Raphaël G. Git Repositories - youtubedl/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,
  15 )
  16 from ..utils import (
  17     encode_dict,
  18     ExtractorError,
  19     int_or_none,
  20     unified_strdate,
  21 )
  22
  23
  24 class SoundcloudIE(InfoExtractor):
  25     """Information extractor for soundcloud.com
  26        To access the media, the uid of the song and a stream token
  27        must be extracted from the page source and the script must make
  28        a request to media.soundcloud.com/crossdomain.xml. Then
  29        the media can be grabbed by requesting from an url composed
  30        of the stream token and uid
  31      """
  32
  33     _VALID_URL = r'''(?x)^(?:https?://)?
  34                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
  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             }
  58         },
  59         # not streamable song
  60         {
  61             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  62             'info_dict': {
  63                 'id': '47127627',
  64                 'ext': 'mp3',
  65                 'title': 'Goldrushed',
  66                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  67                 'uploader': 'The Royal Concept',
  68                 'upload_date': '20120521',
  69                 'duration': 227,
  70             },
  71             'params': {
  72                 # rtmp
  73                 'skip_download': True,
  74             },
  75         },
  76         # private link
  77         {
  78             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  79             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  80             'info_dict': {
  81                 'id': '123998367',
  82                 'ext': 'mp3',
  83                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  84                 'uploader': 'jaimeMF',
  85                 'description': 'test chars:  \"\'/\\ä↭',
  86                 'upload_date': '20131209',
  87                 'duration': 9,
  88             },
  89         },
  90         # private link (alt format)
  91         {
  92             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
  93             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  94             'info_dict': {
  95                 'id': '123998367',
  96                 'ext': 'mp3',
  97                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  98                 'uploader': 'jaimeMF',
  99                 'description': 'test chars:  \"\'/\\ä↭',
 100                 'upload_date': '20131209',
 101                 'duration': 9,
 102             },
 103         },
 104         # downloadable song
 105         {
 106             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
 107             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 108             'info_dict': {
 109                 'id': '128590877',
 110                 'ext': 'mp3',
 111                 'title': 'Bus Brakes',
 112                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 113                 'uploader': 'oddsamples',
 114                 'upload_date': '20140109',
 115                 'duration': 17,
 116             },
 117         },
 118     ]
 119
 120     _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea'
 121     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 122
 123     def report_resolve(self, video_id):
 124         """Report information extraction."""
 125         self.to_screen('%s: Resolving id' % video_id)
 126
 127     @classmethod
 128     def _resolv_url(cls, url):
 129         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 130
 131     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 132         track_id = compat_str(info['id'])
 133         name = full_title or track_id
 134         if quiet:
 135             self.report_extraction(name)
 136
 137         thumbnail = info['artwork_url']
 138         if thumbnail is not None:
 139             thumbnail = thumbnail.replace('-large', '-t500x500')
 140         ext = 'mp3'
 141         result = {
 142             'id': track_id,
 143             'uploader': info['user']['username'],
 144             'upload_date': unified_strdate(info['created_at']),
 145             'title': info['title'],
 146             'description': info['description'],
 147             'thumbnail': thumbnail,
 148             'duration': int_or_none(info.get('duration'), 1000),
 149             'webpage_url': info.get('permalink_url'),
 150         }
 151         formats = []
 152         if info.get('downloadable', False):
 153             # We can build a direct link to the song
 154             format_url = (
 155                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 156                     track_id, self._CLIENT_ID))
 157             formats.append({
 158                 'format_id': 'download',
 159                 'ext': info.get('original_format', 'mp3'),
 160                 'url': format_url,
 161                 'vcodec': 'none',
 162                 'preference': 10,
 163             })
 164
 165         # We have to retrieve the url
 166         streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
 167                        'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
 168         format_dict = self._download_json(
 169             streams_url,
 170             track_id, 'Downloading track url')
 171
