]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/generic.py
bcb0765940df39656be9f78c1ab144976adb5e5e
[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 ..utils import (
11 compat_urllib_error,
12 compat_urllib_parse,
13 compat_urllib_request,
14 compat_urlparse,
15 compat_xml_parse_error,
16
17 ExtractorError,
18 HEADRequest,
19 parse_xml,
20 smuggle_url,
21 unescapeHTML,
22 unified_strdate,
23 url_basename,
24 )
25 from .brightcove import BrightcoveIE
26 from .ooyala import OoyalaIE
27 from .rutv import RUTVIE
28 from .smotri import SmotriIE
29
30
31 class GenericIE(InfoExtractor):
32 IE_DESC = 'Generic downloader that works on some sites'
33 _VALID_URL = r'.*'
34 IE_NAME = 'generic'
35 _TESTS = [
36 {
37 'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
38 'md5': '85b90ccc9d73b4acd9138d3af4c27f89',
39 'info_dict': {
40 'id': '13601338388002',
41 'ext': 'mp4',
42 'uploader': 'www.hodiho.fr',
43 'title': 'R\u00e9gis plante sa Jeep',
44 }
45 },
46 # bandcamp page with custom domain
47 {
48 'add_ie': ['Bandcamp'],
49 'url': 'http://bronyrock.com/track/the-pony-mash',
50 'info_dict': {
51 'id': '3235767654',
52 'ext': 'mp3',
53 'title': 'The Pony Mash',
54 'uploader': 'M_Pallante',
55 },
56 'skip': 'There is a limit of 200 free downloads / month for the test song',
57 },
58 # embedded brightcove video
59 # it also tests brightcove videos that need to set the 'Referer' in the
60 # http requests
61 {
62 'add_ie': ['Brightcove'],
63 'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
64 'info_dict': {
65 'id': '2765128793001',
66 'ext': 'mp4',
67 'title': 'Le cours de bourse : l’analyse technique',
68 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
69 'uploader': 'BFM BUSINESS',
70 },
71 'params': {
72 'skip_download': True,
73 },
74 },
75 {
76 # https://github.com/rg3/youtube-dl/issues/2253
77 'url': 'http://bcove.me/i6nfkrc3',
78 'md5': '0ba9446db037002366bab3b3eb30c88c',
79 'info_dict': {
80 'id': '3101154703001',
81 'ext': 'mp4',
82 'title': 'Still no power',
83 'uploader': 'thestar.com',
84 '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.',
85 },
86 'add_ie': ['Brightcove'],
87 },
88 {
89 'url': 'http://www.championat.com/video/football/v/87/87499.html',
90 'md5': 'fb973ecf6e4a78a67453647444222983',
91 'info_dict': {
92 'id': '3414141473001',
93 'ext': 'mp4',
94 'title': 'Видео. Удаление Дзагоева (ЦСКА)',
95 'description': 'Онлайн-трансляция матча ЦСКА - "Волга"',
96 'uploader': 'Championat',
97 },
98 },
99 # Direct link to a video
100 {
101 'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
102 'md5': '67d406c2bcb6af27fa886f31aa934bbe',
103 'info_dict': {
104 'id': 'trailer',
105 'ext': 'mp4',
106 'title': 'trailer',
107 'upload_date': '20100513',
108 }
109 },
110 # ooyala video
111 {
112 'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
113 'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
114 'info_dict': {
115 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
116 'ext': 'mp4',
117 'title': '2cc213299525360.mov', # that's what we get
118 },
119 },
120 # google redirect
121 {
122 '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',
123 'info_dict': {
124 'id': 'cmQHVoWB5FY',
125 'ext': 'mp4',
126 'upload_date': '20130224',
127 'uploader_id': 'TheVerge',
128 '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.',
129 'uploader': 'The Verge',
130 'title': 'First Firefox OS phones side-by-side',
131 },
132 'params': {
133 'skip_download': False,
134 }
135 },
136 # embed.ly video
137 {
138 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
139 'info_dict': {
140 'id': '9ODmcdjQcHQ',
141 'ext': 'mp4',
142 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
143 'upload_date': '20140225',
144 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
145 'uploader': 'Tested',
146 'uploader_id': 'testedcom',
147 },
148 # No need to test YoutubeIE here
149 'params': {
150 'skip_download': True,
151 },
152 },
153 # funnyordie embed
154 {
155 'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
