3 from __future__ import unicode_literals
8 from .common import InfoExtractor
9 from .youtube import YoutubeIE
13 compat_urllib_request,
15 compat_xml_parse_error,
25 from .brightcove import BrightcoveIE
26 from .ooyala import OoyalaIE
27 from .rutv import RUTVIE
30 class GenericIE(InfoExtractor):
31 IE_DESC = 'Generic downloader that works on some sites'
36 'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
37 'file': '13601338388002.mp4',
38 'md5': '6e15c93721d7ec9e9ca3fdbf07982cfd',
40 'uploader': 'www.hodiho.fr',
41 'title': 'R\u00e9gis plante sa Jeep',
44 # bandcamp page with custom domain
46 'add_ie': ['Bandcamp'],
47 'url': 'http://bronyrock.com/track/the-pony-mash',
48 'file': '3235767654.mp3',
50 'title': 'The Pony Mash',
51 'uploader': 'M_Pallante',
53 'skip': 'There is a limit of 200 free downloads / month for the test song',
55 # embedded brightcove video
56 # it also tests brightcove videos that need to set the 'Referer' in the
59 'add_ie': ['Brightcove'],
60 'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
62 'id': '2765128793001',
64 'title': 'Le cours de bourse : l’analyse technique',
65 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
66 'uploader': 'BFM BUSINESS',
69 'skip_download': True,
73 # https://github.com/rg3/youtube-dl/issues/2253
74 'url': 'http://bcove.me/i6nfkrc3',
75 'file': '3101154703001.mp4',
76 'md5': '0ba9446db037002366bab3b3eb30c88c',
78 'title': 'Still no power',
79 'uploader': 'thestar.com',
80 '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.',
82 'add_ie': ['Brightcove'],
84 # Direct link to a video
86 'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
87 'md5': '67d406c2bcb6af27fa886f31aa934bbe',
92 'upload_date': '20100513',
97 'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
98 'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
100 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
102 'title': '2cc213299525360.mov', # that's what we get
105 # second style of embedded ooyala videos
107 'url': 'http://www.smh.com.au/tv/business/show/financial-review-sunday/behind-the-scenes-financial-review-sunday--4350201.html',
109 'id': '13djJjYjptA1XpPx8r9kuzPyj3UZH0Uk',
111 'title': 'Behind-the-scenes: Financial Review Sunday ',
112 'description': 'Step inside Channel Nine studios for an exclusive tour of its upcoming financial business show.',
116 'skip_download': True,
121 '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',
125 'upload_date': '20130224',
126 'uploader_id': 'TheVerge',
127 '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.',
128 'uploader': 'The Verge',
129 'title': 'First Firefox OS phones side-by-side',
132 'skip_download': False,
137 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
141 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
142 'upload_date': '20140225',
143 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
144 'uploader': 'Tested',
145 'uploader_id': 'testedcom',
147 # No need to test YoutubeIE here
149 'skip_download': True,
154 'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
155 'md5': '7cf780be104d40fea7bae52eed4a470e',
159 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
160 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
165 'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
169 'title': 'Охотское море стало целиком российским',
170 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
174 'skip_download': True,
179 'url': 'http://en.support.wordpress.com/videos/ted-talks/',
180 'md5': 'deeeabcc1085eb2ba205474e7235a3d5',
184 'title': 'My web playroom',
185 'uploader': 'Ze Frank',
186 'description': 'md5:ddb2a40ecd6b6a147e400e535874947b',
189 # nowvideo embed hidden behind percent encoding
191 'url': 'http://www.waoanime.tv/the-super-dimension-fortress-macross-episode-1/',
192 'md5': '2baf4ddd70f697d94b1c18cf796d5107',
194 'id': '06e53103ca9aa',
196 'title': 'Macross Episode 001 Watch Macross Episode 001 onl',
197 'description': 'No description',
202 def report_download_webpage(self, video_id):
203 """Report webpage download."""
204 if not self._downloader.params.get('test', False):
205 self._downloader.report_warning('Falling back on generic information extractor.')
206 super(GenericIE, self).report_download_webpage(video_id)
208 def report_following_redirect(self, new_url):
209 """Report information extraction."""
210 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
212 def _send_head(self, url):
213 """Check if it is a redirect, like url shorteners, in case return the new url."""
