]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/generic.py
7a5bf939237ff45731fd3befca5ad0b7dfc0df1f
[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 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': '166dd577b433b4d4ebfee10b0824d8ff',
135 'info_dict': {
136 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
137 'ext': 'mp4',
138 'title': '2cc213299525360.mov', # that's what we get
139 },
140 'add_ie': ['Ooyala'],
141 },
142 # google redirect
143 {
144 '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',
145 'info_dict': {
146 'id': 'cmQHVoWB5FY',
147 'ext': 'mp4',
148 'upload_date': '20130224',
149 'uploader_id': 'TheVerge',
150 'description': 're:^Chris Ziegler takes a look at the\.*',
151 'uploader': 'The Verge',
152 'title': 'First Firefox OS phones side-by-side',
153 },
154 'params': {
155 'skip_download': False,
156 }
157 },
158 # embed.ly video
159 {
160 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
161 'info_dict': {
162 'id': '9ODmcdjQcHQ',
163 'ext': 'mp4',
164 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
165 'upload_date': '20140225',
166 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
167 'uploader': 'Tested',
168 'uploader_id': 'testedcom',
169 },
170 # No need to test YoutubeIE here
171 'params': {
172 'skip_download': True,
173 },
174 },
175 # funnyordie embed
176 {
177 'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
178 'info_dict': {
179 'id': '18e820ec3f',
180 'ext': 'mp4',
181 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
182 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
183 },
184 },
185 # BBC iPlayer embeds
186 {
187 'url': 'http://www.bbc.co.uk/blogs/adamcurtis/posts/BUGGER',
188 'info_dict': {
189 'title': 'BBC - Blogs - Adam Curtis - BUGGER',
190 },
191 'playlist_mincount': 18,
192 },
193 # RUTV embed
194 {
195 'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
196 'info_dict': {
197 'id': '776940',
198 'ext': 'mp4',
199 'title': 'Охотское море стало целиком российским',
200 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
201 },
202 'params': {
203 # m3u8 download
204 'skip_download': True,
205 },
206 },
207 # Embedded TED video
208 {
209 'url': 'http://en.support.wordpress.com/videos/ted-talks/',
210 'md5': '65fdff94098e4a607385a60c5177c638',
211 'info_dict': {
212 'id': '1969',
213 'ext': 'mp4',
214 'title': 'Hidden miracles of the natural world',
215 'uploader': 'Louie Schwartzberg',
216 'description': 'md5:8145d19d320ff3e52f28401f4c4283b9',
217 }
218 },
219 # Embeded Ustream video
220 {
221 'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
222 'md5': '27b99cdb639c9b12a79bca876a073417',
223 'info_dict': {
224 'id': '45734260',
225 'ext': 'flv',
226 'uploader': 'AU SPA: The NSA and Privacy',
227 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
228 }
229 },
230 # nowvideo embed hidden behind percent encoding
231 {
232 'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
233 'md5': '2baf4ddd70f697d94b1c18cf796d5107',
234 'info_dict': {
235 'id': '06e53103ca9aa',
236 'ext': 'flv',
237 'title': 'Macross Episode 001 Watch Macross Episode 001 onl',
238 'description': 'No description',
239 },
240 },
241 # arte embed
242 {
243 'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
244 'md5': '7653032cbb25bf6c80d80f217055fa43',
245 'info_dict': {
246 'id': '048195-004_PLUS7-F',
247 'ext': 'flv',
248 'title': 'X:enius',
249 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
250 'upload_date': '20140320',
251 },
252 'params': {
253 'skip_download': 'Requires rtmpdump'
254 }
255 },
256 # Condé Nast embed
257 {
258 'url': 'http://www.wired.com/2014/04/honda-asimo/',
259 'md5': 'ba0dfe966fa007657bd1443ee672db0f',
260 'info_dict': {
261 'id': '53501be369702d3275860000',
262 'ext': 'mp4',
263 'title': 'Honda’s New Asimo Robot Is More Human Than Ever',
264 }
265 },
266 # Dailymotion embed
267 {
268 'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
269 'md5': '441aeeb82eb72c422c7f14ec533999cd',
270 'info_dict': {
271 'id': 'k2mm4bCdJ6CQ2i7c8o2',
272 'ext': 'mp4',
273 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
274 'uploader': 'Spi0n',
275 },
276 'add_ie': ['Dailymotion'],
277 },
278 # YouTube embed
279 {
280 'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
281 'info_dict': {
282 'id': 'FXRb4ykk4S0',
283 'ext': 'mp4',
284 'title': 'The NBL Auction 2014',
285 'uploader': 'BADMINTON England',
286 'uploader_id': 'BADMINTONEvents',
287 'upload_date': '20140603',
288 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
289 },
290 'add_ie': ['Youtube'],
291 'params': {
292 'skip_download': True,
293 }
294 },
295 # MTVSercices embed
296 {
297 'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
298 'md5': '35727f82f58c76d996fc188f9755b0d5',
299 'info_dict': {
300 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
301 'ext': 'mp4',
302 'title': 'Review',
303 'description': 'Mario\'s life in the fast lane has never looked so good.',
304 },
305 },
306 # YouTube embed via <data-embed-url="">
307 {
308 'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
309 'info_dict': {
310 'id': '4vAffPZIT44',
311 'ext': 'mp4',
312 'title': 'Asphalt 8: Airborne - Update - Welcome to Dubai!',
313 'uploader': 'Gameloft',
314 'uploader_id': 'gameloft',
315 'upload_date': '20140828',
316 'description': 'md5:c80da9ed3d83ae6d1876c834de03e1c4',
317 },
318 'params': {
319 'skip_download': True,
320 }
321 },
322 # Camtasia studio
323 {
324 'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
325 'playlist': [{
326 'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
327 'info_dict': {
328 'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
329 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
330 'ext': 'flv',
331 'duration': 2235.90,
332 }
333 }, {
334 'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
335 'info_dict': {
336 'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
337 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
338 'ext': 'flv',
339 'duration': 2235.93,
340 }
341 }],
342 'info_dict': {
343 'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
344 }
345 },
346 # Flowplayer
347 {
348 'url': 'http://www.handjobhub.com/video/busty-blonde-siri-tit-fuck-while-wank-6313.html',
349 'md5': '9d65602bf31c6e20014319c7d07fba27',
350 'info_dict': {
351 'id': '5123ea6d5e5a7',
352 'ext': 'mp4',
353 'age_limit': 18,
354 'uploader': 'www.handjobhub.com',
355 'title': 'Busty Blonde Siri Tit Fuck While Wank at HandjobHub.com',
356 }
357 },
358 # RSS feed
359 {
360 'url': 'http://phihag.de/2014/youtube-dl/rss2.xml',
361 'info_dict': {
362 'id': 'http://phihag.de/2014/youtube-dl/rss2.xml',
363 'title': 'Zero Punctuation',
364 'description': 're:'
365 },
366 'playlist_mincount': 11,
367 },
368 # Multiple brightcove videos
369 # https://github.com/rg3/youtube-dl/issues/2283
370 {
371 'url': 'http://www.newyorker.com/online/blogs/newsdesk/2014/01/always-never-nuclear-command-and-control.html',
372 'info_dict': {
373 'id': 'always-never',
374 'title': 'Always / Never - The New Yorker',
375 },
376 'playlist_count': 3,
377 'params': {
378 'extract_flat': False,
379 'skip_download': True,
380 }
381 },
382 # MLB embed
383 {
384 'url': 'http://umpire-empire.com/index.php/topic/58125-laz-decides-no-thats-low/',
385 'md5': '96f09a37e44da40dd083e12d9a683327',
386 'info_dict': {
387 'id': '33322633',
388 'ext': 'mp4',
389 'title': 'Ump changes call to ball',
390 'description': 'md5:71c11215384298a172a6dcb4c2e20685',
391 'duration': 48,
392 'timestamp': 1401537900,
393 'upload_date': '20140531',
394 'thumbnail': 're:^https?://.*\.jpg$',
395 },
396 },
397 # Wistia embed
398 {
399 'url': 'http://education-portal.com/academy/lesson/north-american-exploration-failed-colonies-of-spain-france-england.html#lesson',
400 'md5': '8788b683c777a5cf25621eaf286d0c23',
401 'info_dict': {
402 'id': '1cfaf6b7ea',
403 'ext': 'mov',
404 'title': 'md5:51364a8d3d009997ba99656004b5e20d',
405 'duration': 643.0,
406 'filesize': 182808282,
407 'uploader': 'education-portal.com',
408 },
409 },
410 {
411 'url': 'http://thoughtworks.wistia.com/medias/uxjb0lwrcz',
412 'md5': 'baf49c2baa8a7de5f3fc145a8506dcd4',
413 'info_dict': {
414 'id': 'uxjb0lwrcz',
415 'ext': 'mp4',
416 'title': 'Conversation about Hexagonal Rails Part 1 - ThoughtWorks',
417 'duration': 1715.0,
418 'uploader': 'thoughtworks.wistia.com',
419 },
420 },
421 # Direct download with broken HEAD
422 {
423 'url': 'http://ai-radio.org:8000/radio.opus',
424 'info_dict': {
425 'id': 'radio',
426 'ext': 'opus',
427 'title': 'radio',
428 },
429 'params': {
430 'skip_download': True, # infinite live stream
431 },
432 'expected_warnings': [
433 r'501.*Not Implemented'
434 ],
435 },
436 # Soundcloud embed
437 {
438 'url': 'http://nakedsecurity.sophos.com/2014/10/29/sscc-171-are-you-sure-that-1234-is-a-bad-password-podcast/',
439 'info_dict': {
440 'id': '174391317',
441 'ext': 'mp3',
442 'description': 'md5:ff867d6b555488ad3c52572bb33d432c',
443 'uploader': 'Sophos Security',
444 'title': 'Chet Chat 171 - Oct 29, 2014',
445 'upload_date': '20141029',
446 }
447 },
448 # Livestream embed
449 {
450 'url': 'http://www.esa.int/Our_Activities/Space_Science/Rosetta/Philae_comet_touch-down_webcast',
451 'info_dict': {
452 'id': '67864563',
453 'ext': 'flv',
454 'upload_date': '20141112',
455 'title': 'Rosetta #CometLanding webcast HL 10',
456 }
457 },
458 # LazyYT
459 {
460 'url': 'http://discourse.ubuntu.com/t/unity-8-desktop-mode-windows-on-mir/1986',
461 'info_dict': {
462 'title': 'Unity 8 desktop-mode windows on Mir! - Ubuntu Discourse',
463 },
464 'playlist_mincount': 2,
465 },
466 # Direct link with incorrect MIME type
467 {
468 'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
469 'md5': '4ccbebe5f36706d85221f204d7eb5913',
470 'info_dict': {
471 'url': 'http://ftp.nluug.nl/video/nluug/2014-11-20_nj14/zaal-2/5_Lennart_Poettering_-_Systemd.webm',
472 'id': '5_Lennart_Poettering_-_Systemd',
473 'ext': 'webm',
474 'title': '5_Lennart_Poettering_-_Systemd',
475 'upload_date': '20141120',
476 },
477 'expected_warnings': [
478 'URL could be a direct video link, returning it as such.'
479 ]
480 },
481 # Cinchcast embed
482 {
483 'url': 'http://undergroundwellness.com/podcasts/306-5-steps-to-permanent-gut-healing/',
484 'info_dict': {
485 'id': '7141703',
486 'ext': 'mp3',
487 'upload_date': '20141126',
488 'title': 'Jack Tips: 5 Steps to Permanent Gut Healing',
489 }
490 },
491 ]
492
493 def report_following_redirect(self, new_url):
494 """Report information extraction."""
