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     HEADRequest,
  19     orderedSet,
  20     parse_xml,
  21     smuggle_url,
  22     unescapeHTML,
  23     unified_strdate,
  24     url_basename,
  25 )
  26 from .brightcove import BrightcoveIE
  27 from .ooyala import OoyalaIE
  28 from .rutv import RUTVIE
  29 from .smotri import SmotriIE
  30
  31
  32 class GenericIE(InfoExtractor):
  33     IE_DESC = 'Generic downloader that works on some sites'
  34     _VALID_URL = r'.*'
  35     IE_NAME = 'generic'
  36     _TESTS = [
  37         {
  38             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  39             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  40             'info_dict': {
  41                 'id': '13601338388002',
  42                 'ext': 'mp4',
  43                 'uploader': 'www.hodiho.fr',
  44                 'title': 'R\u00e9gis plante sa Jeep',
  45             }
  46         },
  47         # bandcamp page with custom domain
  48         {
  49             'add_ie': ['Bandcamp'],
  50             'url': 'http://bronyrock.com/track/the-pony-mash',
  51             'info_dict': {
  52                 'id': '3235767654',
  53                 'ext': 'mp3',
  54                 'title': 'The Pony Mash',
  55                 'uploader': 'M_Pallante',
  56             },
  57             'skip': 'There is a limit of 200 free downloads / month for the test song',
  58         },
  59         # embedded brightcove video
  60         # it also tests brightcove videos that need to set the 'Referer' in the
  61         # http requests
  62         {
  63             'add_ie': ['Brightcove'],
  64             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  65             'info_dict': {
  66                 'id': '2765128793001',
  67                 'ext': 'mp4',
  68                 'title': 'Le cours de bourse : l’analyse technique',
  69                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  70                 'uploader': 'BFM BUSINESS',
  71             },
  72             'params': {
  73                 'skip_download': True,
  74             },
  75         },
  76         {
  77             # https://github.com/rg3/youtube-dl/issues/2253
  78             'url': 'http://bcove.me/i6nfkrc3',
  79             'md5': '0ba9446db037002366bab3b3eb30c88c',
  80             'info_dict': {
  81                 'id': '3101154703001',
  82                 'ext': 'mp4',
  83                 'title': 'Still no power',
  84                 'uploader': 'thestar.com',
  85                 '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.',
  86             },
  87             'add_ie': ['Brightcove'],
  88         },
  89         {
  90             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  91             'md5': 'fb973ecf6e4a78a67453647444222983',
  92             'info_dict': {
  93                 'id': '3414141473001',
  94                 'ext': 'mp4',
  95                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  96                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
  97                 'uploader': 'Championat',
  98             },
  99         },
 100         # Direct link to a video
 101         {
 102             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 103             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 104             'info_dict': {
 105                 'id': 'trailer',
 106                 'ext': 'mp4',
 107                 'title': 'trailer',
 108                 'upload_date': '20100513',
 109             }
 110         },
 111         # ooyala video
 112         {
 113             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 114             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 115             'info_dict': {
 116                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 117                 'ext': 'mp4',
 118                 'title': '2cc213299525360.mov',  # that's what we get
 119             },
 120         },
 121         # google redirect
 122         {
 123             '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',
 124             'info_dict': {
 125                 'id': 'cmQHVoWB5FY',
 126                 'ext': 'mp4',
 127                 'upload_date': '20130224',
 128                 'uploader_id': 'TheVerge',
 129                 '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.',
 130                 'uploader': 'The Verge',
 131                 'title': 'First Firefox OS phones side-by-side',
 132             },
 133             'params': {
 134                 'skip_download': False,
 135             }
 136         },
 137         # embed.ly video
 138         {
 139             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 140             'info_dict': {
 141                 'id': '9ODmcdjQcHQ',
 142                 'ext': 'mp4',
 143                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 144                 'upload_date': '20140225',
 145                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 146                 'uploader': 'Tested',
 147                 'uploader_id': 'testedcom',
 148             },
 149             # No need to test YoutubeIE here
 150             'params': {
 151                 'skip_download': True,
 152             },
 153         },
 154         # funnyordie embed
 155         {
 156             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 157             'md5': '7cf780be104d40fea7bae52eed4a470e',
 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': 'deeeabcc1085eb2ba205474e7235a3d5',
 183             'info_dict': {
 184                 'id': '981',
 185                 'ext': 'mp4',
 186                 'title': 'My web playroom',
 187                 'uploader': 'Ze Frank',
 188                 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
 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         # smotri embed
 229         {
 230             'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
 231             'md5': 'ec40048448e9284c9a1de77bb188108b',
 232             'info_dict': {
 233                 'id': 'v27008541fad',
