yt_dlp/extractor/soundcloud.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import itertools
   5 import re
   6 import json
   7 # import random
   8
   9 from .common import (
  10     InfoExtractor,
  11     SearchInfoExtractor
  12 )
  13 from ..compat import (
  14     compat_HTTPError,
  15     compat_kwargs,
  16     compat_str,
  17 )
  18 from ..utils import (
  19     error_to_compat_str,
  20     ExtractorError,
  21     float_or_none,
  22     HEADRequest,
  23     int_or_none,
  24     KNOWN_EXTENSIONS,
  25     mimetype2ext,
  26     remove_end,
  27     parse_qs,
  28     str_or_none,
  29     try_get,
  30     unified_timestamp,
  31     update_url_query,
  32     url_or_none,
  33     urlhandle_detect_ext,
  34     sanitized_Request,
  35 )
  36
  37
  38 class SoundcloudEmbedIE(InfoExtractor):
  39     _VALID_URL = r'https?://(?:w|player|p)\.soundcloud\.com/player/?.*?\burl=(?P<id>.+)'
  40     _TEST = {
  41         # from https://www.soundi.fi/uutiset/ennakkokuuntelussa-timo-kaukolammen-station-to-station-to-station-julkaisua-juhlitaan-tanaan-g-livelabissa/
  42         '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',
  43         'only_matching': True,
  44     }
  45
  46     @staticmethod
  47     def _extract_urls(webpage):
  48         return [m.group('url') for m in re.finditer(
  49             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
  50             webpage)]
  51
  52     def _real_extract(self, url):
  53         query = parse_qs(url)
  54         api_url = query['url'][0]
  55         secret_token = query.get('secret_token')
  56         if secret_token:
  57             api_url = update_url_query(api_url, {'secret_token': secret_token[0]})
  58         return self.url_result(api_url)
  59
  60
  61 class SoundcloudIE(InfoExtractor):
  62     """Information extractor for soundcloud.com
  63        To access the media, the uid of the song and a stream token
  64        must be extracted from the page source and the script must make
  65        a request to media.soundcloud.com/crossdomain.xml. Then
  66        the media can be grabbed by requesting from an url composed
  67        of the stream token and uid
  68      """
  69
  70     _VALID_URL = r'''(?x)^(?:https?://)?
  71                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
  72                             (?!stations/track)
  73                             (?P<uploader>[\w\d-]+)/
  74                             (?!(?:tracks|albums|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
  75                             (?P<title>[\w\d-]+)/?
  76                             (?P<token>[^?]+?)?(?:[?].*)?$)
  77                        |(?:api(?:-v2)?\.soundcloud\.com/tracks/(?P<track_id>\d+)
  78                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
  79                     )
  80                     '''
  81     IE_NAME = 'soundcloud'
  82     _TESTS = [
  83         {
  84             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  85             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  86             'info_dict': {
  87                 'id': '62986583',
  88                 'ext': 'mp3',
  89                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
  90                 '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',
  91                 'uploader': 'E.T. ExTerrestrial Music',
  92                 'uploader_id': '1571244',
  93                 'timestamp': 1349920598,
  94                 'upload_date': '20121011',
  95                 'duration': 143.216,
  96                 'license': 'all-rights-reserved',
  97                 'view_count': int,
  98                 'like_count': int,
  99                 'comment_count': int,
 100                 'repost_count': int,
 101             }
 102         },
 103         # geo-restricted
 104         {
 105             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
 106             'info_dict': {
 107                 'id': '47127627',
 108                 'ext': 'mp3',
 109                 'title': 'Goldrushed',
 110                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
 111                 'uploader': 'The Royal Concept',
 112                 'uploader_id': '9615865',
 113                 'timestamp': 1337635207,
 114                 'upload_date': '20120521',
 115                 'duration': 227.155,
 116                 'license': 'all-rights-reserved',
 117                 'view_count': int,
 118                 'like_count': int,
 119                 'comment_count': int,
 120                 'repost_count': int,
 121             },
 122         },
