Raphaël G. Git Repositories - youtubedl/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_error,
  12     compat_urllib_parse,
  13     compat_urllib_request,
  14     compat_urlparse,
  15     compat_xml_parse_error,
  16
  17     ExtractorError,
  18     HEADRequest,
  19     parse_xml,
  20     smuggle_url,
  21     unescapeHTML,
  22     unified_strdate,
  23     url_basename,
  24 )
  25 from .brightcove import BrightcoveIE
  26 from .ooyala import OoyalaIE
  27 from .rutv import RUTVIE
  28 from .smotri import SmotriIE
  29
  30
  31 class GenericIE(InfoExtractor):
  32     IE_DESC = 'Generic downloader that works on some sites'
  33     _VALID_URL = r'.*'
  34     IE_NAME = 'generic'
  35     _TESTS = [
  36         {
  37             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  38             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  39             'info_dict': {
  40                 'id': '13601338388002',
  41                 'ext': 'mp4',
  42                 'uploader': 'www.hodiho.fr',
  43                 'title': 'R\u00e9gis plante sa Jeep',
  44             }
  45         },
  46         # bandcamp page with custom domain
  47         {
  48             'add_ie': ['Bandcamp'],
  49             'url': 'http://bronyrock.com/track/the-pony-mash',
  50             'info_dict': {
  51                 'id': '3235767654',
  52                 'ext': 'mp3',
  53                 'title': 'The Pony Mash',
  54                 'uploader': 'M_Pallante',
  55             },
  56             'skip': 'There is a limit of 200 free downloads / month for the test song',
  57         },
  58         # embedded brightcove video
  59         # it also tests brightcove videos that need to set the 'Referer' in the
  60         # http requests
  61         {
  62             'add_ie': ['Brightcove'],
  63             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  64             'info_dict': {
  65                 'id': '2765128793001',
  66                 'ext': 'mp4',
  67                 'title': 'Le cours de bourse : l’analyse technique',
  68                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  69                 'uploader': 'BFM BUSINESS',
  70             },
  71             'params': {
  72                 'skip_download': True,
  73             },
  74         },
  75         {
  76             # https://github.com/rg3/youtube-dl/issues/2253
  77             'url': 'http://bcove.me/i6nfkrc3',
  78             'md5': '0ba9446db037002366bab3b3eb30c88c',
  79             'info_dict': {
  80                 'id': '3101154703001',
  81                 'ext': 'mp4',
  82                 'title': 'Still no power',
  83                 'uploader': 'thestar.com',
  84                 '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.',
  85             },
  86             'add_ie': ['Brightcove'],
  87         },
  88         {
  89             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  90             'md5': 'fb973ecf6e4a78a67453647444222983',
  91             'info_dict': {
  92                 'id': '3414141473001',
  93                 'ext': 'mp4',
  94                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
  95                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
  96                 'uploader': 'Championat',
  97             },
  98         },
  99         # Direct link to a video
 100         {
 101             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 102             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 103             'info_dict': {
 104                 'id': 'trailer',
 105                 'ext': 'mp4',
 106                 'title': 'trailer',
 107                 'upload_date': '20100513',
 108             }
 109         },
 110         # ooyala video
 111         {
 112             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 113             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 114             'info_dict': {
 115                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 116                 'ext': 'mp4',
 117                 'title': '2cc213299525360.mov',  # that's what we get
 118             },
 119         },
 120         # google redirect
 121         {
 122             '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',
 123             'info_dict': {
 124                 'id': 'cmQHVoWB5FY',
 125                 'ext': 'mp4',
 126                 'upload_date': '20130224',
 127                 'uploader_id': 'TheVerge',
 128                 '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.',
 129                 'uploader': 'The Verge',
 130                 'title': 'First Firefox OS phones side-by-side',
 131             },
 132             'params': {
 133                 'skip_download': False,
 134             }
 135         },
 136         # embed.ly video
 137         {
 138             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 139             'info_dict': {
 140                 'id': '9ODmcdjQcHQ',
 141                 'ext': 'mp4',
 142                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 143                 'upload_date': '20140225',
 144                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 145                 'uploader': 'Tested',
 146                 'uploader_id': 'testedcom',
 147             },
