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

   1 from __future__ import unicode_literals
   2
   3 import base64
   4 import functools
   5 import itertools
   6 import re
   7
   8 from .common import InfoExtractor
   9 from ..compat import (
  10     compat_chr,
  11     compat_ord,
  12     compat_urllib_parse_unquote,
  13     compat_urlparse,
  14     compat_zip
  15 )
  16 from ..utils import (
  17     clean_html,
  18     ExtractorError,
  19     OnDemandPagedList,
  20     str_to_int,
  21     try_get)
  22
  23
  24 class MixcloudIE(InfoExtractor):
  25     _VALID_URL = r'https?://(?:(?:www|beta|m)\.)?mixcloud\.com/([^/]+)/(?!stream|uploads|favorites|listens|playlists)([^/]+)'
  26     IE_NAME = 'mixcloud'
  27
  28     _TESTS = [{
  29         'url': 'http://www.mixcloud.com/dholbach/cryptkeeper/',
  30         'info_dict': {
  31             'id': 'dholbach-cryptkeeper',
  32             'ext': 'm4a',
  33             'title': 'Cryptkeeper',
  34             'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
  35             'uploader': 'Daniel Holbach',
  36             'uploader_id': 'dholbach',
  37             'thumbnail': r're:https?://.*\.jpg',
  38             'view_count': int,
  39         },
  40     }, {
  41         'url': 'http://www.mixcloud.com/gillespeterson/caribou-7-inch-vinyl-mix-chat/',
  42         'info_dict': {
  43             'id': 'gillespeterson-caribou-7-inch-vinyl-mix-chat',
  44             'ext': 'mp3',
  45             'title': 'Caribou 7 inch Vinyl Mix & Chat',
  46             'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
  47             'uploader': 'Gilles Peterson Worldwide',
  48             'uploader_id': 'gillespeterson',
  49             'thumbnail': 're:https?://.*',
  50             'view_count': int,
  51         },
  52     }, {
  53         'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/',
  54         'only_matching': True,
  55     }]
  56
  57     @staticmethod
  58     def _decrypt_xor_cipher(key, ciphertext):
  59         """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR."""
  60         return ''.join([
  61             compat_chr(compat_ord(ch) ^ compat_ord(k))
  62             for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
  63
  64     @staticmethod
  65     def _decrypt_and_extend(stream_info, url_key, getter, key, formats):
  66         maybe_url = stream_info.get(url_key)
  67         if maybe_url is not None:
  68             decrypted = MixcloudIE._decrypt_xor_cipher(key, base64.b64decode(maybe_url))
  69             formats.extend(getter(decrypted))
  70
  71     def _real_extract(self, url):
  72         mobj = re.match(self._VALID_URL, url)
  73         uploader = mobj.group(1)
  74         cloudcast_name = mobj.group(2)
  75         track_id = compat_urllib_parse_unquote('-'.join((uploader, cloudcast_name)))
  76
  77         webpage = self._download_webpage(url, track_id)
  78
  79         # Legacy path
  80         encrypted_play_info = self._search_regex(
  81             r'm-play-info="([^"]+)"', webpage, 'play info', default=None)
  82
  83         if encrypted_play_info is not None:
  84             # Decode
  85             encrypted_play_info = base64.b64decode(encrypted_play_info)
  86         else:
  87             # New path
  88             full_info_json = self._parse_json(self._html_search_regex(
  89                 r'<script id="relay-data" type="text/x-mixcloud">([^<]+)</script>', webpage, 'play info'), 'play info')
  90             for item in full_info_json:
  91                 item_data = try_get(item, lambda x: x['cloudcast']['data']['cloudcastLookup'])
  92                 if try_get(item_data, lambda x: x['streamInfo']['url']):
  93                     info_json = item_data
  94                     break
  95             else:
  96                 raise ExtractorError('Failed to extract matching stream info')
  97
  98         message = self._html_search_regex(
  99             r'(?s)<div[^>]+class="global-message cloudcast-disabled-notice-light"[^>]*>(.+?)<(?:a|/div)',
 100             webpage, 'error message', default=None)
 101
 102         js_url = self._search_regex(
 103             r'<script[^>]+\bsrc=["\"](https://(?:www\.)?mixcloud\.com/media/js2/www_js_4\.[^>]+\.js)',
 104             webpage, 'js url', default=None)
 105         if js_url is None:
 106             js_url = self._search_regex(
 107                 r'<script[^>]+\bsrc=["\"](https://(?:www\.)?mixcloud\.com/media/js/www\.[^>]+\.js)',
 108                 webpage, 'js url')
 109         js = self._download_webpage(js_url, track_id)
 110         # Known plaintext attack
 111         if encrypted_play_info:
 112             kps = ['{"stream_url":']
 113             kpa_target = encrypted_play_info
 114         else:
 115             kps = ['https://', 'http://']
 116             kpa_target = base64.b64decode(info_json['streamInfo']['url'])
 117         for kp in kps:
 118             partial_key = self._decrypt_xor_cipher(kpa_target, kp)
 119             for quote in ["'", '"']:
 120                 key = self._search_regex(r'{0}({1}[^{0}]*){0}'.format(quote, re.escape(partial_key)), js,
 121                                          "encryption key", default=None)
 122                 if key is not None:
 123                     break
 124             else:
 125                 continue
 126             break
 127         else:
 128             raise ExtractorError('Failed to extract encryption key')
 129
 130         if encrypted_play_info is not None:
 131             play_info = self._parse_json(self._decrypt_xor_cipher(key, encrypted_play_info), 'play info')
 132             if message and 'stream_url' not in play_info:
 133                 raise ExtractorError('%s said: %s' % (self.IE_NAME, message), expected=True)
 134             song_url = play_info['stream_url']
 135             formats = [{
 136