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     is_html,
  21     orderedSet,
  22     parse_xml,
  23     smuggle_url,
  24     unescapeHTML,
  25     unified_strdate,
  26     unsmuggle_url,
  27     UnsupportedError,
  28     url_basename,
  29 )
  30 from .brightcove import BrightcoveIE
  31 from .ooyala import OoyalaIE
  32 from .rutv import RUTVIE
  33 from .smotri import SmotriIE
  34 from .condenast import CondeNastIE
  35
  36
  37 class GenericIE(InfoExtractor):
  38     IE_DESC = 'Generic downloader that works on some sites'
  39     _VALID_URL = r'.*'
  40     IE_NAME = 'generic'
  41     _TESTS = [
  42         {
  43             'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
  44             'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
  45             'info_dict': {
  46                 'id': '13601338388002',
  47                 'ext': 'mp4',
  48                 'uploader': 'www.hodiho.fr',
  49                 'title': 'R\u00e9gis plante sa Jeep',
  50             }
  51         },
  52         # bandcamp page with custom domain
  53         {
  54             'add_ie': ['Bandcamp'],
  55             'url': 'http://bronyrock.com/track/the-pony-mash',
  56             'info_dict': {
  57                 'id': '3235767654',
  58                 'ext': 'mp3',
  59                 'title': 'The Pony Mash',
  60                 'uploader': 'M_Pallante',
  61             },
  62             'skip': 'There is a limit of 200 free downloads / month for the test song',
  63         },
  64         # embedded brightcove video
  65         # it also tests brightcove videos that need to set the 'Referer' in the
  66         # http requests
  67         {
  68             'add_ie': ['Brightcove'],
  69             'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
  70             'info_dict': {
  71                 'id': '2765128793001',
  72                 'ext': 'mp4',
  73                 'title': 'Le cours de bourse : l’analyse technique',
  74                 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
  75                 'uploader': 'BFM BUSINESS',
  76             },
  77             'params': {
  78                 'skip_download': True,
  79             },
  80         },
  81         {
  82             # https://github.com/rg3/youtube-dl/issues/2253
  83             'url': 'http://bcove.me/i6nfkrc3',
  84             'md5': '0ba9446db037002366bab3b3eb30c88c',
  85             'info_dict': {
  86                 'id': '3101154703001',
  87                 'ext': 'mp4',
  88                 'title': 'Still no power',
  89                 'uploader': 'thestar.com',
  90                 '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.',
  91             },
  92             'add_ie': ['Brightcove'],
  93         },
  94         {
  95             'url': 'http://www.championat.com/video/football/v/87/87499.html',
  96             'md5': 'fb973ecf6e4a78a67453647444222983',
  97             'info_dict': {
  98                 'id': '3414141473001',
  99                 'ext': 'mp4',
 100                 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
 101                 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
 102                 'uploader': 'Championat',
 103             },
 104         },
 105         {
 106             # https://github.com/rg3/youtube-dl/issues/3541
 107             'add_ie': ['Brightcove'],
 108             'url': 'http://www.kijk.nl/sbs6/leermijvrouwenkennen/videos/jqMiXKAYan2S/aflevering-1',
 109             'info_dict': {
 110                 'id': '3866516442001',
 111                 'ext': 'mp4',
 112                 'title': 'Leer mij vrouwen kennen: Aflevering 1',
 113                 'description': 'Leer mij vrouwen kennen: Aflevering 1',
 114                 'uploader': 'SBS Broadcasting',
 115             },
 116             'skip': 'Restricted to Netherlands',
 117             'params': {
 118                 'skip_download': True,  # m3u8 download
 119             },
 120         },
 121         # Direct link to a video
 122         {
 123             'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
 124             'md5': '67d406c2bcb6af27fa886f31aa934bbe',
 125             'info_dict': {
 126                 'id': 'trailer',
 127                 'ext': 'mp4',
 128                 'title': 'trailer',
 129                 'upload_date': '20100513',
 130             }
 131         },
 132         # ooyala video
 133         {
 134             'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
 135             'md5': '166dd577b433b4d4ebfee10b0824d8ff',
 136             'info_dict': {
 137                 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
 138                 'ext': 'mp4',
 139                 'title': '2cc213299525360.mov',  # that's what we get
 140             },
 141             'add_ie': ['Ooyala'],
 142         },
 143         # google redirect
 144         {
 145             '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',
 146             'info_dict': {
 147                 'id': 'cmQHVoWB5FY',
 148                 'ext': 'mp4',
 149                 'upload_date': '20130224',
 150                 'uploader_id': 'TheVerge',
 151                 'description': 're:^Chris Ziegler takes a look at the\.*',
 152                 'uploader': 'The Verge',
 153                 'title': 'First Firefox OS phones side-by-side',
 154             },
 155             'params': {
 156                 'skip_download': False,
 157             }
 158         },
 159         # embed.ly video
 160         {
 161             'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
 162             'info_dict': {
 163                 'id': '9ODmcdjQcHQ',
 164                 'ext': 'mp4',
 165                 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
 166                 'upload_date': '20140225',
 167                 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
 168                 'uploader': 'Tested',
 169                 'uploader_id': 'testedcom',
 170             },
 171             # No need to test YoutubeIE here
 172             'params': {
 173                 'skip_download': True,
 174             },
 175         },
 176         # funnyordie embed
 177         {
