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 InfoExtractor
   8 from ..utils import (
   9     compat_str,
  10     compat_urlparse,
  11     compat_urllib_parse,
  12
  13     ExtractorError,
  14     int_or_none,
  15     unified_strdate,
  16 )
  17
  18
  19 class SoundcloudIE(InfoExtractor):
  20     """Information extractor for soundcloud.com
  21        To access the media, the uid of the song and a stream token
  22        must be extracted from the page source and the script must make
  23        a request to media.soundcloud.com/crossdomain.xml. Then
  24        the media can be grabbed by requesting from an url composed
  25        of the stream token and uid
  26      """
  27
  28     _VALID_URL = r'''(?x)^(?:https?://)?
  29                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
  30                             (?P<uploader>[\w\d-]+)/
  31                             (?!sets/|likes/?(?:$|[?#]))
  32                             (?P<title>[\w\d-]+)/?
  33                             (?P<token>[^?]+?)?(?:[?].*)?$)
  34                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
  35                           (?:/?\?secret_token=(?P<secret_token>[^&]+?))?$)
  36                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
  37                     )
  38                     '''
  39     IE_NAME = 'soundcloud'
  40     _TESTS = [
  41         {
  42             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  43             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  44             'info_dict': {
  45                 'id': '62986583',
  46                 'ext': 'mp3',
  47                 'upload_date': '20121011',
  48                 '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',
  49                 'uploader': 'E.T. ExTerrestrial Music',
  50                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
  51                 'duration': 143,
  52             }
  53         },
  54         # not streamable song
  55         {
  56             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  57             'info_dict': {
  58                 'id': '47127627',
  59                 'ext': 'mp3',
  60                 'title': 'Goldrushed',
  61                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  62                 'uploader': 'The Royal Concept',
  63                 'upload_date': '20120521',
  64                 'duration': 227,
  65             },
  66             'params': {
  67                 # rtmp
  68                 'skip_download': True,
  69             },
  70         },
  71         # private link
  72         {
  73             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  74             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  75             'info_dict': {
  76                 'id': '123998367',
  77                 'ext': 'mp3',
  78                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  79                 'uploader': 'jaimeMF',
  80                 'description': 'test chars:  \"\'/\\ä↭',
  81                 'upload_date': '20131209',
  82                 'duration': 9,
  83             },
  84         },
  85         # private link (alt format)
  86         {
  87             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
  88             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  89             'info_dict': {
  90                 'id': '123998367',
  91                 'ext': 'mp3',
  92                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  93                 'uploader': 'jaimeMF',
  94                 'description': 'test chars:  \"\'/\\ä↭',
  95                 'upload_date': '20131209',
  96                 'duration': 9,
  97             },
  98         },
  99         # downloadable song
 100         {
 101             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
 102             'md5': '7624f2351f8a3b2e7cd51522496e7631',
 103             'info_dict': {
 104                 'id': '128590877',
 105                 'ext': 'mp3',
 106                 'title': 'Bus Brakes',
 107                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
 108                 'uploader': 'oddsamples',
 109                 'upload_date': '20140109',
 110                 'duration': 17,
 111             },
 112         },
 113     ]
 114
 115     _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28'
 116     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 117
 118     def report_resolve(self, video_id):
 119         """Report information extraction."""
