X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/46113edab215c2211a604c06245c16d5d4e57dcf..e22397a1c5329e12f3719961c9456e378f94dbb1:/youtube_dl/extractor/soundcloud.py diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py index 54256e1..a0b09f5 100644 --- a/youtube_dl/extractor/soundcloud.py +++ b/youtube_dl/extractor/soundcloud.py @@ -1,21 +1,59 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals -import re import itertools +import re -from .common import InfoExtractor -from ..utils import ( +from .common import ( + InfoExtractor, + SearchInfoExtractor +) +from ..compat import ( + compat_HTTPError, + compat_kwargs, compat_str, compat_urlparse, - compat_urllib_parse, - +) +from ..utils import ( + error_to_compat_str, ExtractorError, + float_or_none, + HEADRequest, int_or_none, - unified_strdate, + KNOWN_EXTENSIONS, + mimetype2ext, + str_or_none, + try_get, + unified_timestamp, + update_url_query, + url_or_none, ) +class SoundcloudEmbedIE(InfoExtractor): + _VALID_URL = r'https?://(?:w|player|p)\.soundcloud\.com/player/?.*?\burl=(?P.+)' + _TEST = { + # from https://www.soundi.fi/uutiset/ennakkokuuntelussa-timo-kaukolammen-station-to-station-to-station-julkaisua-juhlitaan-tanaan-g-livelabissa/ + 'url': 'https://w.soundcloud.com/player/?visual=true&url=https%3A%2F%2Fapi.soundcloud.com%2Fplaylists%2F922213810&show_artwork=true&maxwidth=640&maxheight=960&dnt=1&secret_token=s-ziYey', + 'only_matching': True, + } + + @staticmethod + def _extract_urls(webpage): + return [m.group('url') for m in re.finditer( + r']+src=(["\'])(?P(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1', + webpage)] + + def _real_extract(self, url): + query = compat_urlparse.parse_qs( + compat_urlparse.urlparse(url).query) + api_url = query['url'][0] + secret_token = query.get('secret_token') + if secret_token: + api_url = update_url_query(api_url, {'secret_token': secret_token[0]}) + return self.url_result(api_url) + + class SoundcloudIE(InfoExtractor): """Information extractor for soundcloud.com To access the media, the uid of the song and a stream token @@ -27,13 +65,13 @@ class SoundcloudIE(InfoExtractor): _VALID_URL = r'''(?x)^(?:https?://)? (?:(?:(?:www\.|m\.)?soundcloud\.com/ + (?!stations/track) (?P[\w\d-]+)/ - (?!sets/|likes/?(?:$|[?#])) + (?!(?:tracks|albums|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#])) (?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=.*) + |(?:api(?:-v2)?\.soundcloud\.com/tracks/(?P<track_id>\d+) + (?:/?\?secret_token=(?P<secret_token>[^&]+))?) ) ''' IE_NAME = 'soundcloud' @@ -44,14 +82,21 @@ class SoundcloudIE(InfoExtractor): 'info_dict': { 'id': '62986583', 'ext': 'mp3', - 'upload_date': '20121011', + 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1', '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, + 'uploader_id': '1571244', + 'timestamp': 1349920598, + 'upload_date': '20121011', + 'duration': 143.216, + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, } }, - # not streamable song + # not streamable song, preview { 'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep', 'info_dict': { @@ -60,8 +105,15 @@ class SoundcloudIE(InfoExtractor): 'title': 'Goldrushed', 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com', 'uploader': 'The Royal Concept', + 'uploader_id': '9615865', + 'timestamp': 1337635207, 'upload_date': '20120521', - 'duration': 227, + 'duration': 30, + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, }, 'params': { # rtmp @@ -76,10 +128,17 @@ class SoundcloudIE(InfoExtractor): 'id': '123998367', 'ext': 'mp3', 'title': 'Youtube - Dl Test Video \'\' Ä↭', - 'uploader': 'jaimeMF', 'description': 'test chars: \"\'/\\ä↭', + 'uploader': 'jaimeMF', + 'uploader_id': '69767071', + 'timestamp': 1386604920, 'upload_date': '20131209', - 'duration': 9, + 'duration': 9.927, + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, }, }, # private link (alt format) @@ -90,10 +149,17 @@ class SoundcloudIE(InfoExtractor): 'id': '123998367', 'ext': 'mp3', 'title': 'Youtube - Dl