215 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
217 Subclass the HTTPRedirectHandler to make it use our
218 HEADRequest also on the redirected URL
220 def redirect_request(self, req, fp, code, msg, headers, newurl):
221 if code in (301, 302, 303, 307):
222 newurl = newurl.replace(' ', '%20')
223 newheaders = dict((k,v) for k,v in req.headers.items()
224 if k.lower() not in ("content-length", "content-type"))
226 # This function was deprecated in python 3.3 and removed in 3.4
227 origin_req_host = req.get_origin_req_host()
228 except AttributeError:
229 origin_req_host = req.origin_req_host
230 return HEADRequest(newurl,
232 origin_req_host=origin_req_host,
235 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
237 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
239 Fallback to GET if HEAD is not allowed (405 HTTP error)
241 def http_error_405(self, req, fp, code, msg, headers):
245 newheaders = dict((k,v) for k,v in req.headers.items()
246 if k.lower() not in ("content-length", "content-type"))
247 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
249 origin_req_host=req.get_origin_req_host(),
253 opener = compat_urllib_request.OpenerDirector()
254 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
255 HTTPMethodFallback, HEADRedirectHandler,
256 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
257 opener.add_handler(handler())
259 response = opener.open(HEADRequest(url))
261 raise ExtractorError('Invalid URL protocol')
264 def _extract_rss(self, url, video_id, doc):
265 playlist_title = doc.find('./channel/title').text
266 playlist_desc_el = doc.find('./channel/description')
267 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
271 'url': e.find('link').text,
272 'title': e.find('title').text,
273 } for e in doc.findall('./channel/item')]
278 'title': playlist_title,
279 'description': playlist_desc,
283 def _real_extract(self, url):
284 parsed_url = compat_urlparse.urlparse(url)
285 if not parsed_url.scheme:
286 default_search = self._downloader.params.get('default_search')
287 if default_search is None:
288 default_search = 'auto'
290 if default_search == 'auto':
292 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
293 return self.url_result('http://' + url)
295 return self.url_result('ytsearch:' + url)
297 assert ':' in default_search
298 return self.url_result(default_search + url)
299 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
301 self.to_screen('%s: Requesting header' % video_id)
304 response = self._send_head(url)
307 new_url = response.geturl()
309 self.report_following_redirect(new_url)
310 return self.url_result(new_url)
312 # Check for direct link to a video
313 content_type = response.headers.get('Content-Type', '')
314 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
316 upload_date = response.headers.get('Last-Modified')
318 upload_date = unified_strdate(upload_date)
321 'title': os.path.splitext(url_basename(url))[0],
323 'format_id': m.group('format_id'),
325 'vcodec': 'none' if m.group('type') == 'audio' else None
327 'upload_date': upload_date,
330 except compat_urllib_error.HTTPError:
331 # This may be a stupid server that doesn't like HEAD, our UA, or so
335 webpage = self._download_webpage(url, video_id)
337 # since this is the last-resort InfoExtractor, if
338 # this error is thrown, it'll be thrown here
339 raise ExtractorError('Failed to download URL: %s' % url)
341 self.report_extraction(video_id)
345 doc = parse_xml(webpage)
347 return self._extract_rss(url, video_id, doc)
348 except compat_xml_parse_error:
351 # Sometimes embedded video player is hidden behind percent encoding
352 # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
353 # Unescaping the whole page allows to handle those cases in a generic way
354 webpage = compat_urllib_parse.unquote(webpage)
356 # it's tempting to parse this further, but you would
357 # have to take into account all the variations like
358 # Video Title - Site Name
359 # Site Name | Video Title
360 # Video Title - Tagline | Site Name
361 # and so on and so forth; it's just not practical
362 video_title = self._html_search_regex(
363 r'(?s)<title>(.*?)</title>', webpage, 'video title',
366 # video uploader is domain name
367 video_uploader = self._search_regex(
368 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
370 # Look for BrightCove:
371 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
373 self.to_screen('Brightcove video detected.')
