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 ..networking import HEADRequest
   8 from ..networking.exceptions import HTTPError
   9 from ..utils import (
  10     KNOWN_EXTENSIONS,
  11     ExtractorError,
  12     float_or_none,
  13     int_or_none,
  14     join_nonempty,
  15     mimetype2ext,
  16     parse_qs,
  17     str_or_none,
  18     try_call,
  19     unified_timestamp,
  20     update_url_query,
  21     url_or_none,
  22     urlhandle_detect_ext,
  23 )
  24 from ..utils.traversal import traverse_obj
  25
  26
  27 class SoundcloudEmbedIE(InfoExtractor):
  28     _VALID_URL = r'https?://(?:w|player|p)\.soundcloud\.com/player/?.*?\burl=(?P<id>.+)'
  29     _EMBED_REGEX = [r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1']
  30     _TEST = {
  31         # from https://www.soundi.fi/uutiset/ennakkokuuntelussa-timo-kaukolammen-station-to-station-to-station-julkaisua-juhlitaan-tanaan-g-livelabissa/
  32         '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',
  33         'only_matching': True,
  34     }
  35
  36     def _real_extract(self, url):
  37         query = parse_qs(url)
  38         api_url = query['url'][0]
  39         secret_token = query.get('secret_token')
  40         if secret_token:
  41             api_url = update_url_query(api_url, {'secret_token': secret_token[0]})
  42         return self.url_result(api_url)
  43
  44
  45 class SoundcloudBaseIE(InfoExtractor):
  46     _NETRC_MACHINE = 'soundcloud'
  47
  48     _API_V2_BASE = 'https://api-v2.soundcloud.com/'
  49     _BASE_URL = 'https://soundcloud.com/'
  50     _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'
  51     _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
  52     _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
  53     _API_VERIFY_AUTH_TOKEN = 'https://api-auth.soundcloud.com/connect/session%s'
  54     _HEADERS = {}
  55
  56     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
  57
  58     _ARTWORK_MAP = {
  59         'mini': 16,
  60         'tiny': 20,
  61         'small': 32,
  62         'badge': 47,
  63         't67x67': 67,
  64         'large': 100,
  65         't300x300': 300,
  66         'crop': 400,
  67         't500x500': 500,
  68         'original': 0,
  69     }
  70
  71     _DEFAULT_FORMATS = ['http_aac', 'hls_aac', 'http_opus', 'hls_opus', 'http_mp3', 'hls_mp3']
  72
  73     @functools.cached_property
  74     def _is_requested(self):
  75         return re.compile(r'|'.join(set(
  76             re.escape(pattern).replace(r'\*', r'.*') if pattern != 'default'
  77             else '|'.join(map(re.escape, self._DEFAULT_FORMATS))
  78             for pattern in self._configuration_arg('formats', ['default'], ie_key=SoundcloudIE)
  79         ))).fullmatch
  80
  81     def _store_client_id(self, client_id):
  82         self.cache.store('soundcloud', 'client_id', client_id)
  83
  84     def _update_client_id(self):
  85         webpage = self._download_webpage('https://soundcloud.com/', None)
  86         for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
  87             script = self._download_webpage(src, None, fatal=False)
  88             if script:
  89                 client_id = self._search_regex(
  90                     r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
  91                     script, 'client id', default=None)
  92                 if client_id:
  93                     self._CLIENT_ID = client_id
  94                     self._store_client_id(client_id)
  95                     return
  96         raise ExtractorError('Unable to extract client id')
  97
  98     def _call_api(self, *args, **kwargs):
  99         non_fatal = kwargs.get('fatal') is False
 100         if non_fatal:
 101             del kwargs['fatal']
 102         query = kwargs.get('query', {}).copy()
 103         for _ in range(2):
 104             query['client_id'] = self._CLIENT_ID
 105             kwargs['query'] = query
 106             try:
 107                 return self._download_json(*args, **kwargs)
 108             except ExtractorError as e:
 109                 if isinstance(e.cause, HTTPError) and e.cause.status in (401, 403):
 110                     self._store_client_id(None)
 111                     self._update_client_id()
 112                     continue
 113                 elif non_fatal:
 114                     self.report_warning(str(e))
 115                     return False
 116                 raise
 117
 118     def _initialize_pre_login(self):
 119         self._CLIENT_ID = self.cache.load('soundcloud', 'client_id') or 'a3e059563d7fd3372b49b37f00a00bcf'
 120
 121     def _verify_oauth_token(self, token):
 122         if self._request_webpage(
 123                 self._API_VERIFY_AUTH_TOKEN % (self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID),
 124                 None, note='Verifying login token...', fatal=False,
 125                 data=json.dumps({'session': {'access_token': token}}).encode()):
 126             self._HEADERS['Authorization'] = f'OAuth {token}'
 127             self.report_login()
 128         else:
 129             self.report_warning('Provided authorization token is invalid. Continuing as guest')
 130
 131     def _real_initialize(self):
 132         if self._HEADERS:
 133             return
 134         if token := try_call(lambda: self._get_cookies(self._BASE_URL)['oauth_token'].value):
 135             self._verify_oauth_token(token)
 136
 137     def _perform_login(self, username, password):
 138         if username != 'oauth':
 139             raise ExtractorError(
 140                 'Login using username and password is not currently supported. '
 141                 'Use "--username oauth --password <oauth_token>" to login using an oauth token, '
 142                 f'or else {self._login_hint(method="cookies")}', expected=True)
 143         if self._HEADERS:
 144             return
 145         self._verify_oauth_token(password)
