yt_dlp/extractor/soundcloud.py

   1 import functools
   2 import itertools
   3 import json
   4 import re
   5
   6 from .common import InfoExtractor, SearchInfoExtractor
   7 from ..compat import compat_str
   8 from ..networking import HEADRequest
   9 from ..networking.exceptions import HTTPError
  10 from ..utils import (
  11     KNOWN_EXTENSIONS,
  12     ExtractorError,
  13     error_to_compat_str,
  14     float_or_none,
  15     int_or_none,
  16     join_nonempty,
  17     mimetype2ext,
  18     parse_qs,
  19     str_or_none,
  20     try_call,
  21     unified_timestamp,
  22     update_url_query,
  23     url_or_none,
  24     urlhandle_detect_ext,
  25 )
  26 from ..utils.traversal import traverse_obj
  27
  28
  29 class SoundcloudEmbedIE(InfoExtractor):
  30     _VALID_URL = r'https?://(?:w|player|p)\.soundcloud\.com/player/?.*?\burl=(?P<id>.+)'
  31     _EMBED_REGEX = [r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1']
  32     _TEST = {
  33         # from https://www.soundi.fi/uutiset/ennakkokuuntelussa-timo-kaukolammen-station-to-station-to-station-julkaisua-juhlitaan-tanaan-g-livelabissa/
  34         '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',
  35         'only_matching': True,
  36     }
  37
  38     def _real_extract(self, url):
  39         query = parse_qs(url)
  40         api_url = query['url'][0]
  41         secret_token = query.get('secret_token')
  42         if secret_token:
  43             api_url = update_url_query(api_url, {'secret_token': secret_token[0]})
  44         return self.url_result(api_url)
  45
  46
  47 class SoundcloudBaseIE(InfoExtractor):
  48     _NETRC_MACHINE = 'soundcloud'
  49
  50     _API_V2_BASE = 'https://api-v2.soundcloud.com/'
  51     _BASE_URL = 'https://soundcloud.com/'
  52     _USER_AGENT = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.105 Safari/537.36'
  53     _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
  54     _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
  55     _API_VERIFY_AUTH_TOKEN = 'https://api-auth.soundcloud.com/connect/session%s'
  56     _HEADERS = {}
  57
  58     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
  59
  60     _ARTWORK_MAP = {
  61         'mini': 16,
  62         'tiny': 20,
  63         'small': 32,
  64         'badge': 47,
  65         't67x67': 67,
  66         'large': 100,
  67         't300x300': 300,
  68         'crop': 400,
  69         't500x500': 500,
  70         'original': 0,
  71     }
  72
  73     _DEFAULT_FORMATS = ['http_aac', 'hls_aac', 'http_opus', 'hls_opus', 'http_mp3', 'hls_mp3']
  74
  75     @functools.cached_property
  76     def _is_requested(self):
  77         return re.compile(r'|'.join(set(
  78             re.escape(pattern).replace(r'\*', r'.*') if pattern != 'default'
  79             else '|'.join(map(re.escape, self._DEFAULT_FORMATS))
  80             for pattern in self._configuration_arg('formats', ['default'], ie_key=SoundcloudIE)
  81         ))).fullmatch
  82
  83     def _store_client_id(self, client_id):
  84         self.cache.store('soundcloud', 'client_id', client_id)
  85
  86     def _update_client_id(self):
  87         webpage = self._download_webpage('https://soundcloud.com/', None)
  88         for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
  89             script = self._download_webpage(src, None, fatal=False)
  90             if script:
  91                 client_id = self._search_regex(
  92                     r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
  93                     script, 'client id', default=None)
  94                 if client_id:
  95                     self._CLIENT_ID = client_id
  96                     self._store_client_id(client_id)
  97                     return
  98         raise ExtractorError('Unable to extract client id')
  99
 100     def _download_json(self, *args, **kwargs):
 101         non_fatal = kwargs.get('fatal') is False
 102         if non_fatal:
 103             del kwargs['fatal']
 104         query = kwargs.get('query', {}).copy()
 105         for _ in range(2):
 106             query['client_id'] = self._CLIENT_ID
 107             kwargs['query'] = query
 108             try:
 109                 return super()._download_json(*args, **kwargs)
 110             except ExtractorError as e:
 111                 if isinstance(e.cause, HTTPError) and e.cause.status in (401, 403):
 112                     self._store_client_id(None)
 113                     self._update_client_id()
 114                     continue
 115                 elif non_fatal:
 116                     self.report_warning(error_to_compat_str(e))
 117                     return False
 118                 raise
 119
 120     def _initialize_pre_login(self):
 121         self._CLIENT_ID = self.cache.load('soundcloud', 'client_id') or 'a3e059563d7fd3372b49b37f00a00bcf'
 122
 123     def _verify_oauth_token(self, token):
 124         if self._request_webpage(
 125                 self._API_VERIFY_AUTH_TOKEN % (self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID),
 126                 None, note='Verifying login token...', fatal=False,
 127                 data=json.dumps({'session': {'access_token': token}}).encode()):
 128             self._HEADERS['Authorization'] = f'OAuth {token}'
 129             self.report_login()
 130         else:
 131             self.report_warning('Provided authorization token is invalid. Continuing as guest')
 132
 133     def _real_initialize(self):
 134         if self._HEADERS:
 135             return
 136         if token := try_call(lambda: self._get_cookies(self._BASE_URL)['oauth_token'].value):
 137             self._verify_oauth_token(token)
 138
 139     def _perform_login(self, username, password):
 140         if username != 'oauth':
