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