_ 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             'info_dict': {
 159                 'id': '18e820ec3f',
 160                 'ext': 'mp4',
 161                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 162                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 163             },
 164         },
 165         # RUTV embed
 166         {
 167             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 168             'info_dict': {
 169                 'id': '776940',
 170                 'ext': 'mp4',
 171                 'title': 'Охотское море стало целиком российским',
 172                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 173             },
 174             'params': {
 175                 # m3u8 download
 176                 'skip_download': True,
 177             },
 178         },
 179         # Embedded TED video
 180         {
 181             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 182             'md5': '65fdff94098e4a607385a60c5177c638',
 183             'info_dict': {
 184                 'id': '1969',
 185                 'ext': 'mp4',
 186                 'title': 'Hidden miracles of the natural world',
 187                 'uploader': 'Louie Schwartzberg',
 188                 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
 189             }
 190         },
 191         # Embeded Ustream video
 192         {
 193             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 194             'md5': '27b99cdb639c9b12a79bca876a073417',
 195             'info_dict': {
 196                 'id': '45734260',
 197                 'ext': 'flv',
 198                 'uploader': 'AU SPA:  The NSA and Privacy',
 199                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 200             }
 201         },
 202         # nowvideo embed hidden behind percent encoding
 203         {
 204             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 205             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 206             'info_dict': {
 207                 'id': '06e53103ca9aa',
 208                 'ext': 'flv',
 209                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 210                 'description': 'No description',
 211             },
 212         },
 213         # arte embed
 214         {
 215             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 216             'md5': '7653032cbb25bf6c80d80f217055fa43',
 217             'info_dict': {
 218                 'id': '048195-004_PLUS7-F',
 219                 'ext': 'flv',
 220                 'title': 'X:enius',
 221                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 222                 'upload_date': '20140320',
 223             },
 224             'params': {
 225                 'skip_download': 'Requires rtmpdump'
 226             }
 227         },
 228         # Condé Nast embed
 229         {
 230             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 231             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 232             'info_dict': {
 233                 'id': '53501be369702d3275860000',
 234                 'ext': 'mp4',
 235                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 236             }
 237         },
 238         # Dailymotion embed
 239         {
 240             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 241             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 242             'info_dict': {
 243                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 244                 'ext': 'mp4',
 245                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 246                 'uploader': 'Spi0n',
 247             },
 248             'add_ie': ['Dailymotion'],
 249         },
 250         # YouTube embed
 251         {
 252             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 253             'info_dict': {
 254                 'id': 'FXRb4ykk4S0',
 255                 'ext': 'mp4',
 256                 'title': 'The NBL Auction 2014',
 257                 'uploader': 'BADMINTON England',
 258                 'uploader_id': 'BADMINTONEvents',
 259                 'upload_date': '20140603',
 260                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 261             },
 262             'add_ie': ['Youtube'],
 263             'params': {
 264                 'skip_download': True,
 265             }
 266         },
 267         # MTVSercices embed
 268         {
 269             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 270             'md5': '35727f82f58c76d996fc188f9755b0d5',
 271             'info_dict': {
 272                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 273                 'ext': 'mp4',
 274                 'title': 'Review',
 275                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 276             },
 277         },
 278         # YouTube embed via <data-embed-url="">
 279         {
 280             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 281             'info_dict': {
 282                 'id': '4vAffPZIT44',
 283                 'ext': 'mp4',
 284                 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
 285                 'uploader': 'Gameloft',
 286                 'uploader_id': 'gameloft',
 287                 'upload_date': '20140828',
 288                 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
 289             },
 290             'params': {
 291                 'skip_download': True,
 292             }
 293         },
 294         # Camtasia studio
 295         {
 296             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 297             'playlist': [{
 298                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 299                 'info_dict': {
 300                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 301                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 302                     'ext': 'flv',
 303                     'duration': 2235.90,
 304                 }
 305             }, {
 306                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 307                 'info_dict': {
 308                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 309                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 310                     'ext': 'flv',
 311                     'duration': 2235.93,
 312                 }
 313             }],
 314             'info_dict': {
 315                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 316             }