 141             raise ExtractorError(
 142                 'Login using username and password is not currently supported. '
 143                 'Use "--username oauth --password <oauth_token>" to login using an oauth token, '
 144                 f'or else {self._login_hint(method="cookies")}', expected=True)
 145         if self._HEADERS:
 146             return
 147         self._verify_oauth_token(password)
 148
 149         r'''
 150         def genDevId():
 151             def genNumBlock():
 152                 return ''.join([str(random.randrange(10)) for i in range(6)])
 153             return '-'.join([genNumBlock() for i in range(4)])
 154
 155         payload = {
 156             'client_id': self._CLIENT_ID,
 157             'recaptcha_pubkey': 'null',
 158             'recaptcha_response': 'null',
 159             'credentials': {
 160                 'identifier': username,
 161                 'password': password
 162             },
 163             'signature': self.sign(username, password, self._CLIENT_ID),
 164             'device_id': genDevId(),
 165             'user_agent': self._USER_AGENT
 166         }
 167
 168         response = self._download_json(
 169             self._API_AUTH_URL_PW % (self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID),
 170             None, note='Verifying login token...', fatal=False,
 171             data=json.dumps(payload).encode())
 172
 173         if token := traverse_obj(response, ('session', 'access_token', {str})):
 174             self._HEADERS['Authorization'] = f'OAuth {token}'
 175             self.report_login()
 176             return
 177
 178         raise ExtractorError('Unable to get access token, login may have failed', expected=True)
 179         '''
 180
 181     # signature generation
 182     def sign(self, user, pw, clid):
 183         a = 33
 184         i = 1
 185         s = 440123
 186         w = 117
 187         u = 1800000
 188         l = 1042
 189         b = 37
 190         k = 37
 191         c = 5
 192         n = '0763ed7314c69015fd4a0dc16bbf4b90'  # _KEY
 193         y = '8'  # _REV
 194         r = 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/84.0.4147.105 Safari/537.36'  # _USER_AGENT
 195         e = user  # _USERNAME
 196         t = clid  # _CLIENT_ID
 197
 198         d = '-'.join([str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
 199         p = n + y + d + r + e + t + d + n
 200         h = p
 201
 202         m = 8011470
 203         f = 0
 204
 205         for f in range(f, len(h)):
 206             m = (m >> 1) + ((1 & m) << 23)
 207             m += ord(h[f])
 208             m &= 16777215
 209
 210         # c is not even needed
 211         out = str(y) + ':' + str(d) + ':' + format(m, 'x') + ':' + str(c)
 212
 213         return out
 214
 215     def _extract_info_dict(self, info, full_title=None, secret_token=None, extract_flat=False):
 216         track_id = compat_str(info['id'])
 217         title = info['title']
 218
 219         format_urls = set()
 220         formats = []
 221         query = {'client_id': self._CLIENT_ID}
 222         if secret_token:
 223             query['secret_token'] = secret_token
 224
 225         if not extract_flat and info.get('downloadable') and info.get('has_downloads_left'):
 226             download_url = update_url_query(
 227                 self._API_V2_BASE + 'tracks/' + track_id + '/download', query)
 228             redirect_url = (self._download_json(download_url, track_id, fatal=False) or {}).get('redirectUri')
 229             if redirect_url:
 230                 urlh = self._request_webpage(
 231                     HEADRequest(redirect_url), track_id, 'Checking for original download format', fatal=False)
 232                 if urlh:
 233                     format_url = urlh.url
 234                     format_urls.add(format_url)
 235                     formats.append({
 236                         'format_id': 'download',
 237                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
 238                         'filesize': int_or_none(urlh.headers.get('Content-Length')),
 239                         'url': format_url,
 240                         'quality': 10,
 241                         'format_note': 'Original',
 242                     })
 243
 244         def invalid_url(url):
 245             return not url or url in format_urls
 246
 247         def add_format(f, protocol, is_preview=False):
 248             mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url)
 249             if mobj:
 250                 for k, v in mobj.groupdict().items():
 251                     if not f.get(k):
 252                         f[k] = v
 253             format_id_list = []
 254             if protocol:
 255                 format_id_list.append(protocol)
 256             ext = f.get('ext')
 257             if ext == 'aac':
 258                 f.update({
 259                     'abr': 256,
 260                     'quality': 5,
 261                     'format_note': 'Premium',
 262                 })
 263             for k in ('ext', 'abr'):
 264                 v = str_or_none(f.get(k))
 265                 if v:
 266                     format_id_list.append(v)
 267             preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url'])
 268             if preview:
 269                 format_id_list.append('preview')
 270             abr = f.get('abr')
 271             if abr:
 272                 f['abr'] = int(abr)
 273             if protocol in ('hls', 'hls-aes'):
 274                 protocol = 'm3u8' if ext == 'aac' else 'm3u8_native'
 275             else:
 276                 protocol = 'http'
 277             f.update({
 278                 'format_id': '_'.join(format_id_list),
