youtube_dl/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-dlc-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             # with AAC HQ format available via OAuth token
 255             'url': 'https://soundcloud.com/wandw/the-chainsmokers-ft-daya-dont-let-me-down-ww-remix-1',
 256             'only_matching': True,
 257         },
 258     ]
 259
 260     _API_V2_BASE = 'https://api-v2.soundcloud.com/'
 261     _BASE_URL = 'https://soundcloud.com/'
 262     _IMAGE_REPL_RE = r'-([0-9a-z]+)\.jpg'
 263
 264     _ARTWORK_MAP = {
 265         'mini': 16,
 266         'tiny': 20,
 267         'small': 32,
 268         'badge': 47,
 269         't67x67': 67,
 270         'large': 100,
 271         't300x300': 300,
 272         'crop': 400,
 273         't500x500': 500,
 274         'original': 0,
 275     }
 276
 277     def _store_client_id(self, client_id):
 278         self._downloader.cache.store('soundcloud', 'client_id', client_id)
 279
 280     def _update_client_id(self):
 281         webpage = self._download_webpage('https://soundcloud.com/', None)
 282         for src in reversed(re.findall(r'<script[^>]+src="([^"]+)"', webpage)):
 283             script = self._download_webpage(src, None, fatal=False)
 284             if script:
 285                 client_id = self._search_regex(
 286                     r'client_id\s*:\s*"([0-9a-zA-Z]{32})"',
 287                     script, 'client id', default=None)
 288                 if client_id:
 289                     self._CLIENT_ID = client_id
 290                     self._store_client_id(client_id)
 291                     return
 292         raise ExtractorError('Unable to extract client id')
 293
 294     def _download_json(self, *args, **kwargs):
 295         non_fatal = kwargs.get('fatal') is False
 296         if non_fatal:
 297             del kwargs['fatal']
 298         query = kwargs.get('query', {}).copy()
 299         for _ in range(2):
 300             query['client_id'] = self._CLIENT_ID
 301             kwargs['query'] = query
 302             try:
 303                 return super(SoundcloudIE, self)._download_json(*args, **compat_kwargs(kwargs))
 304             except ExtractorError as e:
 305                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
 306                     self._store_client_id(None)
 307                     self._update_client_id()
 308                     continue
 309                 elif non_fatal:
 310                     self._downloader.report_warning(error_to_compat_str(e))
 311                     return False
 312                 raise
 313
 314     def _real_initialize(self):
 315         self._CLIENT_ID = self._downloader.cache.load('soundcloud', 'client_id') or "T5R4kgWS2PRf6lzLyIravUMnKlbIxQag"  # 'EXLwg5lHTO2dslU5EePe3xkw0m1h86Cd' # 'YUKXoArFcqrlQn9tfNHvvyfnDISj04zk'
 316         self._login()
 317
 318     _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"
 319     _API_AUTH_QUERY_TEMPLATE = '?client_id=%s'
 320     _API_AUTH_URL_PW = 'https://api-auth.soundcloud.com/web-auth/sign-in/password%s'
 321     _access_token = None
 322     _HEADERS = {}
 323     _NETRC_MACHINE = 'soundcloud'
 324
 325     def _login(self):
 326         username, password = self._get_login_info()
 327         if username is None:
 328             return
 329
 330         def genDevId():
 331             def genNumBlock():
 332                 return ''.join([str(random.randrange(10)) for i in range(6)])
 333             return '-'.join([genNumBlock() for i in range(4)])
 334
 335         payload = {
 336             'client_id': self._CLIENT_ID,
 337             'recaptcha_pubkey': 'null',
 338             'recaptcha_response': 'null',
 339             'credentials': {
 340                 'identifier': username,
 341                 'password': password
 342             },
 343             'signature': self.sign(username, password, self._CLIENT_ID),
 344             'device_id': genDevId(),
 345             'user_agent': self._USER_AGENT
 346         }
 347
 348         query = self._API_AUTH_QUERY_TEMPLATE % self._CLIENT_ID
 349         login = sanitized_Request(self._API_AUTH_URL_PW % query, json.dumps(payload).encode('utf-8'))
 350         response = self._download_json(login, None)
