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     compat_urlparse,
  18 )
  19 from ..utils import (
  20     error_to_compat_str,
  21     ExtractorError,
  22     float_or_none,
  23     HEADRequest,
  24     int_or_none,
  25     KNOWN_EXTENSIONS,
  26     mimetype2ext,
  27     str_or_none,
  28     try_get,
  29     unified_timestamp,
  30     update_url_query,
  31     url_or_none,
  32     urlhandle_detect_ext,
  33     sanitized_Request,
  34 )
  35
  36
  37 class SoundcloudEmbedIE(InfoExtractor):
  38     _VALID_URL = r'https?://(?:w|player|p)\.soundcloud\.com/player/?.*?\burl=(?P<id>.+)'
  39     _TEST = {
  40         # from https://www.soundi.fi/uutiset/ennakkokuuntelussa-timo-kaukolammen-station-to-station-to-station-julkaisua-juhlitaan-tanaan-g-livelabissa/
  41         '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',
  42         'only_matching': True,
  43     }
  44
  45     @staticmethod
  46     def _extract_urls(webpage):
  47         return [m.group('url') for m in re.finditer(
  48             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
  49             webpage)]
  50
  51     def _real_extract(self, url):
  52         query = compat_urlparse.parse_qs(
  53             compat_urlparse.urlparse(url).query)
  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/oddsamples/bus-brakes',
 168             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 169             'info_dict': {
 170                 'id': '128590877',
 171                 'ext': 'mp3',
 172                 'title': 'Bus Brakes',
 173                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 174                 'uploader': 'oddsamples',
 175                 'uploader_id': '73680509',
 176                 'timestamp': 1389232924,
 177                 'upload_date': '20140109',
 178                 'duration': 17.346,
 179                 'license': 'cc-by-sa',
 180                 'view_count': int,
 181                 'like_count': int,
 182                 'comment_count': int,
 183                 'repost_count': int,
 184             },
 185         },
 186         # private link, downloadable format
 187         {
 188             'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
 189             'md5': '64a60b16e617d41d0bef032b7f55441e',
 190             'info_dict': {
 191                 'id': '340344461',
 192                 'ext': 'wav',
 193                 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
 194                 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
 195                 'uploader': 'Ori Uplift Music',
 196                 'uploader_id': '12563093',
 197                 'timestamp': 1504206263,
 198                 'upload_date': '20170831',
 199                 'duration': 7449.096,
 200                 'license': 'all-rights-reserved',
 201                 'view_count': int,
 202                 'like_count': int,
 203                 'comment_count': int,
 204                 'repost_count': int,
 205             },
 206         },
 207         # no album art, use avatar pic for thumbnail
 208         {
 209             'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
 210             'md5': '59c7872bc44e5d99b7211891664760c2',
 211             'info_dict': {
 212                 'id': '309699954',
 213                 'ext': 'mp3',
 214                 'title': 'Sideways (Prod. Mad Real)',
 215                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
 216                 'uploader': 'garyvee',
 217                 'uploader_id': '2366352',
 218                 'timestamp': 1488152409,
 219                 'upload_date': '20170226',
 220                 'duration': 207.012,
 221                 'thumbnail': r're:https?://.*\.jpg',
 222                 'license': 'all-rights-reserved',
 223                 'view_count': int,
 224                 'like_count': int,
 225                 'comment_count': int,
 226                 'repost_count': int,
 227             },
 228             'params': {
 229                 'skip_download': True,
 230             },
 231         },
 232         {
 233             'url': 'https://soundcloud.com/giovannisarani/mezzo-valzer',
 234             'md5': 'e22aecd2bc88e0e4e432d7dcc0a1abf7',
 235             'info_dict': {
 236                 'id': '583011102',
 237                 'ext': 'mp3',
 238                 'title': 'Mezzo Valzer',
 239                 'description': 'md5:4138d582f81866a530317bae316e8b61',
 240                 'uploader': 'Micronie',
 241                 'uploader_id': '3352531',
 242                 'timestamp': 1551394171,
 243                 'upload_date': '20190228',
 244                 'duration': 180.157,
 245                 'thumbnail': r're:https?://.*\.jpg',
 246                 'license': 'all-rights-reserved',
 247                 'view_count': int,
 248                 'like_count': int,
 249                 'comment_count': int,
 250                 'repost_count': int,
 251             },
