': unescapeHTML(match.group('url')),
 667                 'ie_key': 'Wistia',
 668                 'uploader': video_uploader,
 669                 'title': video_title,
 670                 'id': video_id,
 671             }
 672         match = re.search(r'(?:id=["\']wistia_|data-wistiaid=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 673         if match:
 674             return {
 675                 '_type': 'url_transparent',
 676                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 677                 'ie_key': 'Wistia',
 678                 'uploader': video_uploader,
 679                 'title': video_title,
 680                 'id': match.group('id')
 681             }
 682
 683         # Look for embedded blip.tv player
 684         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 685         if mobj:
 686             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 687         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 688         if mobj:
 689             return self.url_result(mobj.group(1), 'BlipTV')
 690
 691         # Look for embedded condenast player
 692         matches = re.findall(
 693             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 694             webpage)
 695         if matches:
 696             return {
 697                 '_type': 'playlist',
 698                 'entries': [{
 699                     '_type': 'url',
 700                     'ie_key': 'CondeNast',
 701                     'url': ma,
 702                 } for ma in matches],
 703                 'title': video_title,
 704                 'id': video_id,
 705             }
 706
 707         # Look for Bandcamp pages with custom domain
 708         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 709         if mobj is not None:
 710             burl = unescapeHTML(mobj.group(1))
 711             # Don't set the extractor because it can be a track url or an album
 712             return self.url_result(burl)
 713
 714         # Look for embedded Vevo player
 715         mobj = re.search(
 716             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 717         if mobj is not None:
 718             return self.url_result(mobj.group('url'))
 719
 720         # Look for Ooyala videos
 721         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 722              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 723         if mobj is not None:
 724             return OoyalaIE._build_url_result(mobj.group('ec'))
 725
 726         # Look for Aparat videos
 727         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 728         if mobj is not None:
 729             return self.url_result(mobj.group(1), 'Aparat')
 730
 731         # Look for MPORA videos
 732         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 733         if mobj is not None:
 734             return self.url_result(mobj.group(1), 'Mpora')
 735
 736         # Look for embedded NovaMov-based player
 737         mobj = re.search(
 738             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 739                     (?P<url>http://(?:(?:embed|www)\.)?
 740                         (?:novamov\.com|
 741                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 742                            videoweed\.(?:es|com)|
 743                            movshare\.(?:net|sx|ag)|
 744                            divxstage\.(?:eu|net|ch|co|at|ag))
 745                         /embed\.php.+?)\1''', webpage)
 746         if mobj is not None:
 747             return self.url_result(mobj.group('url'))
 748
 749         # Look for embedded Facebook player
 750         mobj = re.search(
 751             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 752         if mobj is not None:
 753             return self.url_result(mobj.group('url'), 'Facebook')
 754
 755         # Look for embedded VK player
 756         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 757         if mobj is not None:
 758             return self.url_result(mobj.group('url'), 'VK')
 759
 760         # Look for embedded ivi player
 761         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 762         if mobj is not None:
 763             return self.url_result(mobj.group('url'), 'Ivi')
 764
 765         # Look for embedded Huffington Post player
 766         mobj = re.search(
 767             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 768         if mobj is not None:
 769             return self.url_result(mobj.group('url'), 'HuffPost')
 770
 771         # Look for embed.ly
 772         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 773         if mobj is not None:
 774             return self.url_result(mobj.group('url'))
 775         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 776         if mobj is not None:
 777             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 778
 779         # Look for funnyordie embed
 780         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 781         if matches:
 782             return _playlist_from_matches(
 783                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 784
 785         # Look for embedded RUTV player
 786         rutv_url = RUTVIE._extract_url(webpage)
 787         if rutv_url:
 788             return self.url_result(rutv_url, 'RUTV')
 789
 790         # Look for embedded TED player
 791         mobj = re.search(
 792             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 793         if mobj is not None:
 794             return self.url_result(mobj.group('url'), 'TED')
 795
 796         # Look for embedded Ustream videos
 797         mobj = re.search(
 798             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 799         if mobj is not None:
 800             return self.url_result(mobj.group('url'), 'Ustream')
 801
 802         # Look for embedded arte.tv player
 803         mobj = re.search(
 804             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 805             webpage)
 806         if mobj is not None:
