_ 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 ..compat import (
  11     compat_urllib_parse,
  12     compat_urlparse,
  13     compat_xml_parse_error,
  14 )
  15 from ..utils import (
  16     determine_ext,
  17     ExtractorError,
  18     float_or_none,
  19     HEADRequest,
  20     orderedSet,
  21     parse_xml,
  22     smuggle_url,
  23     unescapeHTML,
  24     unified_strdate,
  25     unsmuggle_url,
  26     url_basename,
  27 )
  28 from .brightcove import BrightcoveIE
  29 from .ooyala import OoyalaIE
  30 from .rutv import RUTVIE
  31 from .smotri import SmotriIE
  32 from .condenast import CondeNastIE
  33
  34
  35 class GenericIE(InfoExtractor):
  36     IE_DESC = 'Generic downloader that works on some sites'
  37     _VALID_URL = r'.*'
  38     IE_NAME = 'generic'
  39     _TESTS = [
  40         {
  41             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  42             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  43             'info_dict': {
  44                 'id': '13601338388002',
  45                 'ext': 'mp4',
  46                 'uploader': 'www.hodiho.fr',
  47                 'title': 'R\u00e9gis plante sa Jeep',
  48             }
  49         },
  50         # bandcamp page with custom domain
  51         {
  52             'add_ie': ['Bandcamp'],
  53             'url': 'http://bronyrock.com/track/the-pony-mash',
  54             'info_dict': {
  55                 'id': '3235767654',
  56                 'ext': 'mp3',
  57                 'title': 'The Pony Mash',
  58                 'uploader': 'M_Pallante',
  59             },
  60             'skip': 'There is a limit of 200 free downloads / month for the test song',
  61         },
  62         # embedded brightcove video
  63         # it also tests brightcove videos that need to set the 'Referer' in the
  64         # http requests
  65         {
  66             'add_ie': ['Brightcove'],
  67             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  68             'info_dict': {
  69                 'id': '2765128793001',
  70                 'ext': 'mp4',
  71                 'title': 'Le cours de bourse : l’analyse technique',
  72                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  73                 'uploader': 'BFM BUSINESS',
  74             },
  75             'params': {
  76                 'skip_download': True,
  77             },
  78         },
  79         {
  80             # https://github.com/rg3/youtube-dl/issues/2253
  81             'url': 'http://bcove.me/i6nfkrc3',
  82             'md5': '0ba9446db037002366bab3b3eb30c88c',
  83             'info_dict': {
  84                 'id': '3101154703001',
  85                 'ext': 'mp4',
  86                 'title': 'Still no power',
  87                 'uploader': 'thestar.com',
  88                 '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.',
  89             },
  90             'add_ie': ['Brightcove'],
  91         },
  92         {
  93             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  94             'md5': 'fb973ecf6e4a78a67453647444222983',
  95             'info_dict': {
  96                 'id': '3414141473001',
  97                 'ext': 'mp4',
  98                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  99                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
 100                 'uploader': 'Championat',
 101             },
 102         },
 103         # Direct link to a video
 104         {
 105             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 106             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 107             'info_dict': {
 108                 'id': 'trailer',
 109                 'ext': 'mp4',
 110                 'title': 'trailer',
 111                 'upload_date': '20100513',
 112             }
 113         },
 114         # ooyala video
 115         {
 116             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 117             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 118             'info_dict': {
 119                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 120                 'ext': 'mp4',
 121                 'title': '2cc213299525360.mov',  # that's what we get
 122             },
 123         },
 124         # google redirect
 125         {
 126             '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',
 127             'info_dict': {
 128                 'id': 'cmQHVoWB5FY',
 129                 'ext': 'mp4',
 130                 'upload_date': '20130224',
 131                 'uploader_id': 'TheVerge',
 132                 '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.',
 133                 'uploader': 'The Verge',
 134                 'title': 'First Firefox OS phones side-by-side',
 135             },
 136             'params': {
 137                 'skip_download': False,
 138             }
 139         },
 140         # embed.ly video
 141         {
 142             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 143             'info_dict': {
 144                 'id': '9ODmcdjQcHQ',
 145                 'ext': 'mp4',
 146                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 147                 'upload_date': '20140225',
 148                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 149                 'uploader': 'Tested',
 150                 'uploader_id': 'testedcom',
 151             },
 152             # No need to test YoutubeIE here
