_ Git - youtube-dl/blob - youtube_dl/extractor/soundcloud.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5 import itertools
   6
   7 from .common import (
   8     InfoExtractor,
   9     SearchInfoExtractor
  10 )
  11 from ..compat import (
  12     compat_str,
  13     compat_urlparse,
  14     compat_urllib_parse_urlencode,
  15 )
  16 from ..utils import (
  17     ExtractorError,
  18     int_or_none,
  19     unified_strdate,
  20 )
  21
  22
  23 class SoundcloudIE(InfoExtractor):
  24     """Information extractor for soundcloud.com
  25        To access the media, the uid of the song and a stream token
  26        must be extracted from the page source and the script must make
  27        a request to media.soundcloud.com/crossdomain.xml. Then
  28        the media can be grabbed by requesting from an url composed
  29        of the stream token and uid
  30      """
  31
  32     _VALID_URL = r'''(?x)^(?:https?://)?
  33                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
  34                             (?!stations/track)
  35                             (?P<uploader>[\w\d-]+)/
  36                             (?!(?:tracks|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
  37                             (?P<title>[\w\d-]+)/?
  38                             (?P<token>[^?]+?)?(?:[?].*)?$)
  39                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
  40                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
  41                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
  42                     )
  43                     '''
  44     IE_NAME = 'soundcloud'
  45     _TESTS = [
  46         {
  47             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  48             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  49             'info_dict': {
  50                 'id': '62986583',
  51                 'ext': 'mp3',
  52                 'upload_date': '20121011',
  53                 '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',
  54                 'uploader': 'E.T. ExTerrestrial Music',
  55                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
  56                 'duration': 143,
  57                 'license': 'all-rights-reserved',
  58             }
  59         },
  60         # not streamable song
  61         {
  62             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  63             'info_dict': {
  64                 'id': '47127627',
  65                 'ext': 'mp3',
  66                 'title': 'Goldrushed',
  67                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  68                 'uploader': 'The Royal Concept',
  69                 'upload_date': '20120521',
  70                 'duration': 227,
  71                 'license': 'all-rights-reserved',
  72             },
  73             'params': {
  74                 # rtmp
  75                 'skip_download': True,
  76             },
  77         },
  78         # private link
  79         {
  80             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  81             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  82             'info_dict': {
  83                 'id': '123998367',
  84                 'ext': 'mp3',
  85                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  86                 'uploader': 'jaimeMF',
  87                 'description': 'test chars:  \"\'/\\ä↭',
  88                 'upload_date': '20131209',
  89                 'duration': 9,
  90                 'license': 'all-rights-reserved',
  91             },
  92         },
  93         # private link (alt format)
  94         {
  95             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
  96             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  97             'info_dict': {
  98                 'id': '123998367',
  99                 'ext': 'mp3',
 100                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
 101                 'uploader': 'jaimeMF',
 102                 'description': 'test chars:  \"\'/\\ä↭',
 103                 'upload_date': '20131209',
 104                 'duration': 9,
 105                 'license': 'all-rights-reserved',
 106             },
 107         },
 108         # downloadable song
 109         {
 110             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
 111             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 112             'info_dict': {
 113                 'id': '128590877',
 114                 'ext': 'mp3',
 115                 'title': 'Bus Brakes',
 116                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 117                 'uploader': 'oddsamples',
 118                 'upload_date': '20140109',
 119                 'duration': 17,
 120                 'license': 'cc-by-sa',
 121             },
 122         },
 123     ]
 124
 125     _CLIENT_ID = 'JlZIsxg2hY5WnBgtn3jfS0UYCl0K8DOg'
 126     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 127
 128     @staticmethod
 129     def _extract_urls(webpage):
 130         return [m.group('url') for m in re.finditer(
 131             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
 132             webpage)]
 133
 134     def report_resolve(self, video_id):
 135         """Report information extraction."""
