X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fgeneric.py;h=57a6b1820c90891a9c9c138e27ff042155f36670;hb=aff24732b96b5ec89cb41a05fd132e12c5990caf;hp=2c8fcf5ae5df24a1dedc5e461feb5ac2300688a6;hpb=38db46794f3ccfef09094db9b411e55acd4c1a3d;p=youtube-dl diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py index 2c8fcf5ae..57a6b1820 100644 --- a/youtube_dl/extractor/generic.py +++ b/youtube_dl/extractor/generic.py @@ -11,10 +11,14 @@ from ..utils import ( compat_urlparse, ExtractorError, + HEADRequest, smuggle_url, unescapeHTML, + unified_strdate, + url_basename, ) from .brightcove import BrightcoveIE +from .ooyala import OoyalaIE class GenericIE(InfoExtractor): @@ -33,6 +37,7 @@ class GenericIE(InfoExtractor): }, # embedded vimeo video { + u'add_ie': ['Vimeo'], u'url': u'http://skillsmatter.com/podcast/home/move-semanticsperfect-forwarding-and-rvalue-references', u'file': u'22444065.mp4', u'md5': u'2903896e23df39722c33f015af0666e2', @@ -44,6 +49,7 @@ class GenericIE(InfoExtractor): }, # bandcamp page with custom domain { + u'add_ie': ['Bandcamp'], u'url': u'http://bronyrock.com/track/the-pony-mash', u'file': u'3235767654.mp3', u'info_dict': { @@ -52,6 +58,44 @@ class GenericIE(InfoExtractor): }, u'skip': u'There is a limit of 200 free downloads / month for the test song', }, + # embedded brightcove video + # it also tests brightcove videos that need to set the 'Referer' in the + # http requests + { + u'add_ie': ['Brightcove'], + u'url': u'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/', + u'info_dict': { + u'id': u'2765128793001', + u'ext': u'mp4', + u'title': u'Le cours de bourse : l’analyse technique', + u'description': u'md5:7e9ad046e968cb2d1114004aba466fd9', + u'uploader': u'BFM BUSINESS', + }, + u'params': { + u'skip_download': True, + }, + }, + # Direct link to a video + { + u'url': u'http://media.w3.org/2010/05/sintel/trailer.mp4', + u'file': u'trailer.mp4', + u'md5': u'67d406c2bcb6af27fa886f31aa934bbe', + u'info_dict': { + u'id': u'trailer', + u'title': u'trailer', + u'upload_date': u'20100513', + } + }, + # ooyala video + { + u'url': u'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219', + u'md5': u'5644c6ca5d5782c1d0d350dad9bd840c', + u'info_dict': { + u'id': u'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ', + u'ext': u'mp4', + u'title': u'2cc213299525360.mov', #that's what we get + }, + }, ] def report_download_webpage(self, video_id): @@ -64,23 +108,20 @@ class GenericIE(InfoExtractor): """Report information extraction.""" self._downloader.to_screen(u'[redirect] Following redirect to %s' % new_url) - def _test_redirect(self, url): + def _send_head(self, url): """Check if it is a redirect, like url shorteners, in case return the new url.""" - class HeadRequest(compat_urllib_request.Request): - def get_method(self): - return "HEAD" class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler): """ Subclass the HTTPRedirectHandler to make it use our - HeadRequest also on the redirected URL + HEADRequest also on the redirected URL """ def redirect_request(self, req, fp, code, msg, headers, newurl): if code in (301, 302, 303, 307): newurl = newurl.replace(' ', '%20') newheaders = dict((k,v) for k,v in req.headers.items() if k.lower() not in ("content-length", "content-type")) - return HeadRequest(newurl, + return HEADRequest(newurl, headers=newheaders, origin_req_host=req.get_origin_req_host(), unverifiable=True) @@ -109,32 +150,51 @@ class GenericIE(InfoExtractor): compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]: opener.add_handler(handler()) - response = opener.open(HeadRequest(url)) + response = opener.open(HEADRequest(url)) if response is None: raise ExtractorError(u'Invalid URL protocol') - new_url = response.geturl() - - if url == new_url: - return False - - self.report_following_redirect(new_url) - return new_url + return response def _real_extract(self, url): parsed_url = compat_urlparse.urlparse(url) if not parsed_url.scheme: self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http') return self.url_result('http://' + url) + video_id = os.path.splitext(url.split('/')[-1])[0] + + self.to_screen(u'%s: Requesting header' % video_id) try: - new_url = self._test_redirect(url) - if new_url: - return [self.url_result(new_url)] + response = self._send_head(url) + + # Check for redirect + new_url = response.geturl() + if url != new_url: + self.report_following_redirect(new_url) + return self.url_result(new_url) + + # Check for direct link to a video + content_type = response.headers.get('Content-Type', '') + m = re.match(r'^(?Paudio|video|application(?