 123         # private link
 124         {
 125             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
 126             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 127             'info_dict': {
 128                 'id': '123998367',
 129                 'ext': 'mp3',
 130                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 131                 'description': 'test chars:  \"\'/\\ä↭',
 132                 'uploader': 'jaimeMF',
 133                 'uploader_id': '69767071',
 134                 'timestamp': 1386604920,
 135                 'upload_date': '20131209',
 136                 'duration': 9.927,
 137                 'license': 'all-rights-reserved',
 138                 'view_count': int,
 139                 'like_count': int,
 140                 'comment_count': int,
 141                 'repost_count': int,
 142             },
 143         },
 144         # private link (alt format)
 145         {
 146             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
 147             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
 148             'info_dict': {
 149                 'id': '123998367',
 150                 'ext': 'mp3',
 151                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 152                 'description': 'test chars:  \"\'/\\ä↭',
 153                 'uploader': 'jaimeMF',
 154                 'uploader_id': '69767071',
 155                 'timestamp': 1386604920,
 156                 'upload_date': '20131209',
 157                 'duration': 9.927,
 158                 'license': 'all-rights-reserved',
 159                 'view_count': int,
 160                 'like_count': int,
 161                 'comment_count': int,
 162                 'repost_count': int,
 163             },
 164         },
 165         # downloadable song
 166         {
 167             'url': 'https://soundcloud.com/the80m/the-following',
 168             'md5': '9ffcddb08c87d74fb5808a3c183a1d04',
 169             'info_dict': {
 170                 'id': '343609555',
 171                 'ext': 'wav',
 172             },
 173         },
 174         # private link, downloadable format
 175         {
 176             'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
 177             'md5': '64a60b16e617d41d0bef032b7f55441e',
 178             'info_dict': {
 179                 'id': '340344461',
 180                 'ext': 'wav',
 181                 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
 182                 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
 183                 'uploader': 'Ori Uplift Music',
 184                 'uploader_id': '12563093',
 185                 'timestamp': 1504206263,
 186                 'upload_date': '20170831',
 187                 'duration': 7449.096,
 188                 'license': 'all-rights-reserved',
 189                 'view_count': int,
 190                 'like_count': int,
 191                 'comment_count': int,
 192                 'repost_count': int,
 193             },
 194         },
 195         # no album art, use avatar pic for thumbnail
 196         {
 197             'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
 198             'md5': '59c7872bc44e5d99b7211891664760c2',
 199             'info_dict': {
 200                 'id': '309699954',
 201                 'ext': 'mp3',
 202                 'title': 'Sideways (Prod. Mad Real)',
 203                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
 204                 'uploader': 'garyvee',
 205                 'uploader_id': '2366352',
 206                 'timestamp': 1488152409,
 207                 'upload_date': '20170226',
 208                 'duration': 207.012,
 209                 'thumbnail': r're:https?://.*\.jpg',
 210                 'license': 'all-rights-reserved',
 211                 'view_count': int,
 212                 'like_count': int,
 213                 'comment_count': int,
 214                 'repost_count': int,
 215             },
 216             'params': {
 217                 'skip_download': True,
 218             },
 219         },
 220         {
 221             'url': 'https://soundcloud.com/giovannisarani/mezzo-valzer',
 222             'md5': 'e22aecd2bc88e0e4e432d7dcc0a1abf7',
 223             'info_dict': {
 224                 'id': '583011102',
 225                 'ext': 'mp3',
 226                 'title': 'Mezzo Valzer',
 227                 'description': 'md5:4138d582f81866a530317bae316e8b61',
 228                 'uploader': 'Micronie',
 229                 'uploader_id': '3352531',
 230                 'timestamp': 1551394171,