Test Video \'\' Ä↭', - 'uploader': 'jaimeMF', 'description': 'test chars: \"\'/\\ä↭', + 'uploader': 'jaimeMF', + 'uploader_id': '69767071', + 'timestamp': 1386604920, 'upload_date': '20131209', - 'duration': 9, + 'duration': 9.927, + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, }, }, # downloadable song @@ -106,244 +172,587 @@ class SoundcloudIE(InfoExtractor): 'title': 'Bus Brakes', 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66', 'uploader': 'oddsamples', + 'uploader_id': '73680509', + 'timestamp': 1389232924, 'upload_date': '20140109', - 'duration': 17, + 'duration': 17.346, + 'license': 'cc-by-sa', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, + }, + }, + # private link, downloadable format + { + 'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd', + 'md5': '64a60b16e617d41d0bef032b7f55441e', + 'info_dict': { + 'id': '340344461', + 'ext': 'wav', + 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]', + 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366', + 'uploader': 'Ori Uplift Music', + 'uploader_id': '12563093', + 'timestamp': 1504206263, + 'upload_date': '20170831', + 'duration': 7449.096, + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, + }, + }, + # no album art, use avatar pic for thumbnail + { + 'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real', + 'md5': '59c7872bc44e5d99b7211891664760c2', + 'info_dict': { + 'id': '309699954', + 'ext': 'mp3', + 'title': 'Sideways (Prod. Mad Real)', + 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e', + 'uploader': 'garyvee', + 'uploader_id': '2366352', + 'timestamp': 1488152409, + 'upload_date': '20170226', + 'duration': 207.012, + 'thumbnail': r're:https?://.*\.jpg', + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, + }, + 'params': { + 'skip_download': True, }, }, + # not available via api.soundcloud.com/i1/tracks/id/streams + { + 'url': 'https://soundcloud.com/giovannisarani/mezzo-valzer', + 'md5': 'e22aecd2bc88e0e4e432d7dcc0a1abf7', + 'info_dict': { + 'id': '583011102', + 'ext': 'mp3', + 'title': 'Mezzo Valzer', + 'description': 'md5:4138d582f81866a530317bae316e8b61', + 'uploader': 'Giovanni Sarani', + 'uploader_id': '3352531', + 'timestamp': 1551394171, + 'upload_date': '20190228', + 'duration': 180.157, + 'thumbnail': r're:https?://.*\.jpg', + 'license': 'all-rights-reserved', + 'view_count': int, + 'like_count': int, + 'comment_count': int, + 'repost_count': int, + }, + 'expected_warnings': ['Unable to download JSON metadata'], + } ] - _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28' - _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf' - - def report_resolve(self, video_id): - """Report information extraction.""" - self.to_screen('%s: Resolving id' % video_id) + _API_BASE = 'https://api.soundcloud.com/' + _API_V2_BASE = 'https://api-v2.soundcloud.com/' + _BASE_URL = 'https://soundcloud.com/' + _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg' + + _ARTWORK_MAP = { + 'mini': 16, + 'tiny': 20, + 'small': 32, + 'badge': 47, + 't67x67': 67, + 'large': 100, + 't300x300': 300, + 'crop': 400, + 't500x500': 500, + 'original': 0, + } + + def _store_client_id(self, client_id): + self._downloader.cache.store('soundcloud', 'client_id', client_id) + + def _update_client_id(self): + webpage = self._download_webpage('https://soundcloud.com/', None) + for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)): + script = self._download_webpage(src, None, fatal=False) + if script: + client_id = self._search_regex( + r'client_id\s*:\s*"([0-9a-zA-Z]{32})"', + script, 'client id', default=None) + if client_id: + self._CLIENT_ID = client_id + self._store_client_id(client_id) + return + raise ExtractorError('Unable to extract client id') + + def _download_json(self, *args, **kwargs): + non_fatal = kwargs.get('fatal') is False + if non_fatal: + del kwargs['fatal'] + query = kwargs.get('query', {}).copy() + for _ in range(2): + query['client_id'] = self._CLIENT_ID + kwargs['query'] = query + try: + return