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_error,
  12     compat_urllib_parse,
  13     compat_urllib_request,
  14     compat_urlparse,
  15     compat_xml_parse_error,
  16
  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
  33
  34 class GenericIE(InfoExtractor):
  35     IE_DESC = 'Generic downloader that works on some sites'
  36     _VALID_URL = r'.*'
  37     IE_NAME = 'generic'
  38     _TESTS = [
  39         {
  40             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  41             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  42             'info_dict': {
  43                 'id': '13601338388002',
  44                 'ext': 'mp4',
  45                 'uploader': 'www.hodiho.fr',
  46                 'title': 'R\u00e9gis plante sa Jeep',
  47             }
  48         },
  49         # bandcamp page with custom domain
  50         {
  51             'add_ie': ['Bandcamp'],
  52             'url': 'http://bronyrock.com/track/the-pony-mash',
  53             'info_dict': {
  54                 'id': '3235767654',
  55                 'ext': 'mp3',
  56                 'title': 'The Pony Mash',
  57                 'uploader': 'M_Pallante',
  58             },
  59             'skip': 'There is a limit of 200 free downloads / month for the test song',
  60         },
  61         # embedded brightcove video
  62         # it also tests brightcove videos that need to set the 'Referer' in the
  63         # http requests
  64         {
  65             'add_ie': ['Brightcove'],
  66             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  67             'info_dict': {
  68                 'id': '2765128793001',
  69                 'ext': 'mp4',
  70                 'title': 'Le cours de bourse : l’analyse technique',
  71                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  72                 'uploader': 'BFM BUSINESS',
  73             },
  74             'params': {
  75                 'skip_download': True,
  76             },
  77         },
  78         {
  79             # https://github.com/rg3/youtube-dl/issues/2253
  80             'url': 'http://bcove.me/i6nfkrc3',
  81             'md5': '0ba9446db037002366bab3b3eb30c88c',
  82             'info_dict': {
  83                 'id': '3101154703001',
  84                 'ext': 'mp4',
  85                 'title': 'Still no power',
  86                 'uploader': 'thestar.com',
  87                 'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
  88             },
  89             'add_ie': ['Brightcove'],
  90         },
  91         {
  92             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  93             'md5': 'fb973ecf6e4a78a67453647444222983',
  94             'info_dict': {
  95                 'id': '3414141473001',
  96                 'ext': 'mp4',
  97                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  98                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
  99                 'uploader': 'Championat',
 100             },
 101         },
 102         # Direct link to a video
 103         {
 104             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 105             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 106             'info_dict': {
 107                 'id': 'trailer',
 108                 'ext': 'mp4',
 109                 'title': 'trailer',
 110                 'upload_date': '20100513',
 111             }
 112         },
 113         # ooyala video
 114         {
 115             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 116             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 117             'info_dict': {
 118                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 119                 'ext': 'mp4',
 120                 'title': '2cc213299525360.mov',  # that's what we get
 121             },
 122         },
 123         # google redirect
 124         {
 125             'url': 'http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CCUQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DcmQHVoWB5FY&ei=F-sNU-LLCaXk4QT52ICQBQ&usg=AFQjCNEw4hL29zgOohLXvpJ-Bdh2bils1Q&bvm=bv.61965928,d.bGE',
 126             'info_dict': {
 127                 'id': 'cmQHVoWB5FY',
 128                 'ext': 'mp4',
 129                 'upload_date': '20130224',
 130                 'uploader_id': 'TheVerge',
 131                 'description': 'Chris Ziegler takes a look at the Alcatel OneTouch Fire and the ZTE Open; two of the first Firefox OS handsets to be officially announced.',
 132                 'uploader': 'The Verge',
 133                 'title': 'First Firefox OS phones side-by-side',
 134             },
 135             'params': {
 136                 'skip_download': False,
 137             }
 138         },
 139         # embed.ly video
 140         {
 141             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 142             'info_dict': {
 143                 'id': '9ODmcdjQcHQ',
 144                 'ext': 'mp4',
 145                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 146                 'upload_date': '20140225',
 147                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 148                 'uploader': 'Tested',
 149                 'uploader_id': 'testedcom',
 150             },
 151             # No need to test YoutubeIE here
 152             'params': {
 153                 'skip_download': True,
 154             },
 155         },
 156         # funnyordie embed
 157         {