495 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
496
497 def _extract_rss(self, url, video_id, doc):
498 playlist_title = doc.find('./channel/title').text
499 playlist_desc_el = doc.find('./channel/description')
500 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
501
502 entries = [{
503 '_type': 'url',
504 'url': e.find('link').text,
505 'title': e.find('title').text,
506 } for e in doc.findall('./channel/item')]
507
508 return {
509 '_type': 'playlist',
510 'id': url,
511 'title': playlist_title,
512 'description': playlist_desc,
513 'entries': entries,
514 }
515
516 def _extract_camtasia(self, url, video_id, webpage):
517 """ Returns None if no camtasia video can be found. """
518
519 camtasia_cfg = self._search_regex(
520 r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
521 webpage, 'camtasia configuration file', default=None)
522 if camtasia_cfg is None:
523 return None
524
525 title = self._html_search_meta('DC.title', webpage, fatal=True)
526
527 camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
528 camtasia_cfg = self._download_xml(
529 camtasia_url, video_id,
530 note='Downloading camtasia configuration',
531 errnote='Failed to download camtasia configuration')
532 fileset_node = camtasia_cfg.find('./playlist/array/fileset')
533
534 entries = []
535 for n in fileset_node.getchildren():
536 url_n = n.find('./uri')
537 if url_n is None:
538 continue
539
540 entries.append({
541 'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
542 'title': '%s - %s' % (title, n.tag),
543 'url': compat_urlparse.urljoin(url, url_n.text),
544 'duration': float_or_none(n.find('./duration').text),
545 })
546
547 return {
548 '_type': 'playlist',
549 'entries': entries,
550 'title': title,
551 }
552
553 def _real_extract(self, url):
554 if url.startswith('//'):
555 return {
556 '_type': 'url',
557 'url': self.http_scheme() + url,
558 }
559
560 parsed_url = compat_urlparse.urlparse(url)
561 if not parsed_url.scheme:
562 default_search = self._downloader.params.get('default_search')
563 if default_search is None:
564 default_search = 'fixup_error'
565
566 if default_search in ('auto', 'auto_warning', 'fixup_error'):
567 if '/' in url:
568 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
569 return self.url_result('http://' + url)
570 elif default_search != 'fixup_error':
571 if default_search == 'auto_warning':
572 if re.match(r'^(?:url|URL)$', url):
573 raise ExtractorError(
574 'Invalid URL: %r . Call youtube-dl like this: youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc" ' % url,
575 expected=True)
576 else:
577 self._downloader.report_warning(
578 'Falling back to youtube search for %s . Set --default-search "auto" to suppress this warning.' % url)
579 return self.url_result('ytsearch:' + url)
580
581 if default_search in ('error', 'fixup_error'):
582 raise ExtractorError(
583 '%r is not a valid URL. '
584 'Set --default-search "ytsearch" (or run youtube-dl "ytsearch:%s" ) to search YouTube'
585 % (url, url), expected=True)
586 else:
587 if ':' not in default_search:
588 default_search += ':'
589 return self.url_result(default_search + url)
590
591 url, smuggled_data = unsmuggle_url(url)
592 force_videoid = None
593 is_intentional = smuggled_data and smuggled_data.get('to_generic')
594 if smuggled_data and 'force_videoid' in smuggled_data:
595 force_videoid = smuggled_data['force_videoid']
596 video_id = force_videoid
597 else:
598 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
599
600 self.to_screen('%s: Requesting header' % video_id)
601
602 head_req = HEADRequest(url)
603 head_response = self._request_webpage(
604 head_req, video_id,
605 note=False, errnote='Could not send HEAD request to %s' % url,
606 fatal=False)
607
608 if head_response is not False:
609 # Check for redirect
610 new_url = head_response.geturl()
611 if url != new_url:
612 self.report_following_redirect(new_url)
613 if force_videoid:
614 new_url = smuggle_url(
615 new_url, {'force_videoid': force_videoid})
616 return self.url_result(new_url)
617
618 full_response = None
619 if head_response is False:
620 full_response = self._request_webpage(url, video_id)
621 head_response = full_response
622
623 # Check for direct link to a video
624 content_type = head_response.headers.get('Content-Type', '')
625 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
626 if m:
627 upload_date = unified_strdate(
628 head_response.headers.get('Last-Modified'))
629 return {
630 'id': video_id,
631 'title': os.path.splitext(url_basename(url))[0],
632 'direct': True,
633 'formats': [{
634 'format_id': m.group('format_id'),
635 'url': url,
636 'vcodec': 'none' if m.group('type') == 'audio' else None
637 }],
638 'upload_date': upload_date,
639 }
640
641 if not self._downloader.params.get('test', False) and not is_intentional:
642 self._downloader.report_warning('Falling back on generic information extractor.')