 231                 'upload_date': '20190228',
 232                 'duration': 180.157,
 233                 'thumbnail': r're:https?://.*\.jpg',
 234                 'license': 'all-rights-reserved',
 235                 'view_count': int,
 236                 'like_count': int,
 237                 'comment_count': int,
 238                 'repost_count': int,
 239             },
 240         },
 241         {
 242             # AAC HQ format available (account with active subscription needed)
 243             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
 244             'only_matching': True,
 245         },
 246         {
 247             # Go+ (account with active subscription needed)
 248             'url': 'https://soundcloud.com/taylorswiftofficial/look-what-you-made-me-do',
 249             'only_matching': True,
 250         },
 251     ]
 252
 253     _API_V2_BASE = 'https://api-v2.soundcloud.com/'
 254     _BASE_URL = 'https://soundcloud.com/'
 255     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
 256
 257     _ARTWORK_MAP = {
 258         'mini': 16,
 259         'tiny': 20,
 260         'small': 32,
 261         'badge': 47,
 262         't67x67': 67,
 263         'large': 100,
 264         't300x300': 300,
 265         'crop': 400,
 266         't500x500': 500,
 267         'original': 0,
 268     }
 269
 270     def _store_client_id(self, client_id):
 271         self._downloader.cache.store('soundcloud', 'client_id', client_id)
 272
 273     def _update_client_id(self):
 274         webpage = self._download_webpage('https://soundcloud.com/', None)
 275         for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
 276             script = self._download_webpage(src, None, fatal=False)
 277             if script:
 278                 client_id = self._search_regex(
 279                     r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
 280                     script, 'client id', default=None)
 281                 if client_id:
 282                     self._CLIENT_ID = client_id
 283                     self._store_client_id(client_id)
 284                     return
 285         raise ExtractorError('Unable to extract client id')
 286
 287     def _download_json(self, *args, **kwargs):
 288         non_fatal = kwargs.get('fatal') is False
 289         if non_fatal:
 290             del kwargs['fatal']
 291         query = kwargs.get('query', {}).copy()
 292         for _ in range(2):
 293             query['client_id'] = self._CLIENT_ID
 294             kwargs['query'] = query
 295             try:
 296                 return super(SoundcloudIE, self)._download_json(*args, **compat_kwargs(kwargs))
 297             except ExtractorError as e:
 298                 if isinstance(e.cause, compat_HTTPError) and e.cause.code in (401, 403):
 299                     self._store_client_id(None)
 300                     self._update_client_id()
 301                     continue
 302                 elif non_fatal:
 303                     self.report_warning(error_to_compat_str(e))
 304                     return False
 305                 raise
 306
 307     def _real_initialize(self):
 308         self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or 'a3e059563d7fd3372b49b37f00a00bcf'
 309         self._login()
 310
 311     _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'
 312     _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
 313     _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
 314     _API_VERIFY_AUTH_TOKEN = 'https://api-auth.soundcloud.com/connect/session%s'
 315     _access_token = None
 316     _HEADERS = {}
 317     _NETRC_MACHINE = 'soundcloud'
 318
 319     def _login(self):
 320         username, password = self._get_login_info()
 321         if username is None:
 322             return
 323
 324         if username == 'oauth' and password is not None:
 325             self._access_token = password
 326             query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
 327             payload = {'session': {'access_token': self._access_token}}
 328             token_verification = sanitized_Request(self._API_VERIFY_AUTH_TOKEN % query, json.dumps(payload).encode('utf-8'))
 329             response = self._download_json(token_verification, None, note='Verifying login token...', fatal=False)
 330             if response is not False:
 331                 self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
 332                 self.report_login()
 333             else:
 334                 self.report_warning('Provided authorization token seems to be invalid. Continue as guest')