 136         self.to_screen('%s: Resolving id' % video_id)
 137
 138     @classmethod
 139     def _resolv_url(cls, url):
 140         return 'https://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 141
 142     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 143         track_id = compat_str(info['id'])
 144         name = full_title or track_id
 145         if quiet:
 146             self.report_extraction(name)
 147         thumbnail = info.get('artwork_url')
 148         if isinstance(thumbnail, compat_str):
 149             thumbnail = thumbnail.replace('-large', '-t500x500')
 150         ext = 'mp3'
 151         result = {
 152             'id': track_id,
 153             'uploader': info.get('user', {}).get('username'),
 154             'upload_date': unified_strdate(info.get('created_at')),
 155             'title': info['title'],
 156             'description': info.get('description'),
 157             'thumbnail': thumbnail,
 158             'duration': int_or_none(info.get('duration'), 1000),
 159             'webpage_url': info.get('permalink_url'),
 160             'license': info.get('license'),
 161         }
 162         formats = []
 163         if info.get('downloadable', False):
 164             # We can build a direct link to the song
 165             format_url = (
 166                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 167                     track_id, self._CLIENT_ID))
 168             formats.append({
 169                 'format_id': 'download',
 170                 'ext': info.get('original_format', 'mp3'),
 171                 'url': format_url,
 172                 'vcodec': 'none',
 173                 'preference': 10,
 174             })
 175
 176         # We have to retrieve the url
 177         format_dict = self._download_json(
 178             'https://api.soundcloud.com/i1/tracks/%s/streams' % track_id,
 179             track_id, 'Downloading track url', query={
 180                 'client_id': self._CLIENT_ID,
 181                 'secret_token': secret_token,
 182             })
 183
 184         for key, stream_url in format_dict.items():
 185             abr = int_or_none(self._search_regex(
 186                 r'_(\d+)_url', key, 'audio bitrate', default=None))
 187             if key.startswith('http'):
 188                 stream_formats = [{
 189                     'format_id': key,
 190                     'ext': ext,
 191                     'url': stream_url,
 192                 }]
 193             elif key.startswith('rtmp'):
 194                 # The url doesn't have an rtmp app, we have to extract the playpath
 195                 url, path = stream_url.split('mp3:', 1)
 196                 stream_formats = [{
 197                     'format_id': key,
 198                     'url': url,
 199                     'play_path': 'mp3:' + path,
 200                     'ext': 'flv',
 201                 }]
 202             elif key.startswith('hls'):
 203                 stream_formats = self._extract_m3u8_formats(
 204                     stream_url, track_id, 'mp3', entry_protocol='m3u8_native',
 205                     m3u8_id=key, fatal=False)
 206             else:
 207                 continue
 208
 209             for f in stream_formats:
 210                 f['abr'] = abr
 211
 212             formats.extend(stream_formats)
 213
 214         if not formats:
 215             # We fallback to the stream_url in the original info, this
 216             # cannot be always used, sometimes it can give an HTTP 404 error
 217             formats.append({
 218                 'format_id': 'fallback',
 219                 'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 220                 'ext': ext,
 221             })
 222
 223         for f in formats:
 224             f['vcodec'] = 'none'
 225
 226         self._check_formats(formats, track_id)
 227         self._sort_formats(formats)
 228         result['formats'] = formats
 229
 230         return result
 231
 232     def _real_extract(self, url):
 233         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 234         if mobj is None:
 235             raise ExtractorError('Invalid URL: %s' % url)
 236
 237         track_id = mobj.group('track_id')
 238
 239         if track_id is not None:
 240             info_json_url = 'https://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 241             full_title = track_id
 242             token = mobj.group('secret_token')
 243             if token:
 244                 info_json_url += '&secret_token=' + token
 245         elif mobj.group('player'):
 246             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 247             real_url = query['url'][0]
 248             # If the token is in the query of the original url we have to
 249             # manually add it
 250             if 'secret_token' in query:
 251                 real_url += '?secret_token=' + query['secret_token'][0]
 252             return self.url_result(real_url)
 253         else:
 254             # extract uploader (which is in the url)
 255             uploader = mobj.group('uploader')
 256             # extract simple title (uploader + slug of song title)
 257             slug_title = mobj.group('title')
 258             token = mobj.group('token')
 259             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 260             if token:
 261                 resolve_title += '/%s' % token
 262
 263             self.report_resolve(full_title)
 264
 265             url = 'https://soundcloud.com/%s' % resolve_title
 266             info_json_url = self._resolv_url(url)
 267         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 268
 269         return self._extract_info_dict(info, full_title, secret_token=token)
 270
 271
 272 class SoundcloudPlaylistBaseIE(SoundcloudIE):
 273     @staticmethod
 274     def _extract_id(e):
 275         return compat_str(e['id']) if e.get('id') else None
 276
 277     def _extract_track_entries(self, tracks):
 278         return [
 279             self.url_result(
 280                 track['permalink_url'], SoundcloudIE.ie_key(),