643
644 if not full_response:
645 full_response = self._request_webpage(url, video_id)
646
647 # Maybe it's a direct link to a video?
648 # Be careful not to download the whole thing!
649 first_bytes = full_response.read(512)
650 if not re.match(r'^\s*<', first_bytes.decode('utf-8', 'replace')):
651 self._downloader.report_warning(
652 'URL could be a direct video link, returning it as such.')
653 upload_date = unified_strdate(
654 head_response.headers.get('Last-Modified'))
655 return {
656 'id': video_id,
657 'title': os.path.splitext(url_basename(url))[0],
658 'direct': True,
659 'url': url,
660 'upload_date': upload_date,
661 }
662
663 webpage = self._webpage_read_content(
664 full_response, url, video_id, prefix=first_bytes)
665
666 self.report_extraction(video_id)
667
668 # Is it an RSS feed?
669 try:
670 doc = parse_xml(webpage)
671 if doc.tag == 'rss':
672 return self._extract_rss(url, video_id, doc)
673 except compat_xml_parse_error:
674 pass
675
676 # Is it a Camtasia project?
677 camtasia_res = self._extract_camtasia(url, video_id, webpage)
678 if camtasia_res is not None:
679 return camtasia_res
680
681 # Sometimes embedded video player is hidden behind percent encoding
682 # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
683 # Unescaping the whole page allows to handle those cases in a generic way
684 webpage = compat_urllib_parse.unquote(webpage)
685
686 # it's tempting to parse this further, but you would
687 # have to take into account all the variations like
688 # Video Title - Site Name
689 # Site Name | Video Title
690 # Video Title - Tagline | Site Name
691 # and so on and so forth; it's just not practical
692 video_title = self._html_search_regex(
693 r'(?s)<title>(.*?)</title>', webpage, 'video title',
694 default='video')
695
696 # Try to detect age limit automatically
697 age_limit = self._rta_search(webpage)
698 # And then there are the jokers who advertise that they use RTA,
699 # but actually don't.
700 AGE_LIMIT_MARKERS = [
701 r'Proudly Labeled <a href="http://www.rtalabel.org/" title="Restricted to Adults">RTA</a>',
702 ]
703 if any(re.search(marker, webpage) for marker in AGE_LIMIT_MARKERS):
704 age_limit = 18
705
706 # video uploader is domain name
707 video_uploader = self._search_regex(
708 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
709
710 # Helper method
711 def _playlist_from_matches(matches, getter=None, ie=None):
712 urlrs = orderedSet(
713 self.url_result(self._proto_relative_url(getter(m) if getter else m), ie)
714 for m in matches)
715 return self.playlist_result(
716 urlrs, playlist_id=video_id, playlist_title=video_title)
717
718 # Look for BrightCove:
719 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
720 if bc_urls:
721 self.to_screen('Brightcove video detected.')
722 entries = [{
723 '_type': 'url',
724 'url': smuggle_url(bc_url, {'Referer': url}),
725 'ie_key': 'Brightcove'
726 } for bc_url in bc_urls]
727
728 return {
729 '_type': 'playlist',
730 'title': video_title,
731 'id': video_id,
732 'entries': entries,
733 }
734
735 # Look for embedded (iframe) Vimeo player
736 mobj = re.search(
737 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
738 if mobj:
739 player_url = unescapeHTML(mobj.group('url'))
740 surl = smuggle_url(player_url, {'Referer': url})
741 return self.url_result(surl)
742
743 # Look for embedded (swf embed) Vimeo player
744 mobj = re.search(
745 r'<embed[^>]+?src="((?:https?:)?//(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
746 if mobj:
747 return self.url_result(mobj.group(1))
748
749 # Look for embedded YouTube player
750 matches = re.findall(r'''(?x)
751 (?:
752 <iframe[^>]+?src=|
753 data-video-url=|
754 <embed[^>]+?src=|
755 embedSWF\(?:\s*|
756 new\s+SWFObject\(
757 )
758 (["\'])
759 (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
760 (?:embed|v|p)/.+?)