 351         self._access_token = response.get('session').get('access_token')
 352         if not self._access_token:
 353             self.report_warning('Unable to get access token, login may has failed')
 354         else:
 355             self._HEADERS = {'Authorization': 'OAuth ' + self._access_token}
 356
 357     # signature generation
 358     def sign(self, user, pw, clid):
 359         a = 33
 360         i = 1
 361         s = 440123
 362         w = 117
 363         u = 1800000
 364         l = 1042
 365         b = 37
 366         k = 37
 367         c = 5
 368         n = "0763ed7314c69015fd4a0dc16bbf4b90"  # _KEY
 369         y = "8"  # _REV
 370         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
 371         e = user  # _USERNAME
 372         t = clid  # _CLIENT_ID
 373
 374         d = '-'.join([str(mInt) for mInt in [a, i, s, w, u, l, b, k]])
 375         p = n + y + d + r + e + t + d + n
 376         h = p
 377
 378         m = 8011470
 379         f = 0
 380
 381         for f in range(f, len(h)):
 382             m = (m >> 1) + ((1 & m) << 23)
 383             m += ord(h[f])
 384             m &= 16777215
 385
 386         # c is not even needed
 387         out = str(y) + ':' + str(d) + ':' + format(m, 'x') + ':' + str(c)
 388
 389         return out
 390
 391     @classmethod
 392     def _resolv_url(cls, url):
 393         return SoundcloudIE._API_V2_BASE + 'resolve?url=' + url
 394
 395     def _extract_info_dict(self, info, full_title=None, secret_token=None):
 396         track_id = compat_str(info['id'])
 397         title = info['title']
 398
 399         format_urls = set()
 400         formats = []
 401         query = {'client_id': self._CLIENT_ID}
 402         if secret_token:
 403             query['secret_token'] = secret_token
 404
 405         if info.get('downloadable') and info.get('has_downloads_left'):
 406             download_url = update_url_query(
 407                 self._API_V2_BASE + 'tracks/' + track_id + '/download', query)
 408             redirect_url = (self._download_json(download_url, track_id, fatal=False) or {}).get('redirectUri')
 409             if redirect_url:
 410                 urlh = self._request_webpage(
 411                     HEADRequest(redirect_url), track_id, fatal=False)
 412                 if urlh:
 413                     format_url = urlh.geturl()
 414                     format_urls.add(format_url)
 415                     formats.append({
 416                         'format_id': 'download',
 417                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
 418                         'filesize': int_or_none(urlh.headers.get('Content-Length')),
 419                         'url': format_url,
 420                         'preference': 10,
 421                     })
 422
 423         def invalid_url(url):
 424             return not url or url in format_urls
 425
 426         def add_format(f, protocol, is_preview=False):
 427             mobj = re.search(r'\.(?P<abr>\d+)\.(?P<ext>[0-9a-z]{3,4})(?=[/?])', stream_url)
 428             if mobj:
 429                 for k, v in mobj.groupdict().items():
 430                     if not f.get(k):
 431                         f[k] = v
 432             format_id_list = []
 433             if protocol:
 434                 format_id_list.append(protocol)
 435             ext = f.get('ext')
 436             if ext == 'aac':
 437                 f['abr'] = '256'
 438             for k in ('ext', 'abr'):
 439                 v = f.get(k)
 440                 if v:
 441                     format_id_list.append(v)
 442             preview = is_preview or re.search(r'/(?:preview|playlist)/0/30/', f['url'])
 443             if preview:
 444                 format_id_list.append('preview')
 445             abr = f.get('abr')
 446             if abr:
 447                 f['abr'] = int(abr)
 448             if protocol == 'hls':
 449                 protocol = 'm3u8' if ext == 'aac' else 'm3u8_native'
 450             else:
 451                 protocol = 'http'
 452             f.update({
 453                 'format_id': '_'.join(format_id_list),
 454                 'protocol': protocol,
 455                 'preference': -10 if preview else None,
 456             })
 457             formats.append(f)
 458
 459         # New API
