X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Farte.py;h=9591bad8a66254e90247a204b43b10b6db4f6406;hb=b1d65c33695feec76a801b7647fe8077869c25ad;hp=646377e4b0f16c9b3628987b51c3ff530c12b297;hpb=b84d6e7fc42affddeb1baf989cf394fedc41a96d;p=youtube-dl diff --git a/youtube_dl/extractor/arte.py b/youtube_dl/extractor/arte.py index 646377e4b..9591bad8a 100644 --- a/youtube_dl/extractor/arte.py +++ b/youtube_dl/extractor/arte.py @@ -39,7 +39,10 @@ class ArteTvIE(InfoExtractor): formats = [{ 'forma_id': q.attrib['quality'], - 'url': q.text, + # The playpath starts at 'mp4:', if we don't manually + # split the url, rtmpdump will incorrectly parse them + 'url': q.text.split('mp4:', 1)[0], + 'play_path': 'mp4:' + q.text.split('mp4:', 1)[1], 'ext': 'flv', 'quality': 2 if q.attrib['quality'] == 'hd' else 1, } for q in config.findall('./urls/url')] @@ -74,7 +77,8 @@ class ArteTVPlus7IE(InfoExtractor): return self._extract_from_webpage(webpage, video_id, lang) def _extract_from_webpage(self, webpage, video_id, lang): - json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url') + json_url = self._html_search_regex( + r'arte_vp_url="(.*?)"', webpage, 'json vp url') return self._extract_from_json_url(json_url, video_id, lang) def _extract_from_json_url(self, json_url, video_id, lang): @@ -110,7 +114,7 @@ class ArteTVPlus7IE(InfoExtractor): if not formats: # Some videos are only available in the 'Originalversion' # they aren't tagged as being in French or German - if all(f['versionCode'] == 'VO' for f in all_formats): + if all(f['versionCode'] == 'VO' or f['versionCode'] == 'VA' for f in all_formats): formats = all_formats else: raise ExtractorError(u'The formats list is empty') @@ -120,14 +124,17 @@ class ArteTVPlus7IE(InfoExtractor): return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality']) else: def sort_key(f): + versionCode = f.get('versionCode') + if versionCode is None: + versionCode = '' return ( # Sort first by quality - int(f.get('height',-1)), - int(f.get('bitrate',-1)), + int(f.get('height', -1)), + int(f.get('bitrate', -1)), # The original version with subtitles has lower relevance - re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None, + re.match(r'VO-ST(F|A)', versionCode) is None, # The version with sourds/mal subtitles has also lower relevance - re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None, + re.match(r'VO?(F|A)-STM\1', versionCode) is None, # Prefer http downloads over m3u8 0 if f['url'].endswith('m3u8') else 1, ) @@ -185,9 +192,10 @@ class ArteTVFutureIE(ArteTVPlus7IE): _TEST = { 'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081', 'info_dict': { - 'id': '050940-003', + 'id': '5201', 'ext': 'mp4', 'title': 'Les champignons au secours de la planète', + 'upload_date': '20131101', }, }