 172         for key, stream_url in format_dict.items():
 173             if key.startswith('http'):
 174                 formats.append({
 175                     'format_id': key,
 176                     'ext': ext,
 177                     'url': stream_url,
 178                     'vcodec': 'none',
 179                 })
 180             elif key.startswith('rtmp'):
 181                 # The url doesn't have an rtmp app, we have to extract the playpath
 182                 url, path = stream_url.split('mp3:', 1)
 183                 formats.append({
 184                     'format_id': key,
 185                     'url': url,
 186                     'play_path': 'mp3:' + path,
 187                     'ext': 'flv',
 188                     'vcodec': 'none',
 189                 })
 190
 191             if not formats:
 192                 # We fallback to the stream_url in the original info, this
 193                 # cannot be always used, sometimes it can give an HTTP 404 error
 194                 formats.append({
 195                     'format_id': 'fallback',
 196                     'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 197                     'ext': ext,
 198                     'vcodec': 'none',
 199                 })
 200
 201             for f in formats:
 202                 if f['format_id'].startswith('http'):
 203                     f['protocol'] = 'http'
 204                 if f['format_id'].startswith('rtmp'):
 205                     f['protocol'] = 'rtmp'
 206
 207         self._check_formats(formats, track_id)
 208         self._sort_formats(formats)
 209         result['formats'] = formats
 210
 211         return result
 212
 213     def _real_extract(self, url):
 214         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 215         if mobj is None:
 216             raise ExtractorError('Invalid URL: %s' % url)
 217
 218         track_id = mobj.group('track_id')
 219         token = None
 220         if track_id is not None:
 221             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 222             full_title = track_id
 223             token = mobj.group('secret_token')
 224             if token:
 225                 info_json_url += "&secret_token=" + token
 226         elif mobj.group('player'):
 227             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 228             real_url = query['url'][0]
 229             # If the token is in the query of the original url we have to
 230             # manually add it
 231             if 'secret_token' in query:
 232                 real_url += '?secret_token=' + query['secret_token'][0]
 233             return self.url_result(real_url)
 234         else:
 235             # extract uploader (which is in the url)
 236             uploader = mobj.group('uploader')
 237             # extract simple title (uploader + slug of song title)
 238             slug_title = mobj.group('title')
 239             token = mobj.group('token')
 240             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 241             if token:
 242                 resolve_title += '/%s' % token
 243
 244             self.report_resolve(full_title)
 245
 246             url = 'http://soundcloud.com/%s' % resolve_title
 247             info_json_url = self._resolv_url(url)
 248         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 249
 250         return self._extract_info_dict(info, full_title, secret_token=token)
 251
 252
 253 class SoundcloudSetIE(SoundcloudIE):
 254     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 255     IE_NAME = 'soundcloud:set'
 256     _TESTS = [{
 257         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 258         'info_dict': {
 259             'id': '2284613',
 260             'title': 'The Royal Concept EP',
 261         },
 262         'playlist_mincount': 6,
 263     }]
 264
 265     def _real_extract(self, url):
 266         mobj = re.match(self._VALID_URL, url)
 267
 268         # extract uploader (which is in the url)
 269         uploader = mobj.group('uploader')
 270         # extract simple title (uploader + slug of song title)
 271         slug_title = mobj.group('slug_title')
 272         full_title = '%s/sets/%s' % (uploader, slug_title)
 273         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 274
 275         token = mobj.group('token')
 276         if token:
 277             full_title += '/' + token
 278             url += '/' + token
 279
 280         self.report_resolve(full_title)
 281
 282         resolv_url = self._resolv_url(url)
 283         info = self._download_json(resolv_url, full_title)
 284
 285         if 'errors' in info:
 286             msgs = (compat_str(err['error_message']) for err in info['errors'])
 287             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 288
 289         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in info['tracks']]
 290
 291         return {
 292             '_type': 'playlist',
 293             'entries': entries,
 294             'id': '%s' % info['id'],
 295             'title': info['title'],
 296         }
 297
 298
 299 class SoundcloudUserIE(SoundcloudIE):
 300     _VALID_URL = r'''(?x)
 301                         https?://
 302                             (?:(?:www|m)\.)?soundcloud\.com/
 303                             (?P<user>[^/]+)
 304                             (?:/
 305                                 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
 306                             )?