 252         },
 253         {
 254             # AAC HQ format available (account with active subscription needed)
 255             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
 256             'only_matching': True,
 257         },
 258         {
 259             # Go+ (account with active subscription needed)
 260             'url': 'https://soundcloud.com/taylorswiftofficial/look-what-you-made-me-do',
 261             'only_matching': True,
 262         },
 263     ]
 264
 265     _API_V2_BASE = 'https://api-v2.soundcloud.com/'
 266     _BASE_URL = 'https://soundcloud.com/'
 267     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
 268
 269     _ARTWORK_MAP = {
 270         'mini': 16,
 271         'tiny': 20,
 272         'small': 32,
 273         'badge': 47,
 274         't67x67': 67,
 275         'large': 100,
 276         't300x300': 300,
 277         'crop': 400,
 278         't500x500': 500,
 279         'original': 0,
 280     }
 281
 282     def _store_client_id(self, client_id):
 283         self._downloader.cache.store('soundcloud', 'client_id', client_id)
 284
 285     def _update_client_id(self):
 286         webpage = self._download_webpage('https://soundcloud.com/', None)
 287         for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
 288             script = self._download_webpage(src, None, fatal=False)
 289             if script:
 290                 client_id = self._search_regex(
 291                     r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
 292                     script, 'client id', default=None)
 293                 if client_id:
 294                     self._CLIENT_ID = client_id
 295                     self._store_client_id(client_id)
 296                     return
 297         raise ExtractorError('Unable to extract client id')
 298
 299     def _download_json(self, *args, **kwargs):
 300         non_fatal = kwargs.get('fatal') is False
 301         if non_fatal:
 302             del kwargs['fatal']
 303         query = kwargs.get('query', {}).copy()
 304         for _ in range(2):
 305             query['client_id'] = self._CLIENT_ID
 306             kwargs['query'] = query
 307             try:
 308                 return super(SoundcloudIE, self)._download_json(*args, **compat_kwargs(kwargs))
 309             except ExtractorError as e:
 310                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 311                     self._store_client_id(None)
 312                     self._update_client_id()
 313                     continue
 314                 elif non_fatal:
 315                     self.report_warning(error_to_compat_str(e))
 316                     return False
 317                 raise
 318
 319     def _real_initialize(self):
 320         self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or "T5R4kgWS2PRf6lzLyIravUMnKlbIxQag"  # 'EXLwg5lHTO2dslU5EePe3xkw0m1h86Cd' # 'YUKXoArFcqrlQn9tfNHvvyfnDISj04zk'
 321         self._login()
 322
 323     _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"
 324     _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
 325     _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
 326     _access_token = None
 327     _HEADERS = {}
 328     _NETRC_MACHINE = 'soundcloud'
 329
 330     def _login(self):
 331         username, password = self._get_login_info()
 332         if username is None:
 333             return
 334
 335         def genDevId():
 336             def genNumBlock():
 337                 return ''.join([str(random.randrange(10)) for i in range(6)])
 338             return '-'.join([genNumBlock() for i in range(4)])
 339
 340         payload = {
 341             'client_id': self._CLIENT_ID,
 342             'recaptcha_pubkey': 'null',
 343             'recaptcha_response': 'null',
 344             'credentials': {
 345                 'identifier': username,
 346                 'password': password
 347             },
 348             'signature': self.sign(username, password, self._CLIENT_ID),
 349             'device_id': genDevId(),
 350             'user_agent': self._USER_AGENT
 351         }
 352
 353         query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
 354         login = sanitized_Request(self._API_AUTH_URL_PW % query, json.dumps(payload).encode('utf-8'))
 355         response = self._download_json(login, None)
 356         self._access_token = response.get('session').get('access_token')
 357         if not self._access_token:
 358             self.report_warning('Unable to get access token, login may has failed')
 359         else:
 360             self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
 361
 362     # signature generation
 363     def sign(self, user, pw, clid):
 364         a = 33
 365         i = 1
 366         s = 440123
 367         w = 117
 368         u = 1800000
 369         l = 1042
 370         b = 37
 371         k = 37
 372         c = 5
 373         n = "0763ed7314c69015fd4a0dc16bbf4b90"  # _KEY
 374         y = "8"  # _REV