 148             # No need to test YoutubeIE here
 149             'params': {
 150                 'skip_download': True,
 151             },
 152         },
 153         # funnyordie embed
 154         {
 155             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 156             'md5': '7cf780be104d40fea7bae52eed4a470e',
 157             'info_dict': {
 158                 'id': '18e820ec3f',
 159                 'ext': 'mp4',
 160                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 161                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 162             },
 163         },
 164         # RUTV embed
 165         {
 166             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 167             'info_dict': {
 168                 'id': '776940',
 169                 'ext': 'mp4',
 170                 'title': 'Охотское море стало целиком российским',
 171                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 172             },
 173             'params': {
 174                 # m3u8 download
 175                 'skip_download': True,
 176             },
 177         },
 178         # Embedded TED video
 179         {
 180             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 181             'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
 182             'info_dict': {
 183                 'id': '981',
 184                 'ext': 'mp4',
 185                 'title': 'My web playroom',
 186                 'uploader': 'Ze Frank',
 187                 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
 188             }
 189         },
 190         # Embeded Ustream video
 191         {
 192             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 193             'md5': '27b99cdb639c9b12a79bca876a073417',
 194             'info_dict': {
 195                 'id': '45734260',
 196                 'ext': 'flv',
 197                 'uploader': 'AU SPA:  The NSA and Privacy',
 198                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 199             }
 200         },
 201         # nowvideo embed hidden behind percent encoding
 202         {
 203             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 204             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 205             'info_dict': {
 206                 'id': '06e53103ca9aa',
 207                 'ext': 'flv',
 208                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 209                 'description': 'No description',
 210             },
 211         },
 212         # arte embed
 213         {
 214             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 215             'md5': '7653032cbb25bf6c80d80f217055fa43',
 216             'info_dict': {
 217                 'id': '048195-004_PLUS7-F',
 218                 'ext': 'flv',
 219                 'title': 'X:enius',
 220                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 221                 'upload_date': '20140320',
 222             },
 223             'params': {
 224                 'skip_download': 'Requires rtmpdump'
 225             }
 226         },
 227         # smotri embed
 228         {
 229             'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
 230             'md5': 'ec40048448e9284c9a1de77bb188108b',
 231             'info_dict': {
 232                 'id': 'v27008541fad',
 233                 'ext': 'mp4',
 234                 'title': 'Крым и Севастополь вошли в состав России',
 235                 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
 236                 'duration': 900,
 237                 'upload_date': '20140318',
 238                 'uploader': 'rbctv_2012_4',
 239                 'uploader_id': 'rbctv_2012_4',
 240             },
 241         },
 242         # Condé Nast embed
 243         {
 244             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 245             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 246             'info_dict': {
 247                 'id': '53501be369702d3275860000',
 248                 'ext': 'mp4',
 249                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 250             }
 251         },
 252         # Dailymotion embed
 253         {
 254             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 255             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 256             'info_dict': {
 257                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 258                 'ext': 'mp4',
 259                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 260                 'uploader': 'Spi0n',
 261             },
 262             'add_ie': ['Dailymotion'],
 263         }
 264     ]
 265
 266     def report_download_webpage(self, video_id):
 267         """Report webpage download."""
 268         if not self._downloader.params.get('test', False):
 269             self._downloader.report_warning('Falling back on generic information extractor.')
 270         super(GenericIE, self).report_download_webpage(video_id)
 271
 272     def report_following_redirect(self, new_url):
 273         """Report information extraction."""
 274         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 275
 276     def _send_head(self, url):
 277         """Check if it is a redirect, like url shorteners, in case return the new url."""
