X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/90cd207d52d77e075e62cc256cc39dceee05bbf3..848723ea972c09f28787db91d8c06e98a274ab89:/youtube_dl/extractor/soundcloud.py diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 4717fbb..393b5f1 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -1,3 +1,6 @@ +# encoding: utf-8 +from __future__ import unicode_literals + import json import re import itertools @@ -23,42 +26,72 @@ class SoundcloudIE(InfoExtractor): """ _VALID_URL = r'''^(?:https?://)? - (?:(?:(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)/?(?:[?].*)?$) + (?:(?:(?:www\.|m\.)?soundcloud\.com/ + (?P[\w\d-]+)/ + (?!sets/)(?P[\w\d-]+)/? + (?P<token>[^?]+?)?(?:[?].*)?$) |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)) - |(?P<widget>w.soundcloud.com/player/?.*?url=.*) + |(?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', + 'file': '62986583.mp3', + 'md5': 'ebef0a451b909710ed1d7787dddbf0d7', + 'info_dict': { + "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" } }, # 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', + 'uploader': 'The Royal Concept', + 'upload_date': '20120521', }, - 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', + }, + }, + # downloadable song + { + 'url': 'https://soundcloud.com/simgretina/just-your-problem-baby-1', + 'md5': '56a8b69568acaa967b4c49f9d1d52d19', + 'info_dict': { + 'id': '105614606', + 'ext': 'wav', + 'title': 'Just Your Problem Baby (Acapella)', + 'description': 'Vocals', + 'uploader': 'Sim Gretina', + 'upload_date': '20130815', }, }, ] _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28' + _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf' @classmethod def suitable(cls, url): @@ -72,39 +105,83 @@ class SoundcloudIE(InfoExtractor): 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, } 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 + # 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)) + result['formats'] = [{ + 'format_id': 'download', + 'ext': info.get('original_format', 'mp3'), + 'url': format_url, + 'vcodec': 'none', + }] + else: + # 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)) 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, - }) + streams_url, + track_id, 'Downloading track url') + + 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, + 'vcodec': 'none', + }) + 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, + '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): @@ -113,31 +190,35 @@ class SoundcloudIE(InfoExtractor): raise ExtractorError(u'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'): + elif mobj.group('player'): 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) + 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_json = self._download_webpage(info_json_url, full_title, 'Downloading info JSON') info = json.loads(info_json) - return self._extract_info_dict(info, full_title) + return self._extract_info_dict(info, full_title, secret_token=token) class SoundcloudSetIE(SoundcloudIE): _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$' - IE_NAME = u'soundcloud:set' + IE_NAME = 'soundcloud:set' # it's in tests/test_playlists.py _TESTS = [] @@ -158,7 +239,6 @@ class SoundcloudSetIE(SoundcloudIE): resolv_url = self._resolv_url(url) info_json = self._download_webpage(resolv_url, full_title) - videos = [] info = json.loads(info_json) if 'errors' in info: for err in info['errors']: @@ -174,8 +254,8 @@ class SoundcloudSetIE(SoundcloudIE): class SoundcloudUserIE(SoundcloudIE): - _VALID_URL = r'https?://(www\.)?soundcloud.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$' - IE_NAME = u'soundcloud:user' + _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$' + IE_NAME = 'soundcloud:user' # it's in tests/test_playlists.py _TESTS = [] @@ -187,7 +267,7 @@ class SoundcloudUserIE(SoundcloudIE): url = 'http://soundcloud.com/%s/' % uploader resolv_url = self._resolv_url(url) user_json = self._download_webpage(resolv_url, uploader, - u'Downloading user info') + 'Downloading user info') user = json.loads(user_json) tracks = [] @@ -197,7 +277,7 @@ class SoundcloudUserIE(SoundcloudIE): }) 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)) + '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: