youtube_dl/extractor/generic.py

   1 # encoding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5 import os
   6 import re
   7
   8 from .common import InfoExtractor
   9 from .youtube import YoutubeIE
  10 from ..utils import (
  11     compat_urllib_parse,
  12     compat_urlparse,
  13     compat_xml_parse_error,
  14
  15     ExtractorError,
  16     float_or_none,
  17     HEADRequest,
  18     orderedSet,
  19     parse_xml,
  20     smuggle_url,
  21     unescapeHTML,
  22     unified_strdate,
  23     unsmuggle_url,
  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         # Camtasia studio
 310         {
 311             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 312             'playlist': [{
 313                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 314                 'info_dict': {
 315                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 316                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 317                     'ext': 'flv',
 318                     'duration': 2235.90,
 319                 }
 320             }, {
 321                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 322                 'info_dict': {
 323                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 324                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 325                     'ext': 'flv',
 326                     'duration': 2235.93,
 327                 }
 328             }],
 329             'info_dict': {
 330                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 331             }
 332         },
 333         # Flowplayer
 334         {
 335             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 336             'md5': '9d65602bf31c6e20014319c7d07fba27',
 337             'info_dict': {
 338                 'id': '5123ea6d5e5a7',
 339                 'ext': 'mp4',
 340                 'age_limit': 18,
 341                 'uploader': 'www.handjobhub.com',
 342                 'title': 'Busty Blonde Siri Tit Fuck While Wank at Handjob Hub',
 343             }
 344         },
 345         # RSS feed
 346         {
 347             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 348             'info_dict': {
 349                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 350                 'title': 'Zero Punctuation',
 351                 'description': 're:'
 352             },
 353             'playlist_mincount': 11,
 354         }
 355     ]
 356
 357     def report_download_webpage(self, video_id):
 358         """Report webpage download."""
 359         if not self._downloader.params.get('test', False):
 360             self._downloader.report_warning('Falling back on generic information extractor.')
 361         super(GenericIE, self).report_download_webpage(video_id)
 362
 363     def report_following_redirect(self, new_url):
 364         """Report information extraction."""
 365         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 366
 367     def _extract_rss(self, url, video_id, doc):
 368         playlist_title = doc.find('./channel/title').text
 369         playlist_desc_el = doc.find('./channel/description')
 370         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 371
 372         entries = [{
 373             '_type': 'url',
 374             'url': e.find('link').text,
 375             'title': e.find('title').text,
 376         } for e in doc.findall('./channel/item')]
 377
 378         return {
 379             '_type': 'playlist',
 380             'id': url,
 381             'title': playlist_title,
 382             'description': playlist_desc,
 383             'entries': entries,
 384         }
 385
 386     def _extract_camtasia(self, url, video_id, webpage):
 387         """ Returns None if no camtasia video can be found. """
 388
 389         camtasia_cfg = self._search_regex(
 390             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 391             webpage, 'camtasia configuration file', default=None)
 392         if camtasia_cfg is None:
 393             return None
 394
 395         title = self._html_search_meta('DC.title', webpage, fatal=True)
 396
 397         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 398         camtasia_cfg = self._download_xml(
 399             camtasia_url, video_id,
 400             note='Downloading camtasia configuration',
 401             errnote='Failed to download camtasia configuration')
 402         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 403
 404         entries = []
 405         for n in fileset_node.getchildren():
 406             url_n = n.find('./uri')
 407             if url_n is None:
 408                 continue
 409
 410             entries.append({
 411                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 412                 'title': '%s - %s' % (title, n.tag),
 413                 'url': compat_urlparse.urljoin(url, url_n.text),
 414                 'duration': float_or_none(n.find('./duration').text),
 415             })
 416
 417         return {
 418             '_type': 'playlist',
 419             'entries': entries,
 420             'title': title,
 421         }
 422
 423     def _real_extract(self, url):
 424         if url.startswith('//'):
 425             return {
 426                 '_type': 'url',
 427                 'url': self.http_scheme() + url,
 428             }
 429
 430         parsed_url = compat_urlparse.urlparse(url)
 431         if not parsed_url.scheme:
 432             default_search = self._downloader.params.get('default_search')
 433             if default_search is None:
 434                 default_search = 'fixup_error'
 435
 436             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 437                 if '/' in url:
 438                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 439                     return self.url_result('http://' + url)
 440                 elif default_search != 'fixup_error':
 441                     if default_search == 'auto_warning':
 442                         if re.match(r'^(?:url|URL)$', url):
 443                             raise ExtractorError(
 444                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 445                                 expected=True)
 446                         else:
 447                             self._downloader.report_warning(
 448                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 449                     return self.url_result('ytsearch:' + url)
 450