 178             'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
 179             'info_dict': {
 180                 'id': '18e820ec3f',
 181                 'ext': 'mp4',
 182                 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
 183                 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
 184             },
 185         },
 186         # BBC iPlayer embeds
 187         {
 188             'url': 'http://www.bbc.co.uk/blogs/adamcurtis/posts/BUGGER',
 189             'info_dict': {
 190                 'title': 'BBC - Blogs -  Adam Curtis - BUGGER',
 191             },
 192             'playlist_mincount': 18,
 193         },
 194         # RUTV embed
 195         {
 196             'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
 197             'info_dict': {
 198                 'id': '776940',
 199                 'ext': 'mp4',
 200                 'title': 'Охотское море стало целиком российским',
 201                 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
 202             },
 203             'params': {
 204                 # m3u8 download
 205                 'skip_download': True,
 206             },
 207         },
 208         # Embedded TED video
 209         {
 210             'url': 'http://en.support.wordpress.com/videos/ted-talks/',
 211             'md5': '65fdff94098e4a607385a60c5177c638',
 212             'info_dict': {
 213                 'id': '1969',
 214                 'ext': 'mp4',
 215                 'title': 'Hidden miracles of the natural world',
 216                 'uploader': 'Louie Schwartzberg',
 217                 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
 218             }
 219         },
 220         # Embeded Ustream video
 221         {
 222             'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
 223             'md5': '27b99cdb639c9b12a79bca876a073417',
 224             'info_dict': {
 225                 'id': '45734260',
 226                 'ext': 'flv',
 227                 'uploader': 'AU SPA:  The NSA and Privacy',
 228                 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
 229             }
 230         },
 231         # nowvideo embed hidden behind percent encoding
 232         {
 233             'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
 234             'md5': '2baf4ddd70f697d94b1c18cf796d5107',
 235             'info_dict': {
 236                 'id': '06e53103ca9aa',
 237                 'ext': 'flv',
 238                 'title': 'Macross Episode 001  Watch Macross Episode 001 onl',
 239                 'description': 'No description',
 240             },
 241         },
 242         # arte embed
 243         {
 244             'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
 245             'md5': '7653032cbb25bf6c80d80f217055fa43',
 246             'info_dict': {
 247                 'id': '048195-004_PLUS7-F',
 248                 'ext': 'flv',
 249                 'title': 'X:enius',
 250                 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
 251                 'upload_date': '20140320',
 252             },
 253             'params': {
 254                 'skip_download': 'Requires rtmpdump'
 255             }
 256         },
 257         # Condé Nast embed
 258         {
 259             'url': 'http://www.wired.com/2014/04/honda-asimo/',
 260             'md5': 'ba0dfe966fa007657bd1443ee672db0f',
 261             'info_dict': {
 262                 'id': '53501be369702d3275860000',
 263                 'ext': 'mp4',
 264                 'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
 265             }
 266         },
 267         # Dailymotion embed
 268         {
 269             'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
 270             'md5': '441aeeb82eb72c422c7f14ec533999cd',
 271             'info_dict': {
 272                 'id': 'k2mm4bCdJ6CQ2i7c8o2',
 273                 'ext': 'mp4',
 274                 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
 275                 'uploader': 'Spi0n',
 276             },
 277             'add_ie': ['Dailymotion'],
 278         },
 279         # YouTube embed
 280         {
 281             'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
 282             'info_dict': {
 283                 'id': 'FXRb4ykk4S0',
 284                 'ext': 'mp4',
 285                 'title': 'The NBL Auction 2014',
 286                 'uploader': 'BADMINTON England',
 287                 'uploader_id': 'BADMINTONEvents',
 288                 'upload_date': '20140603',
 289                 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
 290             },
 291             'add_ie': ['Youtube'],
 292             'params': {
 293                 'skip_download': True,
 294             }
 295         },
 296         # MTVSercices embed
 297         {
 298             'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
 299             'md5': '35727f82f58c76d996fc188f9755b0d5',
 300             'info_dict': {
 301                 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
 302                 'ext': 'mp4',
 303                 'title': 'Review',
 304                 'description': 'Mario\'s life in the fast lane has never looked so good.',
 305             },
 306         },
 307         # YouTube embed via <data-embed-url="">
 308         {
 309             'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
 310             'info_dict': {
 311                 'id': '4vAffPZIT44',
 312                 'ext': 'mp4',
 313                 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
 314                 'uploader': 'Gameloft',
 315                 'uploader_id': 'gameloft',
 316                 'upload_date': '20140828',
 317                 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
 318             },
 319             'params': {
 320                 'skip_download': True,
 321             }
 322         },
 323         # Camtasia studio
 324         {
 325             'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
 326             'playlist': [{
 327                 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
 328                 'info_dict': {