761 \1''', webpage)
762 if matches:
763 return _playlist_from_matches(
764 matches, lambda m: unescapeHTML(m[1]))
765
766 # Look for lazyYT YouTube embed
767 matches = re.findall(
768 r'class="lazyYT" data-youtube-id="([^"]+)"', webpage)
769 if matches:
770 return _playlist_from_matches(matches, lambda m: unescapeHTML(m))
771
772 # Look for embedded Dailymotion player
773 matches = re.findall(
774 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
775 if matches:
776 return _playlist_from_matches(
777 matches, lambda m: unescapeHTML(m[1]))
778
779 # Look for embedded Dailymotion playlist player (#3822)
780 m = re.search(
781 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.[a-z]{2,3}/widget/jukebox\?.+?)\1', webpage)
782 if m:
783 playlists = re.findall(
784 r'list\[\]=/playlist/([^/]+)/', unescapeHTML(m.group('url')))
785 if playlists:
786 return _playlist_from_matches(
787 playlists, lambda p: '//dailymotion.com/playlist/%s' % p)
788
789 # Look for embedded Wistia player
790 match = re.search(
791 r'<(?:meta[^>]+?content|iframe[^>]+?src)=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
792 if match:
793 embed_url = self._proto_relative_url(
794 unescapeHTML(match.group('url')))
795 return {
796 '_type': 'url_transparent',
797 'url': embed_url,
798 'ie_key': 'Wistia',
799 'uploader': video_uploader,
800 'title': video_title,
801 'id': video_id,
802 }
803
804 match = re.search(r'(?:id=["\']wistia_|data-wistia-?id=["\']|Wistia\.embed\(["\'])(?P<id>[^"\']+)', webpage)
805 if match:
806 return {
807 '_type': 'url_transparent',
808 'url': 'http://fast.wistia.net/embed/iframe/{0:}'.format(match.group('id')),
809 'ie_key': 'Wistia',
810 'uploader': video_uploader,
811 'title': video_title,
812 'id': match.group('id')
813 }
814
815 # Look for embedded blip.tv player
816 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
817 if mobj:
818 return self.url_result('http://blip.tv/a/a-' + mobj.group(1), 'BlipTV')
819 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
820 if mobj:
821 return self.url_result(mobj.group(1), 'BlipTV')
822
823 # Look for embedded condenast player
824 matches = re.findall(
825 r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
826 webpage)
827 if matches:
828 return {
829 '_type': 'playlist',
830 'entries': [{
831 '_type': 'url',
832 'ie_key': 'CondeNast',
833 'url': ma,
834 } for ma in matches],
835 'title': video_title,
836 'id': video_id,
837 }
838
839 # Look for Bandcamp pages with custom domain
840 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
841 if mobj is not None:
842 burl = unescapeHTML(mobj.group(1))
843 # Don't set the extractor because it can be a track url or an album
844 return self.url_result(burl)
845
846 # Look for embedded Vevo player
847 mobj = re.search(
848 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
849 if mobj is not None:
850 return self.url_result(mobj.group('url'))
851
852 # Look for Ooyala videos
853 mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
854 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
855 if mobj is not None:
856 return OoyalaIE._build_url_result(mobj.group('ec'))
857
858 # Look for Aparat videos
859 mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
860 if mobj is not None:
861 return self.url_result(mobj.group(1), 'Aparat')
862
863 # Look for MPORA videos
864 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
865 if mobj is not None:
866 return self.url_result(mobj.group(1), 'Mpora')
867
868 # Look for embedded NovaMov-based player
869 mobj = re.search(
870 r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
871 (?P<url>http://(?:(?:embed|www)\.)?