 807             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 808
 809         # Look for embedded smotri.com player
 810         smotri_url = SmotriIE._extract_url(webpage)
 811         if smotri_url:
 812             return self.url_result(smotri_url, 'Smotri')
 813
 814         # Look for embeded soundcloud player
 815         mobj = re.search(
 816             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 817             webpage)
 818         if mobj is not None:
 819             url = unescapeHTML(mobj.group('url'))
 820             return self.url_result(url)
 821
 822         # Look for embedded vulture.com player
 823         mobj = re.search(
 824             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 825             webpage)
 826         if mobj is not None:
 827             url = unescapeHTML(mobj.group('url'))
 828             return self.url_result(url, ie='Vulture')
 829
 830         # Look for embedded mtvservices player
 831         mobj = re.search(
 832             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 833             webpage)
 834         if mobj is not None:
 835             url = unescapeHTML(mobj.group('url'))
 836             return self.url_result(url, ie='MTVServicesEmbedded')
 837
 838         # Look for embedded yahoo player
 839         mobj = re.search(
 840             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 841             webpage)
 842         if mobj is not None:
 843             return self.url_result(mobj.group('url'), 'Yahoo')
 844
 845         # Look for embedded sbs.com.au player
 846         mobj = re.search(
 847             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 848             webpage)
 849         if mobj is not None:
 850             return self.url_result(mobj.group('url'), 'SBS')
 851
 852         mobj = re.search(
 853             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 854             webpage)
 855         if mobj is not None:
 856             return self.url_result(mobj.group('url'), 'MLB')
 857
 858         # Start with something easy: JW Player in SWFObject
 859         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 860         if not found:
 861             # Look for gorilla-vid style embedding
 862             found = re.findall(r'''(?sx)
 863                 (?:
 864                     jw_plugins|
 865                     JWPlayerOptions|
 866                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 867                 )
 868                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 869         if not found:
 870             # Broaden the search a little bit
 871             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 872         if not found:
 873             # Broaden the findall a little bit: JWPlayer JS loader
 874             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 875         if not found:
 876             # Flow player
 877             found = re.findall(r'''(?xs)
 878                 flowplayer\("[^"]+",\s*
 879                     \{[^}]+?\}\s*,
 880                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 881                         ["']?url["']?\s*:\s*["']([^"']+)["']
 882             ''', webpage)
 883         if not found:
 884             # Try to find twitter cards info
 885             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 886         if not found:
 887             # We look for Open Graph info:
 888             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 889             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 890             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 891             if m_video_type is not None:
 892                 def check_video(vurl):
 893                     vpath = compat_urlparse.urlparse(vurl).path
 894                     vext = determine_ext(vpath)
 895                     return '.' in vpath and vext not in ('swf', 'png', 'jpg')
 896                 found = list(filter(
 897                     check_video,
 898                     re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
 899         if not found:
 900             # HTML5 video
 901             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)
 902         if not found:
 903             found = re.search(
 904                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 905                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
 906                 webpage)
 907             if found:
 908                 new_url = found.group(1)
 909                 self.report_following_redirect(new_url)
 910                 return {
 911                     '_type': 'url',
 912                     'url': new_url,
 913                 }
 914         if not found:
 915             raise ExtractorError('Unsupported URL: %s' % url)
 916
 917         entries = []
 918         for video_url in found:
 919             video_url = compat_urlparse.urljoin(url, video_url)
 920             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 921
 922             # Sometimes, jwplayer extraction will result in a YouTube URL
 923             if YoutubeIE.suitable(video_url):
 924                 entries.append(self.url_result(video_url, 'Youtube'))
 925                 continue
 926
 927             # here's a fun little line of code for you:
 928             video_id = os.path.splitext(video_id)[0]
 929
 930             entries.append({
 931                 'id': video_id,
 932                 'url': video_url,
 933                 'uploader': video_uploader,
 934                 'title': video_title,
 935                 'age_limit': age_limit,
 936             })
 937
 938         if len(entries) == 1:
 939             return entries[0]
 940         else:
 941             for num, e in enumerate(entries, start=1):
 942                 e['title'] = '%s (%d)' % (e['title'], num)
 943             return {
 944                 '_type': 'playlist',
 945                 'entries': entries,
 946             }
 947