156 'md5': '7cf780be104d40fea7bae52eed4a470e',
157 'info_dict': {
158 'id': '18e820ec3f',
159 'ext': 'mp4',
160 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
161 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
162 },
163 },
164 # RUTV embed
165 {
166 'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
167 'info_dict': {
168 'id': '776940',
169 'ext': 'mp4',
170 'title': 'Охотское море стало целиком российским',
171 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
172 },
173 'params': {
174 # m3u8 download
175 'skip_download': True,
176 },
177 },
178 # Embedded TED video
179 {
180 'url': 'http://en.support.wordpress.com/videos/ted-talks/',
181 'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
182 'info_dict': {
183 'id': '981',
184 'ext': 'mp4',
185 'title': 'My web playroom',
186 'uploader': 'Ze Frank',
187 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
188 }
189 },
190 # Embeded Ustream video
191 {
192 'url': 'http://www.american.edu/spa/pti/nsa-privacy-janus-2014.cfm',
193 'md5': '27b99cdb639c9b12a79bca876a073417',
194 'info_dict': {
195 'id': '45734260',
196 'ext': 'flv',
197 'uploader': 'AU SPA: The NSA and Privacy',
198 'title': 'NSA and Privacy Forum Debate featuring General Hayden and Barton Gellman'
199 }
200 },
201 # nowvideo embed hidden behind percent encoding
202 {
203 'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
204 'md5': '2baf4ddd70f697d94b1c18cf796d5107',
205 'info_dict': {
206 'id': '06e53103ca9aa',
207 'ext': 'flv',
208 'title': 'Macross Episode 001 Watch Macross Episode 001 onl',
209 'description': 'No description',
210 },
211 },
212 # arte embed
213 {
214 'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
215 'md5': '7653032cbb25bf6c80d80f217055fa43',
216 'info_dict': {
217 'id': '048195-004_PLUS7-F',
218 'ext': 'flv',
219 'title': 'X:enius',
220 'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
221 'upload_date': '20140320',
222 },
223 'params': {
224 'skip_download': 'Requires rtmpdump'
225 }
226 },
227 # smotri embed
228 {
229 'url': 'http://rbctv.rbc.ru/archive/news/562949990879132.shtml',
230 'md5': 'ec40048448e9284c9a1de77bb188108b',
231 'info_dict': {
232 'id': 'v27008541fad',
233 'ext': 'mp4',
234 'title': 'Крым и Севастополь вошли в состав России',
235 'description': 'md5:fae01b61f68984c7bd2fa741e11c3175',
236 'duration': 900,
237 'upload_date': '20140318',
238 'uploader': 'rbctv_2012_4',
239 'uploader_id': 'rbctv_2012_4',
240 },
241 },
242 # Condé Nast embed
243 {
244 'url': 'http://www.wired.com/2014/04/honda-asimo/',
245 'md5': 'ba0dfe966fa007657bd1443ee672db0f',
246 'info_dict': {
247 'id': '53501be369702d3275860000',
248 'ext': 'mp4',
249 'title': 'Honda’s New Asimo Robot Is More Human Than Ever',
250 }
251 },
252 # Dailymotion embed
253 {
254 'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
255 'md5': '441aeeb82eb72c422c7f14ec533999cd',
256 'info_dict': {
257 'id': 'k2mm4bCdJ6CQ2i7c8o2',
258 'ext': 'mp4',
259 'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
260 'uploader': 'Spi0n',
261 },
262 'add_ie': ['Dailymotion'],
263 },
264 # YouTube embed
265 {
266 'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
267 'info_dict': {
268 'id': 'FXRb4ykk4S0',
269 'ext': 'mp4',
270 'title': 'The NBL Auction 2014',
271 'uploader': 'BADMINTON England',
272 'uploader_id': 'BADMINTONEvents',
273 'upload_date': '20140603',
274 'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
275 },
276 'add_ie': ['Youtube'],
277 'params': {
278 'skip_download': True,
279 }
280 },
281 # MTVSercices embed
282 {
283 'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
284 'md5': '35727f82f58c76d996fc188f9755b0d5',
285 'info_dict': {
286 'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
287 'ext': 'mp4',
288 'title': 'Review',
289 'description': 'Mario\'s life in the fast lane has never looked so good.',
290 },
291 },
292 ]
293
294 def report_download_webpage(self, video_id):
295 """Report webpage download."""