 317         },
 318         # Flowplayer
 319         {
 320             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 321             'md5': '9d65602bf31c6e20014319c7d07fba27',
 322             'info_dict': {
 323                 'id': '5123ea6d5e5a7',
 324                 'ext': 'mp4',
 325                 'age_limit': 18,
 326                 'uploader': 'www.handjobhub.com',
 327                 'title': 'Busty Blonde Siri Tit Fuck While Wank at Handjob Hub',
 328             }
 329         },
 330         # RSS feed
 331         {
 332             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 333             'info_dict': {
 334                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 335                 'title': 'Zero Punctuation',
 336                 'description': 're:'
 337             },
 338             'playlist_mincount': 11,
 339         },
 340         # Multiple brightcove videos
 341         # https://github.com/rg3/youtube-dl/issues/2283
 342         {
 343             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 344             'info_dict': {
 345                 'id': 'always-never',
 346                 'title': 'Always / Never - The New Yorker',
 347             },
 348             'playlist_count': 3,
 349             'params': {
 350                 'extract_flat': False,
 351                 'skip_download': True,
 352             }
 353         },
 354         # MLB embed
 355         {
 356             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 357             'md5': '96f09a37e44da40dd083e12d9a683327',
 358             'info_dict': {
 359                 'id': '33322633',
 360                 'ext': 'mp4',
 361                 'title': 'Ump changes call to ball',
 362                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 363                 'duration': 48,
 364                 'timestamp': 1401537900,
 365                 'upload_date': '20140531',
 366                 'thumbnail': 're:^https?://.*\.jpg$',
 367             },
 368         },
 369         # Wistia embed
 370         {
 371             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 372             'md5': '8788b683c777a5cf25621eaf286d0c23',
 373             'info_dict': {
 374                 'id': '1cfaf6b7ea',
 375                 'ext': 'mov',
 376                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 377                 'duration': 643.0,
 378                 'filesize': 182808282,
 379                 'uploader': 'education-portal.com',
 380             },
 381         },
 382     ]
 383
 384     def report_following_redirect(self, new_url):
 385         """Report information extraction."""
 386         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 387
 388     def _extract_rss(self, url, video_id, doc):
 389         playlist_title = doc.find('./channel/title').text
 390         playlist_desc_el = doc.find('./channel/description')
 391         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 392
 393         entries = [{
 394             '_type': 'url',
 395             'url': e.find('link').text,
 396             'title': e.find('title').text,
 397         } for e in doc.findall('./channel/item')]
 398
 399         return {
 400             '_type': 'playlist',
 401             'id': url,
 402             'title': playlist_title,
 403             'description': playlist_desc,
 404             'entries': entries,
 405         }
 406
 407     def _extract_camtasia(self, url, video_id, webpage):
 408         """ Returns None if no camtasia video can be found. """
 409
 410         camtasia_cfg = self._search_regex(
 411             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 412             webpage, 'camtasia configuration file', default=None)
 413         if camtasia_cfg is None:
 414             return None
 415
 416         title = self._html_search_meta('DC.title', webpage, fatal=True)
 417
 418         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 419         camtasia_cfg = self._download_xml(
 420             camtasia_url, video_id,
 421             note='Downloading camtasia configuration',
 422             errnote='Failed to download camtasia configuration')
 423         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 424
 425         entries = []
 426         for n in fileset_node.getchildren():
 427             url_n = n.find('./uri')
 428             if url_n is None:
 429                 continue
 430
 431             entries.append({
 432                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 433                 'title': '%s - %s' % (title, n.tag),
 434                 'url': compat_urlparse.urljoin(url, url_n.text),
 435                 'duration': float_or_none(n.find('./duration').text),
 436             })
 437
 438         return {
 439             '_type': 'playlist',
 440             'entries': entries,
 441             'title': title,
 442         }
 443
 444     def _real_extract(self, url):
 445         if url.startswith('//'):
 446             return {
 447                 '_type': 'url',
 448                 'url': self.http_scheme() + url,
 449             }
 450
 451         parsed_url = compat_urlparse.urlparse(url)
 452         if not parsed_url.scheme:
 453             default_search = self._downloader.params.get('default_search')
 454             if default_search is None:
 455                 default_search = 'fixup_error'
 456
 457             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 458                 if '/' in url:
 459                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 460                     return self.url_result('http://' + url)
 461                 elif default_search != 'fixup_error':
 462                     if default_search == 'auto_warning':
 463                         if re.match(r'^(?:url|URL)$', url):
 464                             raise ExtractorError(
 465                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 466                                 expected=True)
 467                         else:
 468                             self._downloader.report_warning(
 469                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 470                     return self.url_result('ytsearch:' + url)
 471
 472             if default_search in ('error', 'fixup_error'):
 473                 raise ExtractorError(
 474                     ('%r is not a valid URL. '
 475                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 476                     ) % (url, url), expected=True)