 153             'params': {
 154                 'skip_download': True,
 155             },
 156         },
 157         # funnyordie embed
 158         {
 159             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 160             'info_dict': {
 161                 'id': '18e820ec3f',
 162                 'ext': 'mp4',
 163                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 164                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 165             },
 166         },
 167         # RUTV embed
 168         {
 169             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 170             'info_dict': {
 171                 'id': '776940',
 172                 'ext': 'mp4',
 173                 'title': 'Охотское море стало целиком российским',
 174                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 175             },
 176             'params': {
 177                 # m3u8 download
 178                 'skip_download': True,
 179             },
 180         },
 181         # Embedded TED video
 182         {
 183             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 184             'md5': '65fdff94098e4a607385a60c5177c638',
 185             'info_dict': {
 186                 'id': '1969',
 187                 'ext': 'mp4',
 188                 'title': 'Hidden miracles of the natural world',
 189                 'uploader': 'Louie Schwartzberg',
 190                 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
 191             }
 192         },
 193         # Embeded Ustream video
 194         {
 195             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 196             'md5': '27b99cdb639c9b12a79bca876a073417',
 197             'info_dict': {
 198                 'id': '45734260',
 199                 'ext': 'flv',
 200                 'uploader': 'AU SPA:  The NSA and Privacy',
 201                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 202             }
 203         },
 204         # nowvideo embed hidden behind percent encoding
 205         {
 206             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 207             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 208             'info_dict': {
 209                 'id': '06e53103ca9aa',
 210                 'ext': 'flv',
 211                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 212                 'description': 'No description',
 213             },
 214         },
 215         # arte embed
 216         {
 217             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 218             'md5': '7653032cbb25bf6c80d80f217055fa43',
 219             'info_dict': {
 220                 'id': '048195-004_PLUS7-F',
 221                 'ext': 'flv',
 222                 'title': 'X:enius',
 223                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 224                 'upload_date': '20140320',
 225             },
 226             'params': {
 227                 'skip_download': 'Requires rtmpdump'
 228             }
 229         },
 230         # Condé Nast embed
 231         {
 232             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 233             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 234             'info_dict': {
 235                 'id': '53501be369702d3275860000',
 236                 'ext': 'mp4',
 237                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 238             }
 239         },
 240         # Dailymotion embed
 241         {
 242             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 243             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 244             'info_dict': {
 245                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 246                 'ext': 'mp4',
 247                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 248                 'uploader': 'Spi0n',
 249             },
 250             'add_ie': ['Dailymotion'],
 251         },
 252         # YouTube embed
 253         {
 254             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 255             'info_dict': {
 256                 'id': 'FXRb4ykk4S0',
 257                 'ext': 'mp4',
 258                 'title': 'The NBL Auction 2014',
 259                 'uploader': 'BADMINTON England',
 260                 'uploader_id': 'BADMINTONEvents',
 261                 'upload_date': '20140603',
 262                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 263             },
 264             'add_ie': ['Youtube'],
 265             'params': {
 266                 'skip_download': True,
 267             }
 268         },
 269         # MTVSercices embed
 270         {
 271             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 272             'md5': '35727f82f58c76d996fc188f9755b0d5',
 273             'info_dict': {
 274                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 275                 'ext': 'mp4',
 276                 'title': 'Review',
 277                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 278             },
 279         },
 280         # YouTube embed via <data-embed-url="">
 281         {
 282             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 283             'info_dict': {
 284                 'id': '4vAffPZIT44',
 285                 'ext': 'mp4',
 286                 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
 287                 'uploader': 'Gameloft',
 288                 'uploader_id': 'gameloft',
