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

   1 import json
   2 import re
   3 import itertools
   4
   5 from .common import InfoExtractor
   6 from ..utils import (
   7     compat_str,
   8     compat_urlparse,
   9     compat_urllib_parse,
  10
  11     ExtractorError,
  12     unified_strdate,
  13 )
  14
  15
  16 class SoundcloudIE(InfoExtractor):
  17     """Information extractor for soundcloud.com
  18        To access the media, the uid of the song and a stream token
  19        must be extracted from the page source and the script must make
  20        a request to media.soundcloud.com/crossdomain.xml. Then
  21        the media can be grabbed by requesting from an url composed
  22        of the stream token and uid
  23      """
  24
  25     _VALID_URL = r'''^(?:https?://)?
  26                     (?:(?:(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)/?(?:[?].*)?$)
  27                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+))
  28                        |(?P<widget>w.soundcloud.com/player/?.*?url=.*)
  29                     )
  30                     '''
  31     IE_NAME = u'soundcloud'
  32     _TEST = {
  33         u'url': u'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
  34         u'file': u'62986583.mp3',
  35         u'md5': u'ebef0a451b909710ed1d7787dddbf0d7',
  36         u'info_dict': {
  37             u"upload_date": u"20121011",
  38             u"description": u"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",
  39             u"uploader": u"E.T. ExTerrestrial Music",
  40             u"title": u"Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1"
  41         }
  42     }
  43
  44     _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28'
  45
  46     @classmethod
  47     def suitable(cls, url):
  48         return re.match(cls._VALID_URL, url, flags=re.VERBOSE) is not None
  49
  50     def report_resolve(self, video_id):
  51         """Report information extraction."""
  52         self.to_screen(u'%s: Resolving id' % video_id)
  53
  54     @classmethod
  55     def _resolv_url(cls, url):
  56         return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
  57
  58     def _extract_info_dict(self, info, full_title=None, quiet=False):
  59         video_id = info['id']
  60         name = full_title or video_id
  61         if quiet == False:
  62             self.report_extraction(name)
  63
  64         thumbnail = info['artwork_url']
  65         if thumbnail is not None:
  66             thumbnail = thumbnail.replace('-large', '-t500x500')
  67         return {
  68             'id':       info['id'],
  69             'url':      info['stream_url'] + '?client_id=' + self._CLIENT_ID,
  70             'uploader': info['user']['username'],
  71             'upload_date': unified_strdate(info['created_at']),
  72             'title':    info['title'],
  73             'ext':      u'mp3',
  74             'description': info['description'],
  75             'thumbnail': thumbnail,
  76         }
  77
  78     def _real_extract(self, url):
  79         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
  80         if mobj is None:
  81             raise ExtractorError(u'Invalid URL: %s' % url)
  82
  83         track_id = mobj.group('track_id')
  84         if track_id is not None:
  85             info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
  86             full_title = track_id
  87         elif mobj.group('widget'):
  88             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
  89             return self.url_result(query['url'][0], ie='Soundcloud')
  90         else:
  91             # extract uploader (which is in the url)
  92             uploader = mobj.group(1)
  93             # extract simple title (uploader + slug of song title)
  94             slug_title =  mobj.group(2)
  95             full_title = '%s/%s' % (uploader, slug_title)
  96
  97             self.report_resolve(full_title)
  98
  99             url = 'http://soundcloud.com/%s/%s' % (uploader, slug_title)
 100             info_json_url = self._resolv_url(url)
 101         info_json = self._download_webpage(info_json_url, full_title, u'Downloading info JSON')
 102
 103         info = json.loads(info_json)
 104         return self._extract_info_dict(info, full_title)
 105
 106 class SoundcloudSetIE(SoundcloudIE):
 107     _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$'
 108     IE_NAME = u'soundcloud:set'
 109     _TEST = {
 110         u"url":"https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep",
 111         u"playlist": [
 112             {
 113                 u"file":"30510138.mp3",
 114                 u"md5":"f9136bf103901728f29e419d2c70f55d",
 115                 u"info_dict": {
 116                     u"upload_date": u"20111213",
 117                     u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
 118                     u"uploader": u"The Royal Concept",
 119                     u"title": u"D-D-Dance"
 120                 }
 121             },
 122             {
 123                 u"file":"47127625.mp3",
 124                 u"md5":"09b6758a018470570f8fd423c9453dd8",