 146
 147         r'''
 148         def genDevId():
 149             def genNumBlock():
 150                 return ''.join([str(random.randrange(10)) for i in range(6)])
 151             return '-'.join([genNumBlock() for i in range(4)])
 152
 153         payload = {
 154             'client_id': self._CLIENT_ID,
 155             'recaptcha_pubkey': 'null',
 156             'recaptcha_response': 'null',
 157             'credentials': {
 158                 'identifier': username,
 159                 'password': password
 160             },
 161             'signature': self.sign(username, password, self._CLIENT_ID),
 162             'device_id': genDevId(),
 163             'user_agent': self._USER_AGENT
 164         }
 165
 166         response = self._call_api(
 167             self._API_AUTH_URL_PW % (self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID),
 168             None, note='Verifying login token...', fatal=False,
 169             data=json.dumps(payload).encode())
 170
 171         if token := traverse_obj(response, ('session', 'access_token', {str})):
 172             self._HEADERS['Authorization'] = f'OAuth {token}'
 173             self.report_login()
 174             return
 175
 176         raise ExtractorError('Unable to get access token, login may have failed', expected=True)
 177         '''
 178
 179     # signature generation
 180     def sign(self, user, pw, clid):
 181         a = 33
 182         i = 1
 183         s = 440123
 184         w = 117
 185         u = 1800000
 186         l = 1042
 187         b = 37
 188         k = 37
 189         c = 5
 190         n = '0763ed7314c69015fd4a0dc16bbf4b90'  # _KEY
 191         y = '8'  # _REV
 192         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
 193         e = user  # _USERNAME
 194         t = clid  # _CLIENT_ID
 195
 196         d = '-'.join([str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
 197         h = n + y + d + r + e + t + d + n
 198
 199         m = 8011470
 200
 201         for f in range(len(h)):
 202             m = (m >> 1) + ((1 & m) << 23)
 203             m += ord(h[f])
 204             m &= 16777215
 205
 206         # c is not even needed
 207         return f'{y}:{d}:{m:x}:{c}'
 208
 209     def _extract_info_dict(self, info, full_title=None, secret_token=None, extract_flat=False):
 210         track_id = str(info['id'])
 211         title = info['title']
 212
 213         format_urls = set()
 214         formats = []
 215         query = {'client_id': self._CLIENT_ID}
 216         if secret_token:
 217             query['secret_token'] = secret_token
 218
 219         if not extract_flat and info.get('downloadable') and info.get('has_downloads_left'):
 220             try:
 221                 # Do not use _call_api(); HTTP Error codes have different meanings for this request
 222                 download_data = self._download_json(
 223                     f'{self._API_V2_BASE}tracks/{track_id}/download', track_id,
 224                     'Downloading original download format info JSON', query=query, headers=self._HEADERS)
 225             except ExtractorError as e:
 226                 if isinstance(e.cause, HTTPError) and e.cause.status == 401:
 227                     self.report_warning(
 228                         'Original download format is only available '
 229                         f'for registered users. {self._login_hint()}')
 230                 elif isinstance(e.cause, HTTPError) and e.cause.status == 403:
 231                     self.write_debug('Original download format is not available for this client')
 232                 else:
 233                     self.report_warning(e.msg)
 234                 download_data = None
 235
 236             if redirect_url := traverse_obj(download_data, ('redirectUri', {url_or_none})):
 237                 urlh = self._request_webpage(
 238                     HEADRequest(redirect_url), track_id, 'Checking original download format availability',
 239                     'Original download format is not available', fatal=False)
 240                 if urlh:
 241                     format_url = urlh.url
 242                     format_urls.add(format_url)
 243                     formats.append({
 244                         'format_id': 'download',
 245                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
 246                         'filesize': int_or_none(urlh.headers.get('Content-Length')),
 247                         'url': format_url,
 248                         'quality': 10,
 249                         'format_note': 'Original',
 250                     })
 251
 252         def invalid_url(url):
 253             return not url or url in format_urls
 254
 255         def add_format(f, protocol, is_preview=False):
 256             mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url)
 257             if mobj:
 258                 for k, v in mobj.groupdict().items():
 259                     if not f.get(k):
 260                         f[k] = v
 261             format_id_list = []
 262             if protocol:
 263                 format_id_list.append(protocol)
 264             ext = f.get('ext')
 265             if ext == 'aac':
 266                 f.update({
 267                     'abr': 256,
 268                     'quality': 5,
 269                     'format_note': 'Premium',
 270                 })
 271             for k in ('ext', 'abr'):
 272                 v = str_or_none(f.get(k))
 273                 if v:
 274                     format_id_list.append(v)
 275             preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url'])
 276             if preview:
 277                 format_id_list.append('preview')
 278             abr = f.get('abr')
 279             if abr:
 280                 f['abr'] = int(abr)
 281             if protocol in ('hls', 'hls-aes'):
 282                 protocol = 'm3u8' if ext == 'aac' else 'm3u8_native'