 307                             /?(?:[?#].*)?$
 308                     '''
 309     IE_NAME = 'soundcloud:user'
 310     _TESTS = [{
 311         'url': 'https://soundcloud.com/the-akashic-chronicler',
 312         'info_dict': {
 313             'id': '114582580',
 314             'title': 'The Akashic Chronicler (All)',
 315         },
 316         'playlist_mincount': 111,
 317     }, {
 318         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
 319         'info_dict': {
 320             'id': '114582580',
 321             'title': 'The Akashic Chronicler (Tracks)',
 322         },
 323         'playlist_mincount': 50,
 324     }, {
 325         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
 326         'info_dict': {
 327             'id': '114582580',
 328             'title': 'The Akashic Chronicler (Playlists)',
 329         },
 330         'playlist_mincount': 3,
 331     }, {
 332         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
 333         'info_dict': {
 334             'id': '114582580',
 335             'title': 'The Akashic Chronicler (Reposts)',
 336         },
 337         'playlist_mincount': 7,
 338     }, {
 339         'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
 340         'info_dict': {
 341             'id': '114582580',
 342             'title': 'The Akashic Chronicler (Likes)',
 343         },
 344         'playlist_mincount': 321,
 345     }, {
 346         'url': 'https://soundcloud.com/grynpyret/spotlight',
 347         'info_dict': {
 348             'id': '7098329',
 349             'title': 'Grynpyret (Spotlight)',
 350         },
 351         'playlist_mincount': 1,
 352     }]
 353
 354     _API_BASE = 'https://api.soundcloud.com'
 355     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 356
 357     _BASE_URL_MAP = {
 358         'all': '%s/profile/soundcloud:users:%%s' % _API_V2_BASE,
 359         'tracks': '%s/users/%%s/tracks' % _API_BASE,
 360         'sets': '%s/users/%%s/playlists' % _API_V2_BASE,
 361         'reposts': '%s/profile/soundcloud:users:%%s/reposts' % _API_V2_BASE,
 362         'likes': '%s/users/%%s/likes' % _API_V2_BASE,
 363         'spotlight': '%s/users/%%s/spotlight' % _API_V2_BASE,
 364     }
 365
 366     _TITLE_MAP = {
 367         'all': 'All',
 368         'tracks': 'Tracks',
 369         'sets': 'Playlists',
 370         'reposts': 'Reposts',
 371         'likes': 'Likes',
 372         'spotlight': 'Spotlight',
 373     }
 374
 375     def _real_extract(self, url):
 376         mobj = re.match(self._VALID_URL, url)
 377         uploader = mobj.group('user')
 378
 379         url = 'http://soundcloud.com/%s/' % uploader
 380         resolv_url = self._resolv_url(url)
 381         user = self._download_json(
 382             resolv_url, uploader, 'Downloading user info')
 383
 384         resource = mobj.group('rsrc') or 'all'
 385         base_url = self._BASE_URL_MAP[resource] % user['id']
 386
 387         next_href = None
 388
 389         entries = []
 390         for i in itertools.count():
 391             if not next_href:
 392                 data = compat_urllib_parse.urlencode({
 393                     'offset': i * 50,
 394                     'limit': 50,
 395                     'client_id': self._CLIENT_ID,
 396                     'linked_partitioning': '1',
 397                     'representation': 'speedy',
 398                 })
 399                 next_href = base_url + '?' + data
 400
 401             response = self._download_json(
 402                 next_href, uploader, 'Downloading track page %s' % (i + 1))
 403
 404             collection = response['collection']
 405
 406             if not collection:
 407                 self.to_screen('%s: End page received' % uploader)
 408                 break
 409
 410             def resolve_permalink_url(candidates):
 411                 for cand in candidates:
 412                     if isinstance(cand, dict):
 413                         permalink_url = cand.get('permalink_url')
 414                         if permalink_url and permalink_url.startswith('http'):
 415                             return permalink_url
 416
 417             for e in collection:
