X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/509eeaaa67b2f901752c50fc6edf41954dbe3085..03342304420e5daeb428ffdcc7bbd2bbfecfa61a:/youtube_dl/extractor/soundcloud.py?ds=sidebyside diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 7aa100f..1efb2b9 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -4,12 +4,17 @@ from __future__ import unicode_literals import re import itertools -from .common import InfoExtractor -from ..utils import ( +from .common import ( + InfoExtractor, + SearchInfoExtractor +) +from ..compat import ( compat_str, compat_urlparse, compat_urllib_parse, - +) +from ..utils import ( + encode_dict, ExtractorError, int_or_none, unified_strdate, @@ -28,9 +33,11 @@ class SoundcloudIE(InfoExtractor): _VALID_URL = r'''(?x)^(?:https?://)? (?:(?:(?:www\.|m\.)?soundcloud\.com/ (?P[\w\d-]+)/ - (?!sets/)(?P[\w\d-]+)/? + (?!(?:tracks|sets(?:/[^/?#]+)?|reposts|likes|spotlight)/?(?:$|[?#])) + (?P<title>[\w\d-]+)/? (?P<token>[^?]+?)?(?:[?].*)?$) - |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)) + |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+) + (?:/?\?secret_token=(?P<secret_token>[^&]+))?) |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*) ) ''' @@ -38,14 +45,15 @@ class SoundcloudIE(InfoExtractor): _TESTS = [ { '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", - "duration": 143, + '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 @@ -79,23 +87,37 @@ class SoundcloudIE(InfoExtractor): '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/simgretina/just-your-problem-baby-1', - 'md5': '56a8b69568acaa967b4c49f9d1d52d19', + 'url': 'https://soundcloud.com/oddsamples/bus-brakes', + 'md5': '7624f2351f8a3b2e7cd51522496e7631', 'info_dict': { - 'id': '105614606', - 'ext': 'wav', - 'title': 'Just Your Problem Baby (Acapella)', - 'description': 'Vocals', - 'uploader': 'Sim Gretina', - 'upload_date': '20130815', - #'duration': 42, + 'id': '128590877', + 'ext': 'mp3', + 'title': 'Bus Brakes', + 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66', + 'uploader': 'oddsamples', + 'upload_date': '20140109', + 'duration': 17, }, }, ] - _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28' + _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea' _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf' def report_resolve(self, video_id): @@ -124,6 +146,7 @@ class SoundcloudIE(InfoExtractor): '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): @@ -141,7 +164,7 @@ class SoundcloudIE(InfoExtractor): # 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)) + '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') @@ -161,7 +184,7 @@ class SoundcloudIE(InfoExtractor): 'format_id': key, 'url': url, 'play_path': 'mp3:' + path, - 'ext': ext, + 'ext': 'flv', 'vcodec': 'none', }) @@ -181,8 +204,9 @@ class SoundcloudIE(InfoExtractor): if f['format_id'].startswith('rtmp'): f['protocol'] = 'rtmp' - self._sort_formats(formats) - result['formats'] = formats + self._check_formats(formats, track_id) + self._sort_formats(formats) + result['formats'] = formats return result @@ -196,21 +220,29 @@ class SoundcloudIE(InfoExtractor): 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 + 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]) + real_url = query['url'][0] + # If the token is in the query of the original url we have to + # manually add it + if 'secret_token' in query: + real_url += '?secret_token=' + query['secret_token'][0] + return self.url_result(real_url) else: # extract uploader (which is in the url) uploader = mobj.group('uploader') # extract simple title (uploader + slug of song title) - slug_title = mobj.group('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' % resolve_title info_json_url = self._resolv_url(url) info = self._download_json(info_json_url, full_title, 'Downloading info JSON') @@ -219,47 +251,126 @@ class SoundcloudIE(InfoExtractor): class SoundcloudSetIE(SoundcloudIE): - _VALID_URL = r'https?://(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)' + _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?' IE_NAME = 'soundcloud:set' - # it's in tests/test_playlists.py - _TESTS = [] + _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('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 = self._download_json(resolv_url, full_title) if 'errors' in info: - for err in info['errors']: - self._downloader.report_error('unable to download video webpage: %s' % compat_str(err['error_message'])) - return + msgs = (compat_str(err['error_message']) for err in info['errors']) + raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs)) - 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'], - } + entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in info['tracks']] + + return { + '_type': 'playlist', + 'entries': entries, + 'id': '%s' % info['id'], + 'title': info['title'], + } class SoundcloudUserIE(SoundcloudIE): - _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$' + _VALID_URL = r'''(?x) + https?:// + (?:(?:www|m)\.)?soundcloud\.com/ + (?P<user>[^/]+) + (?:/ + (?P<rsrc>tracks|sets|reposts|likes|spotlight) + )? + /?(?:[?#].*)?$ + ''' IE_NAME = 'soundcloud:user' - - # it's in tests/test_playlists.py - _TESTS = [] + _TESTS = [{ + 'url': 'https://soundcloud.com/the-akashic-chronicler', + 'info_dict': { + 'id': '114582580', + 'title': 'The Akashic Chronicler (All)', + }, + 'playlist_mincount': 111, + }, { + 'url': 'https://soundcloud.com/the-akashic-chronicler/tracks', + 'info_dict': { + 'id': '114582580', + 'title': 'The Akashic Chronicler (Tracks)', + }, + 'playlist_mincount': 50, + }, { + 'url': 'https://soundcloud.com/the-akashic-chronicler/sets', + 'info_dict': { + 'id': '114582580', + 'title': 'The Akashic Chronicler (Playlists)', + }, + 'playlist_mincount': 3, + }, { + 'url': 'https://soundcloud.com/the-akashic-chronicler/reposts', + 'info_dict': { + 'id': '114582580', + 'title': 'The Akashic Chronicler (Reposts)', + }, + 'playlist_mincount': 7, + }, { + 'url': 'https://soundcloud.com/the-akashic-chronicler/likes', + 'info_dict': { + 'id': '114582580', + 'title': 'The Akashic Chronicler (Likes)', + }, + 'playlist_mincount': 321, + }, { + 'url': 'https://soundcloud.com/grynpyret/spotlight', + 'info_dict': { + 'id': '7098329', + 'title': 'Grynpyret (Spotlight)', + }, + 'playlist_mincount': 1, + }] + + _API_BASE = 'https://api.soundcloud.com' + _API_V2_BASE = 'https://api-v2.soundcloud.com' + + _BASE_URL_MAP = { + 'all': '%s/profile/soundcloud:users:%%s' % _API_V2_BASE, + 'tracks': '%s/users/%%s/tracks' % _API_BASE, + 'sets': '%s/users/%%s/playlists' % _API_V2_BASE, + 'reposts': '%s/profile/soundcloud:users:%%s/reposts' % _API_V2_BASE, + 'likes': '%s/users/%%s/likes' % _API_V2_BASE, + 'spotlight': '%s/users/%%s/spotlight' % _API_V2_BASE, + } + + _TITLE_MAP = { + 'all': 'All', + 'tracks': 'Tracks', + 'sets': 'Playlists', + 'reposts': 'Reposts', + 'likes': 'Likes', + 'spotlight': 'Spotlight', + } def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) @@ -269,48 +380,91 @@ class SoundcloudUserIE(SoundcloudIE): resolv_url = self._resolv_url(url) user = self._download_json( resolv_url, uploader, 'Downloading user info') - base_url = 'http://api.soundcloud.com/users/%s/tracks.json?' % uploader + + resource = mobj.group('rsrc') or 'all' + base_url = self._BASE_URL_MAP[resource] % user['id'] + + COMMON_QUERY = { + 'limit': 50, + 'client_id': self._CLIENT_ID, + 'linked_partitioning': '1', + } + + query = COMMON_QUERY.copy() + query['offset'] = 0 + + next_href = base_url + '?' + compat_urllib_parse.urlencode(query) entries = [] for i in itertools.count(): - data = compat_urllib_parse.urlencode({ - 'offset': i * 50, - 'client_id': self._CLIENT_ID, - }) - new_entries = self._download_json( - base_url + data, uploader, 'Downloading track page %s' % (i + 1)) - entries.extend(self._extract_info_dict(e, quiet=True) for e in new_entries) - if len(new_entries) < 50: + response = self._download_json( + next_href, uploader, 'Downloading track page %s' % (i + 1)) + + collection = response['collection'] + if not collection: + break + + def resolve_permalink_url(candidates): + for cand in candidates: + if isinstance(cand, dict): + permalink_url = cand.get('permalink_url') + if permalink_url and permalink_url.startswith('http'): + return permalink_url + + for e in collection: + permalink_url = resolve_permalink_url((e, e.get('track'), e.get('playlist'))) + if permalink_url: + entries.append(self.url_result(permalink_url)) + + next_href = response.get('next_href') + if not next_href: break + parsed_next_href = compat_urlparse.urlparse(response['next_href']) + qs = compat_urlparse.parse_qs(parsed_next_href.query) + qs.update(COMMON_QUERY) + next_href = compat_urlparse.urlunparse( + parsed_next_href._replace(query=compat_urllib_parse.urlencode(qs, True))) + return { '_type': 'playlist', 'id': compat_str(user['id']), - 'title': user['username'], + 'title': '%s (%s)' % (user['username'], self._TITLE_MAP[resource]), 'entries': entries, } class SoundcloudPlaylistIE(SoundcloudIE): - _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)' + _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$' IE_NAME = 'soundcloud:playlist' - - # it's in tests/test_playlists.py - _TESTS = [] + _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 = compat_urllib_parse.urlencode({ + 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) for t in data['tracks']] + entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in data['tracks']] return { '_type': 'playlist', @@ -319,3 +473,60 @@ class SoundcloudPlaylistIE(SoundcloudIE): 'description': data.get('description'), 'entries': entries, } + + +class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE): + IE_NAME = 'soundcloud:search' + IE_DESC = 'Soundcloud search' + _MAX_RESULTS = float('inf') + _TESTS = [{ + 'url': 'scsearch15:post-avant jazzcore', + 'info_dict': { + 'title': 'post-avant jazzcore', + }, + 'playlist_count': 15, + }] + + _SEARCH_KEY = 'scsearch' + _MAX_RESULTS_PER_PAGE = 200 + _DEFAULT_RESULTS_PER_PAGE = 50 + _API_V2_BASE = 'https://api-v2.soundcloud.com' + + def _get_collection(self, endpoint, collection_id, **query): + limit = min( + query.get('limit', self._DEFAULT_RESULTS_PER_PAGE), + self._MAX_RESULTS_PER_PAGE) + query['limit'] = limit + query['client_id'] = self._CLIENT_ID + query['linked_partitioning'] = '1' + query['offset'] = 0 + data = compat_urllib_parse.urlencode(encode_dict(query)) + next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data) + + collected_results = 0 + + for i in itertools.count(1): + response = self._download_json( + next_url, collection_id, 'Downloading page {0}'.format(i), + 'Unable to download API page') + + collection = response.get('collection', []) + if not collection: + break + + collection = list(filter(bool, collection)) + collected_results += len(collection) + + for item in collection: + yield self.url_result(item['uri'], SoundcloudIE.ie_key()) + + if not collection or collected_results >= limit: + break + + next_url = response.get('next_href') + if not next_url: + break + + def _get_n_results(self, query, n): + tracks = self._get_collection('/search/tracks', query, limit=n, q=query) + return self.playlist_result(tracks, playlist_title=query)