_ Git - youtube-dl/blob - youtube_dl/extractor/generic.py

   1 # encoding: utf-8
   2
   3 from __future__ import unicode_literals
   4
   5 import os
   6 import re
   7
   8 from .common import InfoExtractor
   9 from .youtube import YoutubeIE
  10 from ..compat import (
  11     compat_urllib_parse,
  12     compat_urlparse,
  13     compat_xml_parse_error,
  14 )
  15 from ..utils import (
  16     determine_ext,
  17     ExtractorError,
  18     float_or_none,
  19     HEADRequest,
  20     orderedSet,
  21     parse_xml,
  22     smuggle_url,
  23     unescapeHTML,
  24     unified_strdate,
  25     unsmuggle_url,
  26     UnsupportedError,
  27     url_basename,
  28 )
  29 from .brightcove import BrightcoveIE
  30 from .ooyala import OoyalaIE
  31 from .rutv import RUTVIE
  32 from .smotri import SmotriIE
  33 from .condenast import CondeNastIE
  34
  35
  36 class GenericIE(InfoExtractor):
  37     IE_DESC = 'Generic downloader that works on some sites'
  38     _VALID_URL = r'.*'
  39     IE_NAME = 'generic'
  40     _TESTS = [
  41         {
  42             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  43             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  44             'info_dict': {
  45                 'id': '13601338388002',
  46                 'ext': 'mp4',
  47                 'uploader': 'www.hodiho.fr',
  48                 'title': 'R\u00e9gis plante sa Jeep',
  49             }
  50         },
  51         # bandcamp page with custom domain
  52         {
  53             'add_ie': ['Bandcamp'],
  54             'url': 'http://bronyrock.com/track/the-pony-mash',
  55             'info_dict': {
  56                 'id': '3235767654',
  57                 'ext': 'mp3',
  58                 'title': 'The Pony Mash',
  59                 'uploader': 'M_Pallante',
  60             },
  61             'skip': 'There is a limit of 200 free downloads / month for the test song',
  62         },
  63         # embedded brightcove video
  64         # it also tests brightcove videos that need to set the 'Referer' in the
  65         # http requests
  66         {
  67             'add_ie': ['Brightcove'],
  68             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  69             'info_dict': {
  70                 'id': '2765128793001',
  71                 'ext': 'mp4',
  72                 'title': 'Le cours de bourse : l’analyse technique',
  73                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  74                 'uploader': 'BFM BUSINESS',
  75             },
  76             'params': {
  77                 'skip_download': True,
  78             },
  79         },
  80         {
  81             # https://github.com/rg3/youtube-dl/issues/2253
  82             'url': 'http://bcove.me/i6nfkrc3',
  83             'md5': '0ba9446db037002366bab3b3eb30c88c',
  84             'info_dict': {
  85                 'id': '3101154703001',
  86                 'ext': 'mp4',
  87                 'title': 'Still no power',
  88                 'uploader': 'thestar.com',
  89                 '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.',
  90             },
  91             'add_ie': ['Brightcove'],
  92         },
  93         {
  94             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  95             'md5': 'fb973ecf6e4a78a67453647444222983',
  96             'info_dict': {
  97                 'id': '3414141473001',
  98                 'ext': 'mp4',
  99                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
 100                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
 101                 'uploader': 'Championat',
 102             },
 103         },
 104         {
 105             # https://github.com/rg3/youtube-dl/issues/3541
 106             'add_ie': ['Brightcove'],
 107             'url': 'http://www.kijk.nl/sbs6/leermijvrouwenkennen/videos/jqMiXKAYan2S/aflevering-1',
 108             'info_dict': {
 109                 'id': '3866516442001',
 110                 'ext': 'mp4',
 111                 'title': 'Leer mij vrouwen kennen: Aflevering 1',
 112                 'description': 'Leer mij vrouwen kennen: Aflevering 1',
 113                 'uploader': 'SBS Broadcasting',
 114             },
 115             'skip': 'Restricted to Netherlands',
 116             'params': {
 117                 'skip_download': True,  # m3u8 download
 118             },
 119         },
 120         # Direct link to a video
 121         {
 122             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 123             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 124             'info_dict': {
 125                 'id': 'trailer',
 126                 'ext': 'mp4',
 127                 'title': 'trailer',
 128                 'upload_date': '20100513',
 129             }
 130         },
 131         # ooyala video
 132         {
 133             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 134             'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
 135             'info_dict': {
 136                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 137                 'ext': 'mp4',
 138                 'title': '2cc213299525360.mov',  # that's what we get
 139             },
 140         },
 141         # google redirect
 142         {
 143             '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',
 144             'info_dict': {
 145                 'id': 'cmQHVoWB5FY',
 146                 'ext': 'mp4',
 147                 'upload_date': '20130224',
 148                 'uploader_id': 'TheVerge',
 149                 '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.',
 150                 'uploader': 'The Verge',
