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

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import json
   5 import re
   6 import itertools
   7
   8 from .common import InfoExtractor
   9 from ..utils import (
  10     compat_str,
  11     compat_urlparse,
  12     compat_urllib_parse,
  13
  14     ExtractorError,
  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/)(?P<title>[\w\d-]+)/?
  32                             (?P<token>[^?]+?)?(?:[?].*)?$)
  33                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+))
  34                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
  35                     )
  36                     '''
  37     IE_NAME = 'soundcloud'
  38     _TESTS = [
  39         {
  40             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  41             'file': '62986583.mp3',
  42             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
  43             'info_dict': {
  44                 "upload_date": "20121011",
  45                 "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",
  46                 "uploader": "E.T. ExTerrestrial Music",
  47                 "title": "Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1"
  48             }
  49         },
  50         # not streamable song
  51         {
  52             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
  53             'info_dict': {
  54                 'id': '47127627',
  55                 'ext': 'mp3',
  56                 'title': 'Goldrushed',
  57                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
  58                 'uploader': 'The Royal Concept',
  59                 'upload_date': '20120521',
  60             },
  61             'params': {
  62                 # rtmp
  63                 'skip_download': True,
  64             },
  65         },
  66         # private link
  67         {
  68             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
  69             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
  70             'info_dict': {
  71                 'id': '123998367',
  72                 'ext': 'mp3',
  73                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
  74                 'uploader': 'jaimeMF',
  75                 'description': 'test chars:  \"\'/\\ä↭',
  76                 'upload_date': '20131209',
  77             },
  78         },
  79         # downloadable song
  80         {
  81             'url': 'https://soundcloud.com/simgretina/just-your-problem-baby-1',
  82             'md5': '56a8b69568acaa967b4c49f9d1d52d19',
  83             'info_dict': {
  84                 'id': '105614606',
  85                 'ext': 'wav',
  86                 'title': 'Just Your Problem Baby (Acapella)',
  87                 'description': 'Vocals',
  88                 'uploader': 'Sim Gretina',
  89                 'upload_date': '20130815',
  90             },
  91         },
  92     ]
  93
  94     _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28'
  95     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
  96
  97     def report_resolve(self, video_id):
  98         """Report information extraction."""
  99         self.to_screen('%s: Resolving id' % video_id)
 100
 101     @classmethod
 102     def _resolv_url(cls, url):
 103         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 104
 105     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
 106         track_id = compat_str(info['id'])
 107         name = full_title or track_id
 108         if quiet:
 109             self.report_extraction(name)
 110
 111         thumbnail = info['artwork_url']
 112         if thumbnail is not None:
 113             thumbnail = thumbnail.replace('-large', '-t500x500')
 114         ext = 'mp3'
 115         result = {
 116             'id': track_id,
 117             'uploader': info['user']['username'],
 118             'upload_date': unified_strdate(info['created_at']),
 119             'title': info['title'],
 120             'description': info['description'],
 121             'thumbnail': thumbnail,
 122         }
 123         formats = []
 124         if info.get('downloadable', False):
 125             # We can build a direct link to the song
 126             format_url = (
 127                 'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
 128                     track_id, self._CLIENT_ID))
 129             formats.append({
 130                 'format_id': 'download',
 131                 'ext': info.get('original_format', 'mp3'),
 132                 'url': format_url,
 133                 'vcodec': 'none',
 134                 'preference': 10,
 135             })
 136
 137         # We have to retrieve the url
 138         streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
 139             'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
 140         format_dict = self._download_json(
 141             streams_url,
 142             track_id, 'Downloading track url')
 143
 144         for key, stream_url in format_dict.items():
 145             if key.startswith('http'):
 146                 formats.append({
 147                     'format_id': key,
 148                     'ext': ext,
 149                     'url': stream_url,
 150                     'vcodec': 'none',
 151                 })
 152             elif key.startswith('rtmp'):
 153                 # The url doesn't have an rtmp app, we have to extract the playpath
 154                 url, path = stream_url.split('mp3:', 1)
 155                 formats.append({
 156                     'format_id': key,
 157                     'url': url,
 158                     'play_path': 'mp3:' + path,
 159                     'ext': ext,
 160                     'vcodec': 'none',
 161                 })
 162
 163             if not formats:
 164                 # We fallback to the stream_url in the original info, this