872 (?:novamov\.com|
873 nowvideo\.(?:ch|sx|eu|at|ag|co)|
874 videoweed\.(?:es|com)|
875 movshare\.(?:net|sx|ag)|
876 divxstage\.(?:eu|net|ch|co|at|ag))
877 /embed\.php.+?)\1''', webpage)
878 if mobj is not None:
879 return self.url_result(mobj.group('url'))
880
881 # Look for embedded Facebook player
882 mobj = re.search(
883 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
884 if mobj is not None:
885 return self.url_result(mobj.group('url'), 'Facebook')
886
887 # Look for embedded VK player
888 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
889 if mobj is not None:
890 return self.url_result(mobj.group('url'), 'VK')
891
892 # Look for embedded ivi player
893 mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
894 if mobj is not None:
895 return self.url_result(mobj.group('url'), 'Ivi')
896
897 # Look for embedded Huffington Post player
898 mobj = re.search(
899 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
900 if mobj is not None:
901 return self.url_result(mobj.group('url'), 'HuffPost')
902
903 # Look for embed.ly
904 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
905 if mobj is not None:
906 return self.url_result(mobj.group('url'))
907 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
908 if mobj is not None:
909 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
910
911 # Look for funnyordie embed
912 matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
913 if matches:
914 return _playlist_from_matches(
915 matches, getter=unescapeHTML, ie='FunnyOrDie')
916
917 # Look for BBC iPlayer embed
918 matches = re.findall(r'setPlaylist\("(https?://www\.bbc\.co\.uk/iplayer/[^/]+/[\da-z]{8})"\)', webpage)
919 if matches:
920 return _playlist_from_matches(matches, ie='BBCCoUk')
921
922 # Look for embedded RUTV player
923 rutv_url = RUTVIE._extract_url(webpage)
924 if rutv_url:
925 return self.url_result(rutv_url, 'RUTV')
926
927 # Look for embedded TED player
928 mobj = re.search(
929 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed(?:-ssl)?\.ted\.com/.+?)\1', webpage)
930 if mobj is not None:
931 return self.url_result(mobj.group('url'), 'TED')
932
933 # Look for embedded Ustream videos
934 mobj = re.search(
935 r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
936 if mobj is not None:
937 return self.url_result(mobj.group('url'), 'Ustream')
938
939 # Look for embedded arte.tv player
940 mobj = re.search(
941 r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
942 webpage)
943 if mobj is not None:
944 return self.url_result(mobj.group('url'), 'ArteTVEmbed')
945
946 # Look for embedded smotri.com player
947 smotri_url = SmotriIE._extract_url(webpage)
948 if smotri_url:
949 return self.url_result(smotri_url, 'Smotri')
950
951 # Look for embeded soundcloud player
952 mobj = re.search(
953 r'<iframe\s+(?:[a-zA-Z0-9_-]+="[^"]+"\s+)*src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
954 webpage)
955 if mobj is not None:
956 url = unescapeHTML(mobj.group('url'))
957 return self.url_result(url)
958
959 # Look for embedded vulture.com player
960 mobj = re.search(
961 r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
962 webpage)
963 if mobj is not None:
964 url = unescapeHTML(mobj.group('url'))
965 return self.url_result(url, ie='Vulture')
966
967 # Look for embedded mtvservices player
968 mobj = re.search(
969 r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
970 webpage)
971 if mobj is not None:
972 url = unescapeHTML(mobj.group('url'))
973 return self.url_result(url, ie='MTVServicesEmbedded')
974
975 # Look for embedded yahoo player
976 mobj = re.search(
977 r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
978 webpage)
979 if mobj is not None:
980 return self.url_result(mobj.group('url'), 'Yahoo')
981
982 # Look for embedded sbs.com.au player
983 mobj = re.search(
984 r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
985 webpage)
986 if mobj is not None:
987 return self.url_result(mobj.group('url'), 'SBS')
988
989 # Look for embedded Cinchcast player
990 mobj = re.search(