 151                 'title': 'First Firefox OS phones side-by-side',
 152             },
 153             'params': {
 154                 'skip_download': False,
 155             }
 156         },
 157         # embed.ly video
 158         {
 159             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 160             'info_dict': {
 161                 'id': '9ODmcdjQcHQ',
 162                 'ext': 'mp4',
 163                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 164                 'upload_date': '20140225',
 165                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 166                 'uploader': 'Tested',
 167                 'uploader_id': 'testedcom',
 168             },
 169             # No need to test YoutubeIE here
 170             'params': {
 171                 'skip_download': True,
 172             },
 173         },
 174         # funnyordie embed
 175         {
 176             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 177             'info_dict': {
 178                 'id': '18e820ec3f',
 179                 'ext': 'mp4',
 180                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 181                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 182             },
 183         },
 184         # BBC iPlayer embeds
 185         {
 186             'url': 'http://www.bbc.co.uk/blogs/adamcurtis/posts/BUGGER',
 187             'info_dict': {
 188                 'title': 'BBC - Blogs -  Adam Curtis - BUGGER',
 189             },
 190             'playlist_mincount': 18,
 191         },
 192         # RUTV embed
 193         {
 194             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 195             'info_dict': {
 196                 'id': '776940',
 197                 'ext': 'mp4',
 198                 'title': 'Охотское море стало целиком российским',
 199                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 200             },
 201             'params': {
 202                 # m3u8 download
 203                 'skip_download': True,
 204             },
 205         },
 206         # Embedded TED video
 207         {
 208             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 209             'md5': '65fdff94098e4a607385a60c5177c638',
 210             'info_dict': {
 211                 'id': '1969',
 212                 'ext': 'mp4',
 213                 'title': 'Hidden miracles of the natural world',
 214                 'uploader': 'Louie Schwartzberg',
 215                 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
 216             }
 217         },
 218         # Embeded Ustream video
 219         {
 220             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 221             'md5': '27b99cdb639c9b12a79bca876a073417',
 222             'info_dict': {
 223                 'id': '45734260',
 224                 'ext': 'flv',
 225                 'uploader': 'AU SPA:  The NSA and Privacy',
 226                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 227             }
 228         },
 229         # nowvideo embed hidden behind percent encoding
 230         {
 231             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 232             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 233             'info_dict': {
 234                 'id': '06e53103ca9aa',
 235                 'ext': 'flv',
 236                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 237                 'description': 'No description',
 238             },
 239         },
 240         # arte embed
 241         {
 242             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 243             'md5': '7653032cbb25bf6c80d80f217055fa43',
 244             'info_dict': {
 245                 'id': '048195-004_PLUS7-F',
 246                 'ext': 'flv',
 247                 'title': 'X:enius',
 248                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 249                 'upload_date': '20140320',
 250             },
 251             'params': {
 252                 'skip_download': 'Requires rtmpdump'
 253             }
 254         },
 255         # Condé Nast embed
 256         {
 257             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 258             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 259             'info_dict': {
 260                 'id': '53501be369702d3275860000',
 261                 'ext': 'mp4',
 262                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 263             }
 264         },
 265         # Dailymotion embed
 266         {
 267             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 268             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 269             'info_dict': {
 270                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 271                 'ext': 'mp4',
 272                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 273                 'uploader': 'Spi0n',
 274             },
 275             'add_ie': ['Dailymotion'],
 276         },
 277         # YouTube embed
 278         {
 279             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 280             'info_dict': {
 281                 'id': 'FXRb4ykk4S0',
 282                 'ext': 'mp4',
 283                 'title': 'The NBL Auction 2014',
 284                 'uploader': 'BADMINTON England',
 285                 'uploader_id': 'BADMINTONEvents',
 286                 'upload_date': '20140603',
 287                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 288             },
 289             'add_ie': ['Youtube'],
 290             'params': {
 291                 'skip_download': True,
 292             }
