X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fsoundcloud.py;h=67b2dff9c9ec431bb7ddd501bb804f6cc572cb61;hb=3f8ced5144a76a3f9ab7ee8cd06cc79bb75dc564;hp=80d7e1b543608a36170577c004d032a7c5e75d1b;hpb=aad0d6d5ba13f1f257720d84caf6d3ba12eb3c92;p=youtube-dl diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 80d7e1b54..67b2dff9c 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -1,9 +1,12 @@ import json import re +import itertools from .common import InfoExtractor from ..utils import ( compat_str, + compat_urlparse, + compat_urllib_parse, ExtractorError, unified_strdate, @@ -19,68 +22,134 @@ class SoundcloudIE(InfoExtractor): of the stream token and uid """ - _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)' + _VALID_URL = r'''^(?:https?://)? + (?:(?:(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)/?(?:[?].*)?$) + |(?:api\.soundcloud\.com/tracks/(?P\d+)) + |(?Pw.soundcloud.com/player/?.*?url=.*) + ) + ''' IE_NAME = u'soundcloud' + _TESTS = [ + { + u'url': u'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy', + u'file': u'62986583.mp3', + u'md5': u'ebef0a451b909710ed1d7787dddbf0d7', + u'info_dict': { + u"upload_date": u"20121011", + u"description": u"No Downloads untill we record the finished version this weekend, i was too pumped n i had to post it , earl is prolly gonna b hella p.o'd", + u"uploader": u"E.T. ExTerrestrial Music", + u"title": u"Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1" + } + }, + # not streamable song + { + u'url': u'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep', + u'info_dict': { + u'id': u'47127627', + u'ext': u'mp3', + u'title': u'Goldrushed', + u'uploader': u'The Royal Concept', + u'upload_date': u'20120521', + }, + u'params': { + # rtmp + u'skip_download': True, + }, + }, + ] + + _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28' + _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf' + + @classmethod + def suitable(cls, url): + return re.match(cls._VALID_URL, url, flags=re.VERBOSE) is not None def report_resolve(self, video_id): """Report information extraction.""" self.to_screen(u'%s: Resolving id' % video_id) + @classmethod + def _resolv_url(cls, url): + return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID + + 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: + self.report_extraction(name) + + thumbnail = info['artwork_url'] + if thumbnail is not None: + thumbnail = thumbnail.replace('-large', '-t500x500') + result = { + 'id': track_id, + 'uploader': info['user']['username'], + 'upload_date': unified_strdate(info['created_at']), + 'title': info['title'], + 'ext': info.get('original_format', u'mp3'), + '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) + 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: + # 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, + + return result + def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) + mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE) if mobj is None: raise ExtractorError(u'Invalid URL: %s' % url) - # extract uploader (which is in the url) - uploader = mobj.group(1) - # extract simple title (uploader + slug of song title) - slug_title = mobj.group(2) - full_title = '%s/%s' % (uploader, slug_title) - - self.report_resolve(full_title) - - url = 'http://soundcloud.com/%s/%s' % (uploader, slug_title) - resolv_url = 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=b45b1aa10f1ac2941910a7f0d10f8e28' - info_json = self._download_webpage(resolv_url, full_title, u'Downloading info JSON') + track_id = mobj.group('track_id') + if track_id is not None: + info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID + full_title = track_id + elif mobj.group('widget'): + query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query) + return self.url_result(query['url'][0], ie='Soundcloud') + else: + # extract uploader (which is in the url) + uploader = mobj.group(1) + # extract simple title (uploader + slug of song title) + slug_title = mobj.group(2) + full_title = '%s/%s' % (uploader, slug_title) + + self.report_resolve(full_title) + + url = 'http://soundcloud.com/%s/%s' % (uploader, slug_title) + info_json_url = self._resolv_url(url) + info_json = self._download_webpage(info_json_url, full_title, u'Downloading info JSON') info = json.loads(info_json) - video_id = info['id'] - self.report_extraction(full_title) - - streams_url = 'https://api.sndcdn.com/i1/tracks/' + str(video_id) + '/streams?client_id=b45b1aa10f1ac2941910a7f0d10f8e28' - stream_json = self._download_webpage(streams_url, full_title, - u'Downloading stream definitions', - u'unable to download stream definitions') - - streams = json.loads(stream_json) - mediaURL = streams['http_mp3_128_url'] - upload_date = unified_strdate(info['created_at']) - - return [{ - 'id': info['id'], - 'url': mediaURL, - 'uploader': info['user']['username'], - 'upload_date': upload_date, - 'title': info['title'], - 'ext': u'mp3', - 'description': info['description'], - }] + return self._extract_info_dict(info, full_title) -class SoundcloudSetIE(InfoExtractor): - """Information extractor for soundcloud.com sets - To access the media, the uid of the song and a stream token - must be extracted from the page source and the script must make - a request to media.soundcloud.com/crossdomain.xml. Then - the media can be grabbed by requesting from an url composed - of the stream token and uid - """ - - _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)' +class SoundcloudSetIE(SoundcloudIE): + _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$' IE_NAME = u'soundcloud:set' - - def report_resolve(self, video_id): - """Report information extraction.""" - self.to_screen(u'%s: Resolving id' % video_id) + # it's in tests/test_playlists.py + _TESTS = [] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -96,10 +165,9 @@ class SoundcloudSetIE(InfoExtractor): self.report_resolve(full_title) url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title) - resolv_url = 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=b45b1aa10f1ac2941910a7f0d10f8e28' + resolv_url = self._resolv_url(url) info_json = self._download_webpage(resolv_url, full_title) - videos = [] info = json.loads(info_json) if 'errors' in info: for err in info['errors']: @@ -107,23 +175,46 @@ class SoundcloudSetIE(InfoExtractor): return self.report_extraction(full_title) - for track in info['tracks']: - video_id = track['id'] - - streams_url = 'https://api.sndcdn.com/i1/tracks/' + str(video_id) + '/streams?client_id=b45b1aa10f1ac2941910a7f0d10f8e28' - stream_json = self._download_webpage(streams_url, video_id, u'Downloading track info JSON') - - self.report_extraction(video_id) - streams = json.loads(stream_json) - mediaURL = streams['http_mp3_128_url'] - - videos.append({ - 'id': video_id, - 'url': mediaURL, - 'uploader': track['user']['username'], - 'upload_date': unified_strdate(track['created_at']), - 'title': track['title'], - 'ext': u'mp3', - 'description': track['description'], - }) - return videos + return {'_type': 'playlist', + 'entries': [self._extract_info_dict(track) for track in info['tracks']], + 'id': info['id'], + 'title': info['title'], + } + + +class SoundcloudUserIE(SoundcloudIE): + _VALID_URL = r'https?://(www\.)?soundcloud.com/(?P[^/]+)(/?(tracks/)?)?(\?.*)?$' + IE_NAME = u'soundcloud:user' + + # it's in tests/test_playlists.py + _TESTS = [] + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + uploader = mobj.group('user') + + url = 'http://soundcloud.com/%s/' % uploader + resolv_url = self._resolv_url(url) + user_json = self._download_webpage(resolv_url, uploader, + u'Downloading user info') + user = json.loads(user_json) + + tracks = [] + for i in itertools.count(): + data = compat_urllib_parse.urlencode({'offset': i*50, + 'client_id': self._CLIENT_ID, + }) + tracks_url = 'http://api.soundcloud.com/users/%s/tracks.json?' % user['id'] + data + response = self._download_webpage(tracks_url, uploader, + u'Downloading tracks page %s' % (i+1)) + new_tracks = json.loads(response) + tracks.extend(self._extract_info_dict(track, quiet=True) for track in new_tracks) + if len(new_tracks) < 50: + break + + return { + '_type': 'playlist', + 'id': compat_str(user['id']), + 'title': user['username'], + 'entries': tracks, + }