3 from __future__ import unicode_literals
7 import xml.etree.ElementTree
9 from .common import InfoExtractor
10 from .youtube import YoutubeIE
14 compat_urllib_request,
16 compat_xml_parse_error,
25 from .brightcove import BrightcoveIE
26 from .ooyala import OoyalaIE
29 class GenericIE(InfoExtractor):
30 IE_DESC = 'Generic downloader that works on some sites'
35 'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
36 'file': '13601338388002.mp4',
37 'md5': '6e15c93721d7ec9e9ca3fdbf07982cfd',
39 'uploader': 'www.hodiho.fr',
40 'title': 'R\u00e9gis plante sa Jeep',
43 # bandcamp page with custom domain
45 'add_ie': ['Bandcamp'],
46 'url': 'http://bronyrock.com/track/the-pony-mash',
47 'file': '3235767654.mp3',
49 'title': 'The Pony Mash',
50 'uploader': 'M_Pallante',
52 'skip': 'There is a limit of 200 free downloads / month for the test song',
54 # embedded brightcove video
55 # it also tests brightcove videos that need to set the 'Referer' in the
58 'add_ie': ['Brightcove'],
59 'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
61 'id': '2765128793001',
63 'title': 'Le cours de bourse : l’analyse technique',
64 'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
65 'uploader': 'BFM BUSINESS',
68 'skip_download': True,
72 # https://github.com/rg3/youtube-dl/issues/2253
73 'url': 'http://bcove.me/i6nfkrc3',
74 'file': '3101154703001.mp4',
75 'md5': '0ba9446db037002366bab3b3eb30c88c',
77 'title': 'Still no power',
78 'uploader': 'thestar.com',
79 '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.',
81 'add_ie': ['Brightcove'],
83 # Direct link to a video
85 'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
86 'md5': '67d406c2bcb6af27fa886f31aa934bbe',
91 'upload_date': '20100513',
96 'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
97 'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
99 'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
101 'title': '2cc213299525360.mov', # that's what we get
106 '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',
110 'upload_date': '20130224',
111 'uploader_id': 'TheVerge',
112 '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.',
113 'uploader': 'The Verge',
114 'title': 'First Firefox OS phones side-by-side',
117 'skip_download': False,
122 'url': 'http://www.tested.com/science/weird/460206-tested-grinding-coffee-2000-frames-second/',
126 'title': 'Tested: Grinding Coffee at 2000 Frames Per Second',
127 'upload_date': '20140225',
128 'description': 'md5:06a40fbf30b220468f1e0957c0f558ff',
129 'uploader': 'Tested',
130 'uploader_id': 'testedcom',
132 # No need to test YoutubeIE here
134 'skip_download': True,
139 def report_download_webpage(self, video_id):
140 """Report webpage download."""
141 if not self._downloader.params.get('test', False):
142 self._downloader.report_warning('Falling back on generic information extractor.')
143 super(GenericIE, self).report_download_webpage(video_id)
145 def report_following_redirect(self, new_url):
146 """Report information extraction."""
147 self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
149 def _send_head(self, url):
150 """Check if it is a redirect, like url shorteners, in case return the new url."""