 293         },
 294         # MTVSercices embed
 295         {
 296             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 297             'md5': '35727f82f58c76d996fc188f9755b0d5',
 298             'info_dict': {
 299                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 300                 'ext': 'mp4',
 301                 'title': 'Review',
 302                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 303             },
 304         },
 305         # YouTube embed via <data-embed-url="">
 306         {
 307             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 308             'info_dict': {
 309                 'id': '4vAffPZIT44',
 310                 'ext': 'mp4',
 311                 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
 312                 'uploader': 'Gameloft',
 313                 'uploader_id': 'gameloft',
 314                 'upload_date': '20140828',
 315                 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
 316             },
 317             'params': {
 318                 'skip_download': True,
 319             }
 320         },
 321         # Camtasia studio
 322         {
 323             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 324             'playlist': [{
 325                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 326                 'info_dict': {
 327                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 328                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 329                     'ext': 'flv',
 330                     'duration': 2235.90,
 331                 }
 332             }, {
 333                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 334                 'info_dict': {
 335                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 336                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 337                     'ext': 'flv',
 338                     'duration': 2235.93,
 339                 }
 340             }],
 341             'info_dict': {
 342                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 343             }
 344         },
 345         # Flowplayer
 346         {
 347             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 348             'md5': '9d65602bf31c6e20014319c7d07fba27',
 349             'info_dict': {
 350                 'id': '5123ea6d5e5a7',
 351                 'ext': 'mp4',
 352                 'age_limit': 18,
 353                 'uploader': 'www.handjobhub.com',
 354                 'title': 'Busty Blonde Siri Tit Fuck While Wank at HandjobHub.com',
 355             }
 356         },
 357         # RSS feed
 358         {
 359             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 360             'info_dict': {
 361                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 362                 'title': 'Zero Punctuation',
 363                 'description': 're:'
 364             },
 365             'playlist_mincount': 11,
 366         },
 367         # Multiple brightcove videos
 368         # https://github.com/rg3/youtube-dl/issues/2283
 369         {
 370             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 371             'info_dict': {
 372                 'id': 'always-never',
 373                 'title': 'Always / Never - The New Yorker',
 374             },
 375             'playlist_count': 3,
 376             'params': {
 377                 'extract_flat': False,
 378                 'skip_download': True,
 379             }
 380         },
 381         # MLB embed
 382         {
 383             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 384             'md5': '96f09a37e44da40dd083e12d9a683327',
 385             'info_dict': {
 386                 'id': '33322633',
 387                 'ext': 'mp4',
 388                 'title': 'Ump changes call to ball',
 389                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 390                 'duration': 48,
 391                 'timestamp': 1401537900,
 392                 'upload_date': '20140531',
 393                 'thumbnail': 're:^https?://.*\.jpg$',
 394             },
 395         },
 396         # Wistia embed
 397         {
 398             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 399             'md5': '8788b683c777a5cf25621eaf286d0c23',
 400             'info_dict': {
 401                 'id': '1cfaf6b7ea',
 402                 'ext': 'mov',
 403                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 404                 'duration': 643.0,
 405                 'filesize': 182808282,
 406                 'uploader': 'education-portal.com',
 407             },
 408         },
 409         {
 410             'url': 'http://thoughtworks.wistia.com/medias/uxjb0lwrcz',
 411             'md5': 'baf49c2baa8a7de5f3fc145a8506dcd4',
 412             'info_dict': {
 413                 'id': 'uxjb0lwrcz',
 414                 'ext': 'mp4',
 415                 'title': 'Conversation about Hexagonal Rails Part 1 - ThoughtWorks',
 416                 'duration': 1715.0,
 417                 'uploader': 'thoughtworks.wistia.com',
 418             },
 419         },
 420         # Direct download with broken HEAD
 421         {
 422             'url': 'http://ai-radio.org:8000/radio.opus',
 423             'info_dict': {
 424                 'id': 'radio',
 425                 'ext': 'opus',
 426                 'title': 'radio',
 427             },
 428             'params': {
 429                 'skip_download': True,  # infinite live stream
 430             },
 431             'expected_warnings': [
 432                 r'501.*Not Implemented'
 433             ],