 234                 'ext': 'mp4',
 235                 'title': 'Крым и Севастополь вошли в состав России',
 236                 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
 237                 'duration': 900,
 238                 'upload_date': '20140318',
 239                 'uploader': 'rbctv_2012_4',
 240                 'uploader_id': 'rbctv_2012_4',
 241             },
 242         },
 243         # Condé Nast embed
 244         {
 245             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 246             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 247             'info_dict': {
 248                 'id': '53501be369702d3275860000',
 249                 'ext': 'mp4',
 250                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 251             }
 252         },
 253         # Dailymotion embed
 254         {
 255             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 256             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 257             'info_dict': {
 258                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 259                 'ext': 'mp4',
 260                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 261                 'uploader': 'Spi0n',
 262             },
 263             'add_ie': ['Dailymotion'],
 264         },
 265         # YouTube embed
 266         {
 267             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 268             'info_dict': {
 269                 'id': 'FXRb4ykk4S0',
 270                 'ext': 'mp4',
 271                 'title': 'The NBL Auction 2014',
 272                 'uploader': 'BADMINTON England',
 273                 'uploader_id': 'BADMINTONEvents',
 274                 'upload_date': '20140603',
 275                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 276             },
 277             'add_ie': ['Youtube'],
 278             'params': {
 279                 'skip_download': True,
 280             }
 281         },
 282         # MTVSercices embed
 283         {
 284             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 285             'md5': '35727f82f58c76d996fc188f9755b0d5',
 286             'info_dict': {
 287                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 288                 'ext': 'mp4',
 289                 'title': 'Review',
 290                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 291             },
 292         },
 293         # YouTube embed via <data-embed-url="">
 294         {
 295             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 296             'info_dict': {
 297                 'id': 'jpSGZsgga_I',
 298                 'ext': 'mp4',
 299                 'title': 'Asphalt 8: Airborne - Launch Trailer',
 300                 'uploader': 'Gameloft',
 301                 'uploader_id': 'gameloft',
 302                 'upload_date': '20130821',
 303                 'description': 'md5:87bd95f13d8be3e7da87a5f2c443106a',
 304             },
 305             'params': {
 306                 'skip_download': True,
 307             }
 308         }
 309     ]
 310
 311     def report_download_webpage(self, video_id):
 312         """Report webpage download."""
 313         if not self._downloader.params.get('test', False):
 314             self._downloader.report_warning('Falling back on generic information extractor.')
 315         super(GenericIE, self).report_download_webpage(video_id)
 316
 317     def report_following_redirect(self, new_url):
 318         """Report information extraction."""
 319         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 320
 321     def _send_head(self, url):
 322         """Check if it is a redirect, like url shorteners, in case return the new url."""
 323
 324         class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
 325             """
 326             Subclass the HTTPRedirectHandler to make it use our
 327             HEADRequest also on the redirected URL
 328             """
 329             def redirect_request(self, req, fp, code, msg, headers, newurl):
 330                 if code in (301, 302, 303, 307):
 331                     newurl = newurl.replace(' ', '%20')
 332                     newheaders = dict((k,v) for k,v in req.headers.items()
 333                                       if k.lower() not in ("content-length", "content-type"))
 334                     try:
 335                         # This function was deprecated in python 3.3 and removed in 3.4
 336                         origin_req_host = req.get_origin_req_host()
 337                     except AttributeError:
 338                         origin_req_host = req.origin_req_host
 339                     return HEADRequest(newurl,
 340                                        headers=newheaders,
 341                                        origin_req_host=origin_req_host,
 342                                        unverifiable=True)
 343                 else:
 344                     raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
 345
 346         class HTTPMethodFallback(compat_urllib_request.BaseHandler):
 347             """
 348             Fallback to GET if HEAD is not allowed (405 HTTP error)
 349             """
 350             def http_error_405(self, req, fp, code, msg, headers):
 351                 fp.read()
 352                 fp.close()
 353
 354                 newheaders = dict((k,v) for k,v in req.headers.items()
 355                                   if k.lower() not in ("content-length", "content-type"))
 356                 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
 357                                                  headers=newheaders,
 358                                                  origin_req_host=req.get_origin_req_host(),
 359                                                  unverifiable=True))
 360
 361         # Build our opener
 362         opener = compat_urllib_request.OpenerDirector()
 363         for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
 364                         HTTPMethodFallback, HEADRedirectHandler,
 365                         compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
 366             opener.add_handler(handler())
