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

   1 # encoding: 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                             (?P<uploader>[\w\d-]+)/
  35                             (?!(?:tracks|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
  36                             (?P<title>[\w\d-]+)/?
  37                             (?P<token>[^?]+?)?(?:[?].*)?$)
  38                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
  39                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
  40                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
  41                     )
  42                     '''
  43     IE_NAME = 'soundcloud'
  44     _TESTS = [
  45         {
  46             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  47             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  48             'info_dict': {
  49                 'id': '62986583',
  50                 'ext': 'mp3',
  51                 'upload_date': '20121011',
  52                 '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',
  53                 'uploader': 'E.T. ExTerrestrial Music',
  54                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
  55                 'duration': 143,
  56             }
  57         },
  58         # not streamable song
  59         {
  60             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  61             'info_dict': {
  62                 'id': '47127627',
  63                 'ext': 'mp3',
  64                 'title': 'Goldrushed',
  65                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  66                 'uploader': 'The Royal Concept',
  67                 'upload_date': '20120521',
  68                 'duration': 227,
  69             },
  70             'params': {
  71                 # rtmp
  72                 'skip_download': True,
  73             },
  74         },
  75         # private link
  76         {
  77             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  78             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  79             'info_dict': {
  80                 'id': '123998367',
  81                 'ext': 'mp3',
  82                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  83                 'uploader': 'jaimeMF',
  84                 'description': 'test chars:  \"\'/\\ä↭',
  85                 'upload_date': '20131209',
  86                 'duration': 9,
  87             },
  88         },
  89         # private link (alt format)
  90         {
  91             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
  92             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  93             'info_dict': {
  94                 'id': '123998367',
  95                 'ext': 'mp3',
  96                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  97                 'uploader': 'jaimeMF',
  98                 'description': 'test chars:  \"\'/\\ä↭',
  99                 'upload_date': '20131209',
 100                 'duration': 9,
 101             },
 102         },
 103         # downloadable song
 104         {
 105             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
 106             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 107             'info_dict': {
 108                 'id': '128590877',
 109                 'ext': 'mp3',
 110                 'title': 'Bus Brakes',
 111                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 112                 'uploader': 'oddsamples',
 113                 'upload_date': '20140109',
 114                 'duration': 17,
 115             },
 116         },
 117     ]
 118
 119     _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea'
 120     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 121
 122     @staticmethod
 123     def _extract_urls(webpage):
 124         return [m.group('url') for m in re.finditer(
 125             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
 126             webpage)]
 127
 128     def report_resolve(self, video_id):
 129         """Report information extraction."""
 130         self.to_screen('%s: Resolving id' % video_id)
 131
 132     @classmethod
 133     def _resolv_url(cls, url):
 134         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 135
 136     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 137         track_id = compat_str(info['id'])
 138         name = full_title or track_id
 139         if quiet:
 140             self.report_extraction(name)
 141
 142         thumbnail = info['artwork_url']
 143         if thumbnail is not None:
 144             thumbnail = thumbnail.replace('-large', '-t500x500')
 145         ext = 'mp3'
 146         result = {
 147             'id': track_id,
 148             'uploader': info['user']['username'],
 149             'upload_date': unified_strdate(info['created_at']),
 150             'title': info['title'],
 151             'description': info['description'],
 152             'thumbnail': thumbnail,
 153             'duration': int_or_none(info.get('duration'), 1000),
 154             'webpage_url': info.get('permalink_url'),
 155         }
 156         formats = []
 157         if info.get('downloadable', False):
 158             # We can build a direct link to the song
 159             format_url = (
 160                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 161                     track_id, self._CLIENT_ID))
 162             formats.append({
 163                 'format_id': 'download',
 164                 'ext': info.get('original_format', 'mp3'),
 165                 'url': format_url,
 166                 'vcodec': 'none',
 167                 'preference': 10,
 168             })
 169
 170         # We have to retrieve the url
 171         streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