296 if not self._downloader.params.get('test', False):
297 self._downloader.report_warning('Falling back on generic information extractor.')
298 super(GenericIE, self).report_download_webpage(video_id)
299
300 def report_following_redirect(self, new_url):
301 """Report information extraction."""
302 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
303
304 def _send_head(self, url):
305 """Check if it is a redirect, like url shorteners, in case return the new url."""
306
307 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
308 """
309 Subclass the HTTPRedirectHandler to make it use our
310 HEADRequest also on the redirected URL
311 """
312 def redirect_request(self, req, fp, code, msg, headers, newurl):
313 if code in (301, 302, 303, 307):
314 newurl = newurl.replace(' ', '%20')
315 newheaders = dict((k,v) for k,v in req.headers.items()
316 if k.lower() not in ("content-length", "content-type"))
317 try:
318 # This function was deprecated in python 3.3 and removed in 3.4
319 origin_req_host = req.get_origin_req_host()
320 except AttributeError:
321 origin_req_host = req.origin_req_host
322 return HEADRequest(newurl,
323 headers=newheaders,
324 origin_req_host=origin_req_host,
325 unverifiable=True)
326 else:
327 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
328
329 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
330 """
331 Fallback to GET if HEAD is not allowed (405 HTTP error)
332 """
333 def http_error_405(self, req, fp, code, msg, headers):
334 fp.read()
335 fp.close()
336
337 newheaders = dict((k,v) for k,v in req.headers.items()
338 if k.lower() not in ("content-length", "content-type"))
339 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
340 headers=newheaders,
341 origin_req_host=req.get_origin_req_host(),
342 unverifiable=True))
343
344 # Build our opener
345 opener = compat_urllib_request.OpenerDirector()
346 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
347 HTTPMethodFallback, HEADRedirectHandler,
348 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
349 opener.add_handler(handler())
350
351 response = opener.open(HEADRequest(url))
352 if response is None:
353 raise ExtractorError('Invalid URL protocol')
354 return response
355
356 def _extract_rss(self, url, video_id, doc):
357 playlist_title = doc.find('./channel/title').text
358 playlist_desc_el = doc.find('./channel/description')
359 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
360
361 entries = [{
362 '_type': 'url',
363 'url': e.find('link').text,
364 'title': e.find('title').text,
365 } for e in doc.findall('./channel/item')]
366
367 return {
368 '_type': 'playlist',
369 'id': url,
370 'title': playlist_title,
371 'description': playlist_desc,
372 'entries': entries,
373 }
374
375 def _real_extract(self, url):
376 if url.startswith('//'):
377 return {
378 '_type': 'url',
379 'url': self.http_scheme() + url,
380 }
381
382 parsed_url = compat_urlparse.urlparse(url)
383 if not parsed_url.scheme:
384 default_search = self._downloader.params.get('default_search')
385 if default_search is None:
386 default_search = 'fixup_error'
387
388 if default_search in ('auto', 'auto_warning', 'fixup_error'):
389 if '/' in url:
390 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
391 return self.url_result('http://' + url)
392 elif default_search != 'fixup_error':
393 if default_search == 'auto_warning':
394 if re.match(r'^(?:url|URL)$', url):
395 raise ExtractorError(
396 'Invalid URL: %r . Call youtube-dl like this: youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc" ' % url,
397 expected=True)
398 else:
399 self._downloader.report_warning(
400 'Falling back to youtube search for %s . Set --default-search "auto" to suppress this warning.' % url)
401 return self.url_result('ytsearch:' + url)