 283             else:
 284                 protocol = 'http'
 285             f.update({
 286                 'format_id': '_'.join(format_id_list),
 287                 'protocol': protocol,
 288                 'preference': -10 if preview else None,
 289             })
 290             formats.append(f)
 291
 292         # New API
 293         for t in traverse_obj(info, ('media', 'transcodings', lambda _, v: url_or_none(v['url']))):
 294             if extract_flat:
 295                 break
 296             format_url = t['url']
 297
 298             protocol = traverse_obj(t, ('format', 'protocol', {str}))
 299             if protocol == 'progressive':
 300                 protocol = 'http'
 301             if protocol != 'hls' and '/hls' in format_url:
 302                 protocol = 'hls'
 303             if protocol == 'encrypted-hls' or '/encrypted-hls' in format_url:
 304                 protocol = 'hls-aes'
 305
 306             ext = None
 307             if preset := traverse_obj(t, ('preset', {str_or_none})):
 308                 ext = preset.split('_')[0]
 309             if ext not in KNOWN_EXTENSIONS:
 310                 ext = mimetype2ext(traverse_obj(t, ('format', 'mime_type', {str})))
 311
 312             identifier = join_nonempty(protocol, ext, delim='_')
 313             if not self._is_requested(identifier):
 314                 self.write_debug(f'"{identifier}" is not a requested format, skipping')
 315                 continue
 316
 317             stream = None
 318             for retry in self.RetryManager(fatal=False):
 319                 try:
 320                     stream = self._call_api(
 321                         format_url, track_id, f'Downloading {identifier} format info JSON',
 322                         query=query, headers=self._HEADERS)
 323                 except ExtractorError as e:
 324                     if isinstance(e.cause, HTTPError) and e.cause.status == 429:
 325                         self.report_warning(
 326                             'You have reached the API rate limit, which is ~600 requests per '
 327                             '10 minutes. Use the --extractor-retries and --retry-sleep options '
 328                             'to configure an appropriate retry count and wait time', only_once=True)
 329                         retry.error = e.cause
 330                     else:
 331                         self.report_warning(e.msg)
 332
 333             stream_url = traverse_obj(stream, ('url', {url_or_none}))
 334             if invalid_url(stream_url):
 335                 continue
 336             format_urls.add(stream_url)
 337             add_format({
 338                 'url': stream_url,
 339                 'ext': ext,
 340             }, protocol, t.get('snipped') or '/preview/' in format_url)
 341
 342         for f in formats:
 343             f['vcodec'] = 'none'
 344
 345         if not formats and info.get('policy') == 'BLOCK':
 346             self.raise_geo_restricted(metadata_available=True)
 347
 348         user = info.get('user') or {}
 349
 350         thumbnails = []
 351         artwork_url = info.get('artwork_url')
 352         thumbnail = artwork_url or user.get('avatar_url')
 353         if isinstance(thumbnail, str):
 354             if re.search(self._IMAGE_REPL_RE, thumbnail):
 355                 for image_id, size in self._ARTWORK_MAP.items():
 356                     i = {
 357                         'id': image_id,
 358                         'url': re.sub(self._IMAGE_REPL_RE, f'-{image_id}.jpg', thumbnail),
 359                     }
 360                     if image_id == 'tiny' and not artwork_url:
 361                         size = 18
 362                     elif image_id == 'original':
 363                         i['preference'] = 10
 364                     if size:
 365                         i.update({
 366                             'width': size,
 367                             'height': size,
 368                         })
 369                     thumbnails.append(i)
 370             else:
 371                 thumbnails = [{'url': thumbnail}]
 372
 373         def extract_count(key):
 374             return int_or_none(info.get(f'{key}_count'))
 375
 376         return {
 377             'id': track_id,
 378             'uploader': user.get('username'),
 379             'uploader_id': str_or_none(user.get('id')) or user.get('permalink'),
 380             'uploader_url': user.get('permalink_url'),
 381             'timestamp': unified_timestamp(info.get('created_at')),
 382             'title': title,
 383             'description': info.get('description'),
 384             'thumbnails': thumbnails,
 385             'duration': float_or_none(info.get('duration'), 1000),
 386             'webpage_url': info.get('permalink_url'),
 387             'license': info.get('license'),
 388             'view_count': extract_count('playback'),
 389             'like_count': extract_count('favoritings') or extract_count('likes'),
 390             'comment_count': extract_count('comment'),
 391             'repost_count': extract_count('reposts'),
 392             'genres': traverse_obj(info, ('genre', {str}, {lambda x: x or None}, all)),
 393             'formats': formats if not extract_flat else None,
 394         }
 395
 396     @classmethod
 397     def _resolv_url(cls, url):
 398         return cls._API_V2_BASE + 'resolve?url=' + url
 399
 400
 401 class SoundcloudIE(SoundcloudBaseIE):
 402     """Information extractor for soundcloud.com
 403        To access the media, the uid of the song and a stream token
 404        must be extracted from the page source and the script must make
 405        a request to media.soundcloud.com/crossdomain.xml. Then
 406        the media can be grabbed by requesting from an url composed
 407        of the stream token and uid
 408      """
 409
 410     _VALID_URL = r'''(?x)^(?:https?://)?