 375         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
 376         e = user  # _USERNAME
 377         t = clid  # _CLIENT_ID
 378
 379         d = '-'.join([str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
 380         p = n + y + d + r + e + t + d + n
 381         h = p
 382
 383         m = 8011470
 384         f = 0
 385
 386         for f in range(f, len(h)):
 387             m = (m >> 1) + ((1 & m) << 23)
 388             m += ord(h[f])
 389             m &= 16777215
 390
 391         # c is not even needed
 392         out = str(y) + ':' + str(d) + ':' + format(m, 'x') + ':' + str(c)
 393
 394         return out
 395
 396     @classmethod
 397     def _resolv_url(cls, url):
 398         return SoundcloudIE._API_V2_BASE + 'resolve?url=' + url
 399
 400     def _extract_info_dict(self, info, full_title=None, secret_token=None):
 401         track_id = compat_str(info['id'])
 402         title = info['title']
 403
 404         format_urls = set()
 405         formats = []
 406         query = {'client_id': self._CLIENT_ID}
 407         if secret_token:
 408             query['secret_token'] = secret_token
 409
 410         if info.get('downloadable') and info.get('has_downloads_left'):
 411             download_url = update_url_query(
 412                 self._API_V2_BASE + 'tracks/' + track_id + '/download', query)
 413             redirect_url = (self._download_json(download_url, track_id, fatal=False) or {}).get('redirectUri')
 414             if redirect_url:
 415                 urlh = self._request_webpage(
 416                     HEADRequest(redirect_url), track_id, fatal=False)
 417                 if urlh:
 418                     format_url = urlh.geturl()
 419                     format_urls.add(format_url)
 420                     formats.append({
 421                         'format_id': 'download',
 422                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
 423                         'filesize': int_or_none(urlh.headers.get('Content-Length')),
 424                         'url': format_url,
 425                         'quality': 10,
 426                     })
 427
 428         def invalid_url(url):
 429             return not url or url in format_urls
 430
 431         def add_format(f, protocol, is_preview=False):
 432             mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url)
 433             if mobj:
 434                 for k, v in mobj.groupdict().items():
 435                     if not f.get(k):
 436                         f[k] = v
 437             format_id_list = []
 438             if protocol:
 439                 format_id_list.append(protocol)
 440             ext = f.get('ext')
 441             if ext == 'aac':
 442                 f['abr'] = '256'
 443             for k in ('ext', 'abr'):
 444                 v = f.get(k)
 445                 if v:
 446                     format_id_list.append(v)
 447             preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url'])
 448             if preview:
 449                 format_id_list.append('preview')
 450             abr = f.get('abr')
 451             if abr:
 452                 f['abr'] = int(abr)
 453             if protocol == 'hls':
 454                 protocol = 'm3u8' if ext == 'aac' else 'm3u8_native'
 455             else:
 456                 protocol = 'http'
 457             f.update({
 458                 'format_id': '_'.join(format_id_list),
 459                 'protocol': protocol,
 460                 'preference': -10 if preview else None,
 461             })
 462             formats.append(f)
 463
 464         # New API
 465         transcodings = try_get(
 466             info, lambda x: x['media']['transcodings'], list) or []
 467         for t in transcodings:
 468             if not isinstance(t, dict):
 469                 continue
 470             format_url = url_or_none(t.get('url'))
 471             if not format_url:
 472                 continue
 473             stream = self._download_json(
 474                 format_url, track_id, query=query, fatal=False, headers=self._HEADERS)
 475             if not isinstance(stream, dict):
 476                 continue
 477             stream_url = url_or_none(stream.get('url'))
 478             if invalid_url(stream_url):
 479                 continue
 480             format_urls.add(stream_url)
 481             stream_format = t.get('format') or {}
 482             protocol = stream_format.get('protocol')
 483             if protocol != 'hls' and '/hls' in format_url:
 484                 protocol = 'hls'
 485             ext = None
 486             preset = str_or_none(t.get('preset'))
 487             if preset:
 488                 ext = preset.split('_')[0]
 489             if ext not in KNOWN_EXTENSIONS:
 490                 ext = mimetype2ext(stream_format.get('mime_type'))
 491             add_format({
 492                 'url': stream_url,
 493                 'ext': ext,
 494             }, 'http' if protocol == 'progressive' else protocol,