402
403 if default_search in ('error', 'fixup_error'):
404 raise ExtractorError(
405 ('%r is not a valid URL. '
406 'Set --default-search "ytsearch" (or run youtube-dl "ytsearch:%s" ) to search YouTube'
407 ) % (url, url), expected=True)
408 else:
409 assert ':' in default_search
410 return self.url_result(default_search + url)
411 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
412
413 self.to_screen('%s: Requesting header' % video_id)
414
415 try:
416 response = self._send_head(url)
417
418 # Check for redirect
419 new_url = response.geturl()
420 if url != new_url:
421 self.report_following_redirect(new_url)
422 return self.url_result(new_url)
423
424 # Check for direct link to a video
425 content_type = response.headers.get('Content-Type', '')
426 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
427 if m:
428 upload_date = response.headers.get('Last-Modified')
429 if upload_date:
430 upload_date = unified_strdate(upload_date)
431 return {
432 'id': video_id,
433 'title': os.path.splitext(url_basename(url))[0],
434 'formats': [{
435 'format_id': m.group('format_id'),
436 'url': url,
437 'vcodec': 'none' if m.group('type') == 'audio' else None
438 }],
439 'upload_date': upload_date,
440 }
441
442 except compat_urllib_error.HTTPError:
443 # This may be a stupid server that doesn't like HEAD, our UA, or so
444 pass
445
446 try:
447 webpage = self._download_webpage(url, video_id)
448 except ValueError:
449 # since this is the last-resort InfoExtractor, if
450 # this error is thrown, it'll be thrown here
451 raise ExtractorError('Failed to download URL: %s' % url)
452
453 self.report_extraction(video_id)
454
455 # Is it an RSS feed?
456 try:
457 doc = parse_xml(webpage)
458 if doc.tag == 'rss':
459 return self._extract_rss(url, video_id, doc)
460 except compat_xml_parse_error:
461 pass
462
463 # Sometimes embedded video player is hidden behind percent encoding
464 # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
465 # Unescaping the whole page allows to handle those cases in a generic way
466 webpage = compat_urllib_parse.unquote(webpage)
467
468 # it's tempting to parse this further, but you would
469 # have to take into account all the variations like
470 # Video Title - Site Name
471 # Site Name | Video Title
472 # Video Title - Tagline | Site Name
473 # and so on and so forth; it's just not practical
474 video_title = self._html_search_regex(
475 r'(?s)<title>(.*?)</title>', webpage, 'video title',
476 default='video')
477
478 # video uploader is domain name
479 video_uploader = self._search_regex(
480 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
481
482 # Look for BrightCove:
483 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
484 if bc_urls:
485 self.to_screen('Brightcove video detected.')
486 entries = [{
487 '_type': 'url',
488 'url': smuggle_url(bc_url, {'Referer': url}),
489 'ie_key': 'Brightcove'
490 } for bc_url in bc_urls]
491
492 return {
493 '_type': 'playlist',
494 'title': video_title,
495 'id': video_id,
496 'entries': entries,
497 }
498
499 # Look for embedded (iframe) Vimeo player
500 mobj = re.search(
501 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
502 if mobj:
503 player_url = unescapeHTML(mobj.group('url'))
504 surl = smuggle_url(player_url, {'Referer': url})
505 return self.url_result(surl, 'Vimeo')
506
507 # Look for embedded (swf embed) Vimeo player
508 mobj = re.search(
509 r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
510 if mobj:
511 return self.url_result(mobj.group(1), 'Vimeo')
512
513 # Look for embedded YouTube player
514 matches = re.findall(r'''(?x)
515 (?:
516 <iframe[^>]+?src=|
517 <embed[^>]+?src=|
518 embedSWF\(?:\s*
519 )
520 (["\'])
521 (?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
522 (?:embed|v)/.+?)