=/ogg$))/(?P.+)$', content_type) + if m: + upload_date = response.headers.get('Last-Modified') + if upload_date: + upload_date = unified_strdate(upload_date) + return { + 'id': video_id, + 'title': os.path.splitext(url_basename(url))[0], + 'formats': [{ + 'format_id': m.group('format_id'), + 'url': url, + 'vcodec': u'none' if m.group('type') == 'audio' else None + }], + 'upload_date': upload_date, + } + except compat_urllib_error.HTTPError: # This may be a stupid server that doesn't like HEAD, our UA, or so pass - video_id = url.split('/')[-1] try: webpage = self._download_webpage(url, video_id) except ValueError: @@ -143,14 +203,28 @@ class GenericIE(InfoExtractor): raise ExtractorError(u'Failed to download URL: %s' % url) self.report_extraction(video_id) + + # it's tempting to parse this further, but you would + # have to take into account all the variations like + # Video Title - Site Name + # Site Name | Video Title + # Video Title - Tagline | Site Name + # and so on and so forth; it's just not practical + video_title = self._html_search_regex( + r'(?s)(.*?)', webpage, u'video title', + default=u'video') + + # video uploader is domain name + video_uploader = self._search_regex( + r'^(?:https?://)?([^/]*)/.*', url, u'video uploader') + # Look for BrightCove: - m_brightcove = re.search(r']+?class=([\'"])[^>]*?BrightcoveExperience.*?\1.+?', webpage, re.DOTALL) - if m_brightcove is not None: + bc_url = BrightcoveIE._extract_brightcove_url(webpage) + if bc_url is not None: self.to_screen(u'Brightcove video detected.') - bc_url = BrightcoveIE._build_brighcove_url(m_brightcove.group()) return self.url_result(bc_url, 'Brightcove') - # Look for embedded Vimeo player + # Look for embedded (iframe) Vimeo player mobj = re.search( r']+?src="(https?://player.vimeo.com/video/.+?)"', webpage) if mobj: @@ -158,18 +232,76 @@ class GenericIE(InfoExtractor): surl = smuggle_url(player_url, {'Referer': url}) return self.url_result(surl, 'Vimeo') - # Look for embedded YouTube player + # Look for embedded (swf embed) Vimeo player mobj = re.search( - r']+?src="(https?://(?:www\.)?youtube.com/embed/.+?)"', webpage) + r']+?src="(https?://(?:www\.)?vimeo.com/moogaloop.swf.+?)"', webpage) + if mobj: + return self.url_result(mobj.group(1), 'Vimeo') + + # Look for embedded YouTube player + matches = re.findall(r'''(?x) + (?:]+?src=|embedSWF\(\s*) + (["\'])(?P(?:https?:)?//(?:www\.)?youtube\.com/ + (?:embed|v)/.+?) + \1''', webpage) + if matches: + urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube') + for tuppl in matches] + return self.playlist_result( + urlrs, playlist_id=video_id, playlist_title=video_title) + + # Look for embedded Dailymotion player + matches = re.findall( + r']+?src=(["\'])(?P(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage) + if matches: + urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion') + for tuppl in matches] + return self.playlist_result( + urlrs, playlist_id=video_id, playlist_title=video_title) + + # Look for embedded Wistia player + match = re.search( + r']+?src=(["\'])(?P(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage) + if match: + return { + '_type': 'url_transparent', + 'url': unescapeHTML(match.group('url')), + 'ie_key': 'Wistia', + 'uploader': video_uploader, + 'title': video_title, + 'id': video_id, + } + + # Look for embedded blip.tv player + mobj = re.search(r']*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage) + if mobj: + return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV') + mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage) if mobj: - surl = unescapeHTML(mobj.group(1)) - return self.url_result(surl, 'Youtube') + return self.url_result(mobj.group(1), 'BlipTV') # Look for Bandcamp pages with custom domain mobj = re.search(r']*?content="(.*?bandcamp\.com.*?)"', webpage) if mobj is not None: burl = unescapeHTML(mobj.group(1)) - return self.url_result(burl, 'Bandcamp') + # Don't set the extractor because it can be a track url or an album + return self.url_result(burl) + + # Look for embedded Vevo player + mobj = re.search( + r']+?src=(["\'])(?P(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage) + if mobj is not None: + return self.url_result(mobj.group('url')) + + # Look for Ooyala videos + mobj = re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=([^"&]+)', webpage) + if mobj is not None: + return OoyalaIE._build_url_result(mobj.group(1)) + + # Look for Aparat videos + mobj = re.search(r'