 411                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
 412                             (?!stations/track)
 413                             (?P<uploader>[\w\d-]+)/
 414                             (?!(?:tracks|albums|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
 415                             (?P<title>[\w\d-]+)
 416                             (?:/(?P<token>(?!(?:albums|sets|recommended))[^?]+?))?
 417                             (?:[?].*)?$)
 418                        |(?:api(?:-v2)?\.soundcloud\.com/tracks/(?P<track_id>\d+)
 419                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
 420                     )
 421                     '''
 422     IE_NAME = 'soundcloud'
 423     _TESTS = [
 424         {
 425             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
 426             'md5': 'de9bac153e7427a7333b4b0c1b6a18d2',
 427             'info_dict': {
 428                 'id': '62986583',
 429                 'ext': 'opus',
 430                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
 431                 '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',
 432                 'uploader': 'E.T. ExTerrestrial Music',
 433                 'uploader_id': '1571244',
 434                 'timestamp': 1349920598,
 435                 'upload_date': '20121011',
 436                 'duration': 143.216,
 437                 'license': 'all-rights-reserved',
 438                 'view_count': int,
 439                 'like_count': int,
 440                 'comment_count': int,
 441                 'repost_count': int,
 442                 'thumbnail': 'https://i1.sndcdn.com/artworks-000031955188-rwb18x-original.jpg',
 443                 'uploader_url': 'https://soundcloud.com/ethmusic',
 444                 'genres': [],
 445             },
 446         },
 447         # geo-restricted
 448         {
 449             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
 450             'info_dict': {
 451                 'id': '47127627',
 452                 'ext': 'opus',
 453                 'title': 'Goldrushed',
 454                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
 455                 'uploader': 'The Royal Concept',
 456                 'uploader_id': '9615865',
 457                 'timestamp': 1337635207,
 458                 'upload_date': '20120521',
 459                 'duration': 227.155,
 460                 'license': 'all-rights-reserved',
 461                 'view_count': int,
 462                 'like_count': int,
 463                 'comment_count': int,
 464                 'repost_count': int,
 465                 'uploader_url': 'https://soundcloud.com/the-concept-band',
 466                 'thumbnail': 'https://i1.sndcdn.com/artworks-v8bFHhXm7Au6-0-original.jpg',
 467                 'genres': ['Alternative'],
 468             },
 469         },
 470         # private link
 471         {
 472             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
 473             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 474             'info_dict': {
 475                 'id': '123998367',
 476                 'ext': 'mp3',
 477                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 478                 'description': 'test chars:  "\'/\\ä↭',
 479                 'uploader': 'jaimeMF',
 480                 'uploader_id': '69767071',
 481                 'timestamp': 1386604920,
 482                 'upload_date': '20131209',
 483                 'duration': 9.927,
 484                 'license': 'all-rights-reserved',
 485                 'view_count': int,
 486                 'like_count': int,
 487                 'comment_count': int,
 488                 'repost_count': int,
 489                 'uploader_url': 'https://soundcloud.com/jaimemf',
 490                 'thumbnail': 'https://a1.sndcdn.com/images/default_avatar_large.png',
 491                 'genres': ['youtubedl'],
 492             },
 493         },
 494         # private link (alt format)
 495         {
 496             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
 497             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 498             'info_dict': {
 499                 'id': '123998367',
 500                 'ext': 'mp3',
 501                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 502                 'description': 'test chars:  "\'/\\ä↭',
 503                 'uploader': 'jaimeMF',
 504                 'uploader_id': '69767071',
 505                 'timestamp': 1386604920,
 506                 'upload_date': '20131209',
 507                 'duration': 9.927,
 508                 'license': 'all-rights-reserved',
 509                 'view_count': int,
 510                 'like_count': int,
 511                 'comment_count': int,
 512                 'repost_count': int,
 513                 'uploader_url': 'https://soundcloud.com/jaimemf',
 514                 'thumbnail': 'https://a1.sndcdn.com/images/default_avatar_large.png',
 515                 'genres': ['youtubedl'],
 516             },
 517         },
 518         # downloadable song
 519         {
 520             'url': 'https://soundcloud.com/the80m/the-following',
 521             'md5': '9ffcddb08c87d74fb5808a3c183a1d04',
 522             'info_dict': {
 523                 'id': '343609555',
 524                 'ext': 'wav',
 525                 'title': 'The Following',
 526                 'description': '',
 527                 'uploader': '80M',
 528                 'uploader_id': '312384765',