 279                 'protocol': protocol,
 280                 'preference': -10 if preview else None,
 281             })
 282             formats.append(f)
 283
 284         # New API
 285         for t in traverse_obj(info, ('media', 'transcodings', lambda _, v: url_or_none(v['url']))):
 286             if extract_flat:
 287                 break
 288             format_url = t['url']
 289
 290             protocol = traverse_obj(t, ('format', 'protocol', {str}))
 291             if protocol == 'progressive':
 292                 protocol = 'http'
 293             if protocol != 'hls' and '/hls' in format_url:
 294                 protocol = 'hls'
 295             if protocol == 'encrypted-hls' or '/encrypted-hls' in format_url:
 296                 protocol = 'hls-aes'
 297
 298             ext = None
 299             if preset := traverse_obj(t, ('preset', {str_or_none})):
 300                 ext = preset.split('_')[0]
 301             if ext not in KNOWN_EXTENSIONS:
 302                 ext = mimetype2ext(traverse_obj(t, ('format', 'mime_type', {str})))
 303
 304             identifier = join_nonempty(protocol, ext, delim='_')
 305             if not self._is_requested(identifier):
 306                 self.write_debug(f'"{identifier}" is not a requested format, skipping')
 307                 continue
 308
 309             stream = None
 310             for retry in self.RetryManager(fatal=False):
 311                 try:
 312                     stream = self._download_json(
 313                         format_url, track_id, f'Downloading {identifier} format info JSON',
 314                         query=query, headers=self._HEADERS)
 315                 except ExtractorError as e:
 316                     if isinstance(e.cause, HTTPError) and e.cause.status == 429:
 317                         self.report_warning(
 318                             'You have reached the API rate limit, which is ~600 requests per '
 319                             '10 minutes. Use the --extractor-retries and --retry-sleep options '
 320                             'to configure an appropriate retry count and wait time', only_once=True)
 321                         retry.error = e.cause
 322                     else:
 323                         self.report_warning(e.msg)
 324
 325             stream_url = traverse_obj(stream, ('url', {url_or_none}))
 326             if invalid_url(stream_url):
 327                 continue
 328             format_urls.add(stream_url)
 329             add_format({
 330                 'url': stream_url,
 331                 'ext': ext,
 332             }, protocol, t.get('snipped') or '/preview/' in format_url)
 333
 334         for f in formats:
 335             f['vcodec'] = 'none'
 336
 337         if not formats and info.get('policy') == 'BLOCK':
 338             self.raise_geo_restricted(metadata_available=True)
 339
 340         user = info.get('user') or {}
 341
 342         thumbnails = []
 343         artwork_url = info.get('artwork_url')
 344         thumbnail = artwork_url or user.get('avatar_url')
 345         if isinstance(thumbnail, compat_str):
 346             if re.search(self._IMAGE_REPL_RE, thumbnail):
 347                 for image_id, size in self._ARTWORK_MAP.items():
 348                     i = {
 349                         'id': image_id,
 350                         'url': re.sub(self._IMAGE_REPL_RE, '-%s.jpg' % image_id, thumbnail),
 351                     }
 352                     if image_id == 'tiny' and not artwork_url:
 353                         size = 18
 354                     elif image_id == 'original':
 355                         i['preference'] = 10
 356                     if size:
 357                         i.update({
 358                             'width': size,
 359                             'height': size,
 360                         })
 361                     thumbnails.append(i)
 362             else:
 363                 thumbnails = [{'url': thumbnail}]
 364
 365         def extract_count(key):
 366             return int_or_none(info.get('%s_count' % key))
 367
 368         return {
 369             'id': track_id,
 370             'uploader': user.get('username'),
 371             'uploader_id': str_or_none(user.get('id')) or user.get('permalink'),
 372             'uploader_url': user.get('permalink_url'),
 373             'timestamp': unified_timestamp(info.get('created_at')),
 374             'title': title,
 375             'description': info.get('description'),
 376             'thumbnails': thumbnails,
 377             'duration': float_or_none(info.get('duration'), 1000),
 378             'webpage_url': info.get('permalink_url'),
 379             'license': info.get('license'),
 380             'view_count': extract_count('playback'),
 381             'like_count': extract_count('favoritings') or extract_count('likes'),
 382             'comment_count': extract_count('comment'),
 383             'repost_count': extract_count('reposts'),
 384             'genres': traverse_obj(info, ('genre', {str}, {lambda x: x or None}, all)),
 385             'formats': formats if not extract_flat else None
 386         }
 387
 388     @classmethod
 389     def _resolv_url(cls, url):
 390         return cls._API_V2_BASE + 'resolve?url=' + url
 391
 392
 393 class SoundcloudIE(SoundcloudBaseIE):
 394     """Information extractor for soundcloud.com
 395        To access the media, the uid of the song and a stream token
 396        must be extracted from the page source and the script must make
 397        a request to media.soundcloud.com/crossdomain.xml. Then
 398        the media can be grabbed by requesting from an url composed
 399        of the stream token and uid
 400      """
 401
 402     _VALID_URL = r'''(?x)^(?:https?://)?