 125                 u"info_dict": {
 126                     u"upload_date": u"20120521",
 127                     u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
 128                     u"uploader": u"The Royal Concept",
 129                     u"title": u"The Royal Concept - Gimme Twice"
 130                 }
 131             },
 132             {
 133                 u"file":"47127627.mp3",
 134                 u"md5":"154abd4e418cea19c3b901f1e1306d9c",
 135                 u"info_dict": {
 136                     u"upload_date": u"20120521",
 137                     u"uploader": u"The Royal Concept",
 138                     u"title": u"Goldrushed"
 139                 }
 140             },
 141             {
 142                 u"file":"47127629.mp3",
 143                 u"md5":"2f5471edc79ad3f33a683153e96a79c1",
 144                 u"info_dict": {
 145                     u"upload_date": u"20120521",
 146                     u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
 147                     u"uploader": u"The Royal Concept",
 148                     u"title": u"In the End"
 149                 }
 150             },
 151             {
 152                 u"file":"47127631.mp3",
 153                 u"md5":"f9ba87aa940af7213f98949254f1c6e2",
 154                 u"info_dict": {
 155                     u"upload_date": u"20120521",
 156                     u"description": u"The Royal Concept from Stockholm\r\nFilip / David / Povel / Magnus\r\nwww.theroyalconceptband.com",
 157                     u"uploader": u"The Royal Concept",
 158                     u"title": u"Knocked Up"
 159                 }
 160             },
 161             {
 162                 u"file":"75206121.mp3",
 163                 u"md5":"f9d1fe9406717e302980c30de4af9353",
 164                 u"info_dict": {
 165                     u"upload_date": u"20130116",
 166                     u"description": u"The unreleased track World on Fire premiered on the CW's hit show Arrow (8pm/7pm central).  \r\nAs a gift to our fans we would like to offer you a free download of the track!  ",
 167                     u"uploader": u"The Royal Concept",
 168                     u"title": u"World On Fire"
 169                 }
 170             }
 171         ]
 172     }
 173
 174     def _real_extract(self, url):
 175         mobj = re.match(self._VALID_URL, url)
 176         if mobj is None:
 177             raise ExtractorError(u'Invalid URL: %s' % url)
 178
 179         # extract uploader (which is in the url)
 180         uploader = mobj.group(1)
 181         # extract simple title (uploader + slug of song title)
 182         slug_title =  mobj.group(2)
 183         full_title = '%s/sets/%s' % (uploader, slug_title)
 184
 185         self.report_resolve(full_title)
 186
 187         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 188         resolv_url = self._resolv_url(url)
 189         info_json = self._download_webpage(resolv_url, full_title)
 190
 191         videos = []
 192         info = json.loads(info_json)
 193         if 'errors' in info:
 194             for err in info['errors']:
 195                 self._downloader.report_error(u'unable to download video webpage: %s' % compat_str(err['error_message']))
 196             return
 197
 198         self.report_extraction(full_title)
 199         return {'_type': 'playlist',
 200                 'entries': [self._extract_info_dict(track) for track in info['tracks']],
 201                 'id': info['id'],
 202                 'title': info['title'],
 203                 }
 204
 205
 206 class SoundcloudUserIE(SoundcloudIE):
 207     _VALID_URL = r'https?://(www\.)?soundcloud.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$'
 208     IE_NAME = u'soundcloud:user'
 209
 210     # it's in tests/test_playlists.py
 211     _TEST = None
 212
 213     def _real_extract(self, url):
 214         mobj = re.match(self._VALID_URL, url)
 215         uploader = mobj.group('user')
 216
 217         url = 'http://soundcloud.com/%s/' % uploader
 218         resolv_url = self._resolv_url(url)
 219         user_json = self._download_webpage(resolv_url, uploader,
 220             u'Downloading user info')
 221         user = json.loads(user_json)
 222
 223         tracks = []
 224         for i in itertools.count():
 225             data = compat_urllib_parse.urlencode({'offset': i*50,
 226                                                   'client_id': self._CLIENT_ID,
 227                                                   })
 228             tracks_url = 'http://api.soundcloud.com/users/%s/tracks.json?' % user['id'] + data
 229             response = self._download_webpage(tracks_url, uploader,
 230                 u'Downloading tracks page %s' % (i+1))
 231             new_tracks = json.loads(response)
 232             tracks.extend(self._extract_info_dict(track, quiet=True) for track in new_tracks)
 233             if len(new_tracks) < 50:
 234                 break
 235
 236         return {
 237             '_type': 'playlist',
 238             'id': compat_str(user['id']),
 239             'title': user['username'],
 240             'entries': tracks,
 241         }