 460         transcodings = try_get(
 461             info, lambda x: x['media']['transcodings'], list) or []
 462         for t in transcodings:
 463             if not isinstance(t, dict):
 464                 continue
 465             format_url = url_or_none(t.get('url'))
 466             if not format_url:
 467                 continue
 468             stream = self._download_json(
 469                 format_url, track_id, query=query, fatal=False, headers=self._HEADERS)
 470             if not isinstance(stream, dict):
 471                 continue
 472             stream_url = url_or_none(stream.get('url'))
 473             if invalid_url(stream_url):
 474                 continue
 475             format_urls.add(stream_url)
 476             stream_format = t.get('format') or {}
 477             protocol = stream_format.get('protocol')
 478             if protocol != 'hls' and '/hls' in format_url:
 479                 protocol = 'hls'
 480             ext = None
 481             preset = str_or_none(t.get('preset'))
 482             if preset:
 483                 ext = preset.split('_')[0]
 484             if ext not in KNOWN_EXTENSIONS:
 485                 ext = mimetype2ext(stream_format.get('mime_type'))
 486             add_format({
 487                 'url': stream_url,
 488                 'ext': ext,
 489             }, 'http' if protocol == 'progressive' else protocol,
 490                 t.get('snipped') or '/preview/' in format_url)
 491
 492         for f in formats:
 493             f['vcodec'] = 'none'
 494
 495         if not formats and info.get('policy') == 'BLOCK':
 496             self.raise_geo_restricted()
 497         self._sort_formats(formats)
 498
 499         user = info.get('user') or {}
 500
 501         thumbnails = []
 502         artwork_url = info.get('artwork_url')
 503         thumbnail = artwork_url or user.get('avatar_url')
 504         if isinstance(thumbnail, compat_str):
 505             if re.search(self._IMAGE_REPL_RE, thumbnail):
 506                 for image_id, size in self._ARTWORK_MAP.items():
 507                     i = {
 508                         'id': image_id,
 509                         'url': re.sub(self._IMAGE_REPL_RE, '-%s.jpg' % image_id, thumbnail),
 510                     }
 511                     if image_id == 'tiny' and not artwork_url:
 512                         size = 18
 513                     elif image_id == 'original':
 514                         i['preference'] = 10
 515                     if size:
 516                         i.update({
 517                             'width': size,
 518                             'height': size,
 519                         })
 520                     thumbnails.append(i)
 521             else:
 522                 thumbnails = [{'url': thumbnail}]
 523
 524         def extract_count(key):
 525             return int_or_none(info.get('%s_count' % key))
 526
 527         return {
 528             'id': track_id,
 529             'uploader': user.get('username'),
 530             'uploader_id': str_or_none(user.get('id')) or user.get('permalink'),
 531             'uploader_url': user.get('permalink_url'),
 532             'timestamp': unified_timestamp(info.get('created_at')),
 533             'title': title,
 534             'description': info.get('description'),
 535             'thumbnails': thumbnails,
 536             'duration': float_or_none(info.get('duration'), 1000),
 537             'webpage_url': info.get('permalink_url'),
 538             'license': info.get('license'),
 539             'view_count': extract_count('playback'),
 540             'like_count': extract_count('favoritings') or extract_count('likes'),
 541             'comment_count': extract_count('comment'),
 542             'repost_count': extract_count('reposts'),
 543             'genre': info.get('genre'),
 544             'formats': formats
 545         }
 546
 547     def _real_extract(self, url):
 548         mobj = re.match(self._VALID_URL, url)
 549
 550         track_id = mobj.group('track_id')
 551
 552         query = {}
 553         if track_id:
 554             info_json_url = self._API_V2_BASE + 'tracks/' + track_id
 555             full_title = track_id
 556             token = mobj.group('secret_token')