 477             else:
 478                 assert ':' in default_search
 479                 return self.url_result(default_search + url)
 480
 481         url, smuggled_data = unsmuggle_url(url)
 482         force_videoid = None
 483         is_intentional = smuggled_data and smuggled_data.get('to_generic')
 484         if smuggled_data and 'force_videoid' in smuggled_data:
 485             force_videoid = smuggled_data['force_videoid']
 486             video_id = force_videoid
 487         else:
 488             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 489
 490         self.to_screen('%s: Requesting header' % video_id)
 491
 492         head_req = HEADRequest(url)
 493         response = self._request_webpage(
 494             head_req, video_id,
 495             note=False, errnote='Could not send HEAD request to %s' % url,
 496             fatal=False)
 497
 498         if response is not False:
 499             # Check for redirect
 500             new_url = response.geturl()
 501             if url != new_url:
 502                 self.report_following_redirect(new_url)
 503                 if force_videoid:
 504                     new_url = smuggle_url(
 505                         new_url, {'force_videoid': force_videoid})
 506                 return self.url_result(new_url)
 507
 508             # Check for direct link to a video
 509             content_type = response.headers.get('Content-Type', '')
 510             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 511             if m:
 512                 upload_date = response.headers.get('Last-Modified')
 513                 if upload_date:
 514                     upload_date = unified_strdate(upload_date)
 515                 return {
 516                     'id': video_id,
 517                     'title': os.path.splitext(url_basename(url))[0],
 518                     'formats': [{
 519                         'format_id': m.group('format_id'),
 520                         'url': url,
 521                         'vcodec': 'none' if m.group('type') == 'audio' else None
 522                     }],
 523                     'upload_date': upload_date,
 524                 }
 525
 526         if not self._downloader.params.get('test', False) and not is_intentional:
 527             self._downloader.report_warning('Falling back on generic information extractor.')
 528
 529         try:
 530             webpage = self._download_webpage(url, video_id)
 531         except ValueError:
 532             # since this is the last-resort InfoExtractor, if
 533             # this error is thrown, it'll be thrown here
 534             raise ExtractorError('Failed to download URL: %s' % url)
 535
 536         self.report_extraction(video_id)
 537
 538         # Is it an RSS feed?
 539         try:
 540             doc = parse_xml(webpage)
 541             if doc.tag == 'rss':
 542                 return self._extract_rss(url, video_id, doc)
 543         except compat_xml_parse_error:
 544             pass
 545
 546         # Is it a Camtasia project?
 547         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 548         if camtasia_res is not None:
 549             return camtasia_res
 550
 551         # Sometimes embedded video player is hidden behind percent encoding
 552         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 553         # Unescaping the whole page allows to handle those cases in a generic way
 554         webpage = compat_urllib_parse.unquote(webpage)
 555
 556         # it's tempting to parse this further, but you would
 557         # have to take into account all the variations like
 558         #   Video Title - Site Name
 559         #   Site Name | Video Title
 560         #   Video Title - Tagline | Site Name
 561         # and so on and so forth; it's just not practical
 562         video_title = self._html_search_regex(
 563             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 564             default='video')
 565
 566         # Try to detect age limit automatically
 567         age_limit = self._rta_search(webpage)
 568         # And then there are the jokers who advertise that they use RTA,
 569         # but actually don't.
 570         AGE_LIMIT_MARKERS = [
 571             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 572         ]
 573         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 574             age_limit = 18
 575
 576         # video uploader is domain name
 577         video_uploader = self._search_regex(
 578             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 579
 580         # Helper method
 581         def _playlist_from_matches(matches, getter, ie=None):
 582             urlrs = orderedSet(
 583                 self.url_result(self._proto_relative_url(getter(m)), ie)
 584                 for m in matches)
 585             return self.playlist_result(
 586                 urlrs, playlist_id=video_id, playlist_title=video_title)
 587
 588         # Look for BrightCove:
 589         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 590         if bc_urls:
 591             self.to_screen('Brightcove video detected.')
 592             entries = [{
 593                 '_type': 'url',
 594                 'url': smuggle_url(bc_url, {'Referer': url}),
 595                 'ie_key': 'Brightcove'
 596             } for bc_url in bc_urls]
 597
 598             return {
 599                 '_type': 'playlist',
 600                 'title': video_title,
 601                 'id': video_id,
 602                 'entries': entries,
 603             }
 604
 605         # Look for embedded (iframe) Vimeo player
 606         mobj = re.search(
 607             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 608         if mobj:
 609             player_url = unescapeHTML(mobj.group('url'))
 610             surl = smuggle_url(player_url, {'Referer': url})
 611             return self.url_result(surl, 'Vimeo')
 612
 613         # Look for embedded (swf embed) Vimeo player
 614         mobj = re.search(
 615             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 616         if mobj:
 617             return self.url_result(mobj.group(1), 'Vimeo')
 618
 619         # Look for embedded YouTube player
 620         matches = re.findall(r'''(?x)
 621             (?:
 622                 <iframe[^>]+?src=|
 623                 data-video-url=|
 624                 <embed[^>]+?src=|
 625                 embedSWF\(?:\s*
 626             )
 627             (["\'])
 628                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 629                 (?:embed|v|p)/.+?)