 367
 368         response = opener.open(HEADRequest(url))
 369         if response is None:
 370             raise ExtractorError('Invalid URL protocol')
 371         return response
 372
 373     def _extract_rss(self, url, video_id, doc):
 374         playlist_title = doc.find('./channel/title').text
 375         playlist_desc_el = doc.find('./channel/description')
 376         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 377
 378         entries = [{
 379             '_type': 'url',
 380             'url': e.find('link').text,
 381             'title': e.find('title').text,
 382         } for e in doc.findall('./channel/item')]
 383
 384         return {
 385             '_type': 'playlist',
 386             'id': url,
 387             'title': playlist_title,
 388             'description': playlist_desc,
 389             'entries': entries,
 390         }
 391
 392     def _real_extract(self, url):
 393         if url.startswith('//'):
 394             return {
 395                 '_type': 'url',
 396                 'url': self.http_scheme() + url,
 397             }
 398
 399         parsed_url = compat_urlparse.urlparse(url)
 400         if not parsed_url.scheme:
 401             default_search = self._downloader.params.get('default_search')
 402             if default_search is None:
 403                 default_search = 'fixup_error'
 404
 405             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 406                 if '/' in url:
 407                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 408                     return self.url_result('http://' + url)
 409                 elif default_search != 'fixup_error':
 410                     if default_search == 'auto_warning':
 411                         if re.match(r'^(?:url|URL)$', url):
 412                             raise ExtractorError(
 413                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 414                                 expected=True)
 415                         else:
 416                             self._downloader.report_warning(
 417                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 418                     return self.url_result('ytsearch:' + url)
 419
 420             if default_search in ('error', 'fixup_error'):
 421                 raise ExtractorError(
 422                     ('%r is not a valid URL. '
 423                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 424                     ) % (url, url), expected=True)
 425             else:
 426                 assert ':' in default_search
 427                 return self.url_result(default_search + url)
 428         video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 429
 430         self.to_screen('%s: Requesting header' % video_id)
 431
 432         try:
 433             response = self._send_head(url)
 434
 435             # Check for redirect
 436             new_url = response.geturl()
 437             if url != new_url:
 438                 self.report_following_redirect(new_url)
 439                 return self.url_result(new_url)
 440
 441             # Check for direct link to a video
 442             content_type = response.headers.get('Content-Type', '')
 443             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 444             if m:
 445                 upload_date = response.headers.get('Last-Modified')
 446                 if upload_date:
 447                     upload_date = unified_strdate(upload_date)
 448                 return {
 449                     'id': video_id,
 450                     'title': os.path.splitext(url_basename(url))[0],
 451                     'formats': [{
 452                         'format_id': m.group('format_id'),
 453                         'url': url,
 454                         'vcodec': 'none' if m.group('type') == 'audio' else None
 455                     }],
 456                     'upload_date': upload_date,
 457                 }
 458
 459         except compat_urllib_error.HTTPError:
 460             # This may be a stupid server that doesn't like HEAD, our UA, or so
 461             pass
 462
 463         try:
 464             webpage = self._download_webpage(url, video_id)
 465         except ValueError:
 466             # since this is the last-resort InfoExtractor, if
 467             # this error is thrown, it'll be thrown here
 468             raise ExtractorError('Failed to download URL: %s' % url)
 469
 470         self.report_extraction(video_id)
 471
 472         # Is it an RSS feed?
 473         try:
 474             doc = parse_xml(webpage)
 475             if doc.tag == 'rss':
 476                 return self._extract_rss(url, video_id, doc)
 477         except compat_xml_parse_error:
 478             pass
 479
 480         # Sometimes embedded video player is hidden behind percent encoding
 481         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 482         # Unescaping the whole page allows to handle those cases in a generic way
 483         webpage = compat_urllib_parse.unquote(webpage)
 484
 485         # it's tempting to parse this further, but you would
 486         # have to take into account all the variations like
 487         #   Video Title - Site Name
 488         #   Site Name | Video Title
 489         #   Video Title - Tagline | Site Name
 490         # and so on and so forth; it's just not practical
 491         video_title = self._html_search_regex(
 492             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 493             default='video')
 494
 495         # video uploader is domain name
 496         video_uploader = self._search_regex(
 497             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 498
 499         # Helper method
 500         def _playlist_from_matches(matches, getter, ie=None):
 501             urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
 502             return self.playlist_result(
 503                 urlrs, playlist_id=video_id, playlist_title=video_title)
 504
 505         # Look for BrightCove:
 506         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 507         if bc_urls:
 508             self.to_screen('Brightcove video detected.')