991 r'<iframe[^>]+?src=(["\'])(?P<url>https?://player\.cinchcast\.com/.+?)\1',
992 webpage)
993 if mobj is not None:
994 return self.url_result(mobj.group('url'), 'Cinchcast')
995
996 mobj = re.search(
997 r'<iframe[^>]+?src=(["\'])(?P<url>https?://m(?:lb)?\.mlb\.com/shared/video/embed/embed\.html\?.+?)\1',
998 webpage)
999 if mobj is not None:
1000 return self.url_result(mobj.group('url'), 'MLB')
1001
1002 mobj = re.search(
1003 r'<iframe[^>]+?src=(["\'])(?P<url>%s)\1' % CondeNastIE.EMBED_URL,
1004 webpage)
1005 if mobj is not None:
1006 return self.url_result(self._proto_relative_url(mobj.group('url'), scheme='http:'), 'CondeNast')
1007
1008 mobj = re.search(
1009 r'<iframe[^>]+src="(?P<url>https?://new\.livestream\.com/[^"]+/player[^"]+)"',
1010 webpage)
1011 if mobj is not None:
1012 return self.url_result(mobj.group('url'), 'Livestream')
1013
1014 def check_video(vurl):
1015 vpath = compat_urlparse.urlparse(vurl).path
1016 vext = determine_ext(vpath)
1017 return '.' in vpath and vext not in ('swf', 'png', 'jpg', 'srt', 'sbv', 'sub', 'vtt', 'ttml')
1018
1019 def filter_video(urls):
1020 return list(filter(check_video, urls))
1021
1022 # Start with something easy: JW Player in SWFObject
1023 found = filter_video(re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage))
1024 if not found:
1025 # Look for gorilla-vid style embedding
1026 found = filter_video(re.findall(r'''(?sx)
1027 (?:
1028 jw_plugins|
1029 JWPlayerOptions|
1030 jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
1031 )
1032 .*?file\s*:\s*["\'](.*?)["\']''', webpage))
1033 if not found:
1034 # Broaden the search a little bit
1035 found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
1036 if not found:
1037 # Broaden the findall a little bit: JWPlayer JS loader
1038 found = filter_video(re.findall(
1039 r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage))
1040 if not found:
1041 # Flow player
1042 found = filter_video(re.findall(r'''(?xs)
1043 flowplayer\("[^"]+",\s*
1044 \{[^}]+?\}\s*,
1045 \s*{[^}]+? ["']?clip["']?\s*:\s*\{\s*
1046 ["']?url["']?\s*:\s*["']([^"']+)["']
1047 ''', webpage))
1048 if not found:
1049 # Try to find twitter cards info
1050 found = filter_video(re.findall(
1051 r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage))
1052 if not found:
1053 # We look for Open Graph info:
1054 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
1055 m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
1056 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
1057 if m_video_type is not None:
1058 found = filter_video(re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage))
1059 if not found:
1060 # HTML5 video
1061 found = re.findall(r'(?s)<video[^<]*(?:>.*?<source[^>]*)?\s+src=["\'](.*?)["\']', webpage)
1062 if not found:
1063 found = re.search(
1064 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
1065 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'?([^\'"]+)',
1066 webpage)
1067 if found:
1068 new_url = found.group(1)
1069 self.report_following_redirect(new_url)
1070 return {
1071 '_type': 'url',
1072 'url': new_url,
1073 }
1074 if not found:
1075 raise UnsupportedError(url)
1076
1077 entries = []
1078 for video_url in found:
1079 video_url = compat_urlparse.urljoin(url, video_url)
1080 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
1081
1082 # Sometimes, jwplayer extraction will result in a YouTube URL
1083 if YoutubeIE.suitable(video_url):
1084 entries.append(self.url_result(video_url, 'Youtube'))
1085 continue
1086
1087 # here's a fun little line of code for you:
1088 video_id = os.path.splitext(video_id)[0]
1089
1090 entries.append({
1091 'id': video_id,
1092 'url': video_url,
1093 'uploader': video_uploader,
1094 'title': video_title,
1095 'age_limit': age_limit,
1096 })
1097
1098 if len(entries) == 1:
1099 return entries[0]
1100 else:
1101 for num, e in enumerate(entries, start=1):
1102 e['title'] = '%s (%d)' % (e['title'], num)
1103 return {
1104 '_type': 'playlist',
1105 'entries': entries,
1106 }