 329                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 330                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
 331                     'ext': 'flv',
 332                     'duration': 2235.90,
 333                 }
 334             }, {
 335                 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
 336                 'info_dict': {
 337                     'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
 338                     'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
 339                     'ext': 'flv',
 340                     'duration': 2235.93,
 341                 }
 342             }],
 343             'info_dict': {
 344                 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
 345             }
 346         },
 347         # Flowplayer
 348         {
 349             'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
 350             'md5': '9d65602bf31c6e20014319c7d07fba27',
 351             'info_dict': {
 352                 'id': '5123ea6d5e5a7',
 353                 'ext': 'mp4',
 354                 'age_limit': 18,
 355                 'uploader': 'www.handjobhub.com',
 356                 'title': 'Busty Blonde Siri Tit Fuck While Wank at HandjobHub.com',
 357             }
 358         },
 359         # RSS feed
 360         {
 361             'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 362             'info_dict': {
 363                 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
 364                 'title': 'Zero Punctuation',
 365                 'description': 're:'
 366             },
 367             'playlist_mincount': 11,
 368         },
 369         # Multiple brightcove videos
 370         # https://github.com/rg3/youtube-dl/issues/2283
 371         {
 372             'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
 373             'info_dict': {
 374                 'id': 'always-never',
 375                 'title': 'Always / Never - The New Yorker',
 376             },
 377             'playlist_count': 3,
 378             'params': {
 379                 'extract_flat': False,
 380                 'skip_download': True,
 381             }
 382         },
 383         # MLB embed
 384         {
 385             'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
 386             'md5': '96f09a37e44da40dd083e12d9a683327',
 387             'info_dict': {
 388                 'id': '33322633',
 389                 'ext': 'mp4',
 390                 'title': 'Ump changes call to ball',
 391                 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
 392                 'duration': 48,
 393                 'timestamp': 1401537900,
 394                 'upload_date': '20140531',
 395                 'thumbnail': 're:^https?://.*\.jpg$',
 396             },
 397         },
 398         # Wistia embed
 399         {
 400             'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
 401             'md5': '8788b683c777a5cf25621eaf286d0c23',
 402             'info_dict': {
 403                 'id': '1cfaf6b7ea',
 404                 'ext': 'mov',
 405                 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
 406                 'duration': 643.0,
 407                 'filesize': 182808282,
 408                 'uploader': 'education-portal.com',
 409             },
 410         },
 411         {
 412             'url': 'http://thoughtworks.wistia.com/medias/uxjb0lwrcz',
 413             'md5': 'baf49c2baa8a7de5f3fc145a8506dcd4',
 414             'info_dict': {
 415                 'id': 'uxjb0lwrcz',
 416                 'ext': 'mp4',
 417                 'title': 'Conversation about Hexagonal Rails Part 1 - ThoughtWorks',
 418                 'duration': 1715.0,
 419                 'uploader': 'thoughtworks.wistia.com',
 420             },
 421         },
 422         # Direct download with broken HEAD
 423         {
 424             'url': 'http://ai-radio.org:8000/radio.opus',
 425             'info_dict': {
 426                 'id': 'radio',
 427                 'ext': 'opus',
 428                 'title': 'radio',
 429             },
 430             'params': {
 431                 'skip_download': True,  # infinite live stream
 432             },
 433             'expected_warnings': [
 434                 r'501.*Not Implemented'
 435             ],
 436         },
 437         # Soundcloud embed
 438         {
 439             'url': 'http://nakedsecurity.sophos.com/2014/10/29/sscc-171-are-you-sure-that-1234-is-a-bad-password-podcast/',
 440             'info_dict': {
 441                 'id': '174391317',
 442                 'ext': 'mp3',
 443                 'description': 'md5:ff867d6b555488ad3c52572bb33d432c',
 444                 'uploader': 'Sophos Security',
 445                 'title': 'Chet Chat 171 - Oct 29, 2014',
 446                 'upload_date': '20141029',
 447             }
 448         },
 449         # Livestream embed
 450         {
 451             'url': 'http://www.esa.int/Our_Activities/Space_Science/Rosetta/Philae_comet_touch-down_webcast',
 452             'info_dict': {
 453                 'id': '67864563',
 454                 'ext': 'flv',
 455                 'upload_date': '20141112',
 456                 'title': 'Rosetta #CometLanding webcast HL 10',
 457             }
 458         },
 459         # LazyYT
 460         {
 461             'url': 'http://discourse.ubuntu.com/t/unity-8-desktop-mode-windows-on-mir/1986',
 462             'info_dict': {
 463                 'title': 'Unity 8 desktop-mode windows on Mir! - Ubuntu Discourse',
 464             },
 465             'playlist_mincount': 2,
 466         },
 467         # Direct link with incorrect MIME type
 468         {
 469             'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
 470             'md5': '4ccbebe5f36706d85221f204d7eb5913',
 471             'info_dict': {
 472                 'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
 473                 'id': '5_Lennart_Poettering_-_Systemd',
 474                 'ext': 'webm',
 475                 'title': '5_Lennart_Poettering_-_Systemd',
 476                 'upload_date': '20141120',
 477             },
 478             'expected_warnings': [
 479                 'URL could be a direct video link, returning it as such.'