 630             \1''', webpage)
 631         if matches:
 632             return _playlist_from_matches(
 633                 matches, lambda m: unescapeHTML(m[1]))
 634
 635         # Look for embedded Dailymotion player
 636         matches = re.findall(
 637             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 638         if matches:
 639             return _playlist_from_matches(
 640                 matches, lambda m: unescapeHTML(m[1]))
 641
 642         # Look for embedded Wistia player
 643         match = re.search(
 644             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 645         if match:
 646             return {
 647                 '_type': 'url_transparent',
 648                 'url': unescapeHTML(match.group('url')),
 649                 'ie_key': 'Wistia',
 650                 'uploader': video_uploader,
 651                 'title': video_title,
 652                 'id': video_id,
 653             }
 654         match = re.search(r'(?:id=["\']wistia_|data-wistiaid=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 655         if match:
 656             return {
 657                 '_type': 'url_transparent',
 658                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 659                 'ie_key': 'Wistia',
 660                 'uploader': video_uploader,
 661                 'title': video_title,
 662                 'id': match.group('id')
 663             }
 664
 665         # Look for embedded blip.tv player
 666         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 667         if mobj:
 668             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 669         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 670         if mobj:
 671             return self.url_result(mobj.group(1), 'BlipTV')
 672
 673         # Look for embedded condenast player
 674         matches = re.findall(
 675             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 676             webpage)
 677         if matches:
 678             return {
 679                 '_type': 'playlist',
 680                 'entries': [{
 681                     '_type': 'url',
 682                     'ie_key': 'CondeNast',
 683                     'url': ma,
 684                 } for ma in matches],
 685                 'title': video_title,
 686                 'id': video_id,
 687             }
 688
 689         # Look for Bandcamp pages with custom domain
 690         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 691         if mobj is not None:
 692             burl = unescapeHTML(mobj.group(1))
 693             # Don't set the extractor because it can be a track url or an album
 694             return self.url_result(burl)
 695
 696         # Look for embedded Vevo player
 697         mobj = re.search(
 698             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 699         if mobj is not None:
 700             return self.url_result(mobj.group('url'))
 701
 702         # Look for Ooyala videos
 703         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 704              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 705         if mobj is not None:
 706             return OoyalaIE._build_url_result(mobj.group('ec'))
 707
 708         # Look for Aparat videos
 709         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 710         if mobj is not None:
 711             return self.url_result(mobj.group(1), 'Aparat')
 712
 713         # Look for MPORA videos
 714         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 715         if mobj is not None:
 716             return self.url_result(mobj.group(1), 'Mpora')
 717
 718         # Look for embedded NovaMov-based player
 719         mobj = re.search(
 720             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 721                     (?P<url>http://(?:(?:embed|www)\.)?
 722                         (?:novamov\.com|
 723                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 724                            videoweed\.(?:es|com)|
 725                            movshare\.(?:net|sx|ag)|
 726                            divxstage\.(?:eu|net|ch|co|at|ag))
 727                         /embed\.php.+?)\1''', webpage)
 728         if mobj is not None:
 729             return self.url_result(mobj.group('url'))
 730
 731         # Look for embedded Facebook player
 732         mobj = re.search(
 733             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 734         if mobj is not None:
 735             return self.url_result(mobj.group('url'), 'Facebook')
 736
 737         # Look for embedded VK player
 738         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 739         if mobj is not None:
 740             return self.url_result(mobj.group('url'), 'VK')
 741
 742         # Look for embedded ivi player
 743         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 744         if mobj is not None:
 745             return self.url_result(mobj.group('url'), 'Ivi')
 746
 747         # Look for embedded Huffington Post player
 748         mobj = re.search(
 749             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 750         if mobj is not None:
 751             return self.url_result(mobj.group('url'), 'HuffPost')
 752
 753         # Look for embed.ly
 754         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 755         if mobj is not None:
 756             return self.url_result(mobj.group('url'))
 757         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 758         if mobj is not None:
 759             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 760
 761         # Look for funnyordie embed
 762         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 763         if matches:
 764             return _playlist_from_matches(
 765                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 766
 767         # Look for embedded RUTV player
 768         rutv_url = RUTVIE._extract_url(webpage)
 769         if rutv_url:
 770             return self.url_result(rutv_url, 'RUTV')
 771
 772         # Look for embedded TED player
 773         mobj = re.search(
 774             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 775         if mobj is not None:
 776             return self.url_result(mobj.group('url'), 'TED')