 172                        'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
 173         format_dict = self._download_json(
 174             streams_url,
 175             track_id, 'Downloading track url')
 176
 177         for key, stream_url in format_dict.items():
 178             if key.startswith('http'):
 179                 formats.append({
 180                     'format_id': key,
 181                     'ext': ext,
 182                     'url': stream_url,
 183                     'vcodec': 'none',
 184                 })
 185             elif key.startswith('rtmp'):
 186                 # The url doesn't have an rtmp app, we have to extract the playpath
 187                 url, path = stream_url.split('mp3:', 1)
 188                 formats.append({
 189                     'format_id': key,
 190                     'url': url,
 191                     'play_path': 'mp3:' + path,
 192                     'ext': 'flv',
 193                     'vcodec': 'none',
 194                 })
 195
 196             if not formats:
 197                 # We fallback to the stream_url in the original info, this
 198                 # cannot be always used, sometimes it can give an HTTP 404 error
 199                 formats.append({
 200                     'format_id': 'fallback',
 201                     'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 202                     'ext': ext,
 203                     'vcodec': 'none',
 204                 })
 205
 206             for f in formats:
 207                 if f['format_id'].startswith('http'):
 208                     f['protocol'] = 'http'
 209                 if f['format_id'].startswith('rtmp'):
 210                     f['protocol'] = 'rtmp'
 211
 212         self._check_formats(formats, track_id)
 213         self._sort_formats(formats)
 214         result['formats'] = formats
 215
 216         return result
 217
 218     def _real_extract(self, url):
 219         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 220         if mobj is None:
 221             raise ExtractorError('Invalid URL: %s' % url)
 222
 223         track_id = mobj.group('track_id')
 224         token = None
 225         if track_id is not None:
 226             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 227             full_title = track_id
 228             token = mobj.group('secret_token')
 229             if token:
 230                 info_json_url += '&secret_token=' + token
 231         elif mobj.group('player'):
 232             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 233             real_url = query['url'][0]
 234             # If the token is in the query of the original url we have to
 235             # manually add it
 236             if 'secret_token' in query:
 237                 real_url += '?secret_token=' + query['secret_token'][0]
 238             return self.url_result(real_url)
 239         else:
 240             # extract uploader (which is in the url)
 241             uploader = mobj.group('uploader')
 242             # extract simple title (uploader + slug of song title)
 243             slug_title = mobj.group('title')
 244             token = mobj.group('token')
 245             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 246             if token:
 247                 resolve_title += '/%s' % token
 248
 249             self.report_resolve(full_title)
 250
 251             url = 'http://soundcloud.com/%s' % resolve_title
 252             info_json_url = self._resolv_url(url)
 253         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 254
 255         return self._extract_info_dict(info, full_title, secret_token=token)
 256
 257
 258 class SoundcloudSetIE(SoundcloudIE):
 259     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 260     IE_NAME = 'soundcloud:set'
 261     _TESTS = [{
 262         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 263         'info_dict': {
 264             'id': '2284613',
 265             'title': 'The Royal Concept EP',
 266         },
 267         'playlist_mincount': 6,
 268     }, {
 269         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
 270         'only_matching': True,
 271     }]
 272
 273     def _real_extract(self, url):
 274         mobj = re.match(self._VALID_URL, url)
 275
 276         # extract uploader (which is in the url)
 277         uploader = mobj.group('uploader')
 278         # extract simple title (uploader + slug of song title)
 279         slug_title = mobj.group('slug_title')
 280         full_title = '%s/sets/%s' % (uploader, slug_title)
 281         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 282
 283         token = mobj.group('token')
 284         if token:
 285             full_title += '/' + token
 286             url += '/' + token
 287
 288         self.report_resolve(full_title)
 289
 290         resolv_url = self._resolv_url(url)
 291         info = self._download_json(resolv_url, full_title)
 292
 293         if 'errors' in info:
 294             msgs = (compat_str(err['error_message']) for err in info['errors'])
 295             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 296
 297         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in info['tracks']]
 298
 299         return {
 300             '_type': 'playlist',
 301             'entries': entries,
 302             'id': '%s' % info['id'],
 303             'title': info['title'],
 304         }
 305
 306
 307 class SoundcloudUserIE(SoundcloudIE):
 308     _VALID_URL = r'''(?x)
 309                         https?://
 310                             (?:(?:www|m)\.)?soundcloud\.com/
 311                             (?P<user>[^/]+)
 312                             (?:/
 313                                 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
 314                             )?