 480             ]
 481         },
 482         # Cinchcast embed
 483         {
 484             'url': 'http://undergroundwellness.com/podcasts/306-5-steps-to-permanent-gut-healing/',
 485             'info_dict': {
 486                 'id': '7141703',
 487                 'ext': 'mp3',
 488                 'upload_date': '20141126',
 489                 'title': 'Jack Tips: 5 Steps to Permanent Gut Healing',
 490             }
 491         },
 492         # Cinerama player
 493         {
 494             'url': 'http://www.abc.net.au/7.30/content/2015/s4164797.htm',
 495             'info_dict': {
 496                 'id': '730m_DandD_1901_512k',
 497                 'ext': 'mp4',
 498                 'uploader': 'www.abc.net.au',
 499                 'title': 'Game of Thrones with dice - Dungeons and Dragons fantasy role-playing game gets new life - 19/01/2015',
 500             }
 501         }
 502     ]
 503
 504     def report_following_redirect(self, new_url):
 505         """Report information extraction."""
 506         self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 507
 508     def _extract_rss(self, url, video_id, doc):
 509         playlist_title = doc.find('./channel/title').text
 510         playlist_desc_el = doc.find('./channel/description')
 511         playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
 512
 513         entries = [{
 514             '_type': 'url',
 515             'url': e.find('link').text,
 516             'title': e.find('title').text,
 517         } for e in doc.findall('./channel/item')]
 518
 519         return {
 520             '_type': 'playlist',
 521             'id': url,
 522             'title': playlist_title,
 523             'description': playlist_desc,
 524             'entries': entries,
 525         }
 526
 527     def _extract_camtasia(self, url, video_id, webpage):
 528         """ Returns None if no camtasia video can be found. """
 529
 530         camtasia_cfg = self._search_regex(
 531             r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
 532             webpage, 'camtasia configuration file', default=None)
 533         if camtasia_cfg is None:
 534             return None
 535
 536         title = self._html_search_meta('DC.title', webpage, fatal=True)
 537
 538         camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
 539         camtasia_cfg = self._download_xml(
 540             camtasia_url, video_id,
 541             note='Downloading camtasia configuration',
 542             errnote='Failed to download camtasia configuration')
 543         fileset_node = camtasia_cfg.find('./playlist/array/fileset')
 544
 545         entries = []
 546         for n in fileset_node.getchildren():
 547             url_n = n.find('./uri')
 548             if url_n is None:
 549                 continue
 550
 551             entries.append({
 552                 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
 553                 'title': '%s - %s' % (title, n.tag),
 554                 'url': compat_urlparse.urljoin(url, url_n.text),
 555                 'duration': float_or_none(n.find('./duration').text),
 556             })
 557
 558         return {
 559             '_type': 'playlist',
 560             'entries': entries,
 561             'title': title,
 562         }
 563
 564     def _real_extract(self, url):
 565         if url.startswith('//'):
 566             return {
 567                 '_type': 'url',
 568                 'url': self.http_scheme() + url,
 569             }
 570
 571         parsed_url = compat_urlparse.urlparse(url)
 572         if not parsed_url.scheme:
 573             default_search = self._downloader.params.get('default_search')
 574             if default_search is None:
 575                 default_search = 'fixup_error'
 576
 577             if default_search in ('auto', 'auto_warning', 'fixup_error'):
 578                 if '/' in url:
 579                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
 580                     return self.url_result('http://' + url)
 581                 elif default_search != 'fixup_error':
 582                     if default_search == 'auto_warning':
 583                         if re.match(r'^(?:url|URL)$', url):
 584                             raise ExtractorError(
 585                                 'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
 586                                 expected=True)
 587                         else:
 588                             self._downloader.report_warning(
 589                                 'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
 590                     return self.url_result('ytsearch:' + url)
 591
 592             if default_search in ('error', 'fixup_error'):
 593                 raise ExtractorError(
 594                     '%r is not a valid URL. '
 595                     'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
 596                     % (url, url), expected=True)
 597             else:
 598                 if ':' not in default_search:
 599                     default_search += ':'
 600                 return self.url_result(default_search + url)
 601
 602         url, smuggled_data = unsmuggle_url(url)
 603         force_videoid = None
 604         is_intentional = smuggled_data and smuggled_data.get('to_generic')
 605         if smuggled_data and 'force_videoid' in smuggled_data:
 606             force_videoid = smuggled_data['force_videoid']
 607             video_id = force_videoid
 608         else:
 609             video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
 610
 611         self.to_screen('%s: Requesting header' % video_id)
 612
 613         head_req = HEADRequest(url)
 614         head_response = self._request_webpage(
 615             head_req, video_id,
 616             note=False, errnote='Could not send HEAD request to %s' % url,
 617             fatal=False)
 618
 619         if head_response is not False:
 620             # Check for redirect
 621             new_url = head_response.geturl()
 622             if url != new_url:
 623                 self.report_following_redirect(new_url)
 624                 if force_videoid:
 625                     new_url = smuggle_url(
 626                         new_url, {'force_videoid': force_videoid})
 627                 return self.url_result(new_url)
 628
 629         full_response = None
 630         if head_response is False:
 631             full_response = self._request_webpage(url, video_id)
 632             head_response = full_response
 633
 634         # Check for direct link to a video
 635         content_type = head_response.headers.get('Content-Type', '')
 636         m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
 637         if m:
 638             upload_date = unified_strdate(
 639                 head_response.headers.get('Last-Modified'))
 640             return {
 641                 'id': video_id,
 642                 'title': os.path.splitext(url_basename(url))[0],
 643                 'direct': True,
 644                 'formats': [{
 645                     'format_id': m.group('format_id'),
 646                     'url': url,
 647                     'vcodec': 'none' if m.group('type') == 'audio' else None
 648                 }],
 649                 'upload_date': upload_date,
 650             }
 651
 652         if not self._downloader.params.get('test', False) and not is_intentional:
 653             self._downloader.report_warning('Falling back on generic information extractor.')
 654
 655         if not full_response:
 656             full_response = self._request_webpage(url, video_id)
 657
 658         # Maybe it's a direct link to a video?
 659         # Be careful not to download the whole thing!
 660         first_bytes = full_response.read(512)
 661         if not is_html(first_bytes):
 662             self._downloader.report_warning(
 663                 'URL could be a direct video link, returning it as such.')
 664             upload_date = unified_strdate(
 665                 head_response.headers.get('Last-Modified'))
 666             return {
 667                 'id': video_id,
 668                 'title': os.path.splitext(url_basename(url))[0],
 669                 'direct': True,
 670                 'url': url,
 671                 'upload_date': upload_date,
 672             }
 673
 674         webpage = self._webpage_read_content(
 675             full_response, url, video_id, prefix=first_bytes)
 676
 677         self.report_extraction(video_id)
 678
 679         # Is it an RSS feed?
 680         try:
 681             doc = parse_xml(webpage)
 682             if doc.tag == 'rss':
 683                 return self._extract_rss(url, video_id, doc)
 684         except compat_xml_parse_error:
 685             pass
 686
 687         # Is it a Camtasia project?
 688         camtasia_res = self._extract_camtasia(url, video_id, webpage)
 689         if camtasia_res is not None:
 690             return camtasia_res
 691
 692         # Sometimes embedded video player is hidden behind percent encoding
 693         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
 694         # Unescaping the whole page allows to handle those cases in a generic way
 695         webpage = compat_urllib_parse.unquote(webpage)
 696
 697         # it's tempting to parse this further, but you would
 698         # have to take into account all the variations like
 699         #   Video Title - Site Name
 700         #   Site Name | Video Title
 701         #   Video Title - Tagline | Site Name
 702         # and so on and so forth; it's just not practical
 703         video_title = self._html_search_regex(
 704             r'(?s)<title>(.*?)</title>', webpage, 'video title',
 705             default='video')
 706
 707         # Try to detect age limit automatically
 708         age_limit = self._rta_search(webpage)
 709         # And then there are the jokers who advertise that they use RTA,
 710         # but actually don't.