 335         elif username is not None:
 336             self.report_warning(
 337                 'Login using username and password is not currently supported. '
 338                 'Use "--user oauth --password <oauth_token>" to login using an oauth token')
 339
 340         r'''
 341         def genDevId():
 342             def genNumBlock():
 343                 return ''.join([str(random.randrange(10)) for i in range(6)])
 344             return '-'.join([genNumBlock() for i in range(4)])
 345
 346         payload = {
 347             'client_id': self._CLIENT_ID,
 348             'recaptcha_pubkey': 'null',
 349             'recaptcha_response': 'null',
 350             'credentials': {
 351                 'identifier': username,
 352                 'password': password
 353             },
 354             'signature': self.sign(username, password, self._CLIENT_ID),
 355             'device_id': genDevId(),
 356             'user_agent': self._USER_AGENT
 357         }
 358
 359         query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
 360         login = sanitized_Request(self._API_AUTH_URL_PW % query, json.dumps(payload).encode('utf-8'))
 361         response = self._download_json(login, None)
 362         self._access_token = response.get('session').get('access_token')
 363         if not self._access_token:
 364             self.report_warning('Unable to get access token, login may has failed')
 365         else:
 366             self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
 367         '''
 368
 369     # signature generation
 370     def sign(self, user, pw, clid):
 371         a = 33
 372         i = 1
 373         s = 440123
 374         w = 117
 375         u = 1800000
 376         l = 1042
 377         b = 37
 378         k = 37
 379         c = 5
 380         n = '0763ed7314c69015fd4a0dc16bbf4b90'  # _KEY
 381         y = '8'  # _REV
 382         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
 383         e = user  # _USERNAME
 384         t = clid  # _CLIENT_ID
 385
 386         d = '-'.join([str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
 387         p = n + y + d + r + e + t + d + n
 388         h = p
 389
 390         m = 8011470
 391         f = 0
 392
 393         for f in range(f, len(h)):
 394             m = (m >> 1) + ((1 & m) << 23)
 395             m += ord(h[f])
 396             m &= 16777215
 397
 398         # c is not even needed
 399         out = str(y) + ':' + str(d) + ':' + format(m, 'x') + ':' + str(c)
 400
 401         return out
 402
 403     @classmethod
 404     def _resolv_url(cls, url):
 405         return SoundcloudIE._API_V2_BASE + 'resolve?url=' + url
 406
 407     def _extract_info_dict(self, info, full_title=None, secret_token=None):
 408         track_id = compat_str(info['id'])
 409         title = info['title']
 410
 411         format_urls = set()
 412         formats = []
 413         query = {'client_id': self._CLIENT_ID}
 414         if secret_token:
 415             query['secret_token'] = secret_token
 416
 417         if info.get('downloadable') and info.get('has_downloads_left'):
 418             download_url = update_url_query(
 419                 self._API_V2_BASE + 'tracks/' + track_id + '/download', query)
 420             redirect_url = (self._download_json(download_url, track_id, fatal=False) or {}).get('redirectUri')
 421             if redirect_url:
 422                 urlh = self._request_webpage(
 423                     HEADRequest(redirect_url), track_id, fatal=False)
 424                 if urlh:
 425                     format_url = urlh.geturl()
 426                     format_urls.add(format_url)
 427                     formats.append({
 428                         'format_id': 'download',
 429                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
 430                         'filesize': int_or_none(urlh.headers.get('Content-Length')),
 431                         'url': format_url,
 432                         'quality': 10,
 433                     })
 434
 435         def invalid_url(url):
 436             return not url or url in format_urls
 437
 438         def add_format(f, protocol, is_preview=False):
 439             mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url)
 440             if mobj:
 441                 for k, v in mobj.groupdict().items():
 442                     if not f.get(k):
 443                         f[k] = v
 444             format_id_list = []
 445             if protocol:
 446                 format_id_list.append(protocol)
 447             ext = f.get('ext')
 448             if ext == 'aac':
 449                 f['abr'] = '256'
