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