 403                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
 404                             (?!stations/track)
 405                             (?P<uploader>[\w\d-]+)/
 406                             (?!(?:tracks|albums|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
 407                             (?P<title>[\w\d-]+)
 408                             (?:/(?P<token>(?!(?:albums|sets|recommended))[^?]+?))?
 409                             (?:[?].*)?$)
 410                        |(?:api(?:-v2)?\.soundcloud\.com/tracks/(?P<track_id>\d+)
 411                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
 412                     )
 413                     '''
 414     IE_NAME = 'soundcloud'
 415     _TESTS = [
 416         {
 417             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
 418             'md5': 'de9bac153e7427a7333b4b0c1b6a18d2',
 419             'info_dict': {
 420                 'id': '62986583',
 421                 'ext': 'opus',
 422                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
 423                 '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',
 424                 'uploader': 'E.T. ExTerrestrial Music',
 425                 'uploader_id': '1571244',
 426                 'timestamp': 1349920598,
 427                 'upload_date': '20121011',
 428                 'duration': 143.216,
 429                 'license': 'all-rights-reserved',
 430                 'view_count': int,
 431                 'like_count': int,
 432                 'comment_count': int,
 433                 'repost_count': int,
 434                 'thumbnail': 'https://i1.sndcdn.com/artworks-000031955188-rwb18x-original.jpg',
 435                 'uploader_url': 'https://soundcloud.com/ethmusic',
 436                 'genres': [],
 437             }
 438         },
 439         # geo-restricted
 440         {
 441             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
 442             'info_dict': {
 443                 'id': '47127627',
 444                 'ext': 'opus',
 445                 'title': 'Goldrushed',
 446                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
 447                 'uploader': 'The Royal Concept',
 448                 'uploader_id': '9615865',
 449                 'timestamp': 1337635207,
 450                 'upload_date': '20120521',
 451                 'duration': 227.155,
 452                 'license': 'all-rights-reserved',
 453                 'view_count': int,
 454                 'like_count': int,
 455                 'comment_count': int,
 456                 'repost_count': int,
 457                 'uploader_url': 'https://soundcloud.com/the-concept-band',
 458                 'thumbnail': 'https://i1.sndcdn.com/artworks-v8bFHhXm7Au6-0-original.jpg',
 459                 'genres': ['Alternative'],
 460             },
 461         },
 462         # private link
 463         {
 464             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
 465             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 466             'info_dict': {
 467                 'id': '123998367',
 468                 'ext': 'mp3',
 469                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 470                 'description': 'test chars:  \"\'/\\ä↭',
 471                 'uploader': 'jaimeMF',
 472                 'uploader_id': '69767071',
 473                 'timestamp': 1386604920,
 474                 'upload_date': '20131209',
 475                 'duration': 9.927,
 476                 'license': 'all-rights-reserved',
 477                 'view_count': int,
 478                 'like_count': int,
 479                 'comment_count': int,
 480                 'repost_count': int,
 481                 'uploader_url': 'https://soundcloud.com/jaimemf',
 482                 'thumbnail': 'https://a1.sndcdn.com/images/default_avatar_large.png',
 483                 'genres': ['youtubedl'],
 484             },
 485         },
 486         # private link (alt format)
 487         {
 488             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
 489             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 490             'info_dict': {
 491                 'id': '123998367',
 492                 'ext': 'mp3',
 493                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 494                 'description': 'test chars:  \"\'/\\ä↭',
 495                 'uploader': 'jaimeMF',
 496                 'uploader_id': '69767071',
 497                 'timestamp': 1386604920,
 498                 'upload_date': '20131209',
 499                 'duration': 9.927,
 500                 'license': 'all-rights-reserved',
 501                 'view_count': int,
 502                 'like_count': int,
 503                 'comment_count': int,
 504                 'repost_count': int,
 505                 'uploader_url': 'https://soundcloud.com/jaimemf',
 506                 'thumbnail': 'https://a1.sndcdn.com/images/default_avatar_large.png',
 507                 'genres': ['youtubedl'],
 508             },
 509         },
 510         # downloadable song
 511         {
 512             'url': 'https://soundcloud.com/the80m/the-following',
 513             'md5': '9ffcddb08c87d74fb5808a3c183a1d04',
 514             'info_dict': {
 515                 'id': '343609555',
 516                 'ext': 'wav',
 517                 'title': 'The Following',
 518                 'description': '',
 519                 'uploader': '80M',
 520                 'uploader_id': '312384765',