 711         AGE_LIMIT_MARKERS = [
 712             r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
 713         ]
 714         if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
 715             age_limit = 18
 716
 717         # video uploader is domain name
 718         video_uploader = self._search_regex(
 719             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 720
 721         # Helper method
 722         def _playlist_from_matches(matches, getter=None, ie=None):
 723             urlrs = orderedSet(
 724                 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
 725                 for m in matches)
 726             return self.playlist_result(
 727                 urlrs, playlist_id=video_id, playlist_title=video_title)
 728
 729         # Look for BrightCove:
 730         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
 731         if bc_urls:
 732             self.to_screen('Brightcove video detected.')
 733             entries = [{
 734                 '_type': 'url',
 735                 'url': smuggle_url(bc_url, {'Referer': url}),
 736                 'ie_key': 'Brightcove'
 737             } for bc_url in bc_urls]
 738
 739             return {
 740                 '_type': 'playlist',
 741                 'title': video_title,
 742                 'id': video_id,
 743                 'entries': entries,
 744             }
 745
 746         # Look for embedded (iframe) Vimeo player
 747         mobj = re.search(
 748             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
 749         if mobj:
 750             player_url = unescapeHTML(mobj.group('url'))
 751             surl = smuggle_url(player_url, {'Referer': url})
 752             return self.url_result(surl)
 753
 754         # Look for embedded (swf embed) Vimeo player
 755         mobj = re.search(
 756             r'<embed[^>]+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
 757         if mobj:
 758             return self.url_result(mobj.group(1))
 759
 760         # Look for embedded YouTube player
 761         matches = re.findall(r'''(?x)
 762             (?:
 763                 <iframe[^>]+?src=|
 764                 data-video-url=|
 765                 <embed[^>]+?src=|
 766                 embedSWF\(?:\s*|
 767                 new\s+SWFObject\(
 768             )
 769             (["\'])
 770                 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
 771                 (?:embed|v|p)/.+?)
 772             \1''', webpage)
 773         if matches:
 774             return _playlist_from_matches(
 775                 matches, lambda m: unescapeHTML(m[1]))
 776
 777         # Look for lazyYT YouTube embed
 778         matches = re.findall(
 779             r'class="lazyYT" data-youtube-id="([^"]+)"', webpage)
 780         if matches:
 781             return _playlist_from_matches(matches, lambda m: unescapeHTML(m))
 782
 783         # Look for embedded Dailymotion player
 784         matches = re.findall(
 785             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
 786         if matches:
 787             return _playlist_from_matches(
 788                 matches, lambda m: unescapeHTML(m[1]))
 789
 790         # Look for embedded Dailymotion playlist player (#3822)
 791         m = re.search(
 792             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
 793         if m:
 794             playlists = re.findall(
 795                 r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
 796             if playlists:
 797                 return _playlist_from_matches(
 798                     playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
 799
 800         # Look for embedded Wistia player
 801         match = re.search(
 802             r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
 803         if match:
 804             embed_url = self._proto_relative_url(
 805                 unescapeHTML(match.group('url')))
 806             return {
 807                 '_type': 'url_transparent',
 808                 'url': embed_url,
 809                 'ie_key': 'Wistia',
 810                 'uploader': video_uploader,
 811                 'title': video_title,
 812                 'id': video_id,
 813             }
 814
 815         match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
 816         if match:
 817             return {
 818                 '_type': 'url_transparent',
 819                 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
 820                 'ie_key': 'Wistia',
 821                 'uploader': video_uploader,
 822                 'title': video_title,
 823                 'id': match.group('id')
 824             }
 825
 826         # Look for embedded blip.tv player
 827         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
 828         if mobj:
 829             return self.url_result('http://blip.tv/a/a-' + mobj.group(1), 'BlipTV')
 830         mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
 831         if mobj:
 832             return self.url_result(mobj.group(1), 'BlipTV')
 833
 834         # Look for embedded condenast player
 835         matches = re.findall(
 836             r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
 837             webpage)
 838         if matches:
 839             return {
 840                 '_type': 'playlist',
 841                 'entries': [{
 842                     '_type': 'url',
 843                     'ie_key': 'CondeNast',
 844                     'url': ma,
 845                 } for ma in matches],
 846                 'title': video_title,
 847                 'id': video_id,
 848             }
 849
 850         # Look for Bandcamp pages with custom domain
 851         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
 852         if mobj is not None:
 853             burl = unescapeHTML(mobj.group(1))
 854             # Don't set the extractor because it can be a track url or an album
 855             return self.url_result(burl)
 856
 857         # Look for embedded Vevo player
 858         mobj = re.search(
 859             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
 860         if mobj is not None:
 861             return self.url_result(mobj.group('url'))
 862
 863         # Look for Ooyala videos
 864         mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
 865                 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
 866         if mobj is not None:
 867             return OoyalaIE._build_url_result(mobj.group('ec'))
 868
 869         # Look for Aparat videos
 870         mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
 871         if mobj is not None:
 872             return self.url_result(mobj.group(1), 'Aparat')
 873
 874         # Look for MPORA videos
 875         mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
 876         if mobj is not None:
 877             return self.url_result(mobj.group(1), 'Mpora')
 878
 879         # Look for embedded NovaMov-based player
 880         mobj = re.search(
 881             r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
 882                     (?P<url>http://(?:(?:embed|www)\.)?