 158             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 159             'md5': '7cf780be104d40fea7bae52eed4a470e',
 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': 'deeeabcc1085eb2ba205474e7235a3d5',
 185             'info_dict': {
 186                 'id': '981',
 187                 'ext': 'mp4',
 188                 'title': 'My web playroom',
 189                 'uploader': 'Ze Frank',
 190                 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
 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         # smotri embed
 231         {
 232             'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
 233             'md5': 'ec40048448e9284c9a1de77bb188108b',
 234             'info_dict': {
 235                 'id': 'v27008541fad',
 236                 'ext': 'mp4',
 237                 'title': 'Крым и Севастополь вошли в состав России',
 238                 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
 239                 'duration': 900,
 240                 'upload_date': '20140318',
 241                 'uploader': 'rbctv_2012_4',
 242                 'uploader_id': 'rbctv_2012_4',
 243             },
 244         },
 245         # Condé Nast embed
 246         {
 247             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 248             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 249             'info_dict': {
 250                 'id': '53501be369702d3275860000',
 251                 'ext': 'mp4',
 252                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 253             }
 254         },
 255         # Dailymotion embed
 256         {
 257             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 258             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 259             'info_dict': {
 260                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 261                 'ext': 'mp4',
 262                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 263                 'uploader': 'Spi0n',
 264             },
 265             'add_ie': ['Dailymotion'],
 266         },
 267         # YouTube embed
 268         {
 269             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 270             'info_dict': {
 271                 'id': 'FXRb4ykk4S0',
 272                 'ext': 'mp4',
 273                 'title': 'The NBL Auction 2014',
 274                 'uploader': 'BADMINTON England',
 275                 'uploader_id': 'BADMINTONEvents',
 276                 'upload_date': '20140603',
 277                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 278             },
 279             'add_ie': ['Youtube'],
 280             'params': {
 281                 'skip_download': True,
 282             }
 283         },
 284         # MTVSercices embed
 285         {
 286             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 287             'md5': '35727f82f58c76d996fc188f9755b0d5',
 288             'info_dict': {
 289                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 290                 'ext': 'mp4',
 291                 'title': 'Review',
 292                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 293             },
 294         },
 295         # YouTube embed via <data-embed-url="">
 296         {
 297             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 298             'info_dict': {
 299                 'id': 'jpSGZsgga_I',
 300                 'ext': 'mp4',
 301                 'title': 'Asphalt 8: Airborne - Launch Trailer',
 302                 'uploader': 'Gameloft',
 303                 'uploader_id': 'gameloft',
 304                 'upload_date': '20130821',
 305                 'description': 'md5:87bd95f13d8be3e7da87a5f2c443106a',
 306             },
 307             'params': {
 308                 'skip_download': True,
 309             }
 310         },
 311         # Camtasia studio
 312         {
 313             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 314             'playlist': [{
 315                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 316                 'info_dict': {
 317                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 318                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 319                     'ext': 'flv',
 320                     'duration': 2235.90,
 321                 }
 322             }, {
 323                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 324                 'info_dict': {
 325                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 326                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 327                     'ext': 'flv',
 328                     'duration': 2235.93,
 329                 }
 330             }],
 331             'info_dict': {
 332                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 333             }
 334         }
 335     ]
 336
 337     def report_download_webpage(self, video_id):
 338         """Report webpage download."""
 339         if not self._downloader.params.get('test', False):
 340             self._downloader.report_warning('Falling back on generic information extractor.')
 341         super(GenericIE, self).report_download_webpage(video_id)
 342
 343     def report_following_redirect(self, new_url):
 344         """Report information extraction."""
 345         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 346
 347     def _send_head(self, url):
 348         """Check if it is a redirect, like url shorteners, in case return the new url."""