 450             for k in ('ext', 'abr'):
 451                 v = f.get(k)
 452                 if v:
 453                     format_id_list.append(v)
 454             preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url'])
 455             if preview:
 456                 format_id_list.append('preview')
 457             abr = f.get('abr')
 458             if abr:
 459                 f['abr'] = int(abr)
 460             if protocol == 'hls':
 461                 protocol = 'm3u8' if ext == 'aac' else 'm3u8_native'
 462             else:
 463                 protocol = 'http'
 464             f.update({
 465                 'format_id': '_'.join(format_id_list),
 466                 'protocol': protocol,
 467                 'preference': -10 if preview else None,
 468             })
 469             formats.append(f)
 470
 471         # New API
 472         transcodings = try_get(
 473             info, lambda x: x['media']['transcodings'], list) or []
 474         for t in transcodings:
 475             if not isinstance(t, dict):
 476                 continue
 477             format_url = url_or_none(t.get('url'))
 478             if not format_url:
 479                 continue
 480             stream = self._download_json(
 481                 format_url, track_id, query=query, fatal=False, headers=self._HEADERS)
 482             if not isinstance(stream, dict):
 483                 continue
 484             stream_url = url_or_none(stream.get('url'))
 485             if invalid_url(stream_url):
 486                 continue
 487             format_urls.add(stream_url)
 488             stream_format = t.get('format') or {}
 489             protocol = stream_format.get('protocol')
 490             if protocol != 'hls' and '/hls' in format_url:
 491                 protocol = 'hls'
 492             ext = None
 493             preset = str_or_none(t.get('preset'))
 494             if preset:
 495                 ext = preset.split('_')[0]
 496             if ext not in KNOWN_EXTENSIONS:
 497                 ext = mimetype2ext(stream_format.get('mime_type'))
 498             add_format({
 499                 'url': stream_url,
 500                 'ext': ext,
 501             }, 'http' if protocol == 'progressive' else protocol,
 502                 t.get('snipped') or '/preview/' in format_url)
 503
 504         for f in formats:
 505             f['vcodec'] = 'none'
 506
 507         if not formats and info.get('policy') == 'BLOCK':
 508             self.raise_geo_restricted(metadata_available=True)
 509         self._sort_formats(formats)
 510
 511         user = info.get('user') or {}
 512
 513         thumbnails = []
 514         artwork_url = info.get('artwork_url')
 515         thumbnail = artwork_url or user.get('avatar_url')
 516         if isinstance(thumbnail, compat_str):
 517             if re.search(self._IMAGE_REPL_RE, thumbnail):
 518                 for image_id, size in self._ARTWORK_MAP.items():
 519                     i = {
 520                         'id': image_id,
 521                         'url': re.sub(self._IMAGE_REPL_RE, '-%s.jpg' % image_id, thumbnail),
 522                     }
 523                     if image_id == 'tiny' and not artwork_url:
 524                         size = 18
 525                     elif image_id == 'original':
 526                         i['preference'] = 10
 527                     if size:
 528                         i.update({
 529                             'width': size,
 530                             'height': size,
 531                         })
 532                     thumbnails.append(i)
 533             else:
 534                 thumbnails = [{'url': thumbnail}]
 535
 536         def extract_count(key):
 537             return int_or_none(info.get('%s_count' % key))
 538
 539         return {
 540             'id': track_id,
 541             'uploader': user.get('username'),
 542             'uploader_id': str_or_none(user.get('id')) or user.get('permalink'),
 543             'uploader_url': user.get('permalink_url'),
 544             'timestamp': unified_timestamp(info.get('created_at')),
 545             'title': title,
 546             'description': info.get('description'),
 547             'thumbnails': thumbnails,
 548             'duration': float_or_none(info.get('duration'), 1000),
 549             'webpage_url': info.get('permalink_url'),
 550             'license': info.get('license'),
 551             'view_count': extract_count('playback'),
 552             'like_count': extract_count('favoritings') or extract_count('likes'),
 553             'comment_count': extract_count('comment'),