 529                 'uploader_url': 'https://soundcloud.com/the80m',
 530                 'upload_date': '20170922',
 531                 'timestamp': 1506120436,
 532                 'duration': 397.228,
 533                 'thumbnail': 'https://i1.sndcdn.com/artworks-000243916348-ktoo7d-original.jpg',
 534                 'license': 'all-rights-reserved',
 535                 'like_count': int,
 536                 'comment_count': int,
 537                 'repost_count': int,
 538                 'view_count': int,
 539                 'genres': ['Dance & EDM'],
 540             },
 541         },
 542         # private link, downloadable format
 543         {
 544             'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
 545             'md5': '64a60b16e617d41d0bef032b7f55441e',
 546             'info_dict': {
 547                 'id': '340344461',
 548                 'ext': 'wav',
 549                 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
 550                 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
 551                 'uploader': 'Ori Uplift Music',
 552                 'uploader_id': '12563093',
 553                 'timestamp': 1504206263,
 554                 'upload_date': '20170831',
 555                 'duration': 7449.096,
 556                 'license': 'all-rights-reserved',
 557                 'view_count': int,
 558                 'like_count': int,
 559                 'comment_count': int,
 560                 'repost_count': int,
 561                 'thumbnail': 'https://i1.sndcdn.com/artworks-000240712245-kedn4p-original.jpg',
 562                 'uploader_url': 'https://soundcloud.com/oriuplift',
 563                 'genres': ['Trance'],
 564             },
 565         },
 566         # no album art, use avatar pic for thumbnail
 567         {
 568             'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
 569             'md5': '59c7872bc44e5d99b7211891664760c2',
 570             'info_dict': {
 571                 'id': '309699954',
 572                 'ext': 'mp3',
 573                 'title': 'Sideways (Prod. Mad Real)',
 574                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
 575                 'uploader': 'garyvee',
 576                 'uploader_id': '2366352',
 577                 'timestamp': 1488152409,
 578                 'upload_date': '20170226',
 579                 'duration': 207.012,
 580                 'thumbnail': r're:https?://.*\.jpg',
 581                 'license': 'all-rights-reserved',
 582                 'view_count': int,
 583                 'like_count': int,
 584                 'comment_count': int,
 585                 'repost_count': int,
 586                 'uploader_url': 'https://soundcloud.com/garyvee',
 587                 'genres': [],
 588             },
 589             'params': {
 590                 'skip_download': True,
 591             },
 592         },
 593         {
 594             'url': 'https://soundcloud.com/giovannisarani/mezzo-valzer',
 595             'md5': '8227c3473a4264df6b02ad7e5b7527ac',
 596             'info_dict': {
 597                 'id': '583011102',
 598                 'ext': 'opus',
 599                 'title': 'Mezzo Valzer',
 600                 'description': 'md5:f4d5f39d52e0ccc2b4f665326428901a',
 601                 'uploader': 'Giovanni Sarani',
 602                 'uploader_id': '3352531',
 603                 'timestamp': 1551394171,
 604                 'upload_date': '20190228',
 605                 'duration': 180.157,
 606                 'thumbnail': r're:https?://.*\.jpg',
 607                 'license': 'all-rights-reserved',
 608                 'view_count': int,
 609                 'like_count': int,
 610                 'comment_count': int,
 611                 'repost_count': int,
 612                 'genres': ['Piano'],
 613                 'uploader_url': 'https://soundcloud.com/giovannisarani',
 614             },
 615         },
 616         {
 617             # AAC HQ format available (account with active subscription needed)
 618             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
 619             'only_matching': True,
 620         },
 621         {
 622             # Go+ (account with active subscription needed)
 623             'url': 'https://soundcloud.com/taylorswiftofficial/look-what-you-made-me-do',
 624             'only_matching': True,
 625         },
 626     ]
 627
 628     def _real_extract(self, url):
 629         mobj = self._match_valid_url(url)
 630
 631         track_id = mobj.group('track_id')
 632
 633         query = {}
 634         if track_id:
 635             info_json_url = self._API_V2_BASE + 'tracks/' + track_id
 636             full_title = track_id
 637             token = mobj.group('secret_token')
 638             if token:
 639                 query['secret_token'] = token
 640         else:
 641             full_title = resolve_title = '{}/{}'.format(*mobj.group('uploader', 'title'))
 642             token = mobj.group('token')
 643             if token:
 644                 resolve_title += f'/{token}'
 645             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 646
 647         info = self._call_api(
 648             info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 649
 650         return self._extract_info_dict(info, full_title, token)
 651
 652
 653 class SoundcloudPlaylistBaseIE(SoundcloudBaseIE):
 654     def _extract_set(self, playlist, token=None):
 655         playlist_id = str(playlist['id'])
 656         tracks = playlist.get('tracks') or []
 657         if not all(t.get('permalink_url') for t in tracks) and token:
 658             tracks = self._call_api(
 659                 self._API_V2_BASE + 'tracks', playlist_id,
 660                 'Downloading tracks', query={
 661                     'ids': ','.join([str(t['id']) for t in tracks]),
 662                     'playlistId': playlist_id,
 663                     'playlistSecretToken': token,
 664                 }, headers=self._HEADERS)
 665         entries = []
 666         for track in tracks:
 667             track_id = str_or_none(track.get('id'))
 668             url = track.get('permalink_url')
 669             if not url:
 670                 if not track_id:
 671                     continue
 672                 url = self._API_V2_BASE + 'tracks/' + track_id
 673                 if token:
 674                     url += '?secret_token=' + token
 675             entries.append(self.url_result(
 676                 url, SoundcloudIE.ie_key(), track_id))
 677         return self.playlist_result(
 678             entries, playlist_id,
 679             playlist.get('title'),
 680             playlist.get('description'))
 681
 682
 683 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 684     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[:\w\d-]+)(?:/(?P<token>[^?/]+))?'
 685     IE_NAME = 'soundcloud:set'
 686     _TESTS = [{
 687         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 688         'info_dict': {
 689             'id': '2284613',
 690             'title': 'The Royal Concept EP',
 691             'description': 'md5:71d07087c7a449e8941a70a29e34671e',
 692         },
 693         'playlist_mincount': 5,
 694     }, {
 695         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 696         'only_matching': True,
 697     }, {
 698         'url': 'https://soundcloud.com/discover/sets/weekly::flacmatic',
 699         'only_matching': True,
 700     }, {
 701         'url': 'https://soundcloud.com/discover/sets/charts-top:all-music:de',
 702         'only_matching': True,
 703     }, {
 704         'url': 'https://soundcloud.com/discover/sets/charts-top:hiphoprap:kr',
 705         'only_matching': True,
 706     }]
 707
 708     def _real_extract(self, url):
 709         mobj = self._match_valid_url(url)
 710
 711         full_title = '{}/sets/{}'.format(*mobj.group('uploader', 'slug_title'))
 712         token = mobj.group('token')
 713         if token:
 714             full_title += '/' + token
 715
 716         info = self._call_api(self._resolv_url(
 717             self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 718
 719         if 'errors' in info:
 720             msgs = (str(err['error_message']) for err in info['errors'])
 721             raise ExtractorError('unable to download video webpage: {}'.format(','.join(msgs)))
 722
 723         return self._extract_set(info, token)
 724
 725
 726 class SoundcloudPagedPlaylistBaseIE(SoundcloudBaseIE):
 727     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 728         return {
 729             '_type': 'playlist',
 730             'id': playlist_id,
 731             'title': playlist_title,
 732             'entries': self._entries(base_url, playlist_id),
 733         }
 734
 735     def _entries(self, url, playlist_id):
 736         # Per the SoundCloud documentation, the maximum limit for a linked partitioning query is 200.
 737         # https://developers.soundcloud.com/blog/offset-pagination-deprecated
 738         query = {
 739             'limit': 200,
 740             'linked_partitioning': '1',
 741             'offset': 0,
 742         }
 743
 744         for i in itertools.count():
 745             for retry in self.RetryManager():
 746                 try:
 747                     response = self._call_api(
 748                         url, playlist_id, query=query, headers=self._HEADERS,
 749                         note=f'Downloading track page {i + 1}')
 750                     break
 751                 except ExtractorError as e:
 752                     # Downloading page may result in intermittent 502 HTTP error
 753                     # See https://github.com/yt-dlp/yt-dlp/issues/872
 754                     if not isinstance(e.cause, HTTPError) or e.cause.status != 502:
 755                         raise
 756                     retry.error = e
 757                     continue
 758
 759             def resolve_entry(*candidates):
 760                 for cand in candidates:
 761                     if not isinstance(cand, dict):
 762                         continue
 763                     permalink_url = url_or_none(cand.get('permalink_url'))
 764                     if permalink_url:
 765                         return self.url_result(
 766                             permalink_url,
 767                             SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
 768                             str_or_none(cand.get('id')), cand.get('title'))
 769
 770             for e in response['collection'] or []:
 771                 yield resolve_entry(e, e.get('track'), e.get('playlist'))
 772
 773             url = response.get('next_href')
 774             if not url:
 775                 break
 776             query.pop('offset', None)
 777
 778
 779 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 780     _VALID_URL = r'''(?x)
 781                         https?://
 782                             (?:(?:www|m)\.)?soundcloud\.com/
 783                             (?P<user>[^/]+)
 784                             (?:/
 785                                 (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight)
 786                             )?