 434         },
 435         # Soundcloud embed
 436         {
 437             'url': 'http://nakedsecurity.sophos.com/2014/10/29/sscc-171-are-you-sure-that-1234-is-a-bad-password-podcast/',
 438             'info_dict': {
 439                 'id': '174391317',
 440                 'ext': 'mp3',
 441                 'description': 'md5:ff867d6b555488ad3c52572bb33d432c',
 442                 'uploader': 'Sophos Security',
 443                 'title': 'Chet Chat 171 - Oct 29, 2014',
 444                 'upload_date': '20141029',
 445             }
 446         },
 447         # Livestream embed
 448         {
 449             'url': 'http://www.esa.int/Our_Activities/Space_Science/Rosetta/Philae_comet_touch-down_webcast',
 450             'info_dict': {
 451                 'id': '67864563',
 452                 'ext': 'flv',
 453                 'upload_date': '20141112',
 454                 'title': 'Rosetta #CometLanding webcast HL 10',
 455             }
 456         },
 457         # LazyYT
 458         {
 459             'url': 'http://discourse.ubuntu.com/t/unity-8-desktop-mode-windows-on-mir/1986',
 460             'info_dict': {
 461                 'title': 'Unity 8 desktop-mode windows on Mir! - Ubuntu Discourse',
 462             },
 463             'playlist_mincount': 2,
 464         },
 465         # Direct link with incorrect MIME type
 466         {
 467             'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
 468             'md5': '4ccbebe5f36706d85221f204d7eb5913',
 469             'info_dict': {
 470                 'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
 471                 'id': '5_Lennart_Poettering_-_Systemd',
 472                 'ext': 'webm',
 473                 'title': '5_Lennart_Poettering_-_Systemd',
 474                 'upload_date': '20141120',
 475             },
 476             'expected_warnings': [
 477                 'URL could be a direct video link, returning it as such.'
 478             ]
 479         },
 480         # Cinchcast embed
 481         {
 482             'url': 'http://undergroundwellness.com/podcasts/306-5-steps-to-permanent-gut-healing/',
 483             'info_dict': {
 484                 'id': '7141703',
 485                 'ext': 'mp3',
 486                 'upload_date': '20141126',
 487                 'title': 'Jack Tips: 5 Steps to Permanent Gut Healing',
 488             }
 489         },
 490     ]
 491
 492     def report_following_redirect(self, new_url):
 493         """Report information extraction."""
 494         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 495
 496     def _extract_rss(self, url, video_id, doc):
 497         playlist_title = doc.find('./channel/title').text
 498         playlist_desc_el = doc.find('./channel/description')
 499         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 500
 501         entries = [{
 502             '_type': 'url',
 503             'url': e.find('link').text,
 504             'title': e.find('title').text,
 505         } for e in doc.findall('./channel/item')]
 506
 507         return {
 508             '_type': 'playlist',
 509             'id': url,
 510             'title': playlist_title,
 511             'description': playlist_desc,
 512             'entries': entries,
 513         }
 514
 515     def _extract_camtasia(self, url, video_id, webpage):
 516         """ Returns None if no camtasia video can be found. """
 517
 518         camtasia_cfg = self._search_regex(
 519             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 520             webpage, 'camtasia configuration file', default=None)
 521         if camtasia_cfg is None:
 522             return None
 523
 524         title = self._html_search_meta('DC.title', webpage, fatal=True)
 525
 526         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 527         camtasia_cfg = self._download_xml(
 528             camtasia_url, video_id,
 529             note='Downloading camtasia configuration',
 530             errnote='Failed to download camtasia configuration')
 531         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 532
 533         entries = []
 534         for n in fileset_node.getchildren():
 535             url_n = n.find('./uri')
 536             if url_n is None:
 537                 continue
 538
 539             entries.append({
 540                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 541                 'title': '%s - %s' % (title, n.tag),
 542                 'url': compat_urlparse.urljoin(url, url_n.text),
 543                 'duration': float_or_none(n.find('./duration').text),
 544             })
 545
 546         return {
 547             '_type': 'playlist',
 548             'entries': entries,
 549             'title': title,
 550         }
 551
 552     def _real_extract(self, url):
 553         if url.startswith('//'):
 554             return {
 555                 '_type': 'url',
 556                 'url': self.http_scheme() + url,
 557             }
 558
 559         parsed_url = compat_urlparse.urlparse(url)
 560         if not parsed_url.scheme:
 561             default_search = self._downloader.params.get('default_search')
 562             if default_search is None:
 563                 default_search = 'fixup_error'
 564
 565             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 566                 if '/' in url:
 567                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 568                     return self.url_result('http://' + url)
 569                 elif default_search != 'fixup_error':
 570                     if default_search == 'auto_warning':
 571                         if re.match(r'^(?:url|URL)$', url):
