diff options
Diffstat (limited to 'youtube_dl')
-rwxr-xr-x | youtube_dl/InfoExtractors.py | 38 |
1 files changed, 10 insertions, 28 deletions
diff --git a/youtube_dl/InfoExtractors.py b/youtube_dl/InfoExtractors.py index e380f62a1..f811af5d1 100755 --- a/youtube_dl/InfoExtractors.py +++ b/youtube_dl/InfoExtractors.py @@ -2361,16 +2361,12 @@ class ComedyCentralIE(InfoExtractor): def report_extraction(self, episode_id): self._downloader.to_screen(u'[comedycentral] %s: Extracting information' % episode_id) - def report_config_download(self, episode_id): - self._downloader.to_screen(u'[comedycentral] %s: Downloading configuration' % episode_id) + def report_config_download(self, episode_id, media_id): + self._downloader.to_screen(u'[comedycentral] %s: Downloading configuration for %s' % (episode_id, media_id)) def report_index_download(self, episode_id): self._downloader.to_screen(u'[comedycentral] %s: Downloading show index' % episode_id) - def report_player_url(self, episode_id): - self._downloader.to_screen(u'[comedycentral] %s: Determining player URL' % episode_id) - - def _print_formats(self, formats): print('Available formats:') for x in formats: @@ -2437,15 +2433,6 @@ class ComedyCentralIE(InfoExtractor): else: mMovieParams = [("http://media.mtvnservices.com/" + altMovieParams[0], altMovieParams[0])] - playerUrl_raw = mMovieParams[0][0] - self.report_player_url(epTitle) - try: - urlHandle = compat_urllib_request.urlopen(playerUrl_raw) - playerUrl = urlHandle.geturl() - except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - self._downloader.trouble(u'ERROR: unable to find out player URL: ' + compat_str(err)) - return - uri = mMovieParams[0][1] indexUrl = 'http://shadow.comedycentral.com/feeds/video_player/mrss/?' + compat_urllib_parse.urlencode({'uri': uri}) self.report_index_download(epTitle) @@ -2469,7 +2456,7 @@ class ComedyCentralIE(InfoExtractor): configUrl = ('http://www.comedycentral.com/global/feeds/entertainment/media/mediaGenEntertainment.jhtml?' + compat_urllib_parse.urlencode({'uri': mediaId})) configReq = compat_urllib_request.Request(configUrl) - self.report_config_download(epTitle) + self.report_config_download(epTitle, shortMediaId) try: configXml = compat_urllib_request.urlopen(configReq).read() except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: @@ -2491,7 +2478,7 @@ class ComedyCentralIE(InfoExtractor): return # For now, just pick the highest bitrate - format,video_url = turls[-1] + format,rtmp_video_url = turls[-1] # Get the format arg from the arg stream req_format = self._downloader.params.get('format', None) @@ -2499,16 +2486,14 @@ class ComedyCentralIE(InfoExtractor): # Select format if we can find one for f,v in turls: if f == req_format: - format, video_url = f, v + format, rtmp_video_url = f, v break - # Patch to download from alternative CDN, which does not - # break on current RTMPDump builds - broken_cdn = "rtmpe://viacomccstrmfs.fplive.net/viacomccstrm/gsp.comedystor/" - better_cdn = "rtmpe://cp10740.edgefcs.net/ondemand/mtvnorigin/gsp.comedystor/" - - if video_url.startswith(broken_cdn): - video_url = video_url.replace(broken_cdn, better_cdn) + m = re.match(r'^rtmpe?://.*?/(?P<finalid>gsp.comedystor/.*)$', rtmp_video_url) + if not m: + raise ExtractorError(u'Cannot transform RTMP url') + base = 'http://mtvnmobile.vo.llnwd.net/kip0/_pxn=1+_pxI0=Ripod-h264+_pxL0=undefined+_pxM0=+_pxK=18639+_pxE=mp4/44620/mtvnorigin/' + video_url = base + m.group('finalid') effTitle = showId + u'-' + epTitle info = { @@ -2521,9 +2506,7 @@ class ComedyCentralIE(InfoExtractor): 'format': format, 'thumbnail': None, 'description': officialTitle, - 'player_url': None #playerUrl } - results.append(info) return results @@ -2603,7 +2586,6 @@ class EscapistIE(InfoExtractor): return [info] - class CollegeHumorIE(InfoExtractor): """Information extractor for collegehumor.com""" |