X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/33cd347759d6d999325ebf3c69b7ed5692c343b2..c91254967656b8722087896c531ed6d815cbe670:/youtube_dl/extractor/generic.py diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py index b633e89..10ae062 100644 --- a/youtube_dl/extractor/generic.py +++ b/youtube_dl/extractor/generic.py @@ -8,11 +8,15 @@ from ..utils import ( compat_urllib_error, compat_urllib_parse, compat_urllib_request, + compat_urlparse, ExtractorError, + smuggle_url, + unescapeHTML, ) from .brightcove import BrightcoveIE + class GenericIE(InfoExtractor): IE_DESC = u'Generic downloader that works on some sites' _VALID_URL = r'.*' @@ -21,23 +25,52 @@ class GenericIE(InfoExtractor): { u'url': u'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html', u'file': u'13601338388002.mp4', - u'md5': u'85b90ccc9d73b4acd9138d3af4c27f89', + u'md5': u'6e15c93721d7ec9e9ca3fdbf07982cfd', u'info_dict': { - u"uploader": u"www.hodiho.fr", + u"uploader": u"www.hodiho.fr", u"title": u"R\u00e9gis plante sa Jeep" } }, + # embedded vimeo video { - u'url': u'http://www.8tv.cat/8aldia/videos/xavier-sala-i-martin-aquesta-tarda-a-8-al-dia/', - u'file': u'2371591881001.mp4', - u'md5': u'9e80619e0a94663f0bdc849b4566af19', - u'note': u'Test Brightcove downloads and detection in GenericIE', + 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', u'info_dict': { - u'title': u'Xavier Sala i Martín: “Un banc que no presta és un banc zombi que no serveix per a res”', - u'uploader': u'8TV', - u'description': u'md5:a950cc4285c43e44d763d036710cd9cd', + u'title': u'ACCU 2011: Move Semantics,Perfect Forwarding, and Rvalue references- Scott Meyers- 13/04/2011', + u"uploader_id": u"skillsmatter", + u"uploader": u"Skills Matter", } }, + # 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': { + u'title': u'The Pony Mash', + u'uploader': u'M_Pallante', + }, + 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, + }, + }, ] def report_download_webpage(self, video_id): @@ -107,8 +140,18 @@ class GenericIE(InfoExtractor): return new_url def _real_extract(self, url): - new_url = self._test_redirect(url) - if new_url: return [self.url_result(new_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) + + try: + new_url = self._test_redirect(url) + if new_url: + return [self.url_result(new_url)] + 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: @@ -116,16 +159,58 @@ class GenericIE(InfoExtractor): except ValueError: # since this is the last-resort InfoExtractor, if # this error is thrown, it'll be thrown here - raise ExtractorError(u'Invalid URL: %s' % url) + raise ExtractorError(u'Failed to download URL: %s' % url) self.report_extraction(video_id) - # Look for BrigthCove: - m_brightcove = re.search(r'', webpage, re.DOTALL) - if m_brightcove is not None: + + # 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'(.*)', + webpage, u'video title', default=u'video', flags=re.DOTALL) + + # Look for BrightCove: + 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 + mobj = re.search( + r']+?src="(https?://player.vimeo.com/video/.+?)"', webpage) + if mobj: + player_url = unescapeHTML(mobj.group(1)) + surl = smuggle_url(player_url, {'Referer': url}) + return self.url_result(surl, 'Vimeo') + + # Look for embedded YouTube player + matches = re.findall( + r']+?src=(["\'])(?P(?:https?:)?//(?:www\.)?youtube.com/embed/.+?)\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 Bandcamp pages with custom domain + mobj = re.search(r']*?content="(.*?bandcamp\.com.*?)"', webpage) + if mobj is not None: + burl = unescapeHTML(mobj.group(1)) + # Don't set the extractor because it can be a track url or an album + return self.url_result(burl) + # Start with something easy: JW Player in SWFObject mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage) if mobj is None: @@ -133,7 +218,7 @@ class GenericIE(InfoExtractor): mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage) if mobj is None: # Broaden the search a little bit: JWPlayer JS loader - mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"&]*)', webpage) + mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http[^\'"]*)', webpage) if mobj is None: # Try to find twitter cards info mobj = re.search(r'.*?(.*)', - webpage, u'video title', default=u'video', flags=re.DOTALL) - # video uploader is domain name video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*', url, u'video uploader') - return [{ + return { 'id': video_id, 'url': video_url, 'uploader': video_uploader, 'upload_date': None, 'title': video_title, - 'ext': video_extension, - }] + }