super(SoundcloudIE, self)._download_json(*args, **compat_kwargs(kwargs)) + except ExtractorError as e: + if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401: + self._store_client_id(None) + self._update_client_id() + continue + elif non_fatal: + self._downloader.report_warning(error_to_compat_str(e)) + return False + raise + + def _real_initialize(self): + self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or 'YUKXoArFcqrlQn9tfNHvvyfnDISj04zk' @classmethod def _resolv_url(cls, url): - return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID + return SoundcloudIE._API_V2_BASE + 'resolve?url=' + url - def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None): + def _extract_info_dict(self, info, full_title=None, secret_token=None, version=2): track_id = compat_str(info['id']) - name = full_title or track_id - 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, - 'uploader': info['user']['username'], - 'upload_date': unified_strdate(info['created_at']), - 'title': info['title'], - 'description': info['description'], - 'thumbnail': thumbnail, - 'duration': int_or_none(info.get('duration'), 1000), - 'webpage_url': info.get('permalink_url'), - } + title = info['title'] + track_base_url = self._API_BASE + 'tracks/%s' % track_id + + format_urls = set() formats = [] - if info.get('downloadable', False): - # 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)) + query = {'client_id': self._CLIENT_ID} + if secret_token: + query['secret_token'] = secret_token + + if info.get('downloadable') and info.get('has_downloads_left'): + format_url = update_url_query( + info.get('download_url') or track_base_url + '/download', query) + format_urls.add(format_url) + if version == 2: + v1_info = self._download_json( + track_base_url, track_id, query=query, fatal=False) or {} + else: + v1_info = info formats.append({ 'format_id': 'download', - 'ext': info.get('original_format', 'mp3'), + 'ext': v1_info.get('original_format') or 'mp3', + 'filesize': int_or_none(v1_info.get('original_content_size')), '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 invalid_url(url): + return not url or url in format_urls + + def add_format(f, protocol, is_preview=False): + mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url) + if mobj: + for k, v in mobj.groupdict().items(): + if not f.get(k): + f[k] = v + format_id_list = [] + if protocol: + format_id_list.append(protocol) + for k in ('ext', 'abr'): + v = f.get(k) + if v: + format_id_list.append(v) + preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url']) + if preview: + format_id_list.append('preview') + abr = f.get('abr') + if abr: + f['abr'] = int(abr) + f.update({ + 'format_id': '_'.join(format_id_list), + 'protocol': 'm3u8_native' if protocol == 'hls' else 'http', + 'preference': -10 if preview else None, + }) + formats.append(f) + + # New API + transcodings = try_get( + info, lambda x: x['media']['transcodings'], list) or [] + for t in transcodings: + if not isinstance(t, dict): + continue + format_url = url_or_none(t.get('url')) + if not format_url: + continue + stream = self._download_json( + format_url, track_id, query=query, fatal=False) + if not isinstance(stream, dict): + continue + stream_url = url_or_none(stream.get('url')) + if invalid_url(stream_url): + continue + format_urls.add(stream_url) + stream_format = t.get('format') or {} + protocol = stream_format.get('protocol') + if protocol != 'hls' and '/hls' in format_url: + protocol = 'hls' + ext = None + preset = str_or_none(t.get('preset')) + if preset: + ext = preset.split('_')[0] + if ext not in KNOWN_EXTENSIONS: + ext = mimetype2ext(stream_format.get('mime_type')) + add_format({ + 'url': stream_url, + 'ext': ext, + }, 'http' if protocol == 'progressive' else protocol, + t.get('snipped') or '/preview/' in format_url) + + if not formats: + # Old API, does not work for some tracks (e.g. + # https://soundcloud.com/giovannisarani/mezzo-valzer) + # and might serve preview URLs (e.g. + # http://www.soundcloud.com/snbrn/ele) + format_dict = self._download_json( + track_base_url + '/streams', track_id, + 'Downloading track url', query=query, fatal=False) or {} + + for key, stream_url in format_dict.items(): + if invalid_url(stream_url): + continue + format_urls.add(stream_url) + mobj = re.search(r'(http|hls)_([^_]+)_(\d+)_url', key) + if mobj: + protocol, ext, abr = mobj.groups() + add_format({ + 'abr': abr, + 'ext': ext, + 'url': stream_url, + }, protocol) + + 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 + urlh = self._request_webpage( + HEADRequest(info.get('stream_url') or track_base_url + '/stream'), + track_id, query=query, fatal=False) + if urlh: + stream_url = urlh.geturl() + if not invalid_url(stream_url): + add_format({'url': stream_url}, 'http') + + for f in formats: + f['vcodec'] = 'none' + + self._sort_formats(formats) + + user = info.get('user') or {} + + thumbnails = [] + artwork_url = info.get('artwork_url') + thumbnail = artwork_url or user.get('avatar_url') + if isinstance(thumbnail, compat_str): + if re.search(self._IMAGE_REPL_RE, thumbnail): + for image_id, size in self._ARTWORK_MAP.items(): + i = { + 'id': image_id, + 'url': re.sub(self._IMAGE_REPL_RE, '-%s.jpg' % image_id, thumbnail), + } + if image_id == 'tiny' and not artwork_url: + size = 18 + elif image_id == 'original': + i['preference'] = 10 + if size: + i.update({ + 'width': size, + 'height': size, + }) + thumbnails.append(i) + else: + thumbnails = [{'url': thumbnail}] + + def extract_count(key): + return int_or_none(info.get('%s_count' % key)) + + return { + 'id': track_id, + 'uploader': user.get('username'), + 'uploader_id': str_or_none(user.get('id')) or user.get('permalink'), + 'uploader_url': user.get('permalink_url'), + 'timestamp': unified_timestamp(info.get('created_at')), + 'title': title, + 'description': info.get('description'), + 'thumbnails': thumbnails, + 'duration': float_or_none(info.get('duration'), 1000), + 'webpage_url': info.get('permalink_url'), + 'license': info.get('license'), + 'view_count': extract_count('playback'), + 'like_count': extract_count('favoritings') or extract_count('likes'), + 'comment_count': extract_count('comment'), + 'repost_count': extract_count('reposts'), + 'genre': info.get('genre'), + 'formats': formats + } def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE) - if mobj is None: - raise ExtractorError('Invalid URL: %s' % url) + mobj = re.match(self._VALID_URL, 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 + + query = {} + if track_id: + info_json_url = self._API_V2_BASE + 'tracks/' + track_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]) + query['secret_token'] = token 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') + full_title = resolve_title = '%s/%s' % mobj.group('uploader', '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') + info_json_url = self._resolv_url(self._BASE_URL + resolve_title) + + version = 2 + info = self._download_json( + info_json_url, full_title, 'Downloading info JSON', query=query, fatal=False) + if not info: + info = self._download_json( + info_json_url.replace(self._API_V2_BASE, self._API_BASE), + full_title, 'Downloading info JSON', query=query) + version = 1 - return self._extract_info_dict(info, full_title, secret_token=token) + return self._extract_info_dict(info, full_title, token, version) -class SoundcloudSetIE(SoundcloudIE): - _VALID_URL = r'https?://(?:www\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?' +class SoundcloudPlaylistBaseIE(SoundcloudIE): + def _extract_track_entries(self, tracks, token=None): + entries = [] + for track in tracks: + track_id = str_or_none(track.get('id')) + url = track.get('permalink_url') + if not url: + if not track_id: + continue + url = self._API_V2_BASE + 'tracks/' + track_id + if token: + url += '?secret_token=' + token + entries.append(self.url_result( + url, SoundcloudIE.ie_key(), track_id)) + return entries + + +class SoundcloudSetIE(SoundcloudPlaylistBaseIE): + _VALID_URL = r'https?://(?:(?:www|m)\.)?