 521                 'uploader_url': 'https://soundcloud.com/the80m',
 522                 'upload_date': '20170922',
 523                 'timestamp': 1506120436,
 524                 'duration': 397.228,
 525                 'thumbnail': 'https://i1.sndcdn.com/artworks-000243916348-ktoo7d-original.jpg',
 526                 'license': 'all-rights-reserved',
 527                 'like_count': int,
 528                 'comment_count': int,
 529                 'repost_count': int,
 530                 'view_count': int,
 531                 'genres': ['Dance & EDM'],
 532             },
 533         },
 534         # private link, downloadable format
 535         {
 536             'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
 537             'md5': '64a60b16e617d41d0bef032b7f55441e',
 538             'info_dict': {
 539                 'id': '340344461',
 540                 'ext': 'wav',
 541                 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
 542                 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
 543                 'uploader': 'Ori Uplift Music',
 544                 'uploader_id': '12563093',
 545                 'timestamp': 1504206263,
 546                 'upload_date': '20170831',
 547                 'duration': 7449.096,
 548                 'license': 'all-rights-reserved',
 549                 'view_count': int,
 550                 'like_count': int,
 551                 'comment_count': int,
 552                 'repost_count': int,
 553                 'thumbnail': 'https://i1.sndcdn.com/artworks-000240712245-kedn4p-original.jpg',
 554                 'uploader_url': 'https://soundcloud.com/oriuplift',
 555                 'genres': ['Trance'],
 556             },
 557         },
 558         # no album art, use avatar pic for thumbnail
 559         {
 560             'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
 561             'md5': '59c7872bc44e5d99b7211891664760c2',
 562             'info_dict': {
 563                 'id': '309699954',
 564                 'ext': 'mp3',
 565                 'title': 'Sideways (Prod. Mad Real)',
 566                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
 567                 'uploader': 'garyvee',
 568                 'uploader_id': '2366352',
 569                 'timestamp': 1488152409,
 570                 'upload_date': '20170226',
 571                 'duration': 207.012,
 572                 'thumbnail': r're:https?://.*\.jpg',
 573                 'license': 'all-rights-reserved',
 574                 'view_count': int,
 575                 'like_count': int,
 576                 'comment_count': int,
 577                 'repost_count': int,
 578                 'uploader_url': 'https://soundcloud.com/garyvee',
 579                 'genres': [],
 580             },
 581             'params': {
 582                 'skip_download': True,
 583             },
 584         },
 585         {
 586             'url': 'https://soundcloud.com/giovannisarani/mezzo-valzer',
 587             'md5': '8227c3473a4264df6b02ad7e5b7527ac',
 588             'info_dict': {
 589                 'id': '583011102',
 590                 'ext': 'opus',
 591                 'title': 'Mezzo Valzer',
 592                 'description': 'md5:f4d5f39d52e0ccc2b4f665326428901a',
 593                 'uploader': 'Giovanni Sarani',
 594                 'uploader_id': '3352531',
 595                 'timestamp': 1551394171,
 596                 'upload_date': '20190228',
 597                 'duration': 180.157,
 598                 'thumbnail': r're:https?://.*\.jpg',
 599                 'license': 'all-rights-reserved',
 600                 'view_count': int,
 601                 'like_count': int,
 602                 'comment_count': int,
 603                 'repost_count': int,
 604                 'genres': ['Piano'],
 605                 'uploader_url': 'https://soundcloud.com/giovannisarani',
 606             },
 607         },
 608         {
 609             # AAC HQ format available (account with active subscription needed)
 610             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
 611             'only_matching': True,
 612         },
 613         {
 614             # Go+ (account with active subscription needed)
 615             'url': 'https://soundcloud.com/taylorswiftofficial/look-what-you-made-me-do',
 616             'only_matching': True,
 617         },
 618     ]
 619
 620     def _real_extract(self, url):
 621         mobj = self._match_valid_url(url)
 622
 623         track_id = mobj.group('track_id')
 624
 625         query = {}
 626         if track_id:
 627             info_json_url = self._API_V2_BASE + 'tracks/' + track_id
 628             full_title = track_id
 629             token = mobj.group('secret_token')
 630             if token:
 631                 query['secret_token'] = token
 632         else:
 633             full_title = resolve_title = '%s/%s' % mobj.group('uploader', 'title')
 634             token = mobj.group('token')
 635             if token:
 636                 resolve_title += '/%s' % token
 637             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 638
 639         info = self._download_json(
 640             info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 641
 642         return self._extract_info_dict(info, full_title, token)
 643
 644
 645 class SoundcloudPlaylistBaseIE(SoundcloudBaseIE):
 646     def _extract_set(self, playlist, token=None):
 647         playlist_id = compat_str(playlist['id'])
 648         tracks = playlist.get('tracks') or []
 649         if not all([t.get('permalink_url') for t in tracks]) and token:
 650             tracks = self._download_json(
 651                 self._API_V2_BASE + 'tracks', playlist_id,
 652                 'Downloading tracks', query={
 653                     'ids': ','.join([compat_str(t['id']) for t in tracks]),
 654                     'playlistId': playlist_id,
 655                     'playlistSecretToken': token,
 656                 }, headers=self._HEADERS)
 657         entries = []
 658         for track in tracks:
 659             track_id = str_or_none(track.get('id'))
 660             url = track.get('permalink_url')
 661             if not url:
 662                 if not track_id:
 663                     continue
 664                 url = self._API_V2_BASE + 'tracks/' + track_id
 665                 if token:
 666                     url += '?secret_token=' + token
 667             entries.append(self.url_result(
 668                 url, SoundcloudIE.ie_key(), track_id))
 669         return self.playlist_result(
 670             entries, playlist_id,
 671             playlist.get('title'),
 672             playlist.get('description'))
 673
 674
 675 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 676     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[:\w\d-]+)(?:/(?P<token>[^?/]+))?'
 677     IE_NAME = 'soundcloud:set'
 678     _TESTS = [{
 679         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 680         'info_dict': {
 681             'id': '2284613',
 682             'title': 'The Royal Concept EP',
 683             'description': 'md5:71d07087c7a449e8941a70a29e34671e',
 684         },
 685         'playlist_mincount': 5,
 686     }, {
 687         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 688         'only_matching': True,
 689     }, {
 690         'url': 'https://soundcloud.com/discover/sets/weekly::flacmatic',
 691         'only_matching': True,
 692     }, {
 693         'url': 'https://soundcloud.com/discover/sets/charts-top:all-music:de',
 694         'only_matching': True,
 695     }, {
 696         'url': 'https://soundcloud.com/discover/sets/charts-top:hiphoprap:kr',
 697         'only_matching': True,
 698     }]
 699
 700     def _real_extract(self, url):
 701         mobj = self._match_valid_url(url)
 702
 703         full_title = '%s/sets/%s' % mobj.group('uploader', 'slug_title')
 704         token = mobj.group('token')
 705         if token:
 706             full_title += '/' + token
 707
 708         info = self._download_json(self._resolv_url(
 709             self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 710
 711         if 'errors' in info:
 712             msgs = (compat_str(err['error_message']) for err in info['errors'])
 713             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 714
 715         return self._extract_set(info, token)
 716
 717
 718 class SoundcloudPagedPlaylistBaseIE(SoundcloudBaseIE):
 719     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 720         return {
 721             '_type': 'playlist',
 722             'id': playlist_id,
 723             'title': playlist_title,
 724             'entries': self._entries(base_url, playlist_id),
 725         }
 726
 727     def _entries(self, url, playlist_id):
 728         # Per the SoundCloud documentation, the maximum limit for a linked partitioning query is 200.
 729         # https://developers.soundcloud.com/blog/offset-pagination-deprecated
 730         query = {
 731             'limit': 200,
 732             'linked_partitioning': '1',
 733             'offset': 0,
 734         }
 735
 736         for i in itertools.count():
 737             for retry in self.RetryManager():
 738                 try:
 739                     response = self._download_json(
 740                         url, playlist_id, query=query, headers=self._HEADERS,
 741                         note=f'Downloading track page {i + 1}')
 742                     break
 743                 except ExtractorError as e:
 744                     # Downloading page may result in intermittent 502 HTTP error
 745                     # See https://github.com/yt-dlp/yt-dlp/issues/872
 746                     if not isinstance(e.cause, HTTPError) or e.cause.status != 502:
 747                         raise
 748                     retry.error = e
 749                     continue
 750
 751             def resolve_entry(*candidates):
 752                 for cand in candidates:
 753                     if not isinstance(cand, dict):
 754                         continue
 755                     permalink_url = url_or_none(cand.get('permalink_url'))
 756                     if permalink_url:
 757                         return self.url_result(
 758                             permalink_url,
 759                             SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
 760                             str_or_none(cand.get('id')), cand.get('title'))
 761
 762             for e in response['collection'] or []:
 763                 yield resolve_entry(e, e.get('track'), e.get('playlist'))
 764
 765             url = response.get('next_href')
 766             if not url:
 767                 break
 768             query.pop('offset', None)
 769
 770
 771 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 772     _VALID_URL = r'''(?x)
 773                         https?://
 774                             (?:(?:www|m)\.)?soundcloud\.com/
 775                             (?P<user>[^/]+)
 776                             (?:/
 777                                 (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight)
 778                             )?
