_ 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 InfoExtractor
   8 from ..compat import (
   9     compat_str,
  10     compat_urlparse,
  11     compat_urllib_parse,
  12 )
  13 from ..utils import (
  14     ExtractorError,
  15     int_or_none,
  16     unified_strdate,
  17 )
  18
  19
  20 class SoundcloudIE(InfoExtractor):
  21     """Information extractor for soundcloud.com
  22        To access the media, the uid of the song and a stream token
  23        must be extracted from the page source and the script must make
  24        a request to media.soundcloud.com/crossdomain.xml. Then
  25        the media can be grabbed by requesting from an url composed
  26        of the stream token and uid
  27      """
  28
  29     _VALID_URL = r'''(?x)^(?:https?://)?
  30                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
  31                             (?P<uploader>[\w\d-]+)/
  32                             (?!(?:tracks|sets(?:/[^/?#]+)?|reposts|likes|spotlight)/?(?:$|[?#]))
  33                             (?P<title>[\w\d-]+)/?
  34                             (?P<token>[^?]+?)?(?:[?].*)?$)
  35                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
  36                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
  37                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
  38                     )
  39                     '''
  40     IE_NAME = 'soundcloud'
  41     _TESTS = [
  42         {
  43             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  44             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  45             'info_dict': {
  46                 'id': '62986583',
  47                 'ext': 'mp3',
  48                 'upload_date': '20121011',
  49                 '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',
  50                 'uploader': 'E.T. ExTerrestrial Music',
  51                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
  52                 'duration': 143,
  53             }
  54         },
  55         # not streamable song
  56         {
  57             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  58             'info_dict': {
  59                 'id': '47127627',
  60                 'ext': 'mp3',
  61                 'title': 'Goldrushed',
  62                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  63                 'uploader': 'The Royal Concept',
  64                 'upload_date': '20120521',
  65                 'duration': 227,
  66             },
  67             'params': {
  68                 # rtmp
  69                 'skip_download': True,
  70             },
  71         },
  72         # private link
  73         {
  74             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  75             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  76             'info_dict': {
  77                 'id': '123998367',
  78                 'ext': 'mp3',
  79                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  80                 'uploader': 'jaimeMF',
  81                 'description': 'test chars:  \"\'/\\ä↭',
  82                 'upload_date': '20131209',
  83                 'duration': 9,
  84             },
  85         },
  86         # private link (alt format)
  87         {
  88             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
  89             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  90             'info_dict': {
  91                 'id': '123998367',
  92                 'ext': 'mp3',
  93                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  94                 'uploader': 'jaimeMF',
  95                 'description': 'test chars:  \"\'/\\ä↭',
  96                 'upload_date': '20131209',
  97                 'duration': 9,
  98             },
  99         },
 100         # downloadable song
 101         {
 102             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
 103             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 104             'info_dict': {
 105                 'id': '128590877',
 106                 'ext': 'mp3',
 107                 'title': 'Bus Brakes',
 108                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 109                 'uploader': 'oddsamples',
 110                 'upload_date': '20140109',
 111                 'duration': 17,
 112             },
 113         },
 114     ]
 115
 116     _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea'
 117     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 118
 119     def report_resolve(self, video_id):
 120         """Report information extraction."""
 121         self.to_screen('%s: Resolving id' % video_id)
 122
 123     @classmethod
 124     def _resolv_url(cls, url):
 125         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 126
 127     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 128         track_id = compat_str(info['id'])
 129         name = full_title or track_id
 130         if quiet:
 131             self.report_extraction(name)
 132
 133         thumbnail = info['artwork_url']
 134         if thumbnail is not None:
 135             thumbnail = thumbnail.replace('-large', '-t500x500')
 136         ext = 'mp3'
 137         result = {
 138             'id': track_id,
 139             'uploader': info['user']['username'],
 140             'upload_date': unified_strdate(info['created_at']),
 141             'title': info['title'],
 142             'description': info['description'],
 143             'thumbnail': thumbnail,
 144             'duration': int_or_none(info.get('duration'), 1000),
 145             'webpage_url': info.get('permalink_url'),
 146         }
 147         formats = []
 148         if info.get('downloadable', False):
 149             # We can build a direct link to the song
 150             format_url = (
 151                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 152                     track_id, self._CLIENT_ID))
 153             formats.append({
