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
107 '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',
111 'upload_date': '20130224',
112 'uploader_id': 'TheVerge',
113 '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.',
114 'uploader': 'The Verge',
115 'title': 'First Firefox OS phones side-by-side',
118 'skip_download': False,
123 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
127 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
128 'upload_date': '20140225',
129 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
130 'uploader': 'Tested',
131 'uploader_id': 'testedcom',
133 # No need to test YoutubeIE here
135 'skip_download': True,
140 'url': 'http://www.theguardian.com/world/2014/mar/11/obama-zach-galifianakis-between-two-ferns',
141 'md5': '7cf780be104d40fea7bae52eed4a470e',
145 'title': 'Between Two Ferns with Zach Galifianakis: President Barack Obama',
146 'description': 'Episode 18: President Barack Obama sits down with Zach Galifianakis for his most memorable interview yet.',
151 'url': 'http://www.rg.ru/2014/03/15/reg-dfo/anklav-anons.html',
155 'title': 'Охотское море стало целиком российским',
156 'description': 'md5:5ed62483b14663e2a95ebbe115eb8f43',
160 'skip_download': True,
165 def report_download_webpage(self, video_id):
166 """Report webpage download."""
167 if not self._downloader.params.get('test', False):
168 self._downloader.report_warning('Falling back on generic information extractor.')
169 super(GenericIE, self).report_download_webpage(video_id)
171 def report_following_redirect(self, new_url):
172 """Report information extraction."""
173 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
175 def _send_head(self, url):
176 """Check if it is a redirect, like url shorteners, in case return the new url."""
178 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
180 Subclass the HTTPRedirectHandler to make it use our
181 HEADRequest also on the redirected URL
183 def redirect_request(self, req, fp, code, msg, headers, newurl):
184 if code in (301, 302, 303, 307):
185 newurl = newurl.replace(' ', '%20')
186 newheaders = dict((k,v) for k,v in req.headers.items()
187 if k.lower() not in ("content-length", "content-type"))
189 # This function was deprecated in python 3.3 and removed in 3.4
190 origin_req_host = req.get_origin_req_host()
191 except AttributeError:
192 origin_req_host = req.origin_req_host
193 return HEADRequest(newurl,
195 origin_req_host=origin_req_host,
198 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
200 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
202 Fallback to GET if HEAD is not allowed (405 HTTP error)
204 def http_error_405(self, req, fp, code, msg, headers):
208 newheaders = dict((k,v) for k,v in req.headers.items()
209 if k.lower() not in ("content-length", "content-type"))
210 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
212 origin_req_host=req.get_origin_req_host(),
216 opener = compat_urllib_request.OpenerDirector()
217 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
218 HTTPMethodFallback, HEADRedirectHandler,
219 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
220 opener.add_handler(handler())
222 response = opener.open(HEADRequest(url))
224 raise ExtractorError('Invalid URL protocol')
227 def _extract_rss(self, url, video_id, doc):
228 playlist_title = doc.find('./channel/title').text
229 playlist_desc_el = doc.find('./channel/description')
230 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
234 'url': e.find('link').text,
235 'title': e.find('title').text,
236 } for e in doc.findall('./channel/item')]
241 'title': playlist_title,
242 'description': playlist_desc,
246 def _real_extract(self, url):
247 parsed_url = compat_urlparse.urlparse(url)
248 if not parsed_url.scheme:
249 default_search = self._downloader.params.get('default_search')
250 if default_search is None:
251 default_search = 'auto'
253 if default_search == 'auto':
255 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
256 return self.url_result('http://' + url)
258 return self.url_result('ytsearch:' + url)
260 assert ':' in default_search
261 return self.url_result(default_search + url)
262 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
264 self.to_screen('%s: Requesting header' % video_id)
267 response = self._send_head(url)
270 new_url = response.geturl()
272 self.report_following_redirect(new_url)
273 return self.url_result(new_url)
275 # Check for direct link to a video
276 content_type = response.headers.get('Content-Type', '')
277 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
279 upload_date = response.headers.get('Last-Modified')
281 upload_date = unified_strdate(upload_date)
284 'title': os.path.splitext(url_basename(url))[0],
286 'format_id': m.group('format_id'),
288 'vcodec': 'none' if m.group('type') == 'audio' else None
290 'upload_date': upload_date,
293 except compat_urllib_error.HTTPError:
294 # This may be a stupid server that doesn't like HEAD, our UA, or so
298 webpage = self._download_webpage(url, video_id)
300 # since this is the last-resort InfoExtractor, if
301 # this error is thrown, it'll be thrown here
302 raise ExtractorError('Failed to download URL: %s' % url)
304 self.report_extraction(video_id)
308 doc = parse_xml(webpage)
310 return self._extract_rss(url, video_id, doc)
311 except compat_xml_parse_error:
314 # it's tempting to parse this further, but you would
315 # have to take into account all the variations like
316 # Video Title - Site Name
317 # Site Name | Video Title
318 # Video Title - Tagline | Site Name
319 # and so on and so forth; it's just not practical
320 video_title = self._html_search_regex(
321 r'(?s)<title>(.*?)</title>', webpage, 'video title',
324 # video uploader is domain name
325 video_uploader = self._search_regex(
326 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
328 # Look for BrightCove:
329 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
331 self.to_screen('Brightcove video detected.')
