X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/90cd207d52d77e075e62cc256cc39dceee05bbf3..f46044c66663049e286c20ee015db99d47d9dd8a:/youtube_dl/extractor/soundcloud.py?ds=inline diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 4717fbb..c5284fa 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -1,14 +1,18 @@ -import json +# encoding: utf-8 +from __future__ import unicode_literals + import re import itertools from .common import InfoExtractor -from ..utils import ( +from ..compat import ( compat_str, compat_urlparse, compat_urllib_parse, - +) +from ..utils import ( ExtractorError, + int_or_none, unified_strdate, ) @@ -22,190 +26,359 @@ class SoundcloudIE(InfoExtractor): of the stream token and uid """ - _VALID_URL = r'''^(?:https?://)? - (?:(?:(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)/?(?:[?].*)?$) - |(?:api\.soundcloud\.com/tracks/(?P\d+)) - |(?Pw.soundcloud.com/player/?.*?url=.*) + _VALID_URL = r'''(?x)^(?:https?://)? + (?:(?:(?:www\.|m\.)?soundcloud\.com/ + (?P[\w\d-]+)/ + (?!sets/|likes/?(?:$|[?#])) + (?P[\w\d-]+)/? + (?P<token>[^?]+?)?(?:[?].*)?$) + |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+) + (?:/?\?secret_token=(?P<secret_token>[^&]+))?) + |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*) ) ''' - IE_NAME = u'soundcloud' + IE_NAME = '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" + 'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy', + 'md5': 'ebef0a451b909710ed1d7787dddbf0d7', + 'info_dict': { + 'id': '62986583', + 'ext': 'mp3', + 'upload_date': '20121011', + 'description': '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', + 'uploader': 'E.T. ExTerrestrial Music', + 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1', + 'duration': 143, } }, # 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', + 'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep', + 'info_dict': { + 'id': '47127627', + 'ext': 'mp3', + 'title': 'Goldrushed', + 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com', + 'uploader': 'The Royal Concept', + 'upload_date': '20120521', + 'duration': 227, }, - u'params': { + 'params': { # rtmp - u'skip_download': True, + 'skip_download': True, + }, + }, + # private link + { + 'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp', + 'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604', + 'info_dict': { + 'id': '123998367', + 'ext': 'mp3', + 'title': 'Youtube - Dl Test Video \'\' Ä↭', + 'uploader': 'jaimeMF', + 'description': 'test chars: \"\'/\\ä↭', + 'upload_date': '20131209', + 'duration': 9, + }, + }, + # private link (alt format) + { + 'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp', + 'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604', + 'info_dict': { + 'id': '123998367', + 'ext': 'mp3', + 'title': 'Youtube - Dl Test Video \'\' Ä↭', + 'uploader': 'jaimeMF', + 'description': 'test chars: \"\'/\\ä↭', + 'upload_date': '20131209', + 'duration': 9, + }, + }, + # downloadable song + { + 'url': 'https://soundcloud.com/oddsamples/bus-brakes', + 'md5': '7624f2351f8a3b2e7cd51522496e7631', + 'info_dict': { + 'id': '128590877', + 'ext': 'mp3', + 'title': 'Bus Brakes', + 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66', + 'uploader': 'oddsamples', + 'upload_date': '20140109', + 'duration': 17, }, }, ] _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28' - - @classmethod - def suitable(cls, url): - return re.match(cls._VALID_URL, url, flags=re.VERBOSE) is not None + _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf' def report_resolve(self, video_id): """Report information extraction.""" - self.to_screen(u'%s: Resolving id' % video_id) + self.to_screen('%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): + def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None): 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 = 'mp3' result = { - 'id': track_id, - 'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID, + 'id': track_id, 'uploader': info['user']['username'], 'upload_date': unified_strdate(info['created_at']), - 'title': info['title'], - 'ext': u'mp3', + 'title': info['title'], 'description': info['description'], 