 883                         (?:novamov\.com|
 884                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
 885                            videoweed\.(?:es|com)|
 886                            movshare\.(?:net|sx|ag)|
 887                            divxstage\.(?:eu|net|ch|co|at|ag))
 888                         /embed\.php.+?)\1''', webpage)
 889         if mobj is not None:
 890             return self.url_result(mobj.group('url'))
 891
 892         # Look for embedded Facebook player
 893         mobj = re.search(
 894             r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
 895         if mobj is not None:
 896             return self.url_result(mobj.group('url'), 'Facebook')
 897
 898         # Look for embedded VK player
 899         mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
 900         if mobj is not None:
 901             return self.url_result(mobj.group('url'), 'VK')
 902
 903         # Look for embedded ivi player
 904         mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
 905         if mobj is not None:
 906             return self.url_result(mobj.group('url'), 'Ivi')
 907
 908         # Look for embedded Huffington Post player
 909         mobj = re.search(
 910             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
 911         if mobj is not None:
 912             return self.url_result(mobj.group('url'), 'HuffPost')
 913
 914         # Look for embed.ly
 915         mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
 916         if mobj is not None:
 917             return self.url_result(mobj.group('url'))
 918         mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
 919         if mobj is not None:
 920             return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
 921
 922         # Look for funnyordie embed
 923         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
 924         if matches:
 925             return _playlist_from_matches(
 926                 matches, getter=unescapeHTML, ie='FunnyOrDie')
 927
 928         # Look for BBC iPlayer embed
 929         matches = re.findall(r'setPlaylist\("(https?://www\.bbc\.co\.uk/iplayer/[^/]+/[\da-z]{8})"\)', webpage)
 930         if matches:
 931             return _playlist_from_matches(matches, ie='BBCCoUk')
 932
 933         # Look for embedded RUTV player
 934         rutv_url = RUTVIE._extract_url(webpage)
 935         if rutv_url:
 936             return self.url_result(rutv_url, 'RUTV')
 937
 938         # Look for embedded TED player
 939         mobj = re.search(
 940             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed(?:-ssl)?\.ted\.com/.+?)\1', webpage)
 941         if mobj is not None:
 942             return self.url_result(mobj.group('url'), 'TED')
 943
 944         # Look for embedded Ustream videos
 945         mobj = re.search(
 946             r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
 947         if mobj is not None:
 948             return self.url_result(mobj.group('url'), 'Ustream')
 949
 950         # Look for embedded arte.tv player
 951         mobj = re.search(
 952             r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
 953             webpage)
 954         if mobj is not None:
 955             return self.url_result(mobj.group('url'), 'ArteTVEmbed')
 956
 957         # Look for embedded smotri.com player
 958         smotri_url = SmotriIE._extract_url(webpage)
 959         if smotri_url:
 960             return self.url_result(smotri_url, 'Smotri')
 961
 962         # Look for embeded soundcloud player
 963         mobj = re.search(
 964             r'<iframe\s+(?:[a-zA-Z0-9_-]+="[^"]+"\s+)*src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
 965             webpage)
 966         if mobj is not None:
 967             url = unescapeHTML(mobj.group('url'))
 968             return self.url_result(url)
 969
 970         # Look for embedded vulture.com player
 971         mobj = re.search(
 972             r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
 973             webpage)
 974         if mobj is not None:
 975             url = unescapeHTML(mobj.group('url'))
 976             return self.url_result(url, ie='Vulture')
 977
 978         # Look for embedded mtvservices player
 979         mobj = re.search(
 980             r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
 981             webpage)
 982         if mobj is not None:
 983             url = unescapeHTML(mobj.group('url'))
 984             return self.url_result(url, ie='MTVServicesEmbedded')
 985
 986         # Look for embedded yahoo player
 987         mobj = re.search(
 988             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
 989             webpage)
 990         if mobj is not None:
 991             return self.url_result(mobj.group('url'), 'Yahoo')
 992
 993         # Look for embedded sbs.com.au player
 994         mobj = re.search(
 995             r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
 996             webpage)
 997         if mobj is not None:
 998             return self.url_result(mobj.group('url'), 'SBS')
 999
1000         # Look for embedded Cinchcast player
1001         mobj = re.search(
1002             r'<iframe[^>]+?src=(["\'])(?P<url>https?://player\.cinchcast\.com/.+?)\1',
1003             webpage)