334 'url': smuggle_url(bc_url, {'Referer': url}),
335 'ie_key': 'Brightcove'
336 } for bc_url in bc_urls]
340 'title': video_title,
345 # Look for embedded (iframe) Vimeo player
347 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//player\.vimeo\.com/video/.+?)\1', webpage)
349 player_url = unescapeHTML(mobj.group('url'))
350 surl = smuggle_url(player_url, {'Referer': url})
351 return self.url_result(surl, 'Vimeo')
353 # Look for embedded (swf embed) Vimeo player
355 r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
357 return self.url_result(mobj.group(1), 'Vimeo')
359 # Look for embedded YouTube player
360 matches = re.findall(r'''(?x)
361 (?:<iframe[^>]+?src=|embedSWF\(\s*)
362 (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
366 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
367 for tuppl in matches]
368 return self.playlist_result(
369 urlrs, playlist_id=video_id, playlist_title=video_title)
371 # Look for embedded Dailymotion player
372 matches = re.findall(
373 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
375 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
376 for tuppl in matches]
377 return self.playlist_result(
378 urlrs, playlist_id=video_id, playlist_title=video_title)
380 # Look for embedded Wistia player
382 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
385 '_type': 'url_transparent',
386 'url': unescapeHTML(match.group('url')),
388 'uploader': video_uploader,
389 'title': video_title,
393 # Look for embedded blip.tv player
394 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
396 return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
397 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
399 return self.url_result(mobj.group(1), 'BlipTV')
401 # Look for Bandcamp pages with custom domain
402 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
404 burl = unescapeHTML(mobj.group(1))
405 # Don't set the extractor because it can be a track url or an album
406 return self.url_result(burl)
408 # Look for embedded Vevo player
410 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
412 return self.url_result(mobj.group('url'))
414 # Look for Ooyala videos
415 mobj = re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=([^"&]+)', webpage)
417 return OoyalaIE._build_url_result(mobj.group(1))
419 # Look for Aparat videos
420 mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
422 return self.url_result(mobj.group(1), 'Aparat')
424 # Look for MPORA videos
425 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
427 return self.url_result(mobj.group(1), 'Mpora')
429 # Look for embedded NovaMov player
431 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
433 return self.url_result(mobj.group('url'), 'NovaMov')
435 # Look for embedded NowVideo player
437 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?nowvideo\.(?:ch|sx|eu)/embed\.php.+?)\1', webpage)
439 return self.url_result(mobj.group('url'), 'NowVideo')
441 # Look for embedded Facebook player
443 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
445 return self.url_result(mobj.group('url'), 'Facebook')
447 # Look for embedded VK player
448 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
450 return self.url_result(mobj.group('url'), 'VK')
452 # Look for embedded Huffington Post player
454 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
456 return self.url_result(mobj.group('url'), 'HuffPost')
459 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
461 return self.url_result(mobj.group('url'))
462 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
464 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
466 # Look for funnyordie embed
467 matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
469 urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
471 return self.playlist_result(
472 urlrs, playlist_id=video_id, playlist_title=video_title)
474 # Look for embedded RUTV player
475 rutv_url = RUTVIE._extract_url(webpage)
477 return self.url_result(rutv_url, 'RUTV')
479 # Start with something easy: JW Player in SWFObject
480 mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
482 # Look for gorilla-vid style embedding
483 mobj = re.search(r'(?s)(?:jw_plugins|JWPlayerOptions).*?file\s*:\s*["\'](.*?)["\']', webpage)
485 # Broaden the search a little bit
486 mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
488 # Broaden the search a little bit: JWPlayer JS loader
489 mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
491 # Try to find twitter cards info
492 mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
494 # We look for Open Graph info:
495 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
496 m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
497 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
498 if m_video_type is not None:
499 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
502 mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
505 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
506 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
509 new_url = mobj.group(1)
510 self.report_following_redirect(new_url)
516 raise ExtractorError('Unsupported URL: %s' % url)
518 # It's possible that one of the regexes
519 # matched, but returned an empty group:
520 if mobj.group(1) is None:
521 raise ExtractorError('Did not find a valid video URL at %s' % url)
523 video_url = mobj.group(1)
524 video_url = compat_urlparse.urljoin(url, video_url)
525 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
527 # Sometimes, jwplayer extraction will result in a YouTube URL
528 if YoutubeIE.suitable(video_url):
529 return self.url_result(video_url, 'Youtube')
531 # here's a fun little line of code for you:
532 video_id = os.path.splitext(video_id)[0]
537 'uploader': video_uploader,
538 'title': video_title,