 278
 279         class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
 280             """
 281             Subclass the HTTPRedirectHandler to make it use our
 282             HEADRequest also on the redirected URL
 283             """
 284             def redirect_request(self, req, fp, code, msg, headers, newurl):
 285                 if code in (301, 302, 303, 307):
 286                     newurl = newurl.replace(' ', '%20')
 287                     newheaders = dict((k,v) for k,v in req.headers.items()
 288                                       if k.lower() not in ("content-length", "content-type"))
 289                     try:
 290                         # This function was deprecated in python 3.3 and removed in 3.4
 291                         origin_req_host = req.get_origin_req_host()
 292                     except AttributeError:
 293                         origin_req_host = req.origin_req_host
 294                     return HEADRequest(newurl,
 295                                        headers=newheaders,
 296                                        origin_req_host=origin_req_host,
 297                                        unverifiable=True)
 298                 else:
 299                     raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
 300
 301         class HTTPMethodFallback(compat_urllib_request.BaseHandler):
 302             """
 303             Fallback to GET if HEAD is not allowed (405 HTTP error)
 304             """
 305             def http_error_405(self, req, fp, code, msg, headers):
 306                 fp.read()
 307                 fp.close()
 308
 309                 newheaders = dict((k,v) for k,v in req.headers.items()
 310                                   if k.lower() not in ("content-length", "content-type"))
 311                 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
 312                                                  headers=newheaders,
 313                                                  origin_req_host=req.get_origin_req_host(),
 314                                                  unverifiable=True))
 315
 316         # Build our opener
 317         opener = compat_urllib_request.OpenerDirector()
 318         for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
 319                         HTTPMethodFallback, HEADRedirectHandler,
 320                         compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
 321             opener.add_handler(handler())
 322
 323         response = opener.open(HEADRequest(url))
 324         if response is None:
 325             raise ExtractorError('Invalid URL protocol')
 326         return response
 327
 328     def _extract_rss(self, url, video_id, doc):
 329         playlist_title = doc.find('./channel/title').text
 330         playlist_desc_el = doc.find('./channel/description')
 331         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 332
 333         entries = [{
 334             '_type': 'url',
 335             'url': e.find('link').text,
 336             'title': e.find('title').text,
 337         } for e in doc.findall('./channel/item')]
 338
 339         return {
 340             '_type': 'playlist',
 341             'id': url,
 342             'title': playlist_title,
 343             'description': playlist_desc,
 344             'entries': entries,
 345         }
 346
 347     def _real_extract(self, url):
 348         if url.startswith('//'):
 349             return {
 350                 '_type': 'url',
 351                 'url': self.http_scheme() + url,
 352             }
 353
 354         parsed_url = compat_urlparse.urlparse(url)
 355         if not parsed_url.scheme:
 356             default_search = self._downloader.params.get('default_search')
 357             if default_search is None:
 358                 default_search = 'auto_warning'
 359
 360             if default_search in ('auto', 'auto_warning'):
 361                 if '/' in url:
 362                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 363                     return self.url_result('http://' + url)
 364                 else:
 365                     if default_search == 'auto_warning':
 366                         if re.match(r'^(?:url|URL)$', url):
 367                             raise ExtractorError(
 368                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 369                                 expected=True)
 370                         else:
 371                             self._downloader.report_warning(
 372                                 'Falling back to youtube search for  %s . Set --default-search to "auto" to suppress this warning.' % url)
 373                     return self.url_result('ytsearch:' + url)
 374             else:
 375                 assert ':' in default_search
 376                 return self.url_result(default_search + url)
 377         video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 378
 379         self.to_screen('%s: Requesting header' % video_id)
 380
 381         try:
 382             response = self._send_head(url)
 383
 384             # Check for redirect
 385             new_url = response.geturl()
 386             if url != new_url:
 387                 self.report_following_redirect(new_url)
 388                 return self.url_result(new_url)
 389
 390             # Check for direct link to a video
 391             content_type = response.headers.get('Content-Type', '')
 392             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 393             if m:
 394                 upload_date = response.headers.get('Last-Modified')
 395                 if upload_date:
 396                     upload_date = unified_strdate(upload_date)
 397                 return {
 398                     'id': video_id,
 399                     'title': os.path.splitext(url_basename(url))[0],
 400                     'formats': [{
 401                         'format_id': m.group('format_id'),
 402                         'url': url,
 403                         'vcodec': 'none' if m.group('type') == 'audio' else None
 404                     }],
 405                     'upload_date': upload_date,
 406                 }
 407
 408         except compat_urllib_error.HTTPError:
 409             # This may be a stupid server that doesn't like HEAD, our UA, or so
 410             pass
 411
 412         try:
 413             webpage = self._download_webpage(url, video_id)
 414         except ValueError:
 415             # since this is the last-resort InfoExtractor, if
 416             # this error is thrown, it'll be thrown here
 417             raise ExtractorError('Failed to download URL: %s' % url)
 418
 419         self.report_extraction(video_id)
 420
 421         # Is it an RSS feed?