 349
 350         class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
 351             """
 352             Subclass the HTTPRedirectHandler to make it use our
 353             HEADRequest also on the redirected URL
 354             """
 355             def redirect_request(self, req, fp, code, msg, headers, newurl):
 356                 if code in (301, 302, 303, 307):
 357                     newurl = newurl.replace(' ', '%20')
 358                     newheaders = dict((k,v) for k,v in req.headers.items()
 359                                       if k.lower() not in ("content-length", "content-type"))
 360                     try:
 361                         # This function was deprecated in python 3.3 and removed in 3.4
 362                         origin_req_host = req.get_origin_req_host()
 363                     except AttributeError:
 364                         origin_req_host = req.origin_req_host
 365                     return HEADRequest(newurl,
 366                                        headers=newheaders,
 367                                        origin_req_host=origin_req_host,
 368                                        unverifiable=True)
 369                 else:
 370                     raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
 371
 372         class HTTPMethodFallback(compat_urllib_request.BaseHandler):
 373             """
 374             Fallback to GET if HEAD is not allowed (405 HTTP error)
 375             """
 376             def http_error_405(self, req, fp, code, msg, headers):
 377                 fp.read()
 378                 fp.close()
 379
 380                 newheaders = dict((k,v) for k,v in req.headers.items()
 381                                   if k.lower() not in ("content-length", "content-type"))
 382                 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
 383                                                  headers=newheaders,
 384                                                  origin_req_host=req.get_origin_req_host(),
 385                                                  unverifiable=True))
 386
 387         # Build our opener
 388         opener = compat_urllib_request.OpenerDirector()
 389         for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
 390                         HTTPMethodFallback, HEADRedirectHandler,
 391                         compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
 392             opener.add_handler(handler())
 393
 394         response = opener.open(HEADRequest(url))
 395         if response is None:
 396             raise ExtractorError('Invalid URL protocol')
 397         return response
 398
 399     def _extract_rss(self, url, video_id, doc):
 400         playlist_title = doc.find('./channel/title').text
 401         playlist_desc_el = doc.find('./channel/description')
 402         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 403
 404         entries = [{
 405             '_type': 'url',
 406             'url': e.find('link').text,
 407             'title': e.find('title').text,
 408         } for e in doc.findall('./channel/item')]
 409
 410         return {
 411             '_type': 'playlist',
 412             'id': url,
 413             'title': playlist_title,
 414             'description': playlist_desc,
 415             'entries': entries,
 416         }
 417
 418     def _extract_camtasia(self, url, video_id, webpage):
 419         """ Returns None if no camtasia video can be found. """
 420
 421         camtasia_cfg = self._search_regex(
 422             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 423             webpage, 'camtasia configuration file', default=None)
 424         if camtasia_cfg is None:
 425             return None
 426
 427         title = self._html_search_meta('DC.title', webpage, fatal=True)
 428
 429         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 430         camtasia_cfg = self._download_xml(
 431             camtasia_url, video_id,
 432             note='Downloading camtasia configuration',
 433             errnote='Failed to download camtasia configuration')
 434         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 435
 436         entries = []
 437         for n in fileset_node.getchildren():
 438             url_n = n.find('./uri')
 439             if url_n is None:
 440                 continue
 441
 442             entries.append({
 443                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 444                 'title': '%s - %s' % (title, n.tag),
 445                 'url': compat_urlparse.urljoin(url, url_n.text),
 446                 'duration': float_or_none(n.find('./duration').text),
 447             })
 448
 449         return {
 450             '_type': 'playlist',
 451             'entries': entries,
 452             'title': title,
 453         }
 454
 455     def _real_extract(self, url):
 456         if url.startswith('//'):
 457             return {
 458                 '_type': 'url',
 459                 'url': self.http_scheme() + url,
 460             }
 461
 462         parsed_url = compat_urlparse.urlparse(url)
 463         if not parsed_url.scheme:
 464             default_search = self._downloader.params.get('default_search')
 465             if default_search is None:
 466                 default_search = 'fixup_error'
 467
 468             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 469                 if '/' in url:
 470                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 471                     return self.url_result('http://' + url)
 472                 elif default_search != 'fixup_error':
 473                     if default_search == 'auto_warning':
 474                         if re.match(r'^(?:url|URL)$', url):
 475                             raise ExtractorError(
 476                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 477                                 expected=True)
 478                         else:
 479                             self._downloader.report_warning(
 480                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 481                     return self.url_result('ytsearch:' + url)