152 class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
154 Subclass the HTTPRedirectHandler to make it use our
155 HEADRequest also on the redirected URL
157 def redirect_request(self, req, fp, code, msg, headers, newurl):
158 if code in (301, 302, 303, 307):
159 newurl = newurl.replace(' ', '%20')
160 newheaders = dict((k,v) for k,v in req.headers.items()
161 if k.lower() not in ("content-length", "content-type"))
162 return HEADRequest(newurl,
164 origin_req_host=req.get_origin_req_host(),
167 raise compat_urllib_error.HTTPError(req.get_full_url(), code, msg, headers, fp)
169 class HTTPMethodFallback(compat_urllib_request.BaseHandler):
171 Fallback to GET if HEAD is not allowed (405 HTTP error)
173 def http_error_405(self, req, fp, code, msg, headers):
177 newheaders = dict((k,v) for k,v in req.headers.items()
178 if k.lower() not in ("content-length", "content-type"))
179 return self.parent.open(compat_urllib_request.Request(req.get_full_url(),
181 origin_req_host=req.get_origin_req_host(),
185 opener = compat_urllib_request.OpenerDirector()
186 for handler in [compat_urllib_request.HTTPHandler, compat_urllib_request.HTTPDefaultErrorHandler,
187 HTTPMethodFallback, HEADRedirectHandler,
188 compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
189 opener.add_handler(handler())
191 response = opener.open(HEADRequest(url))
193 raise ExtractorError('Invalid URL protocol')
196 def _extract_rss(self, url, video_id, doc):
197 playlist_title = doc.find('./channel/title').text
198 playlist_desc_el = doc.find('./channel/description')
199 playlist_desc = None if playlist_desc_el is None else playlist_desc_el.text
203 'url': e.find('link').text,
204 'title': e.find('title').text,
205 } for e in doc.findall('./channel/item')]
210 'title': playlist_title,
211 'description': playlist_desc,
215 def _real_extract(self, url):
216 parsed_url = compat_urlparse.urlparse(url)
217 if not parsed_url.scheme:
218 default_search = self._downloader.params.get('default_search')
219 if default_search is None:
220 default_search = 'auto'
222 if default_search == 'auto':
224 self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
225 return self.url_result('http://' + url)
227 return self.url_result('ytsearch:' + url)
229 assert ':' in default_search
230 return self.url_result(default_search + url)
231 video_id = os.path.splitext(url.rstrip('/').split('/')[-1])[0]
233 self.to_screen('%s: Requesting header' % video_id)
236 response = self._send_head(url)
239 new_url = response.geturl()
241 self.report_following_redirect(new_url)
242 return self.url_result(new_url)
244 # Check for direct link to a video
245 content_type = response.headers.get('Content-Type', '')
246 m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
248 upload_date = response.headers.get('Last-Modified')
250 upload_date = unified_strdate(upload_date)
253 'title': os.path.splitext(url_basename(url))[0],
255 'format_id': m.group('format_id'),
257 'vcodec': 'none' if m.group('type') == 'audio' else None
259 'upload_date': upload_date,
262 except compat_urllib_error.HTTPError:
263 # This may be a stupid server that doesn't like HEAD, our UA, or so
267 webpage = self._download_webpage(url, video_id)
269 # since this is the last-resort InfoExtractor, if
270 # this error is thrown, it'll be thrown here
271 raise ExtractorError('Failed to download URL: %s' % url)
273 self.report_extraction(video_id)
277 doc = xml.etree.ElementTree.fromstring(webpage.encode('utf-8'))
279 return self._extract_rss(url, video_id, doc)
280 except compat_xml_parse_error:
283 # it's tempting to parse this further, but you would
284 # have to take into account all the variations like
285 # Video Title - Site Name
286 # Site Name | Video Title
287 # Video Title - Tagline | Site Name
288 # and so on and so forth; it's just not practical
289 video_title = self._html_search_regex(
290 r'(?s)<title>(.*?)</title>', webpage, 'video title',
293 # video uploader is domain name
294 video_uploader = self._search_regex(
295 r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
297 # Look for BrightCove:
298 bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
300 self.to_screen('Brightcove video detected.')