 495                 t.get('snipped') or '/preview/' in format_url)
 496
 497         for f in formats:
 498             f['vcodec'] = 'none'
 499
 500         if not formats and info.get('policy') == 'BLOCK':
 501             self.raise_geo_restricted(metadata_available=True)
 502         self._sort_formats(formats)
 503
 504         user = info.get('user') or {}
 505
 506         thumbnails = []
 507         artwork_url = info.get('artwork_url')
 508         thumbnail = artwork_url or user.get('avatar_url')
 509         if isinstance(thumbnail, compat_str):
 510             if re.search(self._IMAGE_REPL_RE, thumbnail):
 511                 for image_id, size in self._ARTWORK_MAP.items():
 512                     i = {
 513                         'id': image_id,
 514                         'url': re.sub(self._IMAGE_REPL_RE, '-%s.jpg' % image_id, thumbnail),
 515                     }
 516                     if image_id == 'tiny' and not artwork_url:
 517                         size = 18
 518                     elif image_id == 'original':
 519                         i['preference'] = 10
 520                     if size:
 521                         i.update({
 522                             'width': size,
 523                             'height': size,
 524                         })
 525                     thumbnails.append(i)
 526             else:
 527                 thumbnails = [{'url': thumbnail}]
 528
 529         def extract_count(key):
 530             return int_or_none(info.get('%s_count' % key))
 531
 532         return {
 533             'id': track_id,
 534             'uploader': user.get('username'),
 535             'uploader_id': str_or_none(user.get('id')) or user.get('permalink'),
 536             'uploader_url': user.get('permalink_url'),
 537             'timestamp': unified_timestamp(info.get('created_at')),
 538             'title': title,
 539             'description': info.get('description'),
 540             'thumbnails': thumbnails,
 541             'duration': float_or_none(info.get('duration'), 1000),
 542             'webpage_url': info.get('permalink_url'),
 543             'license': info.get('license'),
 544             'view_count': extract_count('playback'),
 545             'like_count': extract_count('favoritings') or extract_count('likes'),
 546             'comment_count': extract_count('comment'),
 547             'repost_count': extract_count('reposts'),
 548             'genre': info.get('genre'),
 549             'formats': formats
 550         }
 551
 552     def _real_extract(self, url):
 553         mobj = re.match(self._VALID_URL, url)
 554
 555         track_id = mobj.group('track_id')
 556
 557         query = {}
 558         if track_id:
 559             info_json_url = self._API_V2_BASE + 'tracks/' + track_id
 560             full_title = track_id
 561             token = mobj.group('secret_token')
 562             if token:
 563                 query['secret_token'] = token
 564         else:
 565             full_title = resolve_title = '%s/%s' % mobj.group('uploader', 'title')
 566             token = mobj.group('token')
 567             if token:
 568                 resolve_title += '/%s' % token
 569             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 570
 571         info = self._download_json(
 572             info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 573
 574         return self._extract_info_dict(info, full_title, token)
 575
 576
 577 class SoundcloudPlaylistBaseIE(SoundcloudIE):
 578     def _extract_set(self, playlist, token=None):
 579         playlist_id = compat_str(playlist['id'])
 580         tracks = playlist.get('tracks') or []
 581         if not all([t.get('permalink_url') for t in tracks]) and token:
 582             tracks = self._download_json(
 583                 self._API_V2_BASE + 'tracks', playlist_id,
 584                 'Downloading tracks', query={
 585                     'ids': ','.join([compat_str(t['id']) for t in tracks]),
 586                     'playlistId': playlist_id,
 587                     'playlistSecretToken': token,
 588                 }, headers=self._HEADERS)
 589         entries = []
 590         for track in tracks:
 591             track_id = str_or_none(track.get('id'))
 592             url = track.get('permalink_url')
 593             if not url:
 594                 if not track_id:
 595                     continue
 596                 url = self._API_V2_BASE + 'tracks/' + track_id
 597                 if token:
 598                     url += '?secret_token=' + token
 599             entries.append(self.url_result(
 600                 url, SoundcloudIE.ie_key(), track_id))
 601         return self.playlist_result(
 602             entries, playlist_id,
 603             playlist.get('title'),
 604             playlist.get('description'))
 605
 606
 607 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 608     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[:\w\d-]+)(?:/(?P<token>[^?/]+))?'