1004         if mobj is not None:
1005             return self.url_result(mobj.group('url'), 'Cinchcast')
1006
1007         mobj = re.search(
1008             r'<iframe[^>]+?src=(["\'])(?P<url>https?://m(?:lb)?\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
1009             webpage)
1010         if mobj is not None:
1011             return self.url_result(mobj.group('url'), 'MLB')
1012
1013         mobj = re.search(
1014             r'<iframe[^>]+?src=(["\'])(?P<url>%s)\1' % CondeNastIE.EMBED_URL,
1015             webpage)
1016         if mobj is not None:
1017             return self.url_result(self._proto_relative_url(mobj.group('url'), scheme='http:'), 'CondeNast')
1018
1019         mobj = re.search(
1020             r'<iframe[^>]+src="(?P<url>https?://new\.livestream\.com/[^"]+/player[^"]+)"',
1021             webpage)
1022         if mobj is not None:
1023             return self.url_result(mobj.group('url'), 'Livestream')
1024
1025         def check_video(vurl):
1026             vpath = compat_urlparse.urlparse(vurl).path
1027             vext = determine_ext(vpath)
1028             return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
1029
1030         def filter_video(urls):
1031             return list(filter(check_video, urls))
1032
1033         # Start with something easy: JW Player in SWFObject
1034         found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
1035         if not found:
1036             # Look for gorilla-vid style embedding
1037             found = filter_video(re.findall(r'''(?sx)
1038                 (?:
1039                     jw_plugins|
1040                     JWPlayerOptions|
1041                     jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
1042                 )
1043                 .*?file\s*:\s*["\'](.*?)["\']''', webpage))
1044         if not found:
1045             # Broaden the search a little bit
1046             found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
1047         if not found:
1048             # Broaden the findall a little bit: JWPlayer JS loader
1049             found = filter_video(re.findall(
1050                 r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
1051         if not found:
1052             # Flow player
1053             found = filter_video(re.findall(r'''(?xs)
1054                 flowplayer\("[^"]+",\s*
1055                     \{[^}]+?\}\s*,
1056                     \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
1057                         ["']?url["']?\s*:\s*["']([^"']+)["']
1058             ''', webpage))
1059         if not found:
1060             # Cinerama player
1061             found = re.findall(
1062                 r"cinerama\.embedPlayer\(\s*\'[^']+\',\s*'([^']+)'", webpage)
1063         if not found:
1064             # Try to find twitter cards info
1065             found = filter_video(re.findall(
1066                 r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
1067         if not found:
1068             # We look for Open Graph info:
1069             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
1070             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
1071             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
1072             if m_video_type is not None:
1073                 found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
1074         if not found:
1075             # HTML5 video
1076             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]*)?\s+src=["\'](.*?)["\']', webpage)
1077         if not found:
1078             found = re.search(
1079                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
1080                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
1081                 webpage)
1082             if found:
1083                 new_url = found.group(1)
1084                 self.report_following_redirect(new_url)
1085                 return {
1086                     '_type': 'url',
1087                     'url': new_url,
1088                 }
1089         if not found:
1090             raise UnsupportedError(url)
1091
1092         entries = []
1093         for video_url in found:
1094             video_url = compat_urlparse.urljoin(url, video_url)
1095             video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
1096
1097             # Sometimes, jwplayer extraction will result in a YouTube URL
1098             if YoutubeIE.suitable(video_url):
1099                 entries.append(self.url_result(video_url, 'Youtube'))
1100                 continue
1101
1102             # here's a fun little line of code for you:
1103             video_id = os.path.splitext(video_id)[0]
1104
1105             entries.append({
1106                 'id': video_id,
1107                 'url': video_url,
1108                 'uploader': video_uploader,
1109                 'title': video_title,
1110                 'age_limit': age_limit,
1111             })
1112
1113         if len(entries) == 1:
1114             return entries[0]
1115         else:
1116             for num, e in enumerate(entries, start=1):
1117                 e['title'] = '%s (%d)' % (e['title'], num)
1118             return {
1119                 '_type': 'playlist',
1120                 'entries': entries,
1121             }