 779                             /?(?:[?#].*)?$
 780                     '''
 781     IE_NAME = 'soundcloud:user'
 782     _TESTS = [{
 783         'url': 'https://soundcloud.com/soft-cell-official',
 784         'info_dict': {
 785             'id': '207965082',
 786             'title': 'Soft Cell (All)',
 787         },
 788         'playlist_mincount': 28,
 789     }, {
 790         'url': 'https://soundcloud.com/soft-cell-official/tracks',
 791         'info_dict': {
 792             'id': '207965082',
 793             'title': 'Soft Cell (Tracks)',
 794         },
 795         'playlist_mincount': 27,
 796     }, {
 797         'url': 'https://soundcloud.com/soft-cell-official/albums',
 798         'info_dict': {
 799             'id': '207965082',
 800             'title': 'Soft Cell (Albums)',
 801         },
 802         'playlist_mincount': 1,
 803     }, {
 804         'url': 'https://soundcloud.com/jcv246/sets',
 805         'info_dict': {
 806             'id': '12982173',
 807             'title': 'Jordi / cv (Sets)',
 808         },
 809         'playlist_mincount': 2,
 810     }, {
 811         'url': 'https://soundcloud.com/jcv246/reposts',
 812         'info_dict': {
 813             'id': '12982173',
 814             'title': 'Jordi / cv (Reposts)',
 815         },
 816         'playlist_mincount': 6,
 817     }, {
 818         'url': 'https://soundcloud.com/clalberg/likes',
 819         'info_dict': {
 820             'id': '11817582',
 821             'title': 'clalberg (Likes)',
 822         },
 823         'playlist_mincount': 5,
 824     }, {
 825         'url': 'https://soundcloud.com/grynpyret/spotlight',
 826         'info_dict': {
 827             'id': '7098329',
 828             'title': 'Grynpyret (Spotlight)',
 829         },
 830         'playlist_mincount': 1,
 831     }]
 832
 833     _BASE_URL_MAP = {
 834         'all': 'stream/users/%s',
 835         'tracks': 'users/%s/tracks',
 836         'albums': 'users/%s/albums',
 837         'sets': 'users/%s/playlists',
 838         'reposts': 'stream/users/%s/reposts',
 839         'likes': 'users/%s/likes',
 840         'spotlight': 'users/%s/spotlight',
 841     }
 842
 843     def _real_extract(self, url):
 844         mobj = self._match_valid_url(url)
 845         uploader = mobj.group('user')
 846
 847         user = self._download_json(
 848             self._resolv_url(self._BASE_URL + uploader),
 849             uploader, 'Downloading user info', headers=self._HEADERS)
 850
 851         resource = mobj.group('rsrc') or 'all'
 852
 853         return self._extract_playlist(
 854             self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'],
 855             str_or_none(user.get('id')),
 856             '%s (%s)' % (user['username'], resource.capitalize()))
 857
 858
 859 class SoundcloudUserPermalinkIE(SoundcloudPagedPlaylistBaseIE):
 860     _VALID_URL = r'https?://api\.soundcloud\.com/users/(?P<id>\d+)'
 861     IE_NAME = 'soundcloud:user:permalink'
 862     _TESTS = [{
 863         'url': 'https://api.soundcloud.com/users/30909869',
 864         'info_dict': {
 865             'id': '30909869',
 866             'title': 'neilcic',
 867         },
 868         'playlist_mincount': 23,
 869     }]
 870
 871     def _real_extract(self, url):
 872         user_id = self._match_id(url)
 873         user = self._download_json(
 874             self._resolv_url(url), user_id, 'Downloading user info', headers=self._HEADERS)
 875
 876         return self._extract_playlist(
 877             f'{self._API_V2_BASE}stream/users/{user["id"]}', str(user['id']), user.get('username'))
 878
 879
 880 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 881     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 882     IE_NAME = 'soundcloud:trackstation'
 883     _TESTS = [{
 884         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 885         'info_dict': {
 886             'id': '286017854',
 887             'title': 'Track station: your text',
 888         },
 889         'playlist_mincount': 47,
 890     }]
 891
 892     def _real_extract(self, url):
 893         track_name = self._match_id(url)
 894
 895         track = self._download_json(self._resolv_url(url), track_name, headers=self._HEADERS)
 896         track_id = self._search_regex(
 897             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 898
 899         return self._extract_playlist(
 900             self._API_V2_BASE + 'stations/%s/tracks' % track['id'],
 901             track_id, 'Track station: %s' % track['title'])
 902
 903
 904 class SoundcloudRelatedIE(SoundcloudPagedPlaylistBaseIE):
 905     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<slug>[\w\d-]+/[\w\d-]+)/(?P<relation>albums|sets|recommended)'