 609     IE_NAME = 'soundcloud:set'
 610     _TESTS = [{
 611         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 612         'info_dict': {
 613             'id': '2284613',
 614             'title': 'The Royal Concept EP',
 615             'description': 'md5:71d07087c7a449e8941a70a29e34671e',
 616         },
 617         'playlist_mincount': 5,
 618     }, {
 619         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 620         'only_matching': True,
 621     }, {
 622         'url': 'https://soundcloud.com/discover/sets/weekly::flacmatic',
 623         'only_matching': True,
 624     }, {
 625         'url': 'https://soundcloud.com/discover/sets/charts-top:all-music:de',
 626         'only_matching': True,
 627     }, {
 628         'url': 'https://soundcloud.com/discover/sets/charts-top:hiphoprap:kr',
 629         'only_matching': True,
 630     }]
 631
 632     def _real_extract(self, url):
 633         mobj = re.match(self._VALID_URL, url)
 634
 635         full_title = '%s/sets/%s' % mobj.group('uploader', 'slug_title')
 636         token = mobj.group('token')
 637         if token:
 638             full_title += '/' + token
 639
 640         info = self._download_json(self._resolv_url(
 641             self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 642
 643         if 'errors' in info:
 644             msgs = (compat_str(err['error_message']) for err in info['errors'])
 645             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 646
 647         return self._extract_set(info, token)
 648
 649
 650 class SoundcloudPagedPlaylistBaseIE(SoundcloudIE):
 651     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 652         # Per the SoundCloud documentation, the maximum limit for a linked partitioning query is 200.
 653         # https://developers.soundcloud.com/blog/offset-pagination-deprecated
 654         COMMON_QUERY = {
 655             'limit': 200,
 656             'linked_partitioning': '1',
 657         }
 658
 659         query = COMMON_QUERY.copy()
 660         query['offset'] = 0
 661
 662         next_href = base_url
 663
 664         entries = []
 665         for i in itertools.count():
 666             response = self._download_json(
 667                 next_href, playlist_id,
 668                 'Downloading track page %s' % (i + 1), query=query, headers=self._HEADERS)
 669
 670             collection = response['collection']
 671
 672             if not isinstance(collection, list):
 673                 collection = []
 674
 675             # Empty collection may be returned, in this case we proceed
 676             # straight to next_href
 677
 678             def resolve_entry(candidates):
 679                 for cand in candidates:
 680                     if not isinstance(cand, dict):
 681                         continue
 682                     permalink_url = url_or_none(cand.get('permalink_url'))
 683                     if not permalink_url:
 684                         continue
 685                     return self.url_result(
 686                         permalink_url,
 687                         SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
 688                         str_or_none(cand.get('id')), cand.get('title'))
 689
 690             for e in collection:
 691                 entry = resolve_entry((e, e.get('track'), e.get('playlist')))
 692                 if entry:
 693                     entries.append(entry)
 694
 695             next_href = response.get('next_href')
 696             if not next_href:
 697                 break
 698
 699             next_href = response['next_href']
 700             parsed_next_href = compat_urlparse.urlparse(next_href)
 701             query = compat_urlparse.parse_qs(parsed_next_href.query)
 702             query.update(COMMON_QUERY)
 703
 704         return {
 705             '_type': 'playlist',
 706             'id': playlist_id,
 707             'title': playlist_title,
 708             'entries': entries,
 709         }
 710
 711
 712 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 713     _VALID_URL = r'''(?x)
 714                         https?://
 715                             (?:(?:www|m)\.)?soundcloud\.com/
 716                             (?P<user>[^/]+)
 717                             (?:/
 718                                 (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight)
 719                             )?
 720                             /?(?:[?#].*)?$
 721                     '''
 722     IE_NAME = 'soundcloud:user'
 723     _TESTS = [{
 724         'url': 'https://soundcloud.com/soft-cell-official',
 725         'info_dict': {
 726             'id': '207965082',
 727             'title': 'Soft Cell (All)',
 728         },
 729         'playlist_mincount': 28,
 730     }, {
 731         'url': 'https://soundcloud.com/soft-cell-official/tracks',
 732         'info_dict': {
 733             'id': '207965082',
 734             'title': 'Soft Cell (Tracks)',
 735         },
 736         'playlist_mincount': 27,
 737     }, {
 738         'url': 'https://soundcloud.com/soft-cell-official/albums',
 739         'info_dict': {
 740             'id': '207965082',
 741             'title': 'Soft Cell (Albums)',
 742         },
 743         'playlist_mincount': 1,
 744     }, {
 745         'url': 'https://soundcloud.com/jcv246/sets',
 746         'info_dict': {
 747             'id': '12982173',
 748             'title': 'Jordi / cv (Sets)',
 749         },
 750         'playlist_mincount': 2,
 751     }, {
 752         'url': 'https://soundcloud.com/jcv246/reposts',