 120         self.to_screen('%s: Resolving id' % video_id)
 121
 122     @classmethod
 123     def _resolv_url(cls, url):
 124         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 125
 126     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 127         track_id = compat_str(info['id'])
 128         name = full_title or track_id
 129         if quiet:
 130             self.report_extraction(name)
 131
 132         thumbnail = info['artwork_url']
 133         if thumbnail is not None:
 134             thumbnail = thumbnail.replace('-large', '-t500x500')
 135         ext = 'mp3'
 136         result = {
 137             'id': track_id,
 138             'uploader': info['user']['username'],
 139             'upload_date': unified_strdate(info['created_at']),
 140             'title': info['title'],
 141             'description': info['description'],
 142             'thumbnail': thumbnail,
 143             'duration': int_or_none(info.get('duration'), 1000),
 144             'webpage_url': info.get('permalink_url'),
 145         }
 146         formats = []
 147         if info.get('downloadable', False):
 148             # We can build a direct link to the song
 149             format_url = (
 150                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 151                     track_id, self._CLIENT_ID))
 152             formats.append({
 153                 'format_id': 'download',
 154                 'ext': info.get('original_format', 'mp3'),
 155                 'url': format_url,
 156                 'vcodec': 'none',
 157                 'preference': 10,
 158             })
 159
 160         # We have to retrieve the url
 161         streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
 162             'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
 163         format_dict = self._download_json(
 164             streams_url,
 165             track_id, 'Downloading track url')
 166
 167         for key, stream_url in format_dict.items():
 168             if key.startswith('http'):
 169                 formats.append({
 170                     'format_id': key,
 171                     'ext': ext,
 172                     'url': stream_url,
 173                     'vcodec': 'none',
 174                 })
 175             elif key.startswith('rtmp'):
 176                 # The url doesn't have an rtmp app, we have to extract the playpath
 177                 url, path = stream_url.split('mp3:', 1)
 178                 formats.append({
 179                     'format_id': key,
 180                     'url': url,
 181                     'play_path': 'mp3:' + path,
 182                     'ext': ext,
 183                     'vcodec': 'none',
 184                 })
 185
 186             if not formats:
 187                 # We fallback to the stream_url in the original info, this
 188                 # cannot be always used, sometimes it can give an HTTP 404 error
 189                 formats.append({
 190                     'format_id': 'fallback',
 191                     'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 192                     'ext': ext,
 193                     'vcodec': 'none',
 194                 })
 195
 196             for f in formats:
 197                 if f['format_id'].startswith('http'):
 198                     f['protocol'] = 'http'
 199                 if f['format_id'].startswith('rtmp'):
 200                     f['protocol'] = 'rtmp'
 201
 202             self._sort_formats(formats)
 203             result['formats'] = formats
 204
 205         return result
 206
 207     def _real_extract(self, url):
 208         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 209         if mobj is None:
 210             raise ExtractorError('Invalid URL: %s' % url)
 211
 212         track_id = mobj.group('track_id')
 213         token = None
 214         if track_id is not None:
 215             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 216             full_title = track_id
 217             token = mobj.group('secret_token')
 218             if token:
 219                 info_json_url += "&secret_token=" + token
 220         elif mobj.group('player'):
 221             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 222             return self.url_result(query['url'][0])
 223         else:
 224             # extract uploader (which is in the url)
 225             uploader = mobj.group('uploader')
 226             # extract simple title (uploader + slug of song title)
 227             slug_title =  mobj.group('title')
 228             token = mobj.group('token')
 229             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 230             if token:
 231                 resolve_title += '/%s' % token
 232
 233             self.report_resolve(full_title)
 234
 235             url = 'http://soundcloud.com/%s' % resolve_title
 236             info_json_url = self._resolv_url(url)
 237         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 238
 239         return self._extract_info_dict(info, full_title, secret_token=token)
 240
 241
 242 class SoundcloudSetIE(SoundcloudIE):
 243     _VALID_URL = r'https?://(?:www\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