 787                             /?(?:[?#].*)?$
 788                     '''
 789     IE_NAME = 'soundcloud:user'
 790     _TESTS = [{
 791         'url': 'https://soundcloud.com/soft-cell-official',
 792         'info_dict': {
 793             'id': '207965082',
 794             'title': 'Soft Cell (All)',
 795         },
 796         'playlist_mincount': 28,
 797     }, {
 798         'url': 'https://soundcloud.com/soft-cell-official/tracks',
 799         'info_dict': {
 800             'id': '207965082',
 801             'title': 'Soft Cell (Tracks)',
 802         },
 803         'playlist_mincount': 27,
 804     }, {
 805         'url': 'https://soundcloud.com/soft-cell-official/albums',
 806         'info_dict': {
 807             'id': '207965082',
 808             'title': 'Soft Cell (Albums)',
 809         },
 810         'playlist_mincount': 1,
 811     }, {
 812         'url': 'https://soundcloud.com/jcv246/sets',
 813         'info_dict': {
 814             'id': '12982173',
 815             'title': 'Jordi / cv (Sets)',
 816         },
 817         'playlist_mincount': 2,
 818     }, {
 819         'url': 'https://soundcloud.com/jcv246/reposts',
 820         'info_dict': {
 821             'id': '12982173',
 822             'title': 'Jordi / cv (Reposts)',
 823         },
 824         'playlist_mincount': 6,
 825     }, {
 826         'url': 'https://soundcloud.com/clalberg/likes',
 827         'info_dict': {
 828             'id': '11817582',
 829             'title': 'clalberg (Likes)',
 830         },
 831         'playlist_mincount': 5,
 832     }, {
 833         'url': 'https://soundcloud.com/grynpyret/spotlight',
 834         'info_dict': {
 835             'id': '7098329',
 836             'title': 'Grynpyret (Spotlight)',
 837         },
 838         'playlist_mincount': 1,
 839     }]
 840
 841     _BASE_URL_MAP = {
 842         'all': 'stream/users/%s',
 843         'tracks': 'users/%s/tracks',
 844         'albums': 'users/%s/albums',
 845         'sets': 'users/%s/playlists',
 846         'reposts': 'stream/users/%s/reposts',
 847         'likes': 'users/%s/likes',
 848         'spotlight': 'users/%s/spotlight',
 849     }
 850
 851     def _real_extract(self, url):
 852         mobj = self._match_valid_url(url)
 853         uploader = mobj.group('user')
 854
 855         user = self._call_api(
 856             self._resolv_url(self._BASE_URL + uploader),
 857             uploader, 'Downloading user info', headers=self._HEADERS)
 858
 859         resource = mobj.group('rsrc') or 'all'
 860
 861         return self._extract_playlist(
 862             self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'],
 863             str_or_none(user.get('id')),
 864             '{} ({})'.format(user['username'], resource.capitalize()))
 865
 866
 867 class SoundcloudUserPermalinkIE(SoundcloudPagedPlaylistBaseIE):
 868     _VALID_URL = r'https?://api\.soundcloud\.com/users/(?P<id>\d+)'
 869     IE_NAME = 'soundcloud:user:permalink'
 870     _TESTS = [{
 871         'url': 'https://api.soundcloud.com/users/30909869',
 872         'info_dict': {
 873             'id': '30909869',
 874             'title': 'neilcic',
 875         },
 876         'playlist_mincount': 23,
 877     }]
 878
 879     def _real_extract(self, url):
 880         user_id = self._match_id(url)
 881         user = self._call_api(
 882             self._resolv_url(url), user_id, 'Downloading user info', headers=self._HEADERS)
 883
 884         return self._extract_playlist(
 885             f'{self._API_V2_BASE}stream/users/{user["id"]}', str(user['id']), user.get('username'))
 886
 887
 888 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 889     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 890     IE_NAME = 'soundcloud:trackstation'
 891     _TESTS = [{
 892         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 893         'info_dict': {
 894             'id': '286017854',
 895             'title': 'Track station: your text',
 896         },
 897         'playlist_mincount': 47,
 898     }]
 899
 900     def _real_extract(self, url):
 901         track_name = self._match_id(url)
 902
 903         track = self._call_api(self._resolv_url(url), track_name, headers=self._HEADERS)
 904         track_id = self._search_regex(
 905             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 906
 907         return self._extract_playlist(
 908             self._API_V2_BASE + 'stations/{}/tracks'.format(track['id']),
 909             track_id, 'Track station: {}'.format(track['title']))
 910
 911
 912 class SoundcloudRelatedIE(SoundcloudPagedPlaylistBaseIE):
 913     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<slug>[\w\d-]+/[\w\d-]+)/(?P<relation>albums|sets|recommended)'