 451             if default_search in ('error', 'fixup_error'):
 452                 raise ExtractorError(
 453                     ('%r is not a valid URL. '
 454                      'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 455                     ) % (url, url), expected=True)
 456             else:
 457                 assert ':' in default_search
 458                 return self.url_result(default_search + url)
 459
 460         url, smuggled_data = unsmuggle_url(url)
 461         force_videoid = None
 462         if smuggled_data and 'force_videoid' in smuggled_data:
 463             force_videoid = smuggled_data['force_videoid']
 464             video_id = force_videoid
 465         else:
 466             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 467
 468         self.to_screen('%s: Requesting header' % video_id)
 469
 470         head_req = HEADRequest(url)
 471         response = self._request_webpage(
 472             head_req, video_id,
 473             note=False, errnote='Could not send HEAD request to %s' % url,
 474             fatal=False)
 475
 476         if response is not False:
 477             # Check for redirect
 478             new_url = response.geturl()
 479             if url != new_url:
 480                 self.report_following_redirect(new_url)
 481                 if force_videoid:
 482                     new_url = smuggle_url(
 483                         new_url, {'force_videoid': force_videoid})
 484                 return self.url_result(new_url)
 485
 486             # Check for direct link to a video
 487             content_type = response.headers.get('Content-Type', '')
 488             m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 489             if m:
 490                 upload_date = response.headers.get('Last-Modified')
 491                 if upload_date:
 492                     upload_date = unified_strdate(upload_date)
 493                 return {
 494                     'id': video_id,
 495                     'title': os.path.splitext(url_basename(url))[0],
 496                     'formats': [{
 497                         'format_id': m.group('format_id'),
 498                         'url': url,
 499                         'vcodec': 'none' if m.group('type') == 'audio' else None
 500                     }],
 501                     'upload_date': upload_date,
 502                 }
 503
 504         try:
 505             webpage = self._download_webpage(url, video_id)
 506         except ValueError:
 507             # since this is the last-resort InfoExtractor, if
 508             # this error is thrown, it'll be thrown here
 509             raise ExtractorError('Failed to download URL: %s' % url)
 510
 511         self.report_extraction(video_id)
 512
 513         # Is it an RSS feed?
 514         try:
 515             doc = parse_xml(webpage)
 516             if doc.tag == 'rss':
 517                 return self._extract_rss(url, video_id, doc)
 518         except compat_xml_parse_error:
 519             pass
 520
 521         # Is it a Camtasia project?
 522         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 523         if camtasia_res is not None:
 524             return camtasia_res
 525
 526         # Sometimes embedded video player is hidden behind percent encoding
 527         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 528         # Unescaping the whole page allows to handle those cases in a generic way
 529         webpage = compat_urllib_parse.unquote(webpage)
 530
 531         # it's tempting to parse this further, but you would
 532         # have to take into account all the variations like
 533         #   Video Title - Site Name
 534         #   Site Name | Video Title
 535         #   Video Title - Tagline | Site Name
 536         # and so on and so forth; it's just not practical
 537         video_title = self._html_search_regex(
 538             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 539             default='video')
 540
 541         # Try to detect age limit automatically
 542         age_limit = self._rta_search(webpage)
 543         # And then there are the jokers who advertise that they use RTA,
 544         # but actually don't.
 545         AGE_LIMIT_MARKERS = [
 546             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 547         ]
 548         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 549             age_limit = 18
 550
 551         # video uploader is domain name
 552         video_uploader = self._search_regex(
 553             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 554
 555         # Helper method
 556         def _playlist_from_matches(matches, getter, ie=None):
 557             urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
 558             return self.playlist_result(
 559                 urlrs, playlist_id=video_id, playlist_title=video_title)
 560
 561         # Look for BrightCove:
 562         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 563         if bc_urls:
 564             self.to_screen('Brightcove video detected.')
 565             entries = [{
 566                 '_type': 'url',
 567                 'url': smuggle_url(bc_url, {'Referer': url}),
 568                 'ie_key': 'Brightcove'
 569             } for bc_url in bc_urls]
 570
 571             return {
 572                 '_type': 'playlist',
 573                 'title': video_title,
 574                 'id': video_id,
 575                 'entries': entries,
 576             }
 577
 578         # Look for embedded (iframe) Vimeo player
 579         mobj = re.search(
 580             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 581         if mobj:
 582             player_url = unescapeHTML(mobj.group('url'))
 583             surl = smuggle_url(player_url, {'Referer': url})
 584             return self.url_result(surl, 'Vimeo')
 585
 586         # Look for embedded (swf embed) Vimeo player
 587         mobj = re.search(
 588             r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 589         if mobj:
 590             return self.url_result(mobj.group(1), 'Vimeo')
 591
 592         # Look for embedded YouTube player
 593         matches = re.findall(r'''(?x)
 594             (?:
 595                 <iframe[^>]+?src=|
 596                 data-video-url=|
 597                 <embed[^>]+?src=|
 598                 embedSWF\(?:\s*
 599             )
 600             (["\'])
 601                 (?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
 602                 (?:embed|v)/.+?)