 315                             /?(?:[?#].*)?$
 316                     '''
 317     IE_NAME = 'soundcloud:user'
 318     _TESTS = [{
 319         'url': 'https://soundcloud.com/the-akashic-chronicler',
 320         'info_dict': {
 321             'id': '114582580',
 322             'title': 'The Akashic Chronicler (All)',
 323         },
 324         'playlist_mincount': 111,
 325     }, {
 326         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
 327         'info_dict': {
 328             'id': '114582580',
 329             'title': 'The Akashic Chronicler (Tracks)',
 330         },
 331         'playlist_mincount': 50,
 332     }, {
 333         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
 334         'info_dict': {
 335             'id': '114582580',
 336             'title': 'The Akashic Chronicler (Playlists)',
 337         },
 338         'playlist_mincount': 3,
 339     }, {
 340         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
 341         'info_dict': {
 342             'id': '114582580',
 343             'title': 'The Akashic Chronicler (Reposts)',
 344         },
 345         'playlist_mincount': 7,
 346     }, {
 347         'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
 348         'info_dict': {
 349             'id': '114582580',
 350             'title': 'The Akashic Chronicler (Likes)',
 351         },
 352         'playlist_mincount': 321,
 353     }, {
 354         'url': 'https://soundcloud.com/grynpyret/spotlight',
 355         'info_dict': {
 356             'id': '7098329',
 357             'title': 'Grynpyret (Spotlight)',
 358         },
 359         'playlist_mincount': 1,
 360     }]
 361
 362     _API_BASE = 'https://api.soundcloud.com'
 363     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 364
 365     _BASE_URL_MAP = {
 366         'all': '%s/profile/soundcloud:users:%%s' % _API_V2_BASE,
 367         'tracks': '%s/users/%%s/tracks' % _API_BASE,
 368         'sets': '%s/users/%%s/playlists' % _API_V2_BASE,
 369         'reposts': '%s/profile/soundcloud:users:%%s/reposts' % _API_V2_BASE,
 370         'likes': '%s/users/%%s/likes' % _API_V2_BASE,
 371         'spotlight': '%s/users/%%s/spotlight' % _API_V2_BASE,
 372     }
 373
 374     _TITLE_MAP = {
 375         'all': 'All',
 376         'tracks': 'Tracks',
 377         'sets': 'Playlists',
 378         'reposts': 'Reposts',
 379         'likes': 'Likes',
 380         'spotlight': 'Spotlight',
 381     }
 382
 383     def _real_extract(self, url):
 384         mobj = re.match(self._VALID_URL, url)
 385         uploader = mobj.group('user')
 386
 387         url = 'http://soundcloud.com/%s/' % uploader
 388         resolv_url = self._resolv_url(url)
 389         user = self._download_json(
 390             resolv_url, uploader, 'Downloading user info')
 391
 392         resource = mobj.group('rsrc') or 'all'
 393         base_url = self._BASE_URL_MAP[resource] % user['id']
 394
 395         COMMON_QUERY = {
 396             'limit': 50,
 397             'client_id': self._CLIENT_ID,
 398             'linked_partitioning': '1',
 399         }
 400
 401         query = COMMON_QUERY.copy()
 402         query['offset'] = 0
 403
 404         next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
 405
 406         entries = []
 407         for i in itertools.count():
 408             response = self._download_json(
 409                 next_href, uploader, 'Downloading track page %s' % (i + 1))
 410
 411             collection = response['collection']
 412             if not collection:
 413                 break
 414
 415             def resolve_permalink_url(candidates):
 416                 for cand in candidates:
 417                     if isinstance(cand, dict):
 418                         permalink_url = cand.get('permalink_url')
 419                         if permalink_url and permalink_url.startswith('http'):
 420                             return permalink_url
 421
 422             for e in collection:
 423                 permalink_url = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
 424                 if permalink_url:
 425                     entries.append(self.url_result(permalink_url))
 426
 427             next_href = response.get('next_href')