 509             entries = [{
 510                 '_type': 'url',
 511                 'url': smuggle_url(bc_url, {'Referer': url}),
 512                 'ie_key': 'Brightcove'
 513             } for bc_url in bc_urls]
 514
 515             return {
 516                 '_type': 'playlist',
 517                 'title': video_title,
 518                 'id': video_id,
 519                 'entries': entries,
 520             }
 521
 522         # Look for embedded (iframe) Vimeo player
 523         mobj = re.search(
 524             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 525         if mobj:
 526             player_url = unescapeHTML(mobj.group('url'))
 527             surl = smuggle_url(player_url, {'Referer': url})
 528             return self.url_result(surl, 'Vimeo')
 529
 530         # Look for embedded (swf embed) Vimeo player
 531         mobj = re.search(
 532             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 533         if mobj:
 534             return self.url_result(mobj.group(1), 'Vimeo')
 535
 536         # Look for embedded YouTube player
 537         matches = re.findall(r'''(?x)
 538             (?:
 539                 <iframe[^>]+?src=|
 540                 data-video-url=|
 541                 <embed[^>]+?src=|
 542                 embedSWF\(?:\s*
 543             )
 544             (["\'])
 545                 (?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
 546                 (?:embed|v)/.+?)
 547             \1''', webpage)
 548         if matches:
 549             return _playlist_from_matches(
 550                 matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
 551
 552         # Look for embedded Dailymotion player
 553         matches = re.findall(
 554             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 555         if matches:
 556             return _playlist_from_matches(
 557                 matches, lambda m: unescapeHTML(m[1]))
 558
 559         # Look for embedded Wistia player
 560         match = re.search(
 561             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 562         if match:
 563             return {
 564                 '_type': 'url_transparent',
 565                 'url': unescapeHTML(match.group('url')),
 566                 'ie_key': 'Wistia',
 567                 'uploader': video_uploader,
 568                 'title': video_title,
 569                 'id': video_id,
 570             }
 571
 572         # Look for embedded blip.tv player
 573         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 574         if mobj:
 575             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 576         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
 577         if mobj:
 578             return self.url_result(mobj.group(1), 'BlipTV')
 579
 580         # Look for embedded condenast player
 581         matches = re.findall(
 582             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 583             webpage)
 584         if matches:
 585             return {
 586                 '_type': 'playlist',
 587                 'entries': [{
 588                     '_type': 'url',
 589                     'ie_key': 'CondeNast',
 590                     'url': ma,
 591                 } for ma in matches],
 592                 'title': video_title,
 593                 'id': video_id,
 594             }
 595
 596         # Look for Bandcamp pages with custom domain
 597         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 598         if mobj is not None:
 599             burl = unescapeHTML(mobj.group(1))
 600             # Don't set the extractor because it can be a track url or an album
 601             return self.url_result(burl)
 602
 603         # Look for embedded Vevo player
 604         mobj = re.search(
 605             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 606         if mobj is not None:
 607             return self.url_result(mobj.group('url'))
 608
 609         # Look for Ooyala videos
 610         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 611              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 612         if mobj is not None:
 613             return OoyalaIE._build_url_result(mobj.group('ec'))
 614
 615         # Look for Aparat videos
 616         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 617         if mobj is not None:
 618             return self.url_result(mobj.group(1), 'Aparat')
 619
 620         # Look for MPORA videos
 621         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 622         if mobj is not None:
 623             return self.url_result(mobj.group(1), 'Mpora')
 624
 625         # Look for embedded NovaMov-based player
 626         mobj = re.search(
 627             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 628                     (?P<url>http://(?:(?:embed|www)\.)?