303 'url': smuggle_url(bc_url, {'Referer': url}),
304 'ie_key': 'Brightcove'
305 } for bc_url in bc_urls]
309 'title': video_title,
314 # Look for embedded (iframe) Vimeo player
316 r'<iframe[^>]+?src="((?:https?:)?//player\.vimeo\.com/video/.+?)"', webpage)
318 player_url = unescapeHTML(mobj.group(1))
319 surl = smuggle_url(player_url, {'Referer': url})
320 return self.url_result(surl, 'Vimeo')
322 # Look for embedded (swf embed) Vimeo player
324 r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
326 return self.url_result(mobj.group(1), 'Vimeo')
328 # Look for embedded YouTube player
329 matches = re.findall(r'''(?x)
330 (?:<iframe[^>]+?src=|embedSWF\(\s*)
331 (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
335 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
336 for tuppl in matches]
337 return self.playlist_result(
338 urlrs, playlist_id=video_id, playlist_title=video_title)
340 # Look for embedded Dailymotion player
341 matches = re.findall(
342 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
344 urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
345 for tuppl in matches]
346 return self.playlist_result(
347 urlrs, playlist_id=video_id, playlist_title=video_title)
349 # Look for embedded Wistia player
351 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
354 '_type': 'url_transparent',
355 'url': unescapeHTML(match.group('url')),
357 'uploader': video_uploader,
358 'title': video_title,
362 # Look for embedded blip.tv player
363 mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
365 return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
366 mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
368 return self.url_result(mobj.group(1), 'BlipTV')
370 # Look for Bandcamp pages with custom domain
371 mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
373 burl = unescapeHTML(mobj.group(1))
374 # Don't set the extractor because it can be a track url or an album
375 return self.url_result(burl)
377 # Look for embedded Vevo player
379 r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
381 return self.url_result(mobj.group('url'))
383 # Look for Ooyala videos
384 mobj = re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=([^"&]+)', webpage)
386 return OoyalaIE._build_url_result(mobj.group(1))
388 # Look for Aparat videos
389 mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
391 return self.url_result(mobj.group(1), 'Aparat')
393 # Look for MPORA videos
394 mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
396 return self.url_result(mobj.group(1), 'Mpora')
398 # Look for embedded NovaMov player
400 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
402 return self.url_result(mobj.group('url'), 'NovaMov')
404 # Look for embedded NowVideo player
406 r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?nowvideo\.(?:ch|sx|eu)/embed\.php.+?)\1', webpage)
408 return self.url_result(mobj.group('url'), 'NowVideo')
410 # Look for embedded Facebook player
412 r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
414 return self.url_result(mobj.group('url'), 'Facebook')
416 # Look for embedded VK player
417 mobj = re.search(r'<iframe[^>]+?src=(["\'])(?P<url>https?://vk\.com/video_ext\.php.+?)\1', webpage)
419 return self.url_result(mobj.group('url'), 'VK')
421 # Look for embedded Huffington Post player
423 r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
425 return self.url_result(mobj.group('url'), 'HuffPost')
428 mobj = re.search(r'class=["\']embedly-card["\'][^>]href=["\'](?P<url>[^"\']+)', webpage)
430 return self.url_result(mobj.group('url'))
431 mobj = re.search(r'class=["\']embedly-embed["\'][^>]src=["\'][^"\']*url=(?P<url>[^&]+)', webpage)
433 return self.url_result(compat_urllib_parse.unquote(mobj.group('url')))
435 # Start with something easy: JW Player in SWFObject
436 mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
438 # Look for gorilla-vid style embedding
439 mobj = re.search(r'(?s)(?:jw_plugins|JWPlayerOptions).*?file\s*:\s*["\'](.*?)["\']', webpage)
441 # Broaden the search a little bit
442 mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
444 # Broaden the search a little bit: JWPlayer JS loader
445 mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
447 # Try to find twitter cards info
448 mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
450 # We look for Open Graph info:
451 # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
452 m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
453 # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
454 if m_video_type is not None:
455 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
458 mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
461 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
462 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
465 new_url = mobj.group(1)
466 self.report_following_redirect(new_url)
472 raise ExtractorError('Unsupported URL: %s' % url)
474 # It's possible that one of the regexes
475 # matched, but returned an empty group:
476 if mobj.group(1) is None:
477 raise ExtractorError('Did not find a valid video URL at %s' % url)
479 video_url = mobj.group(1)
480 video_url = compat_urlparse.urljoin(url, video_url)
481 video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
483 # Sometimes, jwplayer extraction will result in a YouTube URL
484 if YoutubeIE.suitable(video_url):
485 return self.url_result(video_url, 'Youtube')
487 # here's a fun little line of code for you:
488 video_id = os.path.splitext(video_id)[0]
493 'uploader': video_uploader,
494 'title': video_title,