 289                 'upload_date': '20140828',
 290                 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
 291             },
 292             'params': {
 293                 'skip_download': True,
 294             }
 295         },
 296         # Camtasia studio
 297         {
 298             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 299             'playlist': [{
 300                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 301                 'info_dict': {
 302                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 303                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 304                     'ext': 'flv',
 305                     'duration': 2235.90,
 306                 }
 307             }, {
 308                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 309                 'info_dict': {
 310                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 311                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 312                     'ext': 'flv',
 313                     'duration': 2235.93,
 314                 }
 315             }],
 316             'info_dict': {
 317                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 318             }
 319         },
 320         # Flowplayer
 321         {
 322             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 323             'md5': '9d65602bf31c6e20014319c7d07fba27',
 324             'info_dict': {
 325                 'id': '5123ea6d5e5a7',
 326                 'ext': 'mp4',
 327                 'age_limit': 18,
 328                 'uploader': 'www.handjobhub.com',
 329                 'title': 'Busty Blonde Siri Tit Fuck While Wank at HandjobHub.com',
 330             }
 331         },
 332         # RSS feed
 333         {
 334             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 335             'info_dict': {
 336                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 337                 'title': 'Zero Punctuation',
 338                 'description': 're:'
 339             },
 340             'playlist_mincount': 11,
 341         },
 342         # Multiple brightcove videos
 343         # https://github.com/rg3/youtube-dl/issues/2283
 344         {
 345             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 346             'info_dict': {
 347                 'id': 'always-never',
 348                 'title': 'Always / Never - The New Yorker',
 349             },
 350             'playlist_count': 3,
 351             'params': {
 352                 'extract_flat': False,
 353                 'skip_download': True,
 354             }
 355         },
 356         # MLB embed
 357         {
 358             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 359             'md5': '96f09a37e44da40dd083e12d9a683327',
 360             'info_dict': {
 361                 'id': '33322633',
 362                 'ext': 'mp4',
 363                 'title': 'Ump changes call to ball',
 364                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 365                 'duration': 48,
 366                 'timestamp': 1401537900,
 367                 'upload_date': '20140531',
 368                 'thumbnail': 're:^https?://.*\.jpg$',
 369             },
 370         },
 371         # Wistia embed
 372         {
 373             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 374             'md5': '8788b683c777a5cf25621eaf286d0c23',
 375             'info_dict': {
 376                 'id': '1cfaf6b7ea',
 377                 'ext': 'mov',
 378                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 379                 'duration': 643.0,
 380                 'filesize': 182808282,
 381                 'uploader': 'education-portal.com',
 382             },
 383         },
 384         {
 385             'url': 'http://thoughtworks.wistia.com/medias/uxjb0lwrcz',
 386             'md5': 'baf49c2baa8a7de5f3fc145a8506dcd4',
 387             'info_dict': {
 388                 'id': 'uxjb0lwrcz',
 389                 'ext': 'mp4',
 390                 'title': 'Conversation about Hexagonal Rails Part 1 - ThoughtWorks',
 391                 'duration': 1715.0,
 392                 'uploader': 'thoughtworks.wistia.com',
 393             },
 394         },
 395         # Direct download with broken HEAD
 396         {
 397             'url': 'http://ai-radio.org:8000/radio.opus',
 398             'info_dict': {
 399                 'id': 'radio',
 400                 'ext': 'opus',
 401                 'title': 'radio',
 402             },
 403             'params': {
 404                 'skip_download': True,  # infinite live stream
 405             },
 406             'expected_warnings': [
 407                 r'501.*Not Implemented'
 408             ],
 409         },
 410         # Soundcloud embed
 411         {
 412             'url': 'http://nakedsecurity.sophos.com/2014/10/29/sscc-171-are-you-sure-that-1234-is-a-bad-password-podcast/',
 413             'info_dict': {
 414                 'id': '174391317',
 415                 'ext': 'mp3',
 416                 'description': 'md5:ff867d6b555488ad3c52572bb33d432c',
 417                 'uploader': 'Sophos Security',
 418                 'title': 'Chet Chat 171 - Oct 29, 2014',
 419                 'upload_date': '20141029',
 420             }
 421         }
 422     ]
 423
 424     def report_following_redirect(self, new_url):
 425         """Report information extraction."""