376 'url': smuggle_url(bc_url, {'Referer': url}),
377 'ie_key': 'Brightcove'
378 } for bc_url in bc_urls]
382 'title': video_title,
387 # Look for embedded (iframe) Vimeo player
389 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
391 player_url = unescapeHTML(mobj.group('url'))
392 surl = smuggle_url(player_url, {'Referer': url})
393 return self.url_result(surl, 'Vimeo')
395 # Look for embedded (swf embed) Vimeo player
397 r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
399 return self.url_result(mobj.group(1), 'Vimeo')
401 # Look for embedded YouTube player
402 matches = re.findall(r'''(?x)
403 (?:<iframe[^>]+?src=|embedSWF\(\s*)
404 (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
408 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
409 for tuppl in matches]
410 return self.playlist_result(
411 urlrs, playlist_id=video_id, playlist_title=video_title)
413 # Look for embedded Dailymotion player
414 matches = re.findall(
415 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
417 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
418 for tuppl in matches]
419 return self.playlist_result(
420 urlrs, playlist_id=video_id, playlist_title=video_title)
422 # Look for embedded Wistia player
424 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
427 '_type': 'url_transparent',
428 'url': unescapeHTML(match.group('url')),
430 'uploader': video_uploader,
431 'title': video_title,
435 # Look for embedded blip.tv player
436 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
438 return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
439 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
441 return self.url_result(mobj.group(1), 'BlipTV')
443 # Look for Bandcamp pages with custom domain
444 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
446 burl = unescapeHTML(mobj.group(1))
447 # Don't set the extractor because it can be a track url or an album
448 return self.url_result(burl)
450 # Look for embedded Vevo player
452 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
454 return self.url_result(mobj.group('url'))
456 # Look for Ooyala videos
457 mobj = (re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or
458 re.search(r'OO.Player.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage))
460 return OoyalaIE._build_url_result(mobj.group('ec'))
462 # Look for Aparat videos
463 mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
465 return self.url_result(mobj.group(1), 'Aparat')
467 # Look for MPORA videos
468 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
470 return self.url_result(mobj.group(1), 'Mpora')
472 # Look for embedded NovaMov player
474 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
476 return self.url_result(mobj.group('url'), 'NovaMov')
478 # Look for embedded NowVideo player
480 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?nowvideo\.(?:ch|sx|eu)/embed\.php.+?)\1', webpage)
482 return self.url_result(mobj.group('url'), 'NowVideo')
484 # Look for embedded Facebook player
486 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
488 return self.url_result(mobj.group('url'), 'Facebook')
490 # Look for embedded VK player
491 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
493 return self.url_result(mobj.group('url'), 'VK')
495 # Look for embedded Huffington Post player
497 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
499 return self.url_result(mobj.group('url'), 'HuffPost')
502 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
504 return self.url_result(mobj.group('url'))
505 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
507 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
509 # Look for funnyordie embed
510 matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
512 urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
514 return self.playlist_result(
515 urlrs, playlist_id=video_id, playlist_title=video_title)
517 # Look for embedded RUTV player
518 rutv_url = RUTVIE._extract_url(webpage)
520 return self.url_result(rutv_url, 'RUTV')
522 # Look for embedded TED player
524 r'<iframe[^>]+?src=(["\'])(?P<url>http://embed\.ted\.com/.+?)\1', webpage)
526 return self.url_result(mobj.group('url'), 'TED')
528 # Start with something easy: JW Player in SWFObject
529 mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
531 # Look for gorilla-vid style embedding
532 mobj = re.search(r'(?s)(?:jw_plugins|JWPlayerOptions).*?file\s*:\s*["\'](.*?)["\']', webpage)
534 # Broaden the search a little bit
535 mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
537 # Broaden the search a little bit: JWPlayer JS loader
538 mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
541 # Try to find twitter cards info
542 mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
544 # We look for Open Graph info:
545 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
546 m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
547 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
548 if m_video_type is not None:
549 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
552 mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
555 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
556 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
559 new_url = mobj.group(1)
560 self.report_following_redirect(new_url)
566 raise ExtractorError('Unsupported URL: %s' % url)
568 # It's possible that one of the regexes
569 # matched, but returned an empty group:
570 if mobj.group(1) is None:
571 raise ExtractorError('Did not find a valid video URL at %s' % url)
573 video_url = mobj.group(1)
574 video_url = compat_urlparse.urljoin(url, video_url)
575 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
577 # Sometimes, jwplayer extraction will result in a YouTube URL
578 if YoutubeIE.suitable(video_url):
579 return self.url_result(video_url, 'Youtube')
581 # here's a fun little line of code for you:
582 video_id = os.path.splitext(video_id)[0]
587 'uploader': video_uploader,
588 'title': video_title,