 154                 'format_id': 'download',
 155                 'ext': info.get('original_format', 'mp3'),
 156                 'url': format_url,
 157                 'vcodec': 'none',
 158                 'preference': 10,
 159             })
 160
 161         # We have to retrieve the url
 162         streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
 163                        'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
 164         format_dict = self._download_json(
 165             streams_url,
 166             track_id, 'Downloading track url')
 167
 168         for key, stream_url in format_dict.items():
 169             if key.startswith('http'):
 170                 formats.append({
 171                     'format_id': key,
 172                     'ext': ext,
 173                     'url': stream_url,
 174                     'vcodec': 'none',
 175                 })
 176             elif key.startswith('rtmp'):
 177                 # The url doesn't have an rtmp app, we have to extract the playpath
 178                 url, path = stream_url.split('mp3:', 1)
 179                 formats.append({
 180                     'format_id': key,
 181                     'url': url,
 182                     'play_path': 'mp3:' + path,
 183                     'ext': 'flv',
 184                     'vcodec': 'none',
 185                 })
 186
 187             if not formats:
 188                 # We fallback to the stream_url in the original info, this
 189                 # cannot be always used, sometimes it can give an HTTP 404 error
 190                 formats.append({
 191                     'format_id': 'fallback',
 192                     'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 193                     'ext': ext,
 194                     'vcodec': 'none',
 195                 })
 196
 197             for f in formats:
 198                 if f['format_id'].startswith('http'):
 199                     f['protocol'] = 'http'
 200                 if f['format_id'].startswith('rtmp'):
 201                     f['protocol'] = 'rtmp'
 202
 203         self._check_formats(formats, track_id)
 204         self._sort_formats(formats)
 205         result['formats'] = formats
 206
 207         return result
 208
 209     def _real_extract(self, url):
 210         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 211         if mobj is None:
 212             raise ExtractorError('Invalid URL: %s' % url)
 213
 214         track_id = mobj.group('track_id')
 215         token = None
 216         if track_id is not None:
 217             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 218             full_title = track_id
 219             token = mobj.group('secret_token')
 220             if token:
 221                 info_json_url += "&secret_token=" + token
 222         elif mobj.group('player'):
 223             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 224             real_url = query['url'][0]
 225             # If the token is in the query of the original url we have to
 226             # manually add it
 227             if 'secret_token' in query:
 228                 real_url += '?secret_token=' + query['secret_token'][0]
 229             return self.url_result(real_url)
 230         else:
 231             # extract uploader (which is in the url)
 232             uploader = mobj.group('uploader')
 233             # extract simple title (uploader + slug of song title)
 234             slug_title = mobj.group('title')
 235             token = mobj.group('token')
 236             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 237             if token:
 238                 resolve_title += '/%s' % token
 239
 240             self.report_resolve(full_title)
 241
 242             url = 'http://soundcloud.com/%s' % resolve_title
 243             info_json_url = self._resolv_url(url)
 244         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 245
 246         return self._extract_info_dict(info, full_title, secret_token=token)
 247
 248
 249 class SoundcloudSetIE(SoundcloudIE):
 250     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 251     IE_NAME = 'soundcloud:set'
 252     _TESTS = [{
 253         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 254         'info_dict': {
 255             'id': '2284613',
 256             'title': 'The Royal Concept EP',
 257         },
 258         'playlist_mincount': 6,
 259     }]
 260
 261     def _real_extract(self, url):
 262         mobj = re.match(self._VALID_URL, url)
 263
 264         # extract uploader (which is in the url)
 265         uploader = mobj.group('uploader')
 266         # extract simple title (uploader + slug of song title)
 267         slug_title = mobj.group('slug_title')
 268         full_title = '%s/sets/%s' % (uploader, slug_title)
 269         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 270
 271         token = mobj.group('token')
 272         if token:
 273             full_title += '/' + token
 274             url += '/' + token
 275
 276         self.report_resolve(full_title)
 277
 278         resolv_url = self._resolv_url(url)
 279         info = self._download_json(resolv_url, full_title)
 280
 281         if 'errors' in info:
 282             msgs = (compat_str(err['error_message']) for err in info['errors'])
 283             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 284
 285         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in info['tracks']]
 286
 287         return {
 288             '_type': 'playlist',
 289             'entries': entries,
 290             'id': '%s' % info['id'],
 291             'title': info['title'],
 292         }
 293
 294
 295 class SoundcloudUserIE(SoundcloudIE):
 296     _VALID_URL = r'''(?x)
 297                         https?://
 298                             (?:(?:www|m)\.)?soundcloud\.com/
 299                             (?P<user>[^/]+)
 300                             (?:/
 301                                 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
 302                             )?