523 \1''', webpage)
524 if matches:
525 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
526 for tuppl in matches]
527 return self.playlist_result(
528 urlrs, playlist_id=video_id, playlist_title=video_title)
529
530 # Look for embedded Dailymotion player
531 matches = re.findall(
532 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
533 if matches:
534 urlrs = [self.url_result(unescapeHTML(tuppl[1]))
535 for tuppl in matches]
536 return self.playlist_result(
537 urlrs, playlist_id=video_id, playlist_title=video_title)
538
539 # Look for embedded Wistia player
540 match = re.search(
541 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
542 if match:
543 return {
544 '_type': 'url_transparent',
545 'url': unescapeHTML(match.group('url')),
546 'ie_key': 'Wistia',
547 'uploader': video_uploader,
548 'title': video_title,
549 'id': video_id,
550 }
551
552 # Look for embedded blip.tv player
553 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
554 if mobj:
555 return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
556 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
557 if mobj:
558 return self.url_result(mobj.group(1), 'BlipTV')
559
560 # Look for embedded condenast player
561 matches = re.findall(
562 r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
563 webpage)
564 if matches:
565 return {
566 '_type': 'playlist',
567 'entries': [{
568 '_type': 'url',
569 'ie_key': 'CondeNast',
570 'url': ma,
571 } for ma in matches],
572 'title': video_title,
573 'id': video_id,
574 }
575
576 # Look for Bandcamp pages with custom domain
577 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
578 if mobj is not None:
579 burl = unescapeHTML(mobj.group(1))
580 # Don't set the extractor because it can be a track url or an album
581 return self.url_result(burl)
582
583 # Look for embedded Vevo player
584 mobj = re.search(
585 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
586 if mobj is not None:
587 return self.url_result(mobj.group('url'))
588
589 # Look for Ooyala videos
590 mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
591 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
592 if mobj is not None:
593 return OoyalaIE._build_url_result(mobj.group('ec'))
594
595 # Look for Aparat videos
596 mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
597 if mobj is not None:
598 return self.url_result(mobj.group(1), 'Aparat')
599
600 # Look for MPORA videos
601 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
602 if mobj is not None:
603 return self.url_result(mobj.group(1), 'Mpora')
604
605 # Look for embedded NovaMov-based player
606 mobj = re.search(
607 r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
608 (?P<url>http://(?:(?:embed|www)\.)?
609 (?:novamov\.com|
610 nowvideo\.(?:ch|sx|eu|at|ag|co)|
611 videoweed\.(?:es|com)|
612 movshare\.(?:net|sx|ag)|
613 divxstage\.(?:eu|net|ch|co|at|ag))
614 /embed\.php.+?)\1''', webpage)
615 if mobj is not None:
616 return self.url_result(mobj.group('url'))
617
618 # Look for embedded Facebook player
619 mobj = re.search(
620 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
621 if mobj is not None:
622 return self.url_result(mobj.group('url'), 'Facebook')
623
624 # Look for embedded VK player
625 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
626 if mobj is not None:
627 return self.url_result(mobj.group('url'), 'VK')
628
629 # Look for embedded ivi player
630 mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
631 if mobj is not None:
632 return self.url_result(mobj.group('url'), 'Ivi')
633
634 # Look for embedded Huffington Post player
635 mobj = re.search(
636 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
637 if mobj is not None:
638 return self.url_result(mobj.group('url'), 'HuffPost')
639
640 # Look for embed.ly
641 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
642 if mobj is not None:
643 return self.url_result(mobj.group('url'))
644 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
645 if mobj is not None:
646 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
647
648 # Look for funnyordie embed
649 matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
650 if matches:
651 urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