 418                 permalink_url = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
 419                 if permalink_url:
 420                     entries.append(self.url_result(permalink_url))
 421
 422             if 'next_href' in response:
 423                 next_href = response['next_href']
 424                 if not next_href:
 425                     break
 426             else:
 427                 next_href = None
 428
 429         return {
 430             '_type': 'playlist',
 431             'id': compat_str(user['id']),
 432             'title': '%s (%s)' % (user['username'], self._TITLE_MAP[resource]),
 433             'entries': entries,
 434         }
 435
 436
 437 class SoundcloudPlaylistIE(SoundcloudIE):
 438     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 439     IE_NAME = 'soundcloud:playlist'
 440     _TESTS = [{
 441         'url': 'http://api.soundcloud.com/playlists/4110309',
 442         'info_dict': {
 443             'id': '4110309',
 444             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 445             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 446         },
 447         'playlist_count': 6,
 448     }]
 449
 450     def _real_extract(self, url):
 451         mobj = re.match(self._VALID_URL, url)
 452         playlist_id = mobj.group('id')
 453         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 454
 455         data_dict = {
 456             'client_id': self._CLIENT_ID,
 457         }
 458         token = mobj.group('token')
 459
 460         if token:
 461             data_dict['secret_token'] = token
 462
 463         data = compat_urllib_parse.urlencode(data_dict)
 464         data = self._download_json(
 465             base_url + data, playlist_id, 'Downloading playlist')
 466
 467         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in data['tracks']]
 468
 469         return {
 470             '_type': 'playlist',
 471             'id': playlist_id,
 472             'title': data.get('title'),
 473             'description': data.get('description'),
 474             'entries': entries,
 475         }
 476
 477
 478 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
 479     IE_NAME = 'soundcloud:search'
 480     IE_DESC = 'Soundcloud search'
 481     _MAX_RESULTS = float('inf')
 482     _TESTS = [{
 483         'url': 'scsearch15:post-avant jazzcore',
 484         'info_dict': {
 485             'title': 'post-avant jazzcore',
 486         },
 487         'playlist_count': 15,
 488     }]
 489
 490     _SEARCH_KEY = 'scsearch'
 491     _MAX_RESULTS_PER_PAGE = 200
 492     _DEFAULT_RESULTS_PER_PAGE = 50
 493     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 494
 495     def _get_collection(self, endpoint, collection_id, **query):
 496         limit = min(
 497             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
 498             self._MAX_RESULTS_PER_PAGE)
 499         query['limit'] = limit
 500         query['client_id'] = self._CLIENT_ID
 501         query['linked_partitioning'] = '1'
 502         query['offset'] = 0
 503         data = compat_urllib_parse.urlencode(encode_dict(query))
 504         next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
 505
 506         collected_results = 0
 507
 508         for i in itertools.count(1):
 509             response = self._download_json(
 510                 next_url, collection_id, 'Downloading page {0}'.format(i),
 511                 'Unable to download API page')
 512
 513             collection = response.get('collection', [])
 514             if not collection:
 515                 break
 516
 517             collection = list(filter(bool, collection))
 518             collected_results += len(collection)
 519
 520             for item in collection:
 521                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
 522
 523             if not collection or collected_results >= limit:
 524                 break
 525
 526             next_url = response.get('next_href')
 527             if not next_url:
 528                 break
 529
 530     def _get_n_results(self, query, n):
 531         tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
 532         return self.playlist_result(tracks, playlist_title=query)