 426         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 427
 428     def _extract_rss(self, url, video_id, doc):
 429         playlist_title = doc.find('./channel/title').text
 430         playlist_desc_el = doc.find('./channel/description')
 431         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 432
 433         entries = [{
 434             '_type': 'url',
 435             'url': e.find('link').text,
 436             'title': e.find('title').text,
 437         } for e in doc.findall('./channel/item')]
 438
 439         return {
 440             '_type': 'playlist',
 441             'id': url,
 442             'title': playlist_title,
 443             'description': playlist_desc,
 444             'entries': entries,
 445         }
 446
 447     def _extract_camtasia(self, url, video_id, webpage):
 448         """ Returns None if no camtasia video can be found. """
 449
 450         camtasia_cfg = self._search_regex(
 451             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 452             webpage, 'camtasia configuration file', default=None)
 453         if camtasia_cfg is None:
 454             return None
 455
 456         title = self._html_search_meta('DC.title', webpage, fatal=True)
 457
 458         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 459         camtasia_cfg = self._download_xml(
 460             camtasia_url, video_id,
 461             note='Downloading camtasia configuration',
 462             errnote='Failed to download camtasia configuration')
 463         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 464
 465         entries = []
 466         for n in fileset_node.getchildren():
 467             url_n = n.find('./uri')
 468             if url_n is None:
 469                 continue
 470
 471             entries.append({
 472                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 473                 'title': '%s - %s' % (title, n.tag),
 474                 'url': compat_urlparse.urljoin(url, url_n.text),
 475                 'duration': float_or_none(n.find('./duration').text),
 476             })
 477
 478         return {
 479             '_type': 'playlist',
 480             'entries': entries,
 481             'title': title,
 482         }
 483
 484     def _real_extract(self, url):
 485         if url.startswith('//'):
 486             return {
 487                 '_type': 'url',
 488                 'url': self.http_scheme() + url,
 489             }
 490
 491         parsed_url = compat_urlparse.urlparse(url)
 492         if not parsed_url.scheme:
 493             default_search = self._downloader.params.get('default_search')
 494             if default_search is None:
 495                 default_search = 'fixup_error'
 496
 497             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 498                 if '/' in url:
 499                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 500                     return self.url_result('http://' + url)
 501                 elif default_search != 'fixup_error':
 502                     if default_search == 'auto_warning':
 503                         if re.match(r'^(?:url|URL)$', url):
 504                             raise ExtractorError(
 505                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 506                                 expected=True)
 507                         else:
 508                             self._downloader.report_warning(
 509                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 510                     return self.url_result('ytsearch:' + url)
 511
 512             if default_search in ('error', 'fixup_error'):
 513                 raise ExtractorError(
 514                     ('%r is not a valid URL. '
 515                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 516                     ) % (url, url), expected=True)
 517             else:
 518                 if ':' not in default_search:
 519                     default_search += ':'
 520                 return self.url_result(default_search + url)
 521
 522         url, smuggled_data = unsmuggle_url(url)
 523         force_videoid = None
 524         is_intentional = smuggled_data and smuggled_data.get('to_generic')
 525         if smuggled_data and 'force_videoid' in smuggled_data:
 526             force_videoid = smuggled_data['force_videoid']
 527             video_id = force_videoid
 528         else:
 529             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 530
 531         self.to_screen('%s: Requesting header' % video_id)
 532
 533         head_req = HEADRequest(url)
 534         head_response = self._request_webpage(
 535             head_req, video_id,
 536             note=False, errnote='Could not send HEAD request to %s' % url,
 537             fatal=False)
 538
 539         if head_response is not False:
 540             # Check for redirect
 541             new_url = head_response.geturl()
 542             if url != new_url:
 543                 self.report_following_redirect(new_url)
 544                 if force_videoid:
 545                     new_url = smuggle_url(
 546                         new_url, {'force_videoid': force_videoid})
 547                 return self.url_result(new_url)
 548
 549         full_response = None
 550         if head_response is False:
 551             full_response = self._request_webpage(url, video_id)
 552             head_response = full_response
 553
 554         # Check for direct link to a video
 555         content_type = head_response.headers.get('Content-Type', '')
 556         m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 557         if m:
 558             upload_date = unified_strdate(
 559                 head_response.headers.get('Last-Modified'))
 560             return {
 561                 'id': video_id,
 562                 'title': os.path.splitext(url_basename(url))[0],
 563                 'formats': [{
 564                     'format_id': m.group('format_id'),
 565                     'url': url,
 566                     'vcodec': 'none' if m.group('type') == 'audio' else None
 567                 }],
 568                 'upload_date': upload_date,
 569             }
 570
 571         if not self._downloader.params.get('test', False) and not is_intentional:
 572             self._downloader.report_warning('Falling back on generic information extractor.')
 573
 574         if full_response:
 575             webpage = self._webpage_read_content(full_response, url, video_id)
 576         else:
 577             webpage = self._download_webpage(url, video_id)
 578         self.report_extraction(video_id)
 579
 580         # Is it an RSS feed?