 906     IE_NAME = 'soundcloud:related'
 907     _TESTS = [{
 908         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/recommended',
 909         'info_dict': {
 910             'id': '1084577272',
 911             'title': 'Sexapil - Pingers 5 (Recommended)',
 912         },
 913         'playlist_mincount': 50,
 914     }, {
 915         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/albums',
 916         'info_dict': {
 917             'id': '1084577272',
 918             'title': 'Sexapil - Pingers 5 (Albums)',
 919         },
 920         'playlist_mincount': 1,
 921     }, {
 922         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/sets',
 923         'info_dict': {
 924             'id': '1084577272',
 925             'title': 'Sexapil - Pingers 5 (Sets)',
 926         },
 927         'playlist_mincount': 4,
 928     }]
 929
 930     _BASE_URL_MAP = {
 931         'albums': 'tracks/%s/albums',
 932         'sets': 'tracks/%s/playlists_without_albums',
 933         'recommended': 'tracks/%s/related',
 934     }
 935
 936     def _real_extract(self, url):
 937         slug, relation = self._match_valid_url(url).group('slug', 'relation')
 938
 939         track = self._download_json(
 940             self._resolv_url(self._BASE_URL + slug),
 941             slug, 'Downloading track info', headers=self._HEADERS)
 942
 943         if track.get('errors'):
 944             raise ExtractorError(f'{self.IE_NAME} said: %s' % ','.join(
 945                 str(err['error_message']) for err in track['errors']), expected=True)
 946
 947         return self._extract_playlist(
 948             self._API_V2_BASE + self._BASE_URL_MAP[relation] % track['id'], str(track['id']),
 949             '%s (%s)' % (track.get('title') or slug, relation.capitalize()))
 950
 951
 952 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 953     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 954     IE_NAME = 'soundcloud:playlist'
 955     _TESTS = [{
 956         'url': 'https://api.soundcloud.com/playlists/4110309',
 957         'info_dict': {
 958             'id': '4110309',
 959             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 960             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 961         },
 962         'playlist_count': 6,
 963     }]
 964
 965     def _real_extract(self, url):
 966         mobj = self._match_valid_url(url)
 967         playlist_id = mobj.group('id')
 968
 969         query = {}
 970         token = mobj.group('token')
 971         if token:
 972             query['secret_token'] = token
 973
 974         data = self._download_json(
 975             self._API_V2_BASE + 'playlists/' + playlist_id,
 976             playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 977
 978         return self._extract_set(data, token)
 979
 980
 981 class SoundcloudSearchIE(SoundcloudBaseIE, SearchInfoExtractor):
 982     IE_NAME = 'soundcloud:search'
 983     IE_DESC = 'Soundcloud search'
 984     _SEARCH_KEY = 'scsearch'
 985     _TESTS = [{
 986         'url': 'scsearch15:post-avant jazzcore',
 987         'info_dict': {
 988             'id': 'post-avant jazzcore',
 989             'title': 'post-avant jazzcore',
 990         },
 991         'playlist_count': 15,
 992     }]
 993
 994     _MAX_RESULTS_PER_PAGE = 200
 995     _DEFAULT_RESULTS_PER_PAGE = 50
 996
 997     def _get_collection(self, endpoint, collection_id, **query):
 998         limit = min(
 999             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
1000             self._MAX_RESULTS_PER_PAGE)
1001         query.update({
1002             'limit': limit,
1003             'linked_partitioning': 1,
1004             'offset': 0,
1005         })
1006         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
1007
1008         for i in itertools.count(1):
1009             response = self._download_json(
1010                 next_url, collection_id, f'Downloading page {i}',
1011                 'Unable to download API page', headers=self._HEADERS)
1012
1013             for item in response.get('collection') or []:
1014                 if item:
1015                     yield self.url_result(
1016                         item['uri'], SoundcloudIE.ie_key(), **self._extract_info_dict(item, extract_flat=True))
1017
1018             next_url = response.get('next_href')
1019             if not next_url:
1020                 break
1021
1022     def _get_n_results(self, query, n):
1023         return self.playlist_result(itertools.islice(
1024             self._get_collection('search/tracks', query, limit=n, q=query),
1025             0, None if n == float('inf') else n), query, query)