 482
 483             if default_search in ('error', 'fixup_error'):
 484                 raise ExtractorError(
 485                     ('%r is not a valid URL. '
 486                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 487                     ) % (url, url), expected=True)
 488             else:
 489                 assert ':' in default_search
 490                 return self.url_result(default_search + url)
 491
 492         url, smuggled_data = unsmuggle_url(url)
 493         force_videoid = None
 494         if smuggled_data and 'force_videoid' in smuggled_data:
 495             force_videoid = smuggled_data['force_videoid']
 496             video_id = force_videoid
 497         else:
 498             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 499
 500         self.to_screen('%s: Requesting header' % video_id)
 501
 502         try:
 503             response = self._send_head(url)
 504
 505             # Check for redirect
 506             new_url = response.geturl()
 507             if url != new_url:
 508                 self.report_following_redirect(new_url)
 509                 if force_videoid:
 510                     new_url = smuggle_url(
 511                         new_url, {'force_videoid': force_videoid})
 512                 return self.url_result(new_url)
 513
 514             # Check for direct link to a video
 515             content_type = response.headers.get('Content-Type', '')
 516             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 517             if m:
 518                 upload_date = response.headers.get('Last-Modified')
 519                 if upload_date:
 520                     upload_date = unified_strdate(upload_date)
 521                 return {
 522                     'id': video_id,
 523                     'title': os.path.splitext(url_basename(url))[0],
 524                     'formats': [{
 525                         'format_id': m.group('format_id'),
 526                         'url': url,
 527                         'vcodec': 'none' if m.group('type') == 'audio' else None
 528                     }],
 529                     'upload_date': upload_date,
 530                 }
 531
 532         except compat_urllib_error.HTTPError:
 533             # This may be a stupid server that doesn't like HEAD, our UA, or so
 534             pass
 535
 536         try:
 537             webpage = self._download_webpage(url, video_id)
 538         except ValueError:
 539             # since this is the last-resort InfoExtractor, if
 540             # this error is thrown, it'll be thrown here
 541             raise ExtractorError('Failed to download URL: %s' % url)
 542
 543         self.report_extraction(video_id)
 544
 545         # Is it an RSS feed?
 546         try:
 547             doc = parse_xml(webpage)
 548             if doc.tag == 'rss':
 549                 return self._extract_rss(url, video_id, doc)
 550         except compat_xml_parse_error:
 551             pass
 552
 553         # Is it a Camtasia project?
 554         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 555         if camtasia_res is not None:
 556             return camtasia_res
 557
 558         # Sometimes embedded video player is hidden behind percent encoding
 559         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 560         # Unescaping the whole page allows to handle those cases in a generic way
 561         webpage = compat_urllib_parse.unquote(webpage)
 562
 563         # it's tempting to parse this further, but you would
 564         # have to take into account all the variations like
 565         #   Video Title - Site Name
 566         #   Site Name | Video Title
 567         #   Video Title - Tagline | Site Name
 568         # and so on and so forth; it's just not practical
 569         video_title = self._html_search_regex(
 570             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 571             default='video')
 572
 573         # video uploader is domain name
 574         video_uploader = self._search_regex(
 575             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 576
 577         # Helper method
 578         def _playlist_from_matches(matches, getter, ie=None):
 579             urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
 580             return self.playlist_result(
 581                 urlrs, playlist_id=video_id, playlist_title=video_title)
 582
 583         # Look for BrightCove:
 584         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 585         if bc_urls:
 586             self.to_screen('Brightcove video detected.')
 587             entries = [{
 588                 '_type': 'url',
 589                 'url': smuggle_url(bc_url, {'Referer': url}),
 590                 'ie_key': 'Brightcove'
 591             } for bc_url in bc_urls]
 592
 593             return {
 594                 '_type': 'playlist',
 595                 'title': video_title,
 596                 'id': video_id,
 597                 'entries': entries,
 598             }
 599
 600         # Look for embedded (iframe) Vimeo player
 601         mobj = re.search(
 602             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 603         if mobj:
 604             player_url = unescapeHTML(mobj.group('url'))
 605             surl = smuggle_url(player_url, {'Referer': url})
 606             return self.url_result(surl, 'Vimeo')
 607
 608         # Look for embedded (swf embed) Vimeo player
 609         mobj = re.search(
 610             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 611         if mobj:
 612             return self.url_result(mobj.group(1), 'Vimeo')
 613
 614         # Look for embedded YouTube player
 615         matches = re.findall(r'''(?x)
 616             (?:
 617                 <iframe[^>]+?src=|
 618                 data-video-url=|
 619                 <embed[^>]+?src=|
 620                 embedSWF\(?:\s*
 621             )
 622             (["\'])
 623                 (?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
 624                 (?:embed|v)/.+?)