 572                             raise ExtractorError(
 573                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 574                                 expected=True)
 575                         else:
 576                             self._downloader.report_warning(
 577                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 578                     return self.url_result('ytsearch:' + url)
 579
 580             if default_search in ('error', 'fixup_error'):
 581                 raise ExtractorError(
 582                     '%r is not a valid URL. '
 583                     'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 584                     % (url, url), expected=True)
 585             else:
 586                 if ':' not in default_search:
 587                     default_search += ':'
 588                 return self.url_result(default_search + url)
 589
 590         url, smuggled_data = unsmuggle_url(url)
 591         force_videoid = None
 592         is_intentional = smuggled_data and smuggled_data.get('to_generic')
 593         if smuggled_data and 'force_videoid' in smuggled_data:
 594             force_videoid = smuggled_data['force_videoid']
 595             video_id = force_videoid
 596         else:
 597             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 598
 599         self.to_screen('%s: Requesting header' % video_id)
 600
 601         head_req = HEADRequest(url)
 602         head_response = self._request_webpage(
 603             head_req, video_id,
 604             note=False, errnote='Could not send HEAD request to %s' % url,
 605             fatal=False)
 606
 607         if head_response is not False:
 608             # Check for redirect
 609             new_url = head_response.geturl()
 610             if url != new_url:
 611                 self.report_following_redirect(new_url)
 612                 if force_videoid:
 613                     new_url = smuggle_url(
 614                         new_url, {'force_videoid': force_videoid})
 615                 return self.url_result(new_url)
 616
 617         full_response = None
 618         if head_response is False:
 619             full_response = self._request_webpage(url, video_id)
 620             head_response = full_response
 621
 622         # Check for direct link to a video
 623         content_type = head_response.headers.get('Content-Type', '')
 624         m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 625         if m:
 626             upload_date = unified_strdate(
 627                 head_response.headers.get('Last-Modified'))
 628             return {
 629                 'id': video_id,
 630                 'title': os.path.splitext(url_basename(url))[0],
 631                 'direct': True,
 632                 'formats': [{
 633                     'format_id': m.group('format_id'),
 634                     'url': url,
 635                     'vcodec': 'none' if m.group('type') == 'audio' else None
 636                 }],
 637                 'upload_date': upload_date,
 638             }
 639
 640         if not self._downloader.params.get('test', False) and not is_intentional:
 641             self._downloader.report_warning('Falling back on generic information extractor.')
 642
 643         if not full_response:
 644             full_response = self._request_webpage(url, video_id)
 645
 646         # Maybe it's a direct link to a video?
 647         # Be careful not to download the whole thing!
 648         first_bytes = full_response.read(512)
 649         if not re.match(r'^\s*<', first_bytes.decode('utf-8', 'replace')):
 650             self._downloader.report_warning(
 651                 'URL could be a direct video link, returning it as such.')
 652             upload_date = unified_strdate(
 653                 head_response.headers.get('Last-Modified'))
 654             return {
 655                 'id': video_id,
 656                 'title': os.path.splitext(url_basename(url))[0],
 657                 'direct': True,
 658                 'url': url,
 659                 'upload_date': upload_date,
 660             }
 661
 662         webpage = self._webpage_read_content(
 663             full_response, url, video_id, prefix=first_bytes)
 664
 665         self.report_extraction(video_id)
 666
 667         # Is it an RSS feed?
 668         try:
 669             doc = parse_xml(webpage)
 670             if doc.tag == 'rss':
 671                 return self._extract_rss(url, video_id, doc)
 672         except compat_xml_parse_error:
 673             pass
 674
 675         # Is it a Camtasia project?
 676         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 677         if camtasia_res is not None:
 678             return camtasia_res
 679
 680         # Sometimes embedded video player is hidden behind percent encoding
 681         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 682         # Unescaping the whole page allows to handle those cases in a generic way
 683         webpage = compat_urllib_parse.unquote(webpage)
 684
 685         # it's tempting to parse this further, but you would
 686         # have to take into account all the variations like
 687         #   Video Title - Site Name
 688         #   Site Name | Video Title
 689         #   Video Title - Tagline | Site Name
 690         # and so on and so forth; it's just not practical
 691         video_title = self._html_search_regex(
 692             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 693             default='video')
 694
 695         # Try to detect age limit automatically
 696         age_limit = self._rta_search(webpage)
 697         # And then there are the jokers who advertise that they use RTA,
 698         # but actually don't.
 699         AGE_LIMIT_MARKERS = [
 700             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 701         ]
 702         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 703             age_limit = 18
 704
 705         # video uploader is domain name
 706         video_uploader = self._search_regex(
 707             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 708
 709         # Helper method
 710         def _playlist_from_matches(matches, getter=None, ie=None):
 711             urlrs = orderedSet(
 712                 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
 713                 for m in matches)
 714             return self.playlist_result(
 715                 urlrs, playlist_id=video_id, playlist_title=video_title)
 716
 717         # Look for BrightCove:
 718         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 719         if bc_urls:
 720             self.to_screen('Brightcove video detected.')
 721             entries = [{
 722                 '_type': 'url',
 723                 'url': smuggle_url(bc_url, {'Referer': url}),
 724                 'ie_key': 'Brightcove'
 725             } for bc_url in bc_urls]
 726
 727             return {
 728                 '_type': 'playlist',
 729                 'title': video_title,
 730                 'id': video_id,
 731                 'entries': entries,
 732             }
 733
 734         # Look for embedded (iframe) Vimeo player
 735         mobj = re.search(
 736             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 737         if mobj:
 738             player_url = unescapeHTML(mobj.group('url'))
 739             surl = smuggle_url(player_url, {'Referer': url})
 740             return self.url_result(surl)
 741
 742         # Look for embedded (swf embed) Vimeo player
 743         mobj = re.search(
 744             r'<embed[^>]+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 745         if mobj:
 746             return self.url_result(mobj.group(1))
 747
 748         # Look for embedded YouTube player
 749         matches = re.findall(r'''(?x)
 750             (?:
 751                 <iframe[^>]+?src=|
 752                 data-video-url=|
 753                 <embed[^>]+?src=|
 754                 embedSWF\(?:\s*|
 755                 new\s+SWFObject\(
 756             )
 757             (["\'])
 758                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 759                 (?:embed|v|p)/.+?)
 760             \1''', webpage)
 761         if matches:
 762             return _playlist_from_matches(
 763                 matches, lambda m: unescapeHTML(m[1]))
 764
 765         # Look for lazyYT YouTube embed
 766         matches = re.findall(
 767             r'class="lazyYT" data-youtube-id="([^"]+)"', webpage)
 768         if matches:
 769             return _playlist_from_matches(matches, lambda m: unescapeHTML(m))
 770
 771         # Look for embedded Dailymotion player
 772         matches = re.findall(
 773             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 774         if matches:
 775             return _playlist_from_matches(
 776                 matches, lambda m: unescapeHTML(m[1]))
 777
 778         # Look for embedded Dailymotion playlist player (#3822)
 779         m = re.search(
 780             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
 781         if m:
 782             playlists = re.findall(
 783                 r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
 784             if playlists:
 785                 return _playlist_from_matches(
 786                     playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
 787
 788         # Look for embedded Wistia player
 789         match = re.search(
 790             r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 791         if match:
 792             embed_url = self._proto_relative_url(
 793                 unescapeHTML(match.group('url')))
 794             return {
 795                 '_type': 'url_transparent',
 796                 'url': embed_url,
 797                 'ie_key': 'Wistia',
 798                 'uploader': video_uploader,
 799                 'title': video_title,
 800                 'id': video_id,
 801             }
 802
 803         match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 804         if match:
 805             return {
 806                 '_type': 'url_transparent',
 807                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 808                 'ie_key': 'Wistia',
 809                 'uploader': video_uploader,
 810                 'title': video_title,
 811                 'id': match.group('id')
 812             }
 813
 814         # Look for embedded blip.tv player
 815         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 816         if mobj:
 817             return self.url_result('http://blip.tv/a/a-' + mobj.group(1), 'BlipTV')
 818         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 819         if mobj:
 820             return self.url_result(mobj.group(1), 'BlipTV')
 821
 822         # Look for embedded condenast player
 823         matches = re.findall(
 824             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 825             webpage)
 826         if matches:
 827             return {
 828                 '_type': 'playlist',
 829                 'entries': [{
 830                     '_type': 'url',
 831                     'ie_key': 'CondeNast',
 832                     'url': ma,
 833                 } for ma in matches],
 834                 'title': video_title,
 835                 'id': video_id,
 836             }
 837
 838         # Look for Bandcamp pages with custom domain
 839         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 840         if mobj is not None:
 841             burl = unescapeHTML(mobj.group(1))
 842             # Don't set the extractor because it can be a track url or an album
 843             return self.url_result(burl)
 844
 845         # Look for embedded Vevo player
 846         mobj = re.search(
 847             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 848         if mobj is not None:
 849             return self.url_result(mobj.group('url'))
 850
 851         # Look for Ooyala videos
 852         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 853                 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 854         if mobj is not None:
 855             return OoyalaIE._build_url_result(mobj.group('ec'))
 856
 857         # Look for Aparat videos
 858         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 859         if mobj is not None:
 860             return self.url_result(mobj.group(1), 'Aparat')
 861
 862         # Look for MPORA videos
 863         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 864         if mobj is not None:
 865             return self.url_result(mobj.group(1), 'Mpora')
 866
 867         # Look for embedded NovaMov-based player
 868         mobj = re.search(
 869             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 870                     (?P<url>http://(?:(?:embed|www)\.)?