 557             if token:
 558                 query['secret_token'] = token
 559         else:
 560             full_title = resolve_title = '%s/%s' % mobj.group('uploader', 'title')
 561             token = mobj.group('token')
 562             if token:
 563                 resolve_title += '/%s' % token
 564             info_json_url = self._resolv_url(self._BASE_URL + resolve_title)
 565
 566         info = self._download_json(
 567             info_json_url, full_title, 'Downloading info JSON', query=query, headers=self._HEADERS)
 568
 569         return self._extract_info_dict(info, full_title, token)
 570
 571
 572 class SoundcloudPlaylistBaseIE(SoundcloudIE):
 573     def _extract_set(self, playlist, token=None):
 574         playlist_id = compat_str(playlist['id'])
 575         tracks = playlist.get('tracks') or []
 576         if not all([t.get('permalink_url') for t in tracks]) and token:
 577             tracks = self._download_json(
 578                 self._API_V2_BASE + 'tracks', playlist_id,
 579                 'Downloading tracks', query={
 580                     'ids': ','.join([compat_str(t['id']) for t in tracks]),
 581                     'playlistId': playlist_id,
 582                     'playlistSecretToken': token,
 583                 }, headers=self._HEADERS)
 584         entries = []
 585         for track in tracks:
 586             track_id = str_or_none(track.get('id'))
 587             url = track.get('permalink_url')
 588             if not url:
 589                 if not track_id:
 590                     continue
 591                 url = self._API_V2_BASE + 'tracks/' + track_id
 592                 if token:
 593                     url += '?secret_token=' + token
 594             entries.append(self.url_result(
 595                 url, SoundcloudIE.ie_key(), track_id))
 596         return self.playlist_result(
 597             entries, playlist_id,
 598             playlist.get('title'),
 599             playlist.get('description'))
 600
 601
 602 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 603     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 604     IE_NAME = 'soundcloud:set'
 605     _TESTS = [{
 606         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 607         'info_dict': {
 608             'id': '2284613',
 609             'title': 'The Royal Concept EP',
 610             'description': 'md5:71d07087c7a449e8941a70a29e34671e',
 611         },
 612         'playlist_mincount': 5,
 613     }, {
 614         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 615         'only_matching': True,
 616     }]
 617
 618     def _real_extract(self, url):
 619         mobj = re.match(self._VALID_URL, url)
 620
 621         full_title = '%s/sets/%s' % mobj.group('uploader', 'slug_title')
 622         token = mobj.group('token')
 623         if token:
 624             full_title += '/' + token
 625
 626         info = self._download_json(self._resolv_url(
 627             self._BASE_URL + full_title), full_title, headers=self._HEADERS)
 628
 629         if 'errors' in info:
 630             msgs = (compat_str(err['error_message']) for err in info['errors'])
 631             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 632
 633         return self._extract_set(info, token)
 634
 635
 636 class SoundcloudPagedPlaylistBaseIE(SoundcloudIE):
 637     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 638         COMMON_QUERY = {
 639             'limit': 80000,
 640             'linked_partitioning': '1',
 641         }
 642
 643         query = COMMON_QUERY.copy()
 644         query['offset'] = 0
 645
 646         next_href = base_url
 647
 648         entries = []
 649         for i in itertools.count():
 650             response = self._download_json(
 651                 next_href, playlist_id,
 652                 'Downloading track page %s' % (i + 1), query=query, headers=self._HEADERS)
 653
 654             collection = response['collection']
 655
 656             if not isinstance(collection, list):
 657                 collection = []
 658
 659             # Empty collection may be returned, in this case we proceed
 660             # straight to next_href
 661
 662             def resolve_entry(candidates):
 663                 for cand in candidates:
 664                     if not isinstance(cand, dict):
 665                         continue
 666                     permalink_url = url_or_none(cand.get('permalink_url'))
 667                     if not permalink_url:
 668                         continue
 669                     return self.url_result(
 670                         permalink_url,
 671                         SoundcloudIE.ie_key() if SoundcloudIE.suitable(permalink_url) else None,
 672                         str_or_none(cand.get('id')), cand.get('title'))
 673
 674             for e in collection:
 675                 entry = resolve_entry((e, e.get('track'), e.get('playlist')))
 676                 if entry:
 677                     entries.append(entry)
 678
 679             next_href = response.get('next_href')
 680             if not next_href:
 681                 break
 682
 683             next_href = response['next_href']
 684             parsed_next_href = compat_urlparse.urlparse(next_href)
 685             query = compat_urlparse.parse_qs(parsed_next_href.query)
 686             query.update(COMMON_QUERY)
 687
 688         return {
 689             '_type': 'playlist',
 690             'id': playlist_id,
 691             'title': playlist_title,
 692             'entries': entries,
 693         }
 694
 695
 696 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 697     _VALID_URL = r'''(?x)
 698                         https?://
 699                             (?:(?:www|m)\.)?soundcloud\.com/
 700                             (?P<user>[^/]+)
 701                             (?:/
 702                                 (?P<rsrc>tracks|albums|sets|reposts|likes|spotlight)
 703                             )?
 704                             /?(?:[?#].*)?$
 705                     '''
 706     IE_NAME = 'soundcloud:user'
 707     _TESTS = [{
 708         'url': 'https://soundcloud.com/soft-cell-official',
 709         'info_dict': {
 710             'id': '207965082',
 711             'title': 'Soft Cell (All)',
 712         },
 713         'playlist_mincount': 28,
 714     }, {
 715         'url': 'https://soundcloud.com/soft-cell-official/tracks',
 716         'info_dict': {
 717             'id': '207965082',
 718             'title': 'Soft Cell (Tracks)',
 719         },
 720         'playlist_mincount': 27,
 721     }, {
 722         'url': 'https://soundcloud.com/soft-cell-official/albums',
 723         'info_dict': {
 724             'id': '207965082',
 725             'title': 'Soft Cell (Albums)',
 726         },
 727         'playlist_mincount': 1,
 728     }, {
 729         'url': 'https://soundcloud.com/jcv246/sets',
 730         'info_dict': {
 731             'id': '12982173',
 732             'title': 'Jordi / cv (Sets)',
 733         },
 734         'playlist_mincount': 2,
 735     }, {
 736         'url': 'https://soundcloud.com/jcv246/reposts',
 737         'info_dict': {
 738             'id': '12982173',
 739             'title': 'Jordi / cv (Reposts)',
 740         },
 741         'playlist_mincount': 6,
 742     }, {
 743         'url': 'https://soundcloud.com/clalberg/likes',
 744         'info_dict': {
 745             'id': '11817582',
 746             'title': 'clalberg (Likes)',
 747         },
 748         'playlist_mincount': 5,
 749     }, {
 750         'url': 'https://soundcloud.com/grynpyret/spotlight',
 751         'info_dict': {
 752             'id': '7098329',
 753             'title': 'Grynpyret (Spotlight)',
 754         },
 755         'playlist_mincount': 1,
 756     }]
 757
 758     _BASE_URL_MAP = {
 759         'all': 'stream/users/%s',
 760         'tracks': 'users/%s/tracks',
 761         'albums': 'users/%s/albums',
 762         'sets': 'users/%s/playlists',
 763         'reposts': 'stream/users/%s/reposts',
 764         'likes': 'users/%s/likes',
 765         'spotlight': 'users/%s/spotlight',
 766     }
 767
 768     def _real_extract(self, url):
 769         mobj = re.match(self._VALID_URL, url)
 770         uploader = mobj.group('user')
 771
 772         user = self._download_json(
 773             self._resolv_url(self._BASE_URL + uploader),
 774             uploader, 'Downloading user info', headers=self._HEADERS)
 775
 776         resource = mobj.group('rsrc') or 'all'
 777