 625             \1''', webpage)
 626         if matches:
 627             return _playlist_from_matches(
 628                 matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
 629
 630         # Look for embedded Dailymotion player
 631         matches = re.findall(
 632             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 633         if matches:
 634             return _playlist_from_matches(
 635                 matches, lambda m: unescapeHTML(m[1]))
 636
 637         # Look for embedded Wistia player
 638         match = re.search(
 639             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 640         if match:
 641             return {
 642                 '_type': 'url_transparent',
 643                 'url': unescapeHTML(match.group('url')),
 644                 'ie_key': 'Wistia',
 645                 'uploader': video_uploader,
 646                 'title': video_title,
 647                 'id': video_id,
 648             }
 649
 650         # Look for embedded blip.tv player
 651         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 652         if mobj:
 653             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 654         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 655         if mobj:
 656             return self.url_result(mobj.group(1), 'BlipTV')
 657
 658         # Look for embedded condenast player
 659         matches = re.findall(
 660             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 661             webpage)
 662         if matches:
 663             return {
 664                 '_type': 'playlist',
 665                 'entries': [{
 666                     '_type': 'url',
 667                     'ie_key': 'CondeNast',
 668                     'url': ma,
 669                 } for ma in matches],
 670                 'title': video_title,
 671                 'id': video_id,
 672             }
 673
 674         # Look for Bandcamp pages with custom domain
 675         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 676         if mobj is not None:
 677             burl = unescapeHTML(mobj.group(1))
 678             # Don't set the extractor because it can be a track url or an album
 679             return self.url_result(burl)
 680
 681         # Look for embedded Vevo player
 682         mobj = re.search(
 683             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 684         if mobj is not None:
 685             return self.url_result(mobj.group('url'))
 686
 687         # Look for Ooyala videos
 688         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 689              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 690         if mobj is not None:
 691             return OoyalaIE._build_url_result(mobj.group('ec'))
 692
 693         # Look for Aparat videos
 694         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 695         if mobj is not None:
 696             return self.url_result(mobj.group(1), 'Aparat')
 697
 698         # Look for MPORA videos
 699         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 700         if mobj is not None:
 701             return self.url_result(mobj.group(1), 'Mpora')
 702
 703         # Look for embedded NovaMov-based player
 704         mobj = re.search(
 705             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 706                     (?P<url>http://(?:(?:embed|www)\.)?
 707                         (?:novamov\.com|
 708                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 709                            videoweed\.(?:es|com)|
 710                            movshare\.(?:net|sx|ag)|
 711                            divxstage\.(?:eu|net|ch|co|at|ag))
 712                         /embed\.php.+?)\1''', webpage)
 713         if mobj is not None:
 714             return self.url_result(mobj.group('url'))
 715
 716         # Look for embedded Facebook player
 717         mobj = re.search(
 718             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 719         if mobj is not None:
 720             return self.url_result(mobj.group('url'), 'Facebook')
 721
 722         # Look for embedded VK player
 723         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 724         if mobj is not None:
 725             return self.url_result(mobj.group('url'), 'VK')
 726
 727         # Look for embedded ivi player
 728         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 729         if mobj is not None:
 730             return self.url_result(mobj.group('url'), 'Ivi')
 731
 732         # Look for embedded Huffington Post player
 733         mobj = re.search(
 734             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 735         if mobj is not None:
 736             return self.url_result(mobj.group('url'), 'HuffPost')
 737
 738         # Look for embed.ly
 739         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 740         if mobj is not None:
 741             return self.url_result(mobj.group('url'))
 742         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 743         if mobj is not None:
 744             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 745
 746         # Look for funnyordie embed
 747         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 748         if matches:
 749             return _playlist_from_matches(
 750                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 751
 752         # Look for embedded RUTV player
 753         rutv_url = RUTVIE._extract_url(webpage)
 754         if rutv_url:
 755             return self.url_result(rutv_url, 'RUTV')
 756
 757         # Look for embedded TED player
 758         mobj = re.search(
 759             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 760         if mobj is not None:
 761             return self.url_result(mobj.group('url'), 'TED')