 871                         (?:novamov\.com|
 872                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 873                            videoweed\.(?:es|com)|
 874                            movshare\.(?:net|sx|ag)|
 875                            divxstage\.(?:eu|net|ch|co|at|ag))
 876                         /embed\.php.+?)\1''', webpage)
 877         if mobj is not None:
 878             return self.url_result(mobj.group('url'))
 879
 880         # Look for embedded Facebook player
 881         mobj = re.search(
 882             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 883         if mobj is not None:
 884             return self.url_result(mobj.group('url'), 'Facebook')
 885
 886         # Look for embedded VK player
 887         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 888         if mobj is not None:
 889             return self.url_result(mobj.group('url'), 'VK')
 890
 891         # Look for embedded ivi player
 892         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 893         if mobj is not None:
 894             return self.url_result(mobj.group('url'), 'Ivi')
 895
 896         # Look for embedded Huffington Post player
 897         mobj = re.search(
 898             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 899         if mobj is not None:
 900             return self.url_result(mobj.group('url'), 'HuffPost')
 901
 902         # Look for embed.ly
 903         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 904         if mobj is not None:
 905             return self.url_result(mobj.group('url'))
 906         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 907         if mobj is not None:
 908             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 909
 910         # Look for funnyordie embed
 911         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 912         if matches:
 913             return _playlist_from_matches(
 914                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 915
 916         # Look for BBC iPlayer embed
 917         matches = re.findall(r'setPlaylist\("(https?://www\.bbc\.co\.uk/iplayer/[^/]+/[\da-z]{8})"\)', webpage)
 918         if matches:
 919             return _playlist_from_matches(matches, ie='BBCCoUk')
 920
 921         # Look for embedded RUTV player
 922         rutv_url = RUTVIE._extract_url(webpage)
 923         if rutv_url:
 924             return self.url_result(rutv_url, 'RUTV')
 925
 926         # Look for embedded TED player
 927         mobj = re.search(
 928             r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
 929         if mobj is not None:
 930             return self.url_result(mobj.group('url'), 'TED')
 931
 932         # Look for embedded Ustream videos
 933         mobj = re.search(
 934             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 935         if mobj is not None:
 936             return self.url_result(mobj.group('url'), 'Ustream')
 937
 938         # Look for embedded arte.tv player
 939         mobj = re.search(
 940             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 941             webpage)
 942         if mobj is not None:
 943             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 944
 945         # Look for embedded smotri.com player
 946         smotri_url = SmotriIE._extract_url(webpage)
 947         if smotri_url:
 948             return self.url_result(smotri_url, 'Smotri')
 949
 950         # Look for embeded soundcloud player
 951         mobj = re.search(
 952             r'<iframe\s+(?:[a-zA-Z0-9_-]+="[^"]+"\s+)*src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 953             webpage)
 954         if mobj is not None:
 955             url = unescapeHTML(mobj.group('url'))
 956             return self.url_result(url)
 957
 958         # Look for embedded vulture.com player
 959         mobj = re.search(
 960             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 961             webpage)
 962         if mobj is not None:
 963             url = unescapeHTML(mobj.group('url'))
 964             return self.url_result(url, ie='Vulture')
 965
 966         # Look for embedded mtvservices player
 967         mobj = re.search(
 968             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 969             webpage)
 970         if mobj is not None:
 971             url = unescapeHTML(mobj.group('url'))
 972             return self.url_result(url, ie='MTVServicesEmbedded')
 973
 974         # Look for embedded yahoo player
 975         mobj = re.search(
 976             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 977             webpage)
 978         if mobj is not None:
 979             return self.url_result(mobj.group('url'), 'Yahoo')
 980
 981         # Look for embedded sbs.com.au player
 982         mobj = re.search(
 983             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 984             webpage)
 985         if mobj is not None:
 986             return self.url_result(mobj.group('url'), 'SBS')
 987
 988         # Look for embedded Cinchcast player
 989         mobj = re.search(
 990             r'<iframe[^>]+?src=(["\'])(?P<url>https?://player\.cinchcast\.com/.+?)\1',