 603             \1''', webpage)
 604         if matches:
 605             return _playlist_from_matches(
 606                 matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
 607
 608         # Look for embedded Dailymotion player
 609         matches = re.findall(
 610             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 611         if matches:
 612             return _playlist_from_matches(
 613                 matches, lambda m: unescapeHTML(m[1]))
 614
 615         # Look for embedded Wistia player
 616         match = re.search(
 617             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 618         if match:
 619             return {
 620                 '_type': 'url_transparent',
 621                 'url': unescapeHTML(match.group('url')),
 622                 'ie_key': 'Wistia',
 623                 'uploader': video_uploader,
 624                 'title': video_title,
 625                 'id': video_id,
 626             }
 627
 628         # Look for embedded blip.tv player
 629         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 630         if mobj:
 631             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
 632         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 633         if mobj:
 634             return self.url_result(mobj.group(1), 'BlipTV')
 635
 636         # Look for embedded condenast player
 637         matches = re.findall(
 638             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 639             webpage)
 640         if matches:
 641             return {
 642                 '_type': 'playlist',
 643                 'entries': [{
 644                     '_type': 'url',
 645                     'ie_key': 'CondeNast',
 646                     'url': ma,
 647                 } for ma in matches],
 648                 'title': video_title,
 649                 'id': video_id,
 650             }
 651
 652         # Look for Bandcamp pages with custom domain
 653         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 654         if mobj is not None:
 655             burl = unescapeHTML(mobj.group(1))
 656             # Don't set the extractor because it can be a track url or an album
 657             return self.url_result(burl)
 658
 659         # Look for embedded Vevo player
 660         mobj = re.search(
 661             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 662         if mobj is not None:
 663             return self.url_result(mobj.group('url'))
 664
 665         # Look for Ooyala videos
 666         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 667              re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 668         if mobj is not None:
 669             return OoyalaIE._build_url_result(mobj.group('ec'))
 670
 671         # Look for Aparat videos
 672         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 673         if mobj is not None:
 674             return self.url_result(mobj.group(1), 'Aparat')
 675
 676         # Look for MPORA videos
 677         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 678         if mobj is not None:
 679             return self.url_result(mobj.group(1), 'Mpora')
 680
 681         # Look for embedded NovaMov-based player
 682         mobj = re.search(
 683             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 684                     (?P<url>http://(?:(?:embed|www)\.)?
 685                         (?:novamov\.com|
 686                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 687                            videoweed\.(?:es|com)|
 688                            movshare\.(?:net|sx|ag)|
 689                            divxstage\.(?:eu|net|ch|co|at|ag))
 690                         /embed\.php.+?)\1''', webpage)
 691         if mobj is not None:
 692             return self.url_result(mobj.group('url'))
 693
 694         # Look for embedded Facebook player
 695         mobj = re.search(
 696             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 697         if mobj is not None:
 698             return self.url_result(mobj.group('url'), 'Facebook')
 699
 700         # Look for embedded VK player
 701         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 702         if mobj is not None:
 703             return self.url_result(mobj.group('url'), 'VK')
 704
 705         # Look for embedded ivi player
 706         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 707         if mobj is not None:
 708             return self.url_result(mobj.group('url'), 'Ivi')
 709
 710         # Look for embedded Huffington Post player
 711         mobj = re.search(
 712             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 713         if mobj is not None:
 714             return self.url_result(mobj.group('url'), 'HuffPost')
 715
 716         # Look for embed.ly
 717         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 718         if mobj is not None:
 719             return self.url_result(mobj.group('url'))
 720         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 721         if mobj is not None:
 722             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 723
 724         # Look for funnyordie embed
 725         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 726         if matches:
 727             return _playlist_from_matches(
 728                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 729
 730         # Look for embedded RUTV player
 731         rutv_url = RUTVIE._extract_url(webpage)
 732         if rutv_url:
 733             return self.url_result(rutv_url, 'RUTV')
 734
 735         # Look for embedded TED player
 736         mobj = re.search(
 737             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 738         if mobj is not None:
 739             return self.url_result(mobj.group('url'), 'TED')
 740
 741         # Look for embedded Ustream videos
 742         mobj = re.search(
 743             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 744         if mobj is not None:
 745             return self.url_result(mobj.group('url'), 'Ustream')