 165                 # cannot be always used, sometimes it can give an HTTP 404 error
 166                 formats.append({
 167                     'format_id': 'fallback',
 168                     'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
 169                     'ext': ext,
 170                     'vcodec': 'none',
 171                 })
 172
 173             for f in formats:
 174                 if f['format_id'].startswith('http'):
 175                     f['protocol'] = 'http'
 176                 if f['format_id'].startswith('rtmp'):
 177                     f['protocol'] = 'rtmp'
 178
 179             self._sort_formats(formats)
 180             result['formats'] = formats
 181
 182         return result
 183
 184     def _real_extract(self, url):
 185         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
 186         if mobj is None:
 187             raise ExtractorError('Invalid URL: %s' % url)
 188
 189         track_id = mobj.group('track_id')
 190         token = None
 191         if track_id is not None:
 192             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
 193             full_title = track_id
 194         elif mobj.group('player'):
 195             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
 196             return self.url_result(query['url'][0])
 197         else:
 198             # extract uploader (which is in the url)
 199             uploader = mobj.group('uploader')
 200             # extract simple title (uploader + slug of song title)
 201             slug_title =  mobj.group('title')
 202             token = mobj.group('token')
 203             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
 204             if token:
 205                 resolve_title += '/%s' % token
 206
 207             self.report_resolve(full_title)
 208
 209             url = 'http://soundcloud.com/%s' % resolve_title
 210             info_json_url = self._resolv_url(url)
 211         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 212
 213         return self._extract_info_dict(info, full_title, secret_token=token)
 214
 215
 216 class SoundcloudSetIE(SoundcloudIE):
 217     _VALID_URL = r'https?://(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)'
 218     IE_NAME = 'soundcloud:set'
 219     # it's in tests/test_playlists.py
 220     _TESTS = []
 221
 222     def _real_extract(self, url):
 223         mobj = re.match(self._VALID_URL, url)
 224         if mobj is None:
 225             raise ExtractorError('Invalid URL: %s' % url)
 226
 227         # extract uploader (which is in the url)
 228         uploader = mobj.group(1)
 229         # extract simple title (uploader + slug of song title)
 230         slug_title = mobj.group(2)
 231         full_title = '%s/sets/%s' % (uploader, slug_title)
 232
 233         self.report_resolve(full_title)
 234
 235         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 236         resolv_url = self._resolv_url(url)
 237         info = self._download_json(resolv_url, full_title)
 238
 239         if 'errors' in info:
 240             for err in info['errors']:
 241                 self._downloader.report_error('unable to download video webpage: %s' % compat_str(err['error_message']))
 242             return
 243
 244         self.report_extraction(full_title)
 245         return {'_type': 'playlist',
 246                 'entries': [self._extract_info_dict(track) for track in info['tracks']],
 247                 'id': info['id'],
 248                 'title': info['title'],
 249                 }
 250
 251
 252 class SoundcloudUserIE(SoundcloudIE):
 253     _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$'
 254     IE_NAME = 'soundcloud:user'
 255
 256     # it's in tests/test_playlists.py
 257     _TESTS = []
 258
 259     def _real_extract(self, url):
 260         mobj = re.match(self._VALID_URL, url)
 261         uploader = mobj.group('user')
 262
 263         url = 'http://soundcloud.com/%s/' % uploader
 264         resolv_url = self._resolv_url(url)
 265         user = self._download_json(
 266             resolv_url, uploader, 'Downloading user info')
 267         base_url = 'http://api.soundcloud.com/users/%s/tracks.json?' % uploader
 268
 269         entries = []
 270         for i in itertools.count():
 271             data = compat_urllib_parse.urlencode({
 272                 'offset': i * 50,
 273                 'client_id': self._CLIENT_ID,
 274             })
 275             new_entries = self._download_json(
 276                 base_url + data, uploader, 'Downloading track page %s' % (i + 1))
 277             entries.extend(self._extract_info_dict(e, quiet=True) for e in new_entries)
 278             if len(new_entries) < 50:
 279                 break
 280
 281         return {
 282             '_type': 'playlist',
 283             'id': compat_str(user['id']),
 284             'title': user['username'],
 285             'entries': entries,
 286         }
 287
 288
 289 class SoundcloudPlaylistIE(SoundcloudIE):
 290     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)'
 291     IE_NAME = 'soundcloud:playlist'
 292
 293      # it's in tests/test_playlists.py
 294     _TESTS = []
 295
 296     def _real_extract(self, url):
 297         mobj = re.match(self._VALID_URL, url)
 298         playlist_id = mobj.group('id')
 299         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
 300
 301         data = compat_urllib_parse.urlencode({
 302             'client_id': self._CLIENT_ID,
 303         })
 304         data = self._download_json(
 305             base_url + data, playlist_id, 'Downloading playlist')
 306
 307         entries = [
 308             self._extract_info_dict(t, quiet=True) for t in data['tracks']]
 309
 310         return {
 311             '_type': 'playlist',
 312             'id': playlist_id,
 313             'title': data.get('title'),
 314             'description': data.get('description'),
 315             'entries': entries,
 316         }