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, + 'playlist_mincount': 5, + }, { + 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token', + 'only_matching': True, }] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - # extract uploader (which is in the url) - uploader = mobj.group('uploader') - # extract simple title (uploader + slug of song title) - slug_title = mobj.group('slug_title') - full_title = '%s/sets/%s' % (uploader, slug_title) - url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title) - + full_title = '%s/sets/%s' % mobj.group('uploader', 'slug_title') token = mobj.group('token') if token: full_title += '/' + token - url += '/' + token - self.report_resolve(full_title) - - resolv_url = self._resolv_url(url) - info = self._download_json(resolv_url, full_title) + info = self._download_json(self._resolv_url( + self._BASE_URL + full_title), 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)) + + entries = self._extract_track_entries(info['tracks'], token) + + return self.playlist_result( + entries, str_or_none(info.get('id')), info.get('title')) + + +class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE): + def _extract_playlist(self, base_url, playlist_id, playlist_title): + COMMON_QUERY = { + 'limit': 2000000000, + 'linked_partitioning': '1', + } + + query = COMMON_QUERY.copy() + query['offset'] = 0 + + next_href = base_url + + entries = [] + for i in itertools.count(): + response = self._download_json( + next_href, playlist_id, + 'Downloading track page %s' % (i + 1), query=query) + + collection = response['collection'] + + if not isinstance(collection, list): + collection = [] + + # Empty collection may be returned, in this case we proceed + # straight to next_href + + def resolve_entry(candidates): + for cand in candidates: + if not isinstance(cand, dict): + continue + permalink_url = url_or_none(cand.get('permalink_url')) + if not permalink_url: + continue + return self.url_result( + permalink_url, + SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None, + str_or_none(cand.get('id')), cand.get('title')) + + for e in collection: + entry = resolve_entry((e, e.get('track'), e.get('playlist'))) + if entry: + entries.append(entry) + + next_href = response.get('next_href') + if not next_href: + break + + next_href = response['next_href'] + parsed_next_href = compat_urlparse.urlparse(next_href) + query = compat_urlparse.parse_qs(parsed_next_href.query) + query.update(COMMON_QUERY) return { '_type': 'playlist', - 'entries': [self._extract_info_dict(track, secret_token=token) for track in info['tracks']], - 'id': info['id'], - 'title': info['title'], + 'id': playlist_id, + 'title': playlist_title, + 'entries': entries, } -class SoundcloudUserIE(SoundcloudIE): - _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)/?((?P<rsrc>tracks|likes)/?)?(\?.*)?$' +class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE): + _VALID_URL = r'''(?x) + https?:// + (?:(?:www|m)\.)?soundcloud\.com/ + (?P<user>[^/]+) + (?:/ + (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight) + )? + /?(?:[?#].*)?$ + ''' IE_NAME = 'soundcloud:user' _TESTS = [{ - 'url': 'https://soundcloud.com/the-concept-band', + 'url': 'https://soundcloud.com/soft-cell-official', + 'info_dict': { + 'id': '207965082', + 'title': 'Soft Cell (All)', + }, + 'playlist_mincount': 28, + }, { + 'url': 'https://soundcloud.com/soft-cell-official/tracks', 'info_dict': { - 'id': '9615865', - 'title': 'The Royal Concept', + 'id': '207965082', + 'title': 'Soft Cell (Tracks)', }, - 'playlist_mincount': 12 + 'playlist_mincount': 27, }, { - 'url': 'https://soundcloud.com/the-concept-band/likes', + 'url': 'https://soundcloud.com/soft-cell-official/albums', 'info_dict': { - 'id': '9615865', - 'title': 'The Royal Concept', + 'id': '207965082', + 'title': 'Soft Cell (Albums)', + }, + 'playlist_mincount': 1, + }, { + 'url': 'https://soundcloud.com/jcv246/sets', + 'info_dict': { + 'id': '12982173', + 'title': 'Jordi / cv (Sets)', + }, + 'playlist_mincount': 2, + }, { + 