diff options
author | Philipp Hagemeister <[email protected]> | 2013-11-25 20:30:41 +0100 |
---|---|---|
committer | Philipp Hagemeister <[email protected]> | 2013-11-25 20:30:41 +0100 |
commit | 2a15e7063bceed326bcbc0a01ba77324f0373a0c (patch) | |
tree | 78624057f708d98bb89ce4bb267718cc6f14a7cf /youtube_dl/extractor | |
parent | d46cc192d763f66655247ee122e397626481caca (diff) | |
download | youtube-dl-2a15e7063bceed326bcbc0a01ba77324f0373a0c.tar.gz youtube-dl-2a15e7063bceed326bcbc0a01ba77324f0373a0c.zip |
[soundcloud] Prefer HTTP over RTMP (#1798)
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r-- | youtube_dl/extractor/soundcloud.py | 66 |
1 files changed, 48 insertions, 18 deletions
diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 67b2dff9c..ee8da227e 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -76,44 +76,74 @@ class SoundcloudIE(InfoExtractor): def _extract_info_dict(self, info, full_title=None, quiet=False): track_id = compat_str(info['id']) name = full_title or track_id - if quiet == False: + if quiet: self.report_extraction(name) thumbnail = info['artwork_url'] if thumbnail is not None: thumbnail = thumbnail.replace('-large', '-t500x500') + ext = info.get('original_format', u'mp3') result = { - 'id': track_id, + 'id': track_id, 'uploader': info['user']['username'], 'upload_date': unified_strdate(info['created_at']), - 'title': info['title'], - 'ext': info.get('original_format', u'mp3'), + 'title': info['title'], 'description': info['description'], 'thumbnail': thumbnail, } if info.get('downloadable', False): # We can build a direct link to the song - result['url'] = 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(track_id, self._CLIENT_ID) + format_url = ( + u'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format( + track_id, self._CLIENT_ID)) + result['formats'] = [{ + 'format_id': 'download', + 'ext': ext, + 'url': format_url, + }] else: # We have to retrieve the url stream_json = self._download_webpage( 'http://api.soundcloud.com/i1/tracks/{0}/streams?client_id={1}'.format(track_id, self._IPHONE_CLIENT_ID), track_id, u'Downloading track url') - # There should be only one entry in the dictionary - key, stream_url = list(json.loads(stream_json).items())[0] - if key.startswith(u'http'): - result['url'] = stream_url - elif key.startswith(u'rtmp'): - # The url doesn't have an rtmp app, we have to extract the playpath - url, path = stream_url.split('mp3:', 1) - result.update({ - 'url': url, - 'play_path': 'mp3:' + path, - }) - else: + + formats = [] + format_dict = json.loads(stream_json) + for key, stream_url in format_dict.items(): + if key.startswith(u'http'): + formats.append({ + 'format_id': key, + 'ext': ext, + 'url': stream_url, + }) + elif key.startswith(u'rtmp'): + # The url doesn't have an rtmp app, we have to extract the playpath + url, path = stream_url.split('mp3:', 1) + formats.append({ + 'format_id': key, + 'url': url, + 'play_path': 'mp3:' + path, + 'ext': ext, + }) + + if not formats: # We fallback to the stream_url in the original info, this # cannot be always used, sometimes it can give an HTTP 404 error - result['url'] = info['stream_url'] + '?client_id=' + self._CLIENT_ID, + formats.append({ + 'format_id': u'fallback', + 'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID, + 'ext': ext, + }) + + def format_pref(f): + if f['format_id'].startswith('http'): + return 2 + if f['format_id'].startswith('rtmp'): + return 1 + return 0 + + formats.sort(key=format_pref) + result['formats'] = formats return result |