 244     IE_NAME = 'soundcloud:set'
 245     _TESTS = [{
 246         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
 247         'info_dict': {
 248             'title': 'The Royal Concept EP',
 249         },
 250         'playlist_mincount': 6,
 251     }]
 252
 253     def _real_extract(self, url):
 254         mobj = re.match(self._VALID_URL, url)
 255
 256         # extract uploader (which is in the url)
 257         uploader = mobj.group('uploader')
 258         # extract simple title (uploader + slug of song title)
 259         slug_title = mobj.group('slug_title')
 260         full_title = '%s/sets/%s' % (uploader, slug_title)
 261         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 262
 263         token = mobj.group('token')
 264         if token:
 265             full_title += '/' + token
 266             url += '/' + token
 267
 268         self.report_resolve(full_title)
 269
 270         resolv_url = self._resolv_url(url)
 271         info = self._download_json(resolv_url, full_title)
 272
 273         if 'errors' in info:
 274             for err in info['errors']:
 275                 self._downloader.report_error('unable to download video webpage: %s' % compat_str(err['error_message']))
 276             return
 277
 278         return {
 279             '_type': 'playlist',
 280             'entries': [self._extract_info_dict(track, secret_token=token) for track in info['tracks']],
 281             'id': info['id'],
 282             'title': info['title'],
 283         }
 284
 285
 286 class SoundcloudUserIE(SoundcloudIE):
 287     _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)/?((?P<rsrc>tracks|likes)/?)?(\?.*)?$'
 288     IE_NAME = 'soundcloud:user'
 289     _TESTS = [{
 290         'url': 'https://soundcloud.com/the-concept-band',
 291         'info_dict': {
 292             'id': '9615865',
 293             'title': 'The Royal Concept',
 294         },
 295         'playlist_mincount': 12
 296     }, {
 297         'url': 'https://soundcloud.com/the-concept-band/likes',
 298         'info_dict': {
 299             'id': '9615865',
 300             'title': 'The Royal Concept',
 301         },
 302         'playlist_mincount': 1,
 303     }]
 304
 305     def _real_extract(self, url):
 306         mobj = re.match(self._VALID_URL, url)
 307         uploader = mobj.group('user')
 308         resource = mobj.group('rsrc')
 309         if resource is None:
 310             resource = 'tracks'
 311         elif resource == 'likes':
 312             resource = 'favorites'
 313
 314         url = 'http://soundcloud.com/%s/' % uploader
 315         resolv_url = self._resolv_url(url)
 316         user = self._download_json(
 317             resolv_url, uploader, 'Downloading user info')
 318         base_url = 'http://api.soundcloud.com/users/%s/%s.json?' % (uploader, resource)
 319
 320         entries = []
 321         for i in itertools.count():
 322             data = compat_urllib_parse.urlencode({
 323                 'offset': i * 50,
 324                 'limit': 50,
 325                 'client_id': self._CLIENT_ID,
 326             })
 327             new_entries = self._download_json(
 328                 base_url + data, uploader, 'Downloading track page %s' % (i + 1))
 329             if len(new_entries) == 0:
 330                 self.to_screen('%s: End page received' % uploader)
 331                 break
 332             entries.extend(self._extract_info_dict(e, quiet=True) for e in new_entries)
 333
 334         return {
 335             '_type': 'playlist',
 336             'id': compat_str(user['id']),
 337             'title': user['username'],
 338             'entries': entries,
 339         }
 340
 341
 342 class SoundcloudPlaylistIE(SoundcloudIE):
 343     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
 344     IE_NAME = 'soundcloud:playlist'
 345     _TESTS = [{
 346         'url': 'http://api.soundcloud.com/playlists/4110309',
 347         'info_dict': {
 348             'id': '4110309',
 349             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
 350             'description': 're:.*?TILT Brass - Bowery Poetry Club',
 351         },
 352         'playlist_count': 6,
 353     }]
 354
 355     def _real_extract(self, url):
 356         mobj = re.match(self._VALID_URL, url)
 357         playlist_id = mobj.group('id')
 358         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 359
 360         data_dict = {
 361             'client_id': self._CLIENT_ID,
 362         }
 363         token = mobj.group('token')
 364
 365         if token:
 366             data_dict['secret_token'] = token
 367
 368         data = compat_urllib_parse.urlencode(data_dict)
 369         data = self._download_json(
 370             base_url + data, playlist_id, 'Downloading playlist')
 371
 372         entries = [
 373             self._extract_info_dict(t, quiet=True, secret_token=token)
 374                 for t in data['tracks']]
 375
 376         return {
 377             '_type': 'playlist',
 378             'id': playlist_id,
 379             'title': data.get('title'),
 380             'description': data.get('description'),
 381             'entries': entries,
 382         }