 554             'repost_count': extract_count('reposts'),
 555             'genre': info.get('genre'),
 556             'formats': formats
 557         }
 558
 559     def _real_extract(self, url):
 560         mobj = self._match_valid_url(url)
 561
 562         track_id = mobj.group('track_id')
 563
 564         query = {}
 565         if track_id:
 566             info_json_url = self._API_V2_BASE + 'tracks/' + track_id
 567             full_title = track_id
 568             token = mobj.group('secret_token')
 569             if token:
 570                 query['secret_token'] = token
 571         else:
 572             full_title = resolve_title = '%s/%s' % mobj.group('uploader', 'title')
 573             token = mobj.group('token')
 574             if token:
 575                 resolve_title += '/%s' % token
 576             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 577
 578         info = self._download_json(
 579             info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 580
 581         return self._extract_info_dict(info, full_title, token)
 582
 583
 584 class SoundcloudPlaylistBaseIE(SoundcloudIE):
 585     def _extract_set(self, playlist, token=None):
 586         playlist_id = compat_str(playlist['id'])
 587         tracks = playlist.get('tracks') or []
 588         if not all([t.get('permalink_url') for t in tracks]) and token:
 589             tracks = self._download_json(
 590                 self._API_V2_BASE + 'tracks', playlist_id,
 591                 'Downloading tracks', query={
 592                     'ids': ','.join([compat_str(t['id']) for t in tracks]),
 593                     'playlistId': playlist_id,
 594                     'playlistSecretToken': token,
 595                 }, headers=self._HEADERS)
 596         entries = []
 597         for track in tracks:
 598             track_id = str_or_none(track.get('id'))
 599             url = track.get('permalink_url')
 600             if not url:
 601                 if not track_id:
 602                     continue
 603                 url = self._API_V2_BASE + 'tracks/' + track_id
 604                 if token:
 605                     url += '?secret_token=' + token
 606             entries.append(self.url_result(
 607                 url, SoundcloudIE.ie_key(), track_id))
 608         return self.playlist_result(
 609             entries, playlist_id,
 610             playlist.get('title'),
 611             playlist.get('description'))
 612
 613
 614 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 615     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[:\w\d-]+)(?:/(?P<token>[^?/]+))?'
 616     IE_NAME = 'soundcloud:set'
 617     _TESTS = [{
 618         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 619         'info_dict': {
 620             'id': '2284613',
 621             'title': 'The Royal Concept EP',
 622             'description': 'md5:71d07087c7a449e8941a70a29e34671e',
 623         },
 624         'playlist_mincount': 5,
 625     }, {
 626         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 627         'only_matching': True,
 628     }, {
 629         'url': 'https://soundcloud.com/discover/sets/weekly::flacmatic',
 630         'only_matching': True,
 631     }, {
 632         'url': 'https://soundcloud.com/discover/sets/charts-top:all-music:de',
 633         'only_matching': True,
 634     }, {
 635         'url': 'https://soundcloud.com/discover/sets/charts-top:hiphoprap:kr',
 636         'only_matching': True,
 637     }]
 638
 639     def _real_extract(self, url):
 640         mobj = self._match_valid_url(url)
 641
 642         full_title = '%s/sets/%s' % mobj.group('uploader', 'slug_title')
 643         token = mobj.group('token')
 644         if token:
 645             full_title += '/' + token
 646
 647         info = self._download_json(self._resolv_url(
 648             self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 649
 650         if 'errors' in info:
 651             msgs = (compat_str(err['error_message']) for err in info['errors'])
 652             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 653
 654         return self._extract_set(info, token)
 655
 656
 657 class SoundcloudPagedPlaylistBaseIE(SoundcloudIE):
 658     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 659         return {
 660             '_type': 'playlist',
 661             'id': playlist_id,
 662             'title': playlist_title,
 663             'entries': self._entries(base_url, playlist_id),
 664         }
 665
 666     def _entries(self, url, playlist_id):
 667         # Per the SoundCloud documentation, the maximum limit for a linked partitioning query is 200.
 668         # https://developers.soundcloud.com/blog/offset-pagination-deprecated
 669         query = {
 670             'limit': 200,
 671             'linked_partitioning': '1',
 672             'offset': 0,
 673         }
 674
 675         retries = self.get_param('extractor_retries', 3)
 676
 677         for i in itertools.count():
 678             attempt, last_error = -1, None
 679             while attempt < retries:
 680                 attempt += 1
 681                 if last_error:
 682                     self.report_warning('%s. Retrying ...' % remove_end(last_error, '.'), playlist_id)
 683                 try:
 684                     response = self._download_json(
 685                         url, playlist_id, query=query, headers=self._HEADERS,
 686                         note='Downloading track page %s%s' % (i + 1, f' (retry #{attempt})' if attempt else ''))
 687                     break
 688                 except ExtractorError as e:
 689                     # Downloading page may result in intermittent 502 HTTP error
 690                     # See https://github.com/yt-dlp/yt-dlp/issues/872
 691                     if attempt >= retries or not isinstance(e.cause, compat_HTTPError) or e.cause.code != 502:
 692                         raise
 693                     last_error = str(e.cause or e.msg)
 694
 695             def resolve_entry(*candidates):
 696                 for cand in candidates:
 697                     if not isinstance(cand, dict):
 698                         continue
 699                     permalink_url = url_or_none(cand.get('permalink_url'))
 700                     if permalink_url:
 701                         return self.url_result(
 702                             permalink_url,
 703                             SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
 704                             str_or_none(cand.get('id')), cand.get('title'))
 705
 706             for e in response['collection'] or []:
 707                 yield resolve_entry(e, e.get('track'), e.get('playlist'))
 708
 709             url = response.get('next_href')
 710             if not url:
 711                 break
 712             query.pop('offset', None)
 713
 714
 715 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 716     _VALID_URL = r'''(?x)
 717                         https?://
 718                             (?:(?:www|m)\.)?soundcloud\.com/
 719                             (?P<user>[^/]+)
 720                             (?:/
 721                                 (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight)
 722                             )?
 723                             /?(?:[?#].*)?$
 724                     '''
 725     IE_NAME = 'soundcloud:user'
 726     _TESTS = [{
 727         'url': 'https://soundcloud.com/soft-cell-official',
 728         'info_dict': {
 729             'id': '207965082',
 730             'title': 'Soft Cell (All)',
 731         },
 732         'playlist_mincount': 28,
 733     }, {
 734         'url': 'https://soundcloud.com/soft-cell-official/tracks',
 735         'info_dict': {
 736             'id': '207965082',
 737             'title': 'Soft Cell (Tracks)',
 738         },
 739         'playlist_mincount': 27,
 740     }, {
 741         'url': 'https://soundcloud.com/soft-cell-official/albums',
 742         'info_dict': {
 743             'id': '207965082',
 744             'title': 'Soft Cell (Albums)',
 745         },
 746         'playlist_mincount': 1,
 747     }, {
 748         'url': 'https://soundcloud.com/jcv246/sets',
 749         'info_dict': {
 750             'id': '12982173',
 751             'title': 'Jordi / cv (Sets)',
 752         },
 753         'playlist_mincount': 2,
 754     }, {
 755         'url': 'https://soundcloud.com/jcv246/reposts',
 756         'info_dict': {
 757             'id': '12982173',
 758             'title': 'Jordi / cv (Reposts)',
 759         },
 760         'playlist_mincount': 6,
 761     }, {
 762         'url': 'https://soundcloud.com/clalberg/likes',
 763         'info_dict': {
 764             'id': '11817582',
 765             'title': 'clalberg (Likes)',
 766         },
 767         'playlist_mincount': 5,
 768     }, {
 769         'url': 'https://soundcloud.com/grynpyret/spotlight',
 770         'info_dict': {
 771             'id': '7098329',
 772             'title': 'Grynpyret (Spotlight)',
 773         },
 774         'playlist_mincount': 1,
 775     }]
 776
 777     _BASE_URL_MAP = {
 778         'all': 'stream/users/%s',