'thumbnail': thumbnail, + 'duration': int_or_none(info.get('duration'), 1000), + 'webpage_url': info.get('permalink_url'), } + formats = [] if info.get('downloadable', False): - result['url'] = 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(track_id, self._CLIENT_ID) - if not info.get('streamable', False): - # We have to get the rtmp url - stream_json = self._download_webpage( - 'http://api.soundcloud.com/i1/tracks/{0}/streams?client_id={1}'.format(track_id, self._CLIENT_ID), - track_id, u'Downloading track url') - rtmp_url = json.loads(stream_json)['rtmp_mp3_128_url'] - # The url doesn't have an rtmp app, we have to extract the playpath - url, path = rtmp_url.split('mp3:', 1) - result.update({ - 'url': url, - 'play_path': 'mp3:' + path, + # We can build a direct link to the song + format_url = ( + 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format( + track_id, self._CLIENT_ID)) + formats.append({ + 'format_id': 'download', + 'ext': info.get('original_format', 'mp3'), + 'url': format_url, + 'vcodec': 'none', + 'preference': 10, }) + + # We have to retrieve the url + streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?' + 'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token)) + format_dict = self._download_json( + streams_url, + track_id, 'Downloading track url') + + for key, stream_url in format_dict.items(): + if key.startswith('http'): + formats.append({ + 'format_id': key, + 'ext': ext, + 'url': stream_url, + 'vcodec': 'none', + }) + elif key.startswith('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, + 'vcodec': 'none', + }) + + 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 + formats.append({ + 'format_id': 'fallback', + 'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID, + 'ext': ext, + 'vcodec': 'none', + }) + + for f in formats: + if f['format_id'].startswith('http'): + f['protocol'] = 'http' + if f['format_id'].startswith('rtmp'): + f['protocol'] = 'rtmp' + + self._sort_formats(formats) + result['formats'] = formats + return result def _real_extract(self, url): mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE) if mobj is None: - raise ExtractorError(u'Invalid URL: %s' % url) + raise ExtractorError('Invalid URL: %s' % url) track_id = mobj.group('track_id') + token = None 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'): + token = mobj.group('secret_token') + if token: + info_json_url += "&secret_token=" + token + elif mobj.group('player'): query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query) - return self.url_result(query['url'][0], ie='Soundcloud') + return self.url_result(query['url'][0]) else: # extract uploader (which is in the url) - uploader = mobj.group(1) + uploader = mobj.group('uploader') # extract simple title (uploader + slug of song title) - slug_title = mobj.group(2) - full_title = '%s/%s' % (uploader, slug_title) - + slug_title = mobj.group('title') + token = mobj.group('token') + full_title = resolve_title = '%s/%s' % (uploader, slug_title) + if token: + resolve_title += '/%s' % token + self.report_resolve(full_title) - - url = 'http://soundcloud.com/%s/%s' % (uploader, slug_title) + + url = 'http://soundcloud.com/%s' % resolve_title info_json_url = self._resolv_url(url) - info_json = self._download_webpage(info_json_url, full_title, u'Downloading info JSON') + info = self._download_json(info_json_url, full_title, 'Downloading info JSON') + + return self._extract_info_dict(info, full_title, secret_token=token) - info = json.loads(info_json) - return self._extract_info_dict(info, full_title) class SoundcloudSetIE(SoundcloudIE): - _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$' - IE_NAME = u'soundcloud:set' - # it's in tests/test_playlists.py - _TESTS = [] + _VALID_URL = r'https?://(?:www\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?' + IE_NAME = 'soundcloud:set' + _TESTS = [{ + 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep', + 'info_dict': { + 'id': '2284613', + 'title': 'The Royal Concept EP', + }, + 'playlist_mincount': 6, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - if mobj is None: - raise ExtractorError(u'Invalid URL: %s' % url) # extract uploader (which is in the url) - uploader = mobj.group(1) + uploader = mobj.group('uploader') # extract simple title (uploader + slug of song title) - slug_title = mobj.group(2) + slug_title = mobj.group('slug_title') full_title = '%s/sets/%s' % (uploader, slug_title) + url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title) + + token = mobj.group('token') + if token: + full_title += '/' + token + url += '/' + token self.report_resolve(full_title) - url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title) resolv_url = self._resolv_url(url) - info_json = self._download_webpage(resolv_url, full_title) + info = self._download_json(resolv_url, full_title) - videos = [] - info = json.loads(info_json) if 'errors' in info: for err in info['errors']: - self._downloader.report_error(u'unable to download video webpage: %s' % compat_str(err['error_message'])) + self._downloader.report_error('unable to download video webpage: %s' % compat_str(err['error_message'])) return - self.report_extraction(full_title) - return {'_type': 'playlist', - 'entries': [self._extract_info_dict(track) for track in info['tracks']], - 'id': info['id'], - 'title': info['title'], - } + return { + '_type': 'playlist', + 'entries': [self._extract_info_dict(track, secret_token=token) for track in info['tracks']], + 'id': '%s' % info['id'], + 'title': info['title'], + } class SoundcloudUserIE(SoundcloudIE): - _VALID_URL = r'https?://(www\.)?soundcloud.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$' - IE_NAME = u'soundcloud:user' - - # it's in tests/test_playlists.py - _TESTS = [] + _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)/?((?P<rsrc>tracks|likes)/?)?(\?.*)?$' + IE_NAME = 'soundcloud:user' + _TESTS = [{ + 'url': 'https://soundcloud.com/the-concept-band', + 'info_dict': { + 'id': '9615865', + 'title': 'The Royal Concept', + }, + 'playlist_mincount': 12 + }, { + 'url': 'https://soundcloud.com/the-concept-band/likes', + 'info_dict': { + 'id': '9615865', + 'title': 'The Royal Concept', + }, + 'playlist_mincount': 1, + }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) uploader = mobj.group('user') + resource = mobj.group('rsrc') + if resource is None: + resource = 'tracks' + elif resource == 'likes': + resource = 'favorites' 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) + user = self._download_json( + resolv_url, uploader, 'Downloading user info') + base_url = 'http://api.soundcloud.com/users/%s/%s.json?' % (uploader, resource) - tracks = [] + entries = [] 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: + data = compat_urllib_parse.urlencode({ + 'offset': i * 50, + 'limit': 50, + 'client_id': self._CLIENT_ID, + }) + new_entries = self._download_json( + base_url + data, uploader, 'Downloading track page %s' % (i + 1)) + if len(new_entries) == 0: + self.to_screen('%s: End page received' % uploader) break + entries.extend(self._extract_info_dict(e, quiet=True) for e in new_entries) return { '_type': 'playlist', 'id': compat_str(user['id']), 'title': user['username'], - 'entries': tracks, + 'entries': entries, + } + + +class SoundcloudPlaylistIE(SoundcloudIE): + _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$' + IE_NAME = 'soundcloud:playlist' + _TESTS = [{ + 'url': 'http://api.soundcloud.com/playlists/4110309', + 'info_dict': { + 'id': '4110309', + 'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]', + 'description': 're:.*?TILT Brass - Bowery Poetry Club', + }, + 'playlist_count': 6, + }] + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + playlist_id = mobj.group('id') + base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id) + + data_dict = { + 'client_id': self._CLIENT_ID, + } + token = mobj.group('token') + + if token: + data_dict['secret_token'] = token + + data = compat_urllib_parse.urlencode(data_dict) + data = self._download_json( + base_url + data, playlist_id, 'Downloading playlist') + + entries = [ + self._extract_info_dict(t, quiet=True, secret_token=token) + for t in data['tracks']] + + return { + '_type': 'playlist', + 'id': playlist_id, + 'title': data.get('title'), + 'description': data.get('description'), + 'entries': entries, }