 281                 video_id=self._extract_id(track))
 282             for track in tracks if track.get('permalink_url')]
 283
 284
 285 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
 286     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 287     IE_NAME = 'soundcloud:set'
 288     _TESTS = [{
 289         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 290         'info_dict': {
 291             'id': '2284613',
 292             'title': 'The Royal Concept EP',
 293         },
 294         'playlist_mincount': 5,
 295     }, {
 296         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 297         'only_matching': True,
 298     }]
 299
 300     def _real_extract(self, url):
 301         mobj = re.match(self._VALID_URL, url)
 302
 303         # extract uploader (which is in the url)
 304         uploader = mobj.group('uploader')
 305         # extract simple title (uploader + slug of song title)
 306         slug_title = mobj.group('slug_title')
 307         full_title = '%s/sets/%s' % (uploader, slug_title)
 308         url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 309
 310         token = mobj.group('token')
 311         if token:
 312             full_title += '/' + token
 313             url += '/' + token
 314
 315         self.report_resolve(full_title)
 316
 317         resolv_url = self._resolv_url(url)
 318         info = self._download_json(resolv_url, full_title)
 319
 320         if 'errors' in info:
 321             msgs = (compat_str(err['error_message']) for err in info['errors'])
 322             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 323
 324         entries = self._extract_track_entries(info['tracks'])
 325
 326         return {
 327             '_type': 'playlist',
 328             'entries': entries,
 329             'id': '%s' % info['id'],
 330             'title': info['title'],
 331         }
 332
 333
 334 class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE):
 335     _API_BASE = 'https://api.soundcloud.com'
 336     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 337
 338     def _extract_playlist(self, base_url, playlist_id, playlist_title):
 339         COMMON_QUERY = {
 340             'limit': 50,
 341             'client_id': self._CLIENT_ID,
 342             'linked_partitioning': '1',
 343         }
 344
 345         query = COMMON_QUERY.copy()
 346         query['offset'] = 0
 347
 348         next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
 349
 350         entries = []
 351         for i in itertools.count():
 352             response = self._download_json(
 353                 next_href, playlist_id, 'Downloading track page %s' % (i + 1))
 354
 355             collection = response['collection']
 356             if not collection:
 357                 break
 358
 359             def resolve_permalink_url(candidates):
 360                 for cand in candidates:
 361                     if isinstance(cand, dict):
 362                         permalink_url = cand.get('permalink_url')
 363                         entry_id = self._extract_id(cand)
 364                         if permalink_url and permalink_url.startswith('http'):
 365                             return permalink_url, entry_id
 366
 367             for e in collection:
 368                 permalink_url, entry_id = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
 369                 if permalink_url:
 370                     entries.append(self.url_result(permalink_url, video_id=entry_id))
 371
 372             next_href = response.get('next_href')
 373             if not next_href:
 374                 break
 375
 376             parsed_next_href = compat_urlparse.urlparse(response['next_href'])
 377             qs = compat_urlparse.parse_qs(parsed_next_href.query)
 378             qs.update(COMMON_QUERY)
 379             next_href = compat_urlparse.urlunparse(
 380                 parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
 381
 382         return {
 383             '_type': 'playlist',
 384             'id': playlist_id,
 385             'title': playlist_title,
 386             'entries': entries,
 387         }
 388
 389
 390 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
 391     _VALID_URL = r'''(?x)
 392                         https?://
 393                             (?:(?:www|m)\.)?soundcloud\.com/
 394                             (?P<user>[^/]+)
 395                             (?:/
 396                                 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
 397                             )?
 398                             /?(?:[?#].*)?$
 399                     '''
 400     IE_NAME = 'soundcloud:user'
 401     _TESTS = [{
 402         'url': 'https://soundcloud.com/the-akashic-chronicler',
 403         'info_dict': {
 404             'id': '114582580',
 405             'title': 'The Akashic Chronicler (All)',
 406         },
 407         'playlist_mincount': 74,
 408     }, {
 409         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
 410         'info_dict': {
 411             'id': '114582580',
 412             'title': 'The Akashic Chronicler (Tracks)',
 413         },
 414         'playlist_mincount': 37,
 415     }, {
 416         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
 417         'info_dict': {
 418             'id': '114582580',
 419             'title': 'The Akashic Chronicler (Playlists)',
 420         },
 421         'playlist_mincount': 2,
 422     }, {
 423         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
 424         'info_dict': {
 425             'id': '114582580',
 426             'title': 'The Akashic Chronicler (Reposts)',
 427         },
 428         'playlist_mincount': 7,
 429     }, {
 430         'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
 431         'info_dict': {
 432             'id': '114582580',
 433             'title': 'The Akashic Chronicler (Likes)',
 434         },
 435         'playlist_mincount': 321,
 436     }, {
 437         'url': 'https://soundcloud.com/grynpyret/spotlight',
 438         'info_dict': {
 439             'id': '7098329',