 777
 778         # Look for embedded Ustream videos
 779         mobj = re.search(
 780             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 781         if mobj is not None:
 782             return self.url_result(mobj.group('url'), 'Ustream')
 783
 784         # Look for embedded arte.tv player
 785         mobj = re.search(
 786             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 787             webpage)
 788         if mobj is not None:
 789             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 790
 791         # Look for embedded smotri.com player
 792         smotri_url = SmotriIE._extract_url(webpage)
 793         if smotri_url:
 794             return self.url_result(smotri_url, 'Smotri')
 795
 796         # Look for embeded soundcloud player
 797         mobj = re.search(
 798             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 799             webpage)
 800         if mobj is not None:
 801             url = unescapeHTML(mobj.group('url'))
 802             return self.url_result(url)
 803
 804         # Look for embedded vulture.com player
 805         mobj = re.search(
 806             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 807             webpage)
 808         if mobj is not None:
 809             url = unescapeHTML(mobj.group('url'))
 810             return self.url_result(url, ie='Vulture')
 811
 812         # Look for embedded mtvservices player
 813         mobj = re.search(
 814             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 815             webpage)
 816         if mobj is not None:
 817             url = unescapeHTML(mobj.group('url'))
 818             return self.url_result(url, ie='MTVServicesEmbedded')
 819
 820         # Look for embedded yahoo player
 821         mobj = re.search(
 822             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 823             webpage)
 824         if mobj is not None:
 825             return self.url_result(mobj.group('url'), 'Yahoo')
 826
 827         # Look for embedded sbs.com.au player
 828         mobj = re.search(
 829             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 830             webpage)
 831         if mobj is not None:
 832             return self.url_result(mobj.group('url'), 'SBS')
 833
 834         mobj = re.search(
 835             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 836             webpage)
 837         if mobj is not None:
 838             return self.url_result(mobj.group('url'), 'MLB')
 839
 840         # Start with something easy: JW Player in SWFObject
 841         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 842         if not found:
 843             # Look for gorilla-vid style embedding
 844             found = re.findall(r'''(?sx)
 845                 (?:
 846                     jw_plugins|
 847                     JWPlayerOptions|
 848                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 849                 )
 850                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 851         if not found:
 852             # Broaden the search a little bit
 853             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 854         if not found:
 855             # Broaden the findall a little bit: JWPlayer JS loader
 856             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 857         if not found:
 858             # Flow player
 859             found = re.findall(r'''(?xs)
 860                 flowplayer\("[^"]+",\s*
 861                     \{[^}]+?\}\s*,
 862                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 863                         ["']?url["']?\s*:\s*["']([^"']+)["']
 864             ''', webpage)
 865         if not found:
 866             # Try to find twitter cards info
 867             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 868         if not found:
 869             # We look for Open Graph info:
 870             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 871             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 872             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 873             if m_video_type is not None:
 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')
 878                 found = list(filter(
 879                     check_video,
 880                     re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
 881         if not found:
 882             # HTML5 video
 883             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)
 884         if not found:
 885             found = re.search(
 886                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 887                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
 888                 webpage)
 889             if found:
 890                 new_url = found.group(1)
 891                 self.report_following_redirect(new_url)
 892                 return {
 893                     '_type': 'url',
 894                     'url': new_url,
 895                 }
 896         if not found:
 897             raise ExtractorError('Unsupported URL: %s' % url)
 898
 899         entries = []
 900         for video_url in found:
 901             video_url = compat_urlparse.urljoin(url, video_url)
 902             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 903
 904             # Sometimes, jwplayer extraction will result in a YouTube URL
 905             if YoutubeIE.suitable(video_url):
 906                 entries.append(self.url_result(video_url, 'Youtube'))
 907                 continue
 908
 909             # here's a fun little line of code for you:
 910             video_id = os.path.splitext(video_id)[0]
 911
 912             entries.append({
 913                 'id': video_id,
 914                 'url': video_url,
 915                 'uploader': video_uploader,
 916                 'title': video_title,
 917                 'age_limit': age_limit,
 918             })
 919
 920         if len(entries) == 1:
 921             return entries[0]
 922         else:
 923             for num, e in enumerate(entries, start=1):
 924                 e['title'] = '%s (%d)' % (e['title'], num)
 925             return {
 926                 '_type': 'playlist',
 927                 'entries': entries,
 928             }
 929