 753         'info_dict': {
 754             'id': '12982173',
 755             'title': 'Jordi / cv (Reposts)',
 756         },
 757         'playlist_mincount': 6,
 758     }, {
 759         'url': 'https://soundcloud.com/clalberg/likes',
 760         'info_dict': {
 761             'id': '11817582',
 762             'title': 'clalberg (Likes)',
 763         },
 764         'playlist_mincount': 5,
 765     }, {
 766         'url': 'https://soundcloud.com/grynpyret/spotlight',
 767         'info_dict': {
 768             'id': '7098329',
 769             'title': 'Grynpyret (Spotlight)',
 770         },
 771         'playlist_mincount': 1,
 772     }]
 773
 774     _BASE_URL_MAP = {
 775         'all': 'stream/users/%s',
 776         'tracks': 'users/%s/tracks',
 777         'albums': 'users/%s/albums',
 778         'sets': 'users/%s/playlists',
 779         'reposts': 'stream/users/%s/reposts',
 780         'likes': 'users/%s/likes',
 781         'spotlight': 'users/%s/spotlight',
 782     }
 783
 784     def _real_extract(self, url):
 785         mobj = re.match(self._VALID_URL, url)
 786         uploader = mobj.group('user')
 787
 788         user = self._download_json(
 789             self._resolv_url(self._BASE_URL + uploader),
 790             uploader, 'Downloading user info', headers=self._HEADERS)
 791
 792         resource = mobj.group('rsrc') or 'all'
 793
 794         return self._extract_playlist(
 795             self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'],
 796             str_or_none(user.get('id')),
 797             '%s (%s)' % (user['username'], resource.capitalize()))
 798
 799
 800 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 801     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 802     IE_NAME = 'soundcloud:trackstation'
 803     _TESTS = [{
 804         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 805         'info_dict': {
 806             'id': '286017854',
 807             'title': 'Track station: your text',
 808         },
 809         'playlist_mincount': 47,
 810     }]
 811
 812     def _real_extract(self, url):
 813         track_name = self._match_id(url)
 814
 815         track = self._download_json(self._resolv_url(url), track_name, headers=self._HEADERS)
 816         track_id = self._search_regex(
 817             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 818
 819         return self._extract_playlist(
 820             self._API_V2_BASE + 'stations/%s/tracks' % track['id'],
 821             track_id, 'Track station: %s' % track['title'])
 822
 823
 824 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 825     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 826     IE_NAME = 'soundcloud:playlist'
 827     _TESTS = [{
 828         'url': 'https://api.soundcloud.com/playlists/4110309',
 829         'info_dict': {
 830             'id': '4110309',
 831             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 832             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 833         },
 834         'playlist_count': 6,
 835     }]
 836
 837     def _real_extract(self, url):
 838         mobj = re.match(self._VALID_URL, url)
 839         playlist_id = mobj.group('id')
 840
 841         query = {}
 842         token = mobj.group('token')
 843         if token:
 844             query['secret_token'] = token
 845
 846         data = self._download_json(
 847             self._API_V2_BASE + 'playlists/' + playlist_id,
 848             playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 849
 850         return self._extract_set(data, token)
 851
 852
 853 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 854     IE_NAME = 'soundcloud:search'
 855     IE_DESC = 'Soundcloud search'
 856     _MAX_RESULTS = float('inf')
 857     _TESTS = [{
 858         'url': 'scsearch15:post-avant jazzcore',
 859         'info_dict': {
 860             'title': 'post-avant jazzcore',
 861         },
 862         'playlist_count': 15,
 863     }]
 864
 865     _SEARCH_KEY = 'scsearch'
 866     _MAX_RESULTS_PER_PAGE = 200
 867     _DEFAULT_RESULTS_PER_PAGE = 50
 868
 869     def _get_collection(self, endpoint, collection_id, **query):
 870         limit = min(
 871             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 872             self._MAX_RESULTS_PER_PAGE)
 873         query.update({
 874             'limit': limit,
 875             'linked_partitioning': 1,
 876             'offset': 0,
 877         })
 878         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
 879
 880         collected_results = 0
 881
 882         for i in itertools.count(1):
 883             response = self._download_json(
 884                 next_url, collection_id, 'Downloading page {0}'.format(i),
 885                 'Unable to download API page', headers=self._HEADERS)
 886
 887             collection = response.get('collection', [])
 888             if not collection:
 889                 break
 890
 891             collection = list(filter(bool, collection))
 892             collected_results += len(collection)
 893
 894             for item in collection:
 895                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 896
 897             if not collection or collected_results >= limit:
 898                 break
 899
 900             next_url = response.get('next_href')
 901             if not next_url:
 902                 break
 903
 904     def _get_n_results(self, query, n):
 905         tracks = self._get_collection('search/tracks', query, limit=n, q=query)
 906         return self.playlist_result(tracks, playlist_title=query)