 991             webpage)
 992         if mobj is not None:
 993             return self.url_result(mobj.group('url'), 'Cinchcast')
 994
 995         mobj = re.search(
 996             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m(?:lb)?\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
 997             webpage)
 998         if mobj is not None:
 999             return self.url_result(mobj.group('url'), 'MLB')
1000
1001         mobj = re.search(
1002             r'<iframe[^>]+?src=(["\'])(?P<url>%s)\1' % CondeNastIE.EMBED_URL,
1003             webpage)
1004         if mobj is not None:
1005             return self.url_result(self._proto_relative_url(mobj.group('url'), scheme='http:'), 'CondeNast')
1006
1007         mobj = re.search(
1008             r'<iframe[^>]+src="(?P<url>https?://new\.livestream\.com/[^"]+/player[^"]+)"',
1009             webpage)
1010         if mobj is not None:
1011             return self.url_result(mobj.group('url'), 'Livestream')
1012
1013         def check_video(vurl):
1014             vpath = compat_urlparse.urlparse(vurl).path
1015             vext = determine_ext(vpath)
1016             return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
1017
1018         def filter_video(urls):
1019             return list(filter(check_video, urls))
1020
1021         # Start with something easy: JW Player in SWFObject
1022         found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
1023         if not found:
1024             # Look for gorilla-vid style embedding
1025             found = filter_video(re.findall(r'''(?sx)
1026                 (?:
1027                     jw_plugins|
1028                     JWPlayerOptions|
1029                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
1030                 )
1031                 .*?file\s*:\s*["\'](.*?)["\']''', webpage))
1032         if not found:
1033             # Broaden the search a little bit
1034             found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
1035         if not found:
1036             # Broaden the findall a little bit: JWPlayer JS loader
1037             found = filter_video(re.findall(
1038                 r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
1039         if not found:
1040             # Flow player
1041             found = filter_video(re.findall(r'''(?xs)
1042                 flowplayer\("[^"]+",\s*
1043                     \{[^}]+?\}\s*,
1044                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
1045                         ["']?url["']?\s*:\s*["']([^"']+)["']
1046             ''', webpage))
1047         if not found:
1048             # Try to find twitter cards info
1049             found = filter_video(re.findall(
1050                 r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
1051         if not found:
1052             # We look for Open Graph info:
1053             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
1054             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
1055             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
1056             if m_video_type is not None:
1057                 found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
1058         if not found:
1059             # HTML5 video
1060             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]*)?\s+src=["\'](.*?)["\']', webpage)
1061         if not found:
1062             found = re.search(
1063                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
1064                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
1065                 webpage)
1066             if found:
1067                 new_url = found.group(1)
1068                 self.report_following_redirect(new_url)
1069                 return {
1070                     '_type': 'url',
1071                     'url': new_url,
1072                 }
1073         if not found:
1074             raise UnsupportedError(url)
1075
1076         entries = []
1077         for video_url in found:
1078             video_url = compat_urlparse.urljoin(url, video_url)
1079             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
1080
1081             # Sometimes, jwplayer extraction will result in a YouTube URL
1082             if YoutubeIE.suitable(video_url):
1083                 entries.append(self.url_result(video_url, 'Youtube'))
1084                 continue
1085
1086             # here's a fun little line of code for you:
1087             video_id = os.path.splitext(video_id)[0]
1088
1089             entries.append({
1090                 'id': video_id,
1091                 'url': video_url,
1092                 'uploader': video_uploader,
1093                 'title': video_title,
1094                 'age_limit': age_limit,
1095             })
1096
1097         if len(entries) == 1:
1098             return entries[0]
1099         else:
1100             for num, e in enumerate(entries, start=1):
1101                 e['title'] = '%s (%d)' % (e['title'], num)
1102             return {
1103                 '_type': 'playlist',
1104                 'entries': entries,
1105             }