 303                             /?(?:[?#].*)?$
 304                     '''
 305     IE_NAME = 'soundcloud:user'
 306     _TESTS = [{
 307         'url': 'https://soundcloud.com/the-akashic-chronicler',
 308         'info_dict': {
 309             'id': '114582580',
 310             'title': 'The Akashic Chronicler (All)',
 311         },
 312         'playlist_mincount': 111,
 313     }, {
 314         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
 315         'info_dict': {
 316             'id': '114582580',
 317             'title': 'The Akashic Chronicler (Tracks)',
 318         },
 319         'playlist_mincount': 50,
 320     }, {
 321         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
 322         'info_dict': {
 323             'id': '114582580',
 324             'title': 'The Akashic Chronicler (Playlists)',
 325         },
 326         'playlist_mincount': 3,
 327     }, {
 328         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
 329         'info_dict': {
 330             'id': '114582580',
 331             'title': 'The Akashic Chronicler (Reposts)',
 332         },
 333         'playlist_mincount': 7,
 334     }, {
 335         'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
 336         'info_dict': {
 337             'id': '114582580',
 338             'title': 'The Akashic Chronicler (Likes)',
 339         },
 340         'playlist_mincount': 321,
 341     }, {
 342         'url': 'https://soundcloud.com/grynpyret/spotlight',
 343         'info_dict': {
 344             'id': '7098329',
 345             'title': 'Grynpyret (Spotlight)',
 346         },
 347         'playlist_mincount': 1,
 348     }]
 349
 350     _API_BASE = 'https://api.soundcloud.com'
 351     _API_V2_BASE = 'https://api-v2.soundcloud.com'
 352
 353     _BASE_URL_MAP = {
 354         'all': '%s/profile/soundcloud:users:%%s' % _API_V2_BASE,
 355         'tracks': '%s/users/%%s/tracks' % _API_BASE,
 356         'sets': '%s/users/%%s/playlists' % _API_V2_BASE,
 357         'reposts': '%s/profile/soundcloud:users:%%s/reposts' % _API_V2_BASE,
 358         'likes': '%s/users/%%s/likes' % _API_V2_BASE,
 359         'spotlight': '%s/users/%%s/spotlight' % _API_V2_BASE,
 360     }
 361
 362     _TITLE_MAP = {
 363         'all': 'All',
 364         'tracks': 'Tracks',
 365         'sets': 'Playlists',
 366         'reposts': 'Reposts',
 367         'likes': 'Likes',
 368         'spotlight': 'Spotlight',
 369     }
 370
 371     def _real_extract(self, url):
 372         mobj = re.match(self._VALID_URL, url)
 373         uploader = mobj.group('user')
 374
 375         url = 'http://soundcloud.com/%s/' % uploader
 376         resolv_url = self._resolv_url(url)
 377         user = self._download_json(
 378             resolv_url, uploader, 'Downloading user info')
 379
 380         resource = mobj.group('rsrc') or 'all'
 381         base_url = self._BASE_URL_MAP[resource] % user['id']
 382
 383         next_href = None
 384
 385         entries = []
 386         for i in itertools.count():
 387             if not next_href:
 388                 data = compat_urllib_parse.urlencode({
 389                     'offset': i * 50,
 390                     'limit': 50,
 391                     'client_id': self._CLIENT_ID,
 392                     'linked_partitioning': '1',
 393                     'representation': 'speedy',
 394                 })
 395                 next_href = base_url + '?' + data
 396
 397             response = self._download_json(
 398                 next_href, uploader, 'Downloading track page %s' % (i + 1))
 399
 400             collection = response['collection']
 401
 402             if not collection:
 403                 self.to_screen('%s: End page received' % uploader)
 404                 break
 405
 406             def resolve_permalink_url(candidates):
 407                 for cand in candidates:
 408                     if isinstance(cand, dict):
 409                         permalink_url = cand.get('permalink_url')
 410                         if permalink_url and permalink_url.startswith('http'):
 411                             return permalink_url
 412
 413             for e in collection:
 414                 permalink_url = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
 415                 if permalink_url:
 416                     entries.append(self.url_result(permalink_url))
 417
 418             if 'next_href' in response:
 419                 next_href = response['next_href']
 420                 if not next_href:
 421                     break
 422             else:
 423                 next_href = None
 424
 425         return {
 426             '_type': 'playlist',
 427             'id': compat_str(user['id']),
 428             'title': '%s (%s)' % (user['username'], self._TITLE_MAP[resource]),
 429             'entries': entries,
 430         }
 431
 432
 433 class SoundcloudPlaylistIE(SoundcloudIE):
 434     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 435     IE_NAME = 'soundcloud:playlist'
 436     _TESTS = [{
 437         'url': 'http://api.soundcloud.com/playlists/4110309',
 438         'info_dict': {
 439             'id': '4110309',
 440             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 441             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 442         },
 443         'playlist_count': 6,
 444     }]
 445
 446     def _real_extract(self, url):
 447         mobj = re.match(self._VALID_URL, url)
 448         playlist_id = mobj.group('id')
 449         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 450
 451         data_dict = {
 452             'client_id': self._CLIENT_ID,
 453         }
 454         token = mobj.group('token')
 455
 456         if token:
 457             data_dict['secret_token'] = token
 458
 459         data = compat_urllib_parse.urlencode(data_dict)
 460         data = self._download_json(
 461             base_url + data, playlist_id, 'Downloading playlist')
 462
 463         entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in data['tracks']]
 464
 465         return {
 466             '_type': 'playlist',
 467             'id': playlist_id,
 468             'title': data.get('title'),
 469             'description': data.get('description'),
 470             'entries': entries,
 471         }