'url': 'https://soundcloud.com/jcv246/reposts', + 'info_dict': { + 'id': '12982173', + 'title': 'Jordi / cv (Reposts)', + }, + 'playlist_mincount': 6, + }, { + 'url': 'https://soundcloud.com/clalberg/likes', + 'info_dict': { + 'id': '11817582', + 'title': 'clalberg (Likes)', + }, + 'playlist_mincount': 5, + }, { + 'url': 'https://soundcloud.com/grynpyret/spotlight', + 'info_dict': { + 'id': '7098329', + 'title': 'Grynpyret (Spotlight)', }, 'playlist_mincount': 1, }] + _BASE_URL_MAP = { + 'all': 'stream/users/%s', + 'tracks': 'users/%s/tracks', + 'albums': 'users/%s/albums', + 'sets': 'users/%s/playlists', + 'reposts': 'stream/users/%s/reposts', + 'likes': 'users/%s/likes', + 'spotlight': 'users/%s/spotlight', + } + 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 = self._download_json( - resolv_url, uploader, 'Downloading user info') - base_url = 'http://api.soundcloud.com/users/%s/%s.json?' % (uploader, resource) + self._resolv_url(self._BASE_URL + uploader), + uploader, 'Downloading user info') - entries = [] - for i in itertools.count(): - 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) + resource = mobj.group('rsrc') or 'all' - return { - '_type': 'playlist', - 'id': compat_str(user['id']), - 'title': user['username'], - 'entries': entries, - } + return self._extract_playlist( + self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'], + str_or_none(user.get('id')), + '%s (%s)' % (user['username'], resource.capitalize())) -class SoundcloudPlaylistIE(SoundcloudIE): - _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$' +class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE): + _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)' + IE_NAME = 'soundcloud:trackstation' + _TESTS = [{ + 'url': 'https://soundcloud.com/stations/track/officialsundial/your-text', + 'info_dict': { + 'id': '286017854', + 'title': 'Track station: your text', + }, + 'playlist_mincount': 47, + }] + + def _real_extract(self, url): + track_name = self._match_id(url) + + track = self._download_json(self._resolv_url(url), track_name) + track_id = self._search_regex( + r'soundcloud:track-stations:(\d+)', track['id'], 'track id') + + return self._extract_playlist( + self._API_V2_BASE + 'stations/%s/tracks' % track['id'], + track_id, 'Track station: %s' % track['title']) + + +class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE): + _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$' IE_NAME = 'soundcloud:playlist' _TESTS = [{ - 'url': 'http://api.soundcloud.com/playlists/4110309', + 'url': 'https://api.soundcloud.com/playlists/4110309', 'info_dict': { 'id': '4110309', 'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]', @@ -355,28 +764,73 @@ class SoundcloudPlaylistIE(SoundcloudIE): 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, - } + query = {} token = mobj.group('token') - if token: - data_dict['secret_token'] = token + query['secret_token'] = token - data = compat_urllib_parse.urlencode(data_dict) data = self._download_json( - base_url + data, playlist_id, 'Downloading playlist') + self._API_V2_BASE + 'playlists/' + playlist_id, + playlist_id, 'Downloading playlist', query=query) - entries = [ - self._extract_info_dict(t, quiet=True, secret_token=token) - for t in data['tracks']] + entries = self._extract_track_entries(data['tracks'], token) - return { - '_type': 'playlist', - 'id': playlist_id, - 'title': data.get('title'), - 'description': data.get('description'), - 'entries': entries, - } + return self.playlist_result( + entries, playlist_id, data.get('title'), data.get('description')) + + +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 + + def _get_collection(self, endpoint, collection_id, **query): + limit = min( + query.get('limit', self._DEFAULT_RESULTS_PER_PAGE), + self._MAX_RESULTS_PER_PAGE) + query.update({ + 'limit': limit, + 'linked_partitioning': 1, + 'offset': 0, + }) + next_url = update_url_query(self._API_V2_BASE + endpoint, query) + + 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)