 428             if not next_href:
 429                 break
 430
 431             parsed_next_href = compat_urlparse.urlparse(response['next_href'])
 432             qs = compat_urlparse.parse_qs(parsed_next_href.query)
 433             qs.update(COMMON_QUERY)
 434             next_href = compat_urlparse.urlunparse(
 435                 parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
 436
 437         return {
 438             '_type': 'playlist',
 439             'id': compat_str(user['id']),
 440             'title': '%s (%s)' % (user['username'], self._TITLE_MAP[resource]),
 441             'entries': entries,
 442         }
 443
 444
 445 class SoundcloudPlaylistIE(SoundcloudIE):
 446     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 447     IE_NAME = 'soundcloud:playlist'
 448     _TESTS = [{
 449         'url': 'http://api.soundcloud.com/playlists/4110309',
 450         'info_dict': {
 451             'id': '4110309',
 452             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 453             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 454         },
 455         'playlist_count': 6,
 456     }]
 457
 458     def _real_extract(self, url):
 459         mobj = re.match(self._VALID_URL, url)
 460         playlist_id = mobj.group('id')
 461         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 462
 463         data_dict = {
 464             'client_id': self._CLIENT_ID,
 465         }
 466         token = mobj.group('token')
 467
 468         if token:
 469             data_dict['secret_token'] = token
 470
 471         data = compat_urllib_parse_urlencode(data_dict)
 472         data = self._download_json(
 473             base_url + data, playlist_id, 'Downloading playlist')
 474
 475         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in data['tracks']]
 476
 477         return {
 478             '_type': 'playlist',
 479             'id': playlist_id,
 480             'title': data.get('title'),
 481             'description': data.get('description'),
 482             'entries': entries,
 483         }
 484
 485
 486 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 487     IE_NAME = 'soundcloud:search'
 488     IE_DESC = 'Soundcloud search'
 489     _MAX_RESULTS = float('inf')
 490     _TESTS = [{
 491         'url': 'scsearch15:post-avant jazzcore',
 492         'info_dict': {
 493             'title': 'post-avant jazzcore',
 494         },
 495         'playlist_count': 15,
 496     }]
 497
 498     _SEARCH_KEY = 'scsearch'
 499     _MAX_RESULTS_PER_PAGE = 200
 500     _DEFAULT_RESULTS_PER_PAGE = 50
 501     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 502
 503     def _get_collection(self, endpoint, collection_id, **query):
 504         limit = min(
 505             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 506             self._MAX_RESULTS_PER_PAGE)
 507         query['limit'] = limit
 508         query['client_id'] = self._CLIENT_ID
 509         query['linked_partitioning'] = '1'
 510         query['offset'] = 0
 511         data = compat_urllib_parse_urlencode(query)
 512         next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
 513
 514         collected_results = 0
 515
 516         for i in itertools.count(1):
 517             response = self._download_json(
 518                 next_url, collection_id, 'Downloading page {0}'.format(i),
 519                 'Unable to download API page')
 520
 521             collection = response.get('collection', [])
 522             if not collection:
 523                 break
 524
 525             collection = list(filter(bool, collection))
 526             collected_results += len(collection)
 527
 528             for item in collection:
 529                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 530
 531             if not collection or collected_results >= limit:
 532                 break
 533
 534             next_url = response.get('next_href')
 535             if not next_url:
 536                 break
 537
 538     def _get_n_results(self, query, n):
 539         tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
 540         return self.playlist_result(tracks, playlist_title=query)