 581         try:
 582             doc = parse_xml(webpage)
 583             if doc.tag == 'rss':
 584                 return self._extract_rss(url, video_id, doc)
 585         except compat_xml_parse_error:
 586             pass
 587
 588         # Is it a Camtasia project?
 589         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 590         if camtasia_res is not None:
 591             return camtasia_res
 592
 593         # Sometimes embedded video player is hidden behind percent encoding
 594         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 595         # Unescaping the whole page allows to handle those cases in a generic way
 596         webpage = compat_urllib_parse.unquote(webpage)
 597
 598         # it's tempting to parse this further, but you would
 599         # have to take into account all the variations like
 600         #   Video Title - Site Name
 601         #   Site Name | Video Title
 602         #   Video Title - Tagline | Site Name
 603         # and so on and so forth; it's just not practical
 604         video_title = self._html_search_regex(
 605             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 606             default='video')
 607
 608         # Try to detect age limit automatically
 609         age_limit = self._rta_search(webpage)
 610         # And then there are the jokers who advertise that they use RTA,
 611         # but actually don't.
 612         AGE_LIMIT_MARKERS = [
 613             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 614         ]
 615         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 616             age_limit = 18
 617
 618         # video uploader is domain name
 619         video_uploader = self._search_regex(
 620             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 621
 622         # Helper method
 623         def _playlist_from_matches(matches, getter, ie=None):
 624             urlrs = orderedSet(
 625                 self.url_result(self._proto_relative_url(getter(m)), ie)
 626                 for m in matches)
 627             return self.playlist_result(
 628                 urlrs, playlist_id=video_id, playlist_title=video_title)
 629
 630         # Look for BrightCove:
 631         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 632         if bc_urls:
 633             self.to_screen('Brightcove video detected.')
 634             entries = [{
 635                 '_type': 'url',
 636                 'url': smuggle_url(bc_url, {'Referer': url}),
 637                 'ie_key': 'Brightcove'
 638             } for bc_url in bc_urls]
 639
 640             return {
 641                 '_type': 'playlist',
 642                 'title': video_title,
 643                 'id': video_id,
 644                 'entries': entries,
 645             }
 646
 647         # Look for embedded (iframe) Vimeo player
 648         mobj = re.search(
 649             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 650         if mobj:
 651             player_url = unescapeHTML(mobj.group('url'))
 652             surl = smuggle_url(player_url, {'Referer': url})
 653             return self.url_result(surl)
 654
 655         # Look for embedded (swf embed) Vimeo player
 656         mobj = re.search(
 657             r'<embed[^>]+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 658         if mobj:
 659             return self.url_result(mobj.group(1))
 660
 661         # Look for embedded YouTube player
 662         matches = re.findall(r'''(?x)
 663             (?:
 664                 <iframe[^>]+?src=|
 665                 data-video-url=|
 666                 <embed[^>]+?src=|
 667                 embedSWF\(?:\s*|
 668                 new\s+SWFObject\(
 669             )
 670             (["\'])
 671                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 672                 (?:embed|v|p)/.+?)
 673             \1''', webpage)
 674         if matches:
 675             return _playlist_from_matches(
 676                 matches, lambda m: unescapeHTML(m[1]))
 677
 678         # Look for embedded Dailymotion player
 679         matches = re.findall(
 680             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 681         if matches:
 682             return _playlist_from_matches(
 683                 matches, lambda m: unescapeHTML(m[1]))
 684
 685         # Look for embedded Dailymotion playlist player (#3822)
 686         m = re.search(
 687             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
 688         if m:
 689             playlists = re.findall(
 690                 r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
 691             if playlists:
 692                 return _playlist_from_matches(
 693                     playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
 694
 695         # Look for embedded Wistia player
 696         match = re.search(
 697             r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 698         if match:
 699             embed_url = self._proto_relative_url(
 700                 unescapeHTML(match.group('url')))
 701             return {
 702                 '_type': 'url_transparent',
 703                 'url': embed_url,
 704                 'ie_key': 'Wistia',
 705                 'uploader': video_uploader,
 706                 'title': video_title,
 707                 'id': video_id,
 708             }
 709
 710         match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 711         if match:
 712             return {
 713                 '_type': 'url_transparent',
 714                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 715                 'ie_key': 'Wistia',
 716                 'uploader': video_uploader,
 717                 'title': video_title,
 718                 'id': match.group('id')
 719             }
 720
 721         # Look for embedded blip.tv player
 722         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 723         if mobj:
 724             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 725         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 726         if mobj:
 727             return self.url_result(mobj.group(1), 'BlipTV')
 728
 729         # Look for embedded condenast player
 730         matches = re.findall(
 731             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 732             webpage)
 733         if matches:
 734             return {
 735                 '_type': 'playlist',
 736                 'entries': [{
 737                     '_type': 'url',
 738                     'ie_key': 'CondeNast',
 739                     'url': ma,
 740                 } for ma in matches],
 741                 'title': video_title,
 742                 'id': video_id,
 743             }
 744
 745         # Look for Bandcamp pages with custom domain
 746         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 747         if mobj is not None:
 748             burl = unescapeHTML(mobj.group(1))
 749             # Don't set the extractor because it can be a track url or an album
 750             return self.url_result(burl)
 751
 752         # Look for embedded Vevo player
 753         mobj = re.search(
 754             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 755         if mobj is not None:
 756             return self.url_result(mobj.group('url'))
 757
 758         # Look for Ooyala videos
 759         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 760              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 761         if mobj is not None:
 762             return OoyalaIE._build_url_result(mobj.group('ec'))
 763
 764         # Look for Aparat videos
 765         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 766         if mobj is not None:
 767             return self.url_result(mobj.group(1), 'Aparat')
 768
 769         # Look for MPORA videos
 770         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 771         if mobj is not None:
 772             return self.url_result(mobj.group(1), 'Mpora')
 773
 774         # Look for embedded NovaMov-based player
 775         mobj = re.search(
 776             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 777                     (?P<url>http://(?:(?:embed|www)\.)?
 778                         (?:novamov\.com|
 779                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 780                            videoweed\.(?:es|com)|
 781                            movshare\.(?:net|sx|ag)|
 782                            divxstage\.(?:eu|net|ch|co|at|ag))
 783                         /embed\.php.+?)\1''', webpage)
 784         if mobj is not None:
 785             return self.url_result(mobj.group('url'))
 786
 787         # Look for embedded Facebook player
 788         mobj = re.search(
 789             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 790         if mobj is not None:
 791             return self.url_result(mobj.group('url'), 'Facebook')
 792
 793         # Look for embedded VK player
 794         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 795         if mobj is not None:
 796             return self.url_result(mobj.group('url'), 'VK')
 797
 798         # Look for embedded ivi player
 799         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 800         if mobj is not None:
 801             return self.url_result(mobj.group('url'), 'Ivi')
 802
 803         # Look for embedded Huffington Post player
 804         mobj = re.search(
 805             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 806         if mobj is not None:
 807             return self.url_result(mobj.group('url'), 'HuffPost')
 808
 809         # Look for embed.ly
 810         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 811         if mobj is not None:
 812             return self.url_result(mobj.group('url'))
 813         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 814         if mobj is not None:
 815             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 816
 817         # Look for funnyordie embed
 818         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 819         if matches:
 820             return _playlist_from_matches(
 821                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 822
 823         # Look for embedded RUTV player
 824         rutv_url = RUTVIE._extract_url(webpage)
 825         if rutv_url:
 826             return self.url_result(rutv_url, 'RUTV')
 827
 828         # Look for embedded TED player
 829         mobj = re.search(
 830             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 831         if mobj is not None:
 832             return self.url_result(mobj.group('url'), 'TED')
 833
 834         # Look for embedded Ustream videos
 835         mobj = re.search(
 836             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 837         if mobj is not None:
 838             return self.url_result(mobj.group('url'), 'Ustream')
 839
 840         # Look for embedded arte.tv player
 841         mobj = re.search(
 842             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 843             webpage)
 844         if mobj is not None:
 845             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 846
 847         # Look for embedded smotri.com player
 848         smotri_url = SmotriIE._extract_url(webpage)
 849         if smotri_url:
 850             return self.url_result(smotri_url, 'Smotri')
 851
 852         # Look for embeded soundcloud player
 853         mobj = re.search(
 854             r'<iframe\s+(?:[a-zA-Z0-9_-]+="[^"]+"\s+)*src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 855             webpage)
 856         if mobj is not None:
 857             url = unescapeHTML(mobj.group('url'))
 858             return self.url_result(url)
 859
 860         # Look for embedded vulture.com player