 629                         (?:novamov\.com|
 630                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 631                            videoweed\.(?:es|com)|
 632                            movshare\.(?:net|sx|ag)|
 633                            divxstage\.(?:eu|net|ch|co|at|ag))
 634                         /embed\.php.+?)\1''', webpage)
 635         if mobj is not None:
 636             return self.url_result(mobj.group('url'))
 637
 638         # Look for embedded Facebook player
 639         mobj = re.search(
 640             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 641         if mobj is not None:
 642             return self.url_result(mobj.group('url'), 'Facebook')
 643
 644         # Look for embedded VK player
 645         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 646         if mobj is not None:
 647             return self.url_result(mobj.group('url'), 'VK')
 648
 649         # Look for embedded ivi player
 650         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 651         if mobj is not None:
 652             return self.url_result(mobj.group('url'), 'Ivi')
 653
 654         # Look for embedded Huffington Post player
 655         mobj = re.search(
 656             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 657         if mobj is not None:
 658             return self.url_result(mobj.group('url'), 'HuffPost')
 659
 660         # Look for embed.ly
 661         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 662         if mobj is not None:
 663             return self.url_result(mobj.group('url'))
 664         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 665         if mobj is not None:
 666             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 667
 668         # Look for funnyordie embed
 669         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 670         if matches:
 671             return _playlist_from_matches(
 672                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 673
 674         # Look for embedded RUTV player
 675         rutv_url = RUTVIE._extract_url(webpage)
 676         if rutv_url:
 677             return self.url_result(rutv_url, 'RUTV')
 678
 679         # Look for embedded TED player
 680         mobj = re.search(
 681             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 682         if mobj is not None:
 683             return self.url_result(mobj.group('url'), 'TED')
 684
 685         # Look for embedded Ustream videos
 686         mobj = re.search(
 687             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 688         if mobj is not None:
 689             return self.url_result(mobj.group('url'), 'Ustream')
 690
 691         # Look for embedded arte.tv player
 692         mobj = re.search(
 693             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 694             webpage)
 695         if mobj is not None:
 696             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 697
 698         # Look for embedded smotri.com player
 699         smotri_url = SmotriIE._extract_url(webpage)
 700         if smotri_url:
 701             return self.url_result(smotri_url, 'Smotri')
 702
 703         # Look for embeded soundcloud player
 704         mobj = re.search(
 705             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 706             webpage)
 707         if mobj is not None:
 708             url = unescapeHTML(mobj.group('url'))
 709             return self.url_result(url)
 710
 711         # Look for embedded vulture.com player
 712         mobj = re.search(
 713             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 714             webpage)
 715         if mobj is not None:
 716             url = unescapeHTML(mobj.group('url'))
 717             return self.url_result(url, ie='Vulture')
 718
 719         # Look for embedded mtvservices player
 720         mobj = re.search(
 721             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 722             webpage)
 723         if mobj is not None:
 724             url = unescapeHTML(mobj.group('url'))
 725             return self.url_result(url, ie='MTVServicesEmbedded')
 726
 727         # Look for embedded yahoo player
 728         mobj = re.search(
 729             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 730             webpage)
 731         if mobj is not None:
 732             return self.url_result(mobj.group('url'), 'Yahoo')
 733
 734         # Look for embedded sbs.com.au player
 735         mobj = re.search(
 736             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 737             webpage)
 738         if mobj is not None:
 739             return self.url_result(mobj.group('url'), 'SBS')
 740
 741         # Start with something easy: JW Player in SWFObject
 742         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 743         if not found:
 744             # Look for gorilla-vid style embedding
 745             found = re.findall(r'''(?sx)
 746                 (?:
 747                     jw_plugins|
 748                     JWPlayerOptions|
 749                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 750                 )
 751                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 752         if not found:
 753             # Broaden the search a little bit
 754             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 755         if not found:
 756             # Broaden the findall a little bit: JWPlayer JS loader
 757             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 758         if not found:
 759             # Try to find twitter cards info
 760             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 761         if not found:
 762             # We look for Open Graph info:
 763             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 764             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 765             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 766             if m_video_type is not None:
 767                 found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
 768         if not found:
 769             # HTML5 video
 770             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
 771         if not found:
 772             found = re.search(
 773                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 774                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
 775                 webpage)
 776             if found:
 777                 new_url = found.group(1)
 778                 self.report_following_redirect(new_url)
 779                 return {
 780                     '_type': 'url',
 781                     'url': new_url,
 782                 }
 783         if not found:
 784             raise ExtractorError('Unsupported URL: %s' % url)
 785
 786         entries = []
 787         for video_url in found:
 788             video_url = compat_urlparse.urljoin(url, video_url)
 789             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 790
 791             # Sometimes, jwplayer extraction will result in a YouTube URL
 792             if YoutubeIE.suitable(video_url):
 793                 entries.append(self.url_result(video_url, 'Youtube'))
 794                 continue
 795
 796             # here's a fun little line of code for you:
 797             video_id = os.path.splitext(video_id)[0]
 798
 799             entries.append({
 800                 'id': video_id,
 801                 'url': video_url,
 802                 'uploader': video_uploader,
 803                 'title': video_title,
 804             })
 805
 806         if len(entries) == 1:
 807             return entries[0]
 808         else:
 809             for num, e in enumerate(entries, start=1):
 810                 e['title'] = '%s (%d)' % (e['title'], num)
 811             return {
 812                 '_type': 'playlist',
 813                 'entries': entries,
 814             }
 815