 746
 747         # Look for embedded arte.tv player
 748         mobj = re.search(
 749             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 750             webpage)
 751         if mobj is not None:
 752             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 753
 754         # Look for embedded smotri.com player
 755         smotri_url = SmotriIE._extract_url(webpage)
 756         if smotri_url:
 757             return self.url_result(smotri_url, 'Smotri')
 758
 759         # Look for embeded soundcloud player
 760         mobj = re.search(
 761             r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 762             webpage)
 763         if mobj is not None:
 764             url = unescapeHTML(mobj.group('url'))
 765             return self.url_result(url)
 766
 767         # Look for embedded vulture.com player
 768         mobj = re.search(
 769             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 770             webpage)
 771         if mobj is not None:
 772             url = unescapeHTML(mobj.group('url'))
 773             return self.url_result(url, ie='Vulture')
 774
 775         # Look for embedded mtvservices player
 776         mobj = re.search(
 777             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 778             webpage)
 779         if mobj is not None:
 780             url = unescapeHTML(mobj.group('url'))
 781             return self.url_result(url, ie='MTVServicesEmbedded')
 782
 783         # Look for embedded yahoo player
 784         mobj = re.search(
 785             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 786             webpage)
 787         if mobj is not None:
 788             return self.url_result(mobj.group('url'), 'Yahoo')
 789
 790         # Look for embedded sbs.com.au player
 791         mobj = re.search(
 792             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 793             webpage)
 794         if mobj is not None:
 795             return self.url_result(mobj.group('url'), 'SBS')
 796
 797         # Start with something easy: JW Player in SWFObject
 798         found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
 799         if not found:
 800             # Look for gorilla-vid style embedding
 801             found = re.findall(r'''(?sx)
 802                 (?:
 803                     jw_plugins|
 804                     JWPlayerOptions|
 805                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
 806                 )
 807                 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
 808         if not found:
 809             # Broaden the search a little bit
 810             found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
 811         if not found:
 812             # Broaden the findall a little bit: JWPlayer JS loader
 813             found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
 814         if not found:
 815             # Flow player
 816             found = re.findall(r'''(?xs)
 817                 flowplayer\("[^"]+",\s*
 818                     \{[^}]+?\}\s*,
 819                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
 820                         ["']?url["']?\s*:\s*["']([^"']+)["']
 821             ''', webpage)
 822         if not found:
 823             # Try to find twitter cards info
 824             found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
 825         if not found:
 826             # We look for Open Graph info:
 827             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
 828             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
 829             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
 830             if m_video_type is not None:
 831                 def check_video(vurl):
 832                     vpath = compat_urlparse.urlparse(vurl).path
 833                     vext = determine_ext(vpath)
 834                     return '.' in vpath and vext not in ('swf', 'png', 'jpg')
 835                 found = list(filter(
 836                     check_video,
 837                     re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
 838         if not found:
 839             # HTML5 video
 840             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]+)? src="([^"]+)"', webpage)
 841         if not found:
 842             found = re.search(
 843                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
 844                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
 845                 webpage)
 846             if found:
 847                 new_url = found.group(1)
 848                 self.report_following_redirect(new_url)
 849                 return {
 850                     '_type': 'url',
 851                     'url': new_url,
 852                 }
 853         if not found:
 854             raise ExtractorError('Unsupported URL: %s' % url)
 855
 856         entries = []
 857         for video_url in found:
 858             video_url = compat_urlparse.urljoin(url, video_url)
 859             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 860
 861             # Sometimes, jwplayer extraction will result in a YouTube URL
 862             if YoutubeIE.suitable(video_url):
 863                 entries.append(self.url_result(video_url, 'Youtube'))
 864                 continue
 865
 866             # here's a fun little line of code for you:
 867             video_id = os.path.splitext(video_id)[0]
 868
 869             entries.append({
 870                 'id': video_id,
 871                 'url': video_url,
 872                 'uploader': video_uploader,
 873                 'title': video_title,
 874                 'age_limit': age_limit,
 875             })
 876
 877         if len(entries) == 1:
 878             return entries[0]
 879         else:
 880             for num, e in enumerate(entries, start=1):
 881                 e['title'] = '%s (%d)' % (e['title'], num)
 882             return {
 883                 '_type': 'playlist',
 884                 'entries': entries,
 885             }
 886