 779         'tracks': 'users/%s/tracks',
 780         'albums': 'users/%s/albums',
 781         'sets': 'users/%s/playlists',
 782         'reposts': 'stream/users/%s/reposts',
 783         'likes': 'users/%s/likes',
 784         'spotlight': 'users/%s/spotlight',
 785     }
 786
 787     def _real_extract(self, url):
 788         mobj = self._match_valid_url(url)
 789         uploader = mobj.group('user')
 790
 791         user = self._download_json(
 792             self._resolv_url(self._BASE_URL + uploader),
 793             uploader, 'Downloading user info', headers=self._HEADERS)
 794
 795         resource = mobj.group('rsrc') or 'all'
 796
 797         return self._extract_playlist(
 798             self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'],
 799             str_or_none(user.get('id')),
 800             '%s (%s)' % (user['username'], resource.capitalize()))
 801
 802
 803 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 804     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 805     IE_NAME = 'soundcloud:trackstation'
 806     _TESTS = [{
 807         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 808         'info_dict': {
 809             'id': '286017854',
 810             'title': 'Track station: your text',
 811         },
 812         'playlist_mincount': 47,
 813     }]
 814
 815     def _real_extract(self, url):
 816         track_name = self._match_id(url)
 817
 818         track = self._download_json(self._resolv_url(url), track_name, headers=self._HEADERS)
 819         track_id = self._search_regex(
 820             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 821
 822         return self._extract_playlist(
 823             self._API_V2_BASE + 'stations/%s/tracks' % track['id'],
 824             track_id, 'Track station: %s' % track['title'])
 825
 826
 827 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 828     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 829     IE_NAME = 'soundcloud:playlist'
 830     _TESTS = [{
 831         'url': 'https://api.soundcloud.com/playlists/4110309',
 832         'info_dict': {
 833             'id': '4110309',
 834             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 835             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 836         },
 837         'playlist_count': 6,
 838     }]
 839
 840     def _real_extract(self, url):
 841         mobj = self._match_valid_url(url)
 842         playlist_id = mobj.group('id')
 843
 844         query = {}
 845         token = mobj.group('token')
 846         if token:
 847             query['secret_token'] = token
 848
 849         data = self._download_json(
 850             self._API_V2_BASE + 'playlists/' + playlist_id,
 851             playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 852
 853         return self._extract_set(data, token)
 854
 855
 856 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 857     IE_NAME = 'soundcloud:search'
 858     IE_DESC = 'Soundcloud search, "scsearch" keyword'
 859     _MAX_RESULTS = float('inf')
 860     _TESTS = [{
 861         'url': 'scsearch15:post-avant jazzcore',
 862         'info_dict': {
 863             'title': 'post-avant jazzcore',
 864         },
 865         'playlist_count': 15,
 866     }]
 867
 868     _SEARCH_KEY = 'scsearch'
 869     _MAX_RESULTS_PER_PAGE = 200
 870     _DEFAULT_RESULTS_PER_PAGE = 50
 871
 872     def _get_collection(self, endpoint, collection_id, **query):
 873         limit = min(
 874             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 875             self._MAX_RESULTS_PER_PAGE)
 876         query.update({
 877             'limit': limit,
 878             'linked_partitioning': 1,
 879             'offset': 0,
 880         })
 881         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
 882
 883         for i in itertools.count(1):
 884             response = self._download_json(
 885                 next_url, collection_id, f'Downloading page {i}',
 886                 'Unable to download API page', headers=self._HEADERS)
 887
 888             for item in response.get('collection') or []:
 889                 if item:
 890                     yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 891
 892             next_url = response.get('next_href')
 893             if not next_url:
 894                 break
 895
 896     def _get_n_results(self, query, n):
 897         tracks = self._get_collection('search/tracks', query, limit=n, q=query)
 898         return self.playlist_result(tracks, query, query)