 861         mobj = re.search(
 862             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 863             webpage)
 864         if mobj is not None:
 865             url = unescapeHTML(mobj.group('url'))
 866             return self.url_result(url, ie='Vulture')
 867
 868         # Look for embedded mtvservices player
 869         mobj = re.search(
 870             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 871             webpage)
 872         if mobj is not None:
 873             url = unescapeHTML(mobj.group('url'))
 874             return self.url_result(url, ie='MTVServicesEmbedded')
 875
 876         # Look for embedded yahoo player
 877         mobj = re.search(
 878             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 879             webpage)
 880         if mobj is not None:
 881             return self.url_result(mobj.group('url'), 'Yahoo')
 882
 883         # Look for embedded sbs.com.au player
 884         mobj = re.search(
 885             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 886             webpage)
 887         if mobj is not None:
 888             return self.url_result(mobj.group('url'), 'SBS')
 889
 890         mobj = re.search(
 891             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m(?:lb)?\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 892             webpage)
 893         if mobj is not None:
 894             return self.url_result(mobj.group('url'), 'MLB')
 895
 896         mobj = re.search(
 897             r'<iframe[^>]+?src=(["\'])(?P<url>%s)\1' % CondeNastIE.EMBED_URL,
 898             webpage)
 899         if mobj is not None:
 900             return self.url_result(self._proto_relative_url(mobj.group('url'), scheme='http:'), 'CondeNast')
 901
 902         def check_video(vurl):
 903             vpath = compat_urlparse.urlparse(vurl).path
 904             vext = determine_ext(vpath)
 905             return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
 906
 907         def filter_video(urls):
 908             return list(filter(check_video, urls))
 909
 910         # Start with something easy: JW Player in SWFObject
 911         found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
 912         if not found:
 913             # Look for gorilla-vid style embedding
 914             found = filter_video(re.findall(r'''(?sx)
 915                 (?:
 916                     jw_plugins|
 917                     JWPlayerOptions|
 918                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 919                 )
 920                 .*?file\s*:\s*["\'](.*?)["\']''', webpage))
 921         if not found:
 922             # Broaden the search a little bit
 923             found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
 924         if not found:
 925             # Broaden the findall a little bit: JWPlayer JS loader
 926             found = filter_video(re.findall(
 927                 r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
 928         if not found:
 929             # Flow player
 930             found = filter_video(re.findall(r'''(?xs)
 931                 flowplayer\("[^"]+",\s*
 932                     \{[^}]+?\}\s*,
 933                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 934                         ["']?url["']?\s*:\s*["']([^"']+)["']
 935             ''', webpage))
 936         if not found:
 937             # Try to find twitter cards info
 938             found = filter_video(re.findall(
 939                 r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
 940         if not found:
 941             # We look for Open Graph info:
 942             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 943             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 944             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 945             if m_video_type is not None:
 946                 found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
 947         if not found:
 948             # HTML5 video
 949             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]*)?\s+src="([^"]+)"', webpage)
 950         if not found:
 951             found = re.search(
 952                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 953                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
 954                 webpage)
 955             if found:
 956                 new_url = found.group(1)
 957                 self.report_following_redirect(new_url)
 958                 return {
 959                     '_type': 'url',
 960                     'url': new_url,
 961                 }
 962         if not found:
 963             raise ExtractorError('Unsupported URL: %s' % url)
 964
 965         entries = []
 966         for video_url in found:
 967             video_url = compat_urlparse.urljoin(url, video_url)
 968             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 969
 970             # Sometimes, jwplayer extraction will result in a YouTube URL
 971             if YoutubeIE.suitable(video_url):
 972                 entries.append(self.url_result(video_url, 'Youtube'))
 973                 continue
 974
 975             # here's a fun little line of code for you:
 976             video_id = os.path.splitext(video_id)[0]
 977
 978             entries.append({
 979                 'id': video_id,
 980                 'url': video_url,
 981                 'uploader': video_uploader,
 982                 'title': video_title,
 983                 'age_limit': age_limit,
 984             })
 985
 986         if len(entries) == 1:
 987             return entries[0]
 988         else:
 989             for num, e in enumerate(entries, start=1):
 990                 e['title'] = '%s (%d)' % (e['title'], num)
 991             return {
 992                 '_type': 'playlist',
 993                 'entries': entries,
 994             }
 995