 762
 763         # Look for embedded Ustream videos
 764         mobj = re.search(
 765             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 766         if mobj is not None:
 767             return self.url_result(mobj.group('url'), 'Ustream')
 768
 769         # Look for embedded arte.tv player
 770         mobj = re.search(
 771             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 772             webpage)
 773         if mobj is not None:
 774             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 775
 776         # Look for embedded smotri.com player
 777         smotri_url = SmotriIE._extract_url(webpage)
 778         if smotri_url:
 779             return self.url_result(smotri_url, 'Smotri')
 780
 781         # Look for embeded soundcloud player
 782         mobj = re.search(
 783             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 784             webpage)
 785         if mobj is not None:
 786             url = unescapeHTML(mobj.group('url'))
 787             return self.url_result(url)
 788
 789         # Look for embedded vulture.com player
 790         mobj = re.search(
 791             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 792             webpage)
 793         if mobj is not None:
 794             url = unescapeHTML(mobj.group('url'))
 795             return self.url_result(url, ie='Vulture')
 796
 797         # Look for embedded mtvservices player
 798         mobj = re.search(
 799             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 800             webpage)
 801         if mobj is not None:
 802             url = unescapeHTML(mobj.group('url'))
 803             return self.url_result(url, ie='MTVServicesEmbedded')
 804
 805         # Look for embedded yahoo player
 806         mobj = re.search(
 807             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 808             webpage)
 809         if mobj is not None:
 810             return self.url_result(mobj.group('url'), 'Yahoo')
 811
 812         # Look for embedded sbs.com.au player
 813         mobj = re.search(
 814             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 815             webpage)
 816         if mobj is not None:
 817             return self.url_result(mobj.group('url'), 'SBS')
 818
 819         # Start with something easy: JW Player in SWFObject
 820         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 821         if not found:
 822             # Look for gorilla-vid style embedding
 823             found = re.findall(r'''(?sx)
 824                 (?:
 825                     jw_plugins|
 826                     JWPlayerOptions|
 827                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 828                 )
 829                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 830         if not found:
 831             # Broaden the search a little bit
 832             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 833         if not found:
 834             # Broaden the findall a little bit: JWPlayer JS loader
 835             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 836         if not found:
 837             # Try to find twitter cards info
 838             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 839         if not found:
 840             # We look for Open Graph info:
 841             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 842             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 843             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 844             if m_video_type is not None:
 845                 def check_video(vurl):
 846                     vpath = compat_urlparse.urlparse(vurl).path
 847                     return '.' in vpath and not vpath.endswith('.swf')
 848                 found = list(filter(
 849                     check_video,
 850                     re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
 851         if not found:
 852             # HTML5 video
 853             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
 854         if not found:
 855             found = re.search(
 856                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 857                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
 858                 webpage)
 859             if found:
 860                 new_url = found.group(1)
 861                 self.report_following_redirect(new_url)
 862                 return {
 863                     '_type': 'url',
 864                     'url': new_url,
 865                 }
 866         if not found:
 867             raise ExtractorError('Unsupported URL: %s' % url)
 868
 869         entries = []
 870         for video_url in found:
 871             video_url = compat_urlparse.urljoin(url, video_url)
 872             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 873
 874             # Sometimes, jwplayer extraction will result in a YouTube URL
 875             if YoutubeIE.suitable(video_url):
 876                 entries.append(self.url_result(video_url, 'Youtube'))
 877                 continue
 878
 879             # here's a fun little line of code for you:
 880             video_id = os.path.splitext(video_id)[0]
 881
 882             entries.append({
 883                 'id': video_id,
 884                 'url': video_url,
 885                 'uploader': video_uploader,
 886                 'title': video_title,
 887             })
 888
 889         if len(entries) == 1:
 890             return entries[0]
 891         else:
 892             for num, e in enumerate(entries, start=1):
 893                 e['title'] = '%s (%d)' % (e['title'], num)
 894             return {
 895                 '_type': 'playlist',
 896                 'entries': entries,
 897             }
 898