 914     IE_NAME = 'soundcloud:related'
 915     _TESTS = [{
 916         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/recommended',
 917         'info_dict': {
 918             'id': '1084577272',
 919             'title': 'Sexapil - Pingers 5 (Recommended)',
 920         },
 921         'playlist_mincount': 50,
 922     }, {
 923         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/albums',
 924         'info_dict': {
 925             'id': '1084577272',
 926             'title': 'Sexapil - Pingers 5 (Albums)',
 927         },
 928         'playlist_mincount': 1,
 929     }, {
 930         'url': 'https://soundcloud.com/wajang/sexapil-pingers-5/sets',
 931         'info_dict': {
 932             'id': '1084577272',
 933             'title': 'Sexapil - Pingers 5 (Sets)',
 934         },
 935         'playlist_mincount': 4,
 936     }]
 937
 938     _BASE_URL_MAP = {
 939         'albums': 'tracks/%s/albums',
 940         'sets': 'tracks/%s/playlists_without_albums',
 941         'recommended': 'tracks/%s/related',
 942     }
 943
 944     def _real_extract(self, url):
 945         slug, relation = self._match_valid_url(url).group('slug', 'relation')
 946
 947         track = self._call_api(
 948             self._resolv_url(self._BASE_URL + slug),
 949             slug, 'Downloading track info', headers=self._HEADERS)
 950
 951         if track.get('errors'):
 952             raise ExtractorError(f'{self.IE_NAME} said: %s' % ','.join(
 953                 str(err['error_message']) for err in track['errors']), expected=True)
 954
 955         return self._extract_playlist(
 956             self._API_V2_BASE + self._BASE_URL_MAP[relation] % track['id'], str(track['id']),
 957             '{} ({})'.format(track.get('title') or slug, relation.capitalize()))
 958
 959
 960 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 961     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 962     IE_NAME = 'soundcloud:playlist'
 963     _TESTS = [{
 964         'url': 'https://api.soundcloud.com/playlists/4110309',
 965         'info_dict': {
 966             'id': '4110309',
 967             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 968             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 969         },
 970         'playlist_count': 6,
 971     }]
 972
 973     def _real_extract(self, url):
 974         mobj = self._match_valid_url(url)
 975         playlist_id = mobj.group('id')
 976
 977         query = {}
 978         token = mobj.group('token')
 979         if token:
 980             query['secret_token'] = token
 981
 982         data = self._call_api(
 983             self._API_V2_BASE + 'playlists/' + playlist_id,
 984             playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 985
 986         return self._extract_set(data, token)
 987
 988
 989 class SoundcloudSearchIE(SoundcloudBaseIE, SearchInfoExtractor):
 990     IE_NAME = 'soundcloud:search'
 991     IE_DESC = 'Soundcloud search'
 992     _SEARCH_KEY = 'scsearch'
 993     _TESTS = [{
 994         'url': 'scsearch15:post-avant jazzcore',
 995         'info_dict': {
 996             'id': 'post-avant jazzcore',
 997             'title': 'post-avant jazzcore',
 998         },
 999         'playlist_count': 15,
1000     }]
1001
1002     _MAX_RESULTS_PER_PAGE = 200
1003     _DEFAULT_RESULTS_PER_PAGE = 50
1004
1005     def _get_collection(self, endpoint, collection_id, **query):
1006         limit = min(
1007             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
1008             self._MAX_RESULTS_PER_PAGE)
1009         query.update({
1010             'limit': limit,
1011             'linked_partitioning': 1,
1012             'offset': 0,
1013         })
1014         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
1015
1016         for i in itertools.count(1):
1017             response = self._call_api(
1018                 next_url, collection_id, f'Downloading page {i}',
1019                 'Unable to download API page', headers=self._HEADERS)
1020
1021             for item in response.get('collection') or []:
1022                 if item:
1023                     yield self.url_result(
1024                         item['uri'], SoundcloudIE.ie_key(), **self._extract_info_dict(item, extract_flat=True))
1025
1026             next_url = response.get('next_href')
1027             if not next_url:
1028                 break
1029
1030     def _get_n_results(self, query, n):
1031         return self.playlist_result(itertools.islice(
1032             self._get_collection('search/tracks', query, limit=n, q=query),
1033             0, None if n == float('inf') else n), query, query)