 440             'title': 'Grynpyret (Spotlight)',
 441         },
 442         'playlist_mincount': 1,
 443     }]
 444
 445     _BASE_URL_MAP = {
 446         'all': '%s/profile/soundcloud:users:%%s' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
 447         'tracks': '%s/users/%%s/tracks' % SoundcloudPagedPlaylistBaseIE._API_BASE,
 448         'sets': '%s/users/%%s/playlists' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
 449         'reposts': '%s/profile/soundcloud:users:%%s/reposts' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
 450         'likes': '%s/users/%%s/likes' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
 451         'spotlight': '%s/users/%%s/spotlight' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
 452     }
 453
 454     _TITLE_MAP = {
 455         'all': 'All',
 456         'tracks': 'Tracks',
 457         'sets': 'Playlists',
 458         'reposts': 'Reposts',
 459         'likes': 'Likes',
 460         'spotlight': 'Spotlight',
 461     }
 462
 463     def _real_extract(self, url):
 464         mobj = re.match(self._VALID_URL, url)
 465         uploader = mobj.group('user')
 466
 467         url = 'https://soundcloud.com/%s/' % uploader
 468         resolv_url = self._resolv_url(url)
 469         user = self._download_json(
 470             resolv_url, uploader, 'Downloading user info')
 471
 472         resource = mobj.group('rsrc') or 'all'
 473
 474         return self._extract_playlist(
 475             self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
 476             '%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
 477
 478
 479 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
 480     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
 481     IE_NAME = 'soundcloud:trackstation'
 482     _TESTS = [{
 483         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
 484         'info_dict': {
 485             'id': '286017854',
 486             'title': 'Track station: your-text',
 487         },
 488         'playlist_mincount': 47,
 489     }]
 490
 491     def _real_extract(self, url):
 492         track_name = self._match_id(url)
 493
 494         webpage = self._download_webpage(url, track_name)
 495
 496         track_id = self._search_regex(
 497             r'soundcloud:track-stations:(\d+)', webpage, 'track id')
 498
 499         return self._extract_playlist(
 500             '%s/stations/soundcloud:track-stations:%s/tracks'
 501             % (self._API_V2_BASE, track_id),
 502             track_id, 'Track station: %s' % track_name)
 503
 504
 505 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
 506     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 507     IE_NAME = 'soundcloud:playlist'
 508     _TESTS = [{
 509         'url': 'https://api.soundcloud.com/playlists/4110309',
 510         'info_dict': {
 511             'id': '4110309',
 512             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 513             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 514         },
 515         'playlist_count': 6,
 516     }]
 517
 518     def _real_extract(self, url):
 519         mobj = re.match(self._VALID_URL, url)
 520         playlist_id = mobj.group('id')
 521         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 522
 523         data_dict = {
 524             'client_id': self._CLIENT_ID,
 525         }
 526         token = mobj.group('token')
 527
 528         if token:
 529             data_dict['secret_token'] = token
 530
 531         data = compat_urllib_parse_urlencode(data_dict)
 532         data = self._download_json(
 533             base_url + data, playlist_id, 'Downloading playlist')
 534
 535         entries = self._extract_track_entries(data['tracks'])
 536
 537         return {
 538             '_type': 'playlist',
 539             'id': playlist_id,
 540             'title': data.get('title'),
 541             'description': data.get('description'),
 542             'entries': entries,
 543         }
 544
 545
 546 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 547     IE_NAME = 'soundcloud:search'
 548     IE_DESC = 'Soundcloud search'
 549     _MAX_RESULTS = float('inf')
 550     _TESTS = [{
 551         'url': 'scsearch15:post-avant jazzcore',
 552         'info_dict': {
 553             'title': 'post-avant jazzcore',
 554         },
 555         'playlist_count': 15,
 556     }]
 557
 558     _SEARCH_KEY = 'scsearch'
 559     _MAX_RESULTS_PER_PAGE = 200
 560     _DEFAULT_RESULTS_PER_PAGE = 50
 561     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 562
 563     def _get_collection(self, endpoint, collection_id, **query):
 564         limit = min(
 565             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 566             self._MAX_RESULTS_PER_PAGE)
 567         query['limit'] = limit
 568         query['client_id'] = self._CLIENT_ID
 569         query['linked_partitioning'] = '1'
 570         query['offset'] = 0
 571         data = compat_urllib_parse_urlencode(query)
 572         next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
 573
 574         collected_results = 0
 575
 576         for i in itertools.count(1):
 577             response = self._download_json(
 578                 next_url, collection_id, 'Downloading page {0}'.format(i),
 579                 'Unable to download API page')
 580
 581             collection = response.get('collection', [])
 582             if not collection:
 583                 break
 584
 585             collection = list(filter(bool, collection))
 586             collected_results += len(collection)
 587
 588             for item in collection:
 589                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 590
 591             if not collection or collected_results >= limit:
 592                 break
 593
 594             next_url = response.get('next_href')
 595             if not next_url:
 596                 break
 597
 598     def _get_n_results(self, query, n):
 599         tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
 600         return self.playlist_result(tracks, playlist_title=query)