 422         try:
 423             doc = parse_xml(webpage)
 424             if doc.tag == 'rss':
 425                 return self._extract_rss(url, video_id, doc)
 426         except compat_xml_parse_error:
 427             pass
 428
 429         # Sometimes embedded video player is hidden behind percent encoding
 430         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 431         # Unescaping the whole page allows to handle those cases in a generic way
 432         webpage = compat_urllib_parse.unquote(webpage)
 433
 434         # it's tempting to parse this further, but you would
 435         # have to take into account all the variations like
 436         #   Video Title - Site Name
 437         #   Site Name | Video Title
 438         #   Video Title - Tagline | Site Name
 439         # and so on and so forth; it's just not practical
 440         video_title = self._html_search_regex(
 441             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 442             default='video')
 443
 444         # video uploader is domain name
 445         video_uploader = self._search_regex(
 446             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 447
 448         # Look for BrightCove:
 449         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 450         if bc_urls:
 451             self.to_screen('Brightcove video detected.')
 452             entries = [{
 453                 '_type': 'url',
 454                 'url': smuggle_url(bc_url, {'Referer': url}),
 455                 'ie_key': 'Brightcove'
 456             } for bc_url in bc_urls]
 457
 458             return {
 459                 '_type': 'playlist',
 460                 'title': video_title,
 461                 'id': video_id,
 462                 'entries': entries,
 463             }
 464
 465         # Look for embedded (iframe) Vimeo player
 466         mobj = re.search(
 467             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 468         if mobj:
 469             player_url = unescapeHTML(mobj.group('url'))
 470             surl = smuggle_url(player_url, {'Referer': url})
 471             return self.url_result(surl, 'Vimeo')
 472
 473         # Look for embedded (swf embed) Vimeo player
 474         mobj = re.search(
 475             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 476         if mobj:
 477             return self.url_result(mobj.group(1), 'Vimeo')
 478
 479         # Look for embedded YouTube player
 480         matches = re.findall(r'''(?x)
 481             (?:<iframe[^>]+?src=|embedSWF\(\s*)
 482             (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
 483                 (?:embed|v)/.+?)
 484             \1''', webpage)
 485         if matches:
 486             urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
 487                      for tuppl in matches]
 488             return self.playlist_result(
 489                 urlrs, playlist_id=video_id, playlist_title=video_title)
 490
 491         # Look for embedded Dailymotion player
 492         matches = re.findall(
 493             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 494         if matches:
 495             urlrs = [self.url_result(unescapeHTML(tuppl[1]))
 496                      for tuppl in matches]
 497             return self.playlist_result(
 498                 urlrs, playlist_id=video_id, playlist_title=video_title)
 499
 500         # Look for embedded Wistia player
 501         match = re.search(
 502             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 503         if match:
 504             return {
 505                 '_type': 'url_transparent',
 506                 'url': unescapeHTML(match.group('url')),
 507                 'ie_key': 'Wistia',
 508                 'uploader': video_uploader,
 509                 'title': video_title,
 510                 'id': video_id,
 511             }
 512
 513         # Look for embedded blip.tv player
 514         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 515         if mobj:
 516             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 517         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
 518         if mobj:
 519             return self.url_result(mobj.group(1), 'BlipTV')
 520
 521         # Look for embedded condenast player
 522         matches = re.findall(
 523             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 524             webpage)
 525         if matches:
 526             return {
 527                 '_type': 'playlist',
 528                 'entries': [{
 529                     '_type': 'url',
 530                     'ie_key': 'CondeNast',
 531                     'url': ma,
 532                 } for ma in matches],
 533                 'title': video_title,
 534                 'id': video_id,
 535             }
 536
 537         # Look for Bandcamp pages with custom domain
 538         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 539         if mobj is not None:
 540             burl = unescapeHTML(mobj.group(1))
 541             # Don't set the extractor because it can be a track url or an album
 542             return self.url_result(burl)
 543
 544         # Look for embedded Vevo player
 545         mobj = re.search(
 546             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 547         if mobj is not None:
 548             return self.url_result(mobj.group('url'))
 549
 550         # Look for Ooyala videos
 551         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 552              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 553         if mobj is not None:
 554             return OoyalaIE._build_url_result(mobj.group('ec'))
 555
 556         # Look for Aparat videos
 557         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 558         if mobj is not None:
 559             return self.url_result(mobj.group(1), 'Aparat')
 560
 561         # Look for MPORA videos
 562         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 563         if mobj is not None:
 564             return self.url_result(mobj.group(1), 'Mpora')
 565
 566         # Look for embedded NovaMov-based player
 567         mobj = re.search(
 568             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 569                     (?P<url>http://(?:(?:embed|www)\.)?