 778         return self._extract_playlist(
 779             self._API_V2_BASE + self._BASE_URL_MAP[resource] % user['id'],
 780             str_or_none(user.get('id')),
 781             '%s (%s)' % (user['username'], resource.capitalize()))
 782
 783
 784 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 785     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 786     IE_NAME = 'soundcloud:trackstation'
 787     _TESTS = [{
 788         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 789         'info_dict': {
 790             'id': '286017854',
 791             'title': 'Track station: your text',
 792         },
 793         'playlist_mincount': 47,
 794     }]
 795
 796     def _real_extract(self, url):
 797         track_name = self._match_id(url)
 798
 799         track = self._download_json(self._resolv_url(url), track_name, headers=self._HEADERS)
 800         track_id = self._search_regex(
 801             r'soundcloud:track-stations:(\d+)', track['id'], 'track id')
 802
 803         return self._extract_playlist(
 804             self._API_V2_BASE + 'stations/%s/tracks' % track['id'],
 805             track_id, 'Track station: %s' % track['title'])
 806
 807
 808 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 809     _VALID_URL = r'https?://api(?:-v2)?\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 810     IE_NAME = 'soundcloud:playlist'
 811     _TESTS = [{
 812         'url': 'https://api.soundcloud.com/playlists/4110309',
 813         'info_dict': {
 814             'id': '4110309',
 815             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 816             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 817         },
 818         'playlist_count': 6,
 819     }]
 820
 821     def _real_extract(self, url):
 822         mobj = re.match(self._VALID_URL, url)
 823         playlist_id = mobj.group('id')
 824
 825         query = {}
 826         token = mobj.group('token')
 827         if token:
 828             query['secret_token'] = token
 829
 830         data = self._download_json(
 831             self._API_V2_BASE + 'playlists/' + playlist_id,
 832             playlist_id, 'Downloading playlist', query=query, headers=self._HEADERS)
 833
 834         return self._extract_set(data, token)
 835
 836
 837 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 838     IE_NAME = 'soundcloud:search'
 839     IE_DESC = 'Soundcloud search'
 840     _MAX_RESULTS = float('inf')
 841     _TESTS = [{
 842         'url': 'scsearch15:post-avant jazzcore',
 843         'info_dict': {
 844             'title': 'post-avant jazzcore',
 845         },
 846         'playlist_count': 15,
 847     }]
 848
 849     _SEARCH_KEY = 'scsearch'
 850     _MAX_RESULTS_PER_PAGE = 200
 851     _DEFAULT_RESULTS_PER_PAGE = 50
 852
 853     def _get_collection(self, endpoint, collection_id, **query):
 854         limit = min(
 855             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 856             self._MAX_RESULTS_PER_PAGE)
 857         query.update({
 858             'limit': limit,
 859             'linked_partitioning': 1,
 860             'offset': 0,
 861         })
 862         next_url = update_url_query(self._API_V2_BASE + endpoint, query)
 863
 864         collected_results = 0
 865
 866         for i in itertools.count(1):
 867             response = self._download_json(
 868                 next_url, collection_id, 'Downloading page {0}'.format(i),
 869                 'Unable to download API page', headers=self._HEADERS)
 870
 871             collection = response.get('collection', [])
 872             if not collection:
 873                 break
 874
 875             collection = list(filter(bool, collection))
 876             collected_results += len(collection)
 877
 878             for item in collection:
 879                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 880
 881             if not collection or collected_results >= limit:
 882                 break
 883
 884             next_url = response.get('next_href')
 885             if not next_url:
 886                 break
 887
 888     def _get_n_results(self, query, n):
 889         tracks = self._get_collection('search/tracks', query, limit=n, q=query)
 890         return self.playlist_result(tracks, playlist_title=query)