652 for eurl in matches]
653 return self.playlist_result(
654 urlrs, playlist_id=video_id, playlist_title=video_title)
655
656 # Look for embedded RUTV player
657 rutv_url = RUTVIE._extract_url(webpage)
658 if rutv_url:
659 return self.url_result(rutv_url, 'RUTV')
660
661 # Look for embedded TED player
662 mobj = re.search(
663 r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
664 if mobj is not None:
665 return self.url_result(mobj.group('url'), 'TED')
666
667 # Look for embedded Ustream videos
668 mobj = re.search(
669 r'<iframe[^>]+?src=(["\'])(?P<url>http://www\.ustream\.tv/embed/.+?)\1', webpage)
670 if mobj is not None:
671 return self.url_result(mobj.group('url'), 'Ustream')
672
673 # Look for embedded arte.tv player
674 mobj = re.search(
675 r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
676 webpage)
677 if mobj is not None:
678 return self.url_result(mobj.group('url'), 'ArteTVEmbed')
679
680 # Look for embedded smotri.com player
681 smotri_url = SmotriIE._extract_url(webpage)
682 if smotri_url:
683 return self.url_result(smotri_url, 'Smotri')
684
685 # Look for embeded soundcloud player
686 mobj = re.search(
687 r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
688 webpage)
689 if mobj is not None:
690 url = unescapeHTML(mobj.group('url'))
691 return self.url_result(url)
692
693 # Look for embedded vulture.com player
694 mobj = re.search(
695 r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
696 webpage)
697 if mobj is not None:
698 url = unescapeHTML(mobj.group('url'))
699 return self.url_result(url, ie='Vulture')
700
701 # Look for embedded mtvservices player
702 mobj = re.search(
703 r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
704 webpage)
705 if mobj is not None:
706 url = unescapeHTML(mobj.group('url'))
707 return self.url_result(url, ie='MTVServicesEmbedded')
708
709 # Start with something easy: JW Player in SWFObject
710 found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
711 if not found:
712 # Look for gorilla-vid style embedding
713 found = re.findall(r'''(?sx)
714 (?:
715 jw_plugins|
716 JWPlayerOptions|
717 jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
718 )
719 .*?file\s*:\s*["\'](.*?)["\']''', webpage)
720 if not found:
721 # Broaden the search a little bit
722 found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
723 if not found:
724 # Broaden the findall a little bit: JWPlayer JS loader
725 found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
726 if not found:
727 # Try to find twitter cards info
728 found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
729 if not found:
730 # We look for Open Graph info:
731 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
732 m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
733 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
734 if m_video_type is not None:
735 found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
736 if not found:
737 # HTML5 video
738 found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
739 if not found:
740 found = re.search(
741 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
742 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
743 webpage)
744 if found:
745 new_url = found.group(1)
746 self.report_following_redirect(new_url)
747 return {
748 '_type': 'url',
749 'url': new_url,
750 }
751 if not found:
752 raise ExtractorError('Unsupported URL: %s' % url)
753
754 entries = []
755 for video_url in found:
756 video_url = compat_urlparse.urljoin(url, video_url)
757 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
758
759 # Sometimes, jwplayer extraction will result in a YouTube URL
760 if YoutubeIE.suitable(video_url):
761 entries.append(self.url_result(video_url, 'Youtube'))
762 continue
763
764 # here's a fun little line of code for you:
765 video_id = os.path.splitext(video_id)[0]
766
767 entries.append({
768 'id': video_id,
769 'url': video_url,
770 'uploader': video_uploader,
771 'title': video_title,
772 })
773
774 if len(entries) == 1:
775 return entries[0]
776 else:
777 for num, e in enumerate(entries, start=1):
778 e['title'] = '%s (%d)' % (e['title'], num)
779 return {
780 '_type': 'playlist',
781 'entries': entries,
782 }
783