 570                         (?:novamov\.com|
 571                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 572                            videoweed\.(?:es|com)|
 573                            movshare\.(?:net|sx|ag)|
 574                            divxstage\.(?:eu|net|ch|co|at|ag))
 575                         /embed\.php.+?)\1''', webpage)
 576         if mobj is not None:
 577             return self.url_result(mobj.group('url'))
 578
 579         # Look for embedded Facebook player
 580         mobj = re.search(
 581             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 582         if mobj is not None:
 583             return self.url_result(mobj.group('url'), 'Facebook')
 584
 585         # Look for embedded VK player
 586         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 587         if mobj is not None:
 588             return self.url_result(mobj.group('url'), 'VK')
 589
 590         # Look for embedded Huffington Post player
 591         mobj = re.search(
 592             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 593         if mobj is not None:
 594             return self.url_result(mobj.group('url'), 'HuffPost')
 595
 596         # Look for embed.ly
 597         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 598         if mobj is not None:
 599             return self.url_result(mobj.group('url'))
 600         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 601         if mobj is not None:
 602             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 603
 604         # Look for funnyordie embed
 605         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 606         if matches:
 607             urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
 608                      for eurl in matches]
 609             return self.playlist_result(
 610                 urlrs, playlist_id=video_id, playlist_title=video_title)
 611
 612         # Look for embedded RUTV player
 613         rutv_url = RUTVIE._extract_url(webpage)
 614         if rutv_url:
 615             return self.url_result(rutv_url, 'RUTV')
 616
 617         # Look for embedded TED player
 618         mobj = re.search(
 619             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 620         if mobj is not None:
 621             return self.url_result(mobj.group('url'), 'TED')
 622
 623         # Look for embedded Ustream videos
 624         mobj = re.search(
 625             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 626         if mobj is not None:
 627             return self.url_result(mobj.group('url'), 'Ustream')
 628
 629         # Look for embedded arte.tv player
 630         mobj = re.search(
 631             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 632             webpage)
 633         if mobj is not None:
 634             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 635
 636         # Look for embedded smotri.com player
 637         smotri_url = SmotriIE._extract_url(webpage)
 638         if smotri_url:
 639             return self.url_result(smotri_url, 'Smotri')
 640
 641         # Look for embeded soundcloud player
 642         mobj = re.search(
 643             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 644             webpage)
 645         if mobj is not None:
 646             url = unescapeHTML(mobj.group('url'))
 647             return self.url_result(url)
 648
 649         # Start with something easy: JW Player in SWFObject
 650         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 651         if not found:
 652             # Look for gorilla-vid style embedding
 653             found = re.findall(r'''(?sx)
 654                 (?:
 655                     jw_plugins|
 656                     JWPlayerOptions|
 657                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 658                 )
 659                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 660         if not found:
 661             # Broaden the search a little bit
 662             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 663         if not found:
 664             # Broaden the findall a little bit: JWPlayer JS loader
 665             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 666         if not found:
 667             # Try to find twitter cards info
 668             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 669         if not found:
 670             # We look for Open Graph info:
 671             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 672             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 673             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 674             if m_video_type is not None:
 675                 found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
 676         if not found:
 677             # HTML5 video
 678             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
 679         if not found:
 680             found = re.search(
 681                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 682                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
 683                 webpage)
 684             if found:
 685                 new_url = found.group(1)
 686                 self.report_following_redirect(new_url)
 687                 return {
 688                     '_type': 'url',
 689                     'url': new_url,
 690                 }
 691         if not found:
 692             raise ExtractorError('Unsupported URL: %s' % url)
 693
 694         entries = []
 695         for video_url in found:
 696             video_url = compat_urlparse.urljoin(url, video_url)
 697             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 698
 699             # Sometimes, jwplayer extraction will result in a YouTube URL
 700             if YoutubeIE.suitable(video_url):
 701                 entries.append(self.url_result(video_url, 'Youtube'))
 702                 continue
 703
 704             # here's a fun little line of code for you:
 705             video_id = os.path.splitext(video_id)[0]
 706
 707             entries.append({
 708                 'id': video_id,
 709                 'url': video_url,
 710                 'uploader': video_uploader,
 711                 'title': video_title,
 712             })
 713
 714         if len(entries) == 1:
 715             return entries[0]
 716         else:
 717             for num, e in enumerate(entries, start=1):
 718                 e['title'] = '%s (%d)' % (e['title'], num)
 719             return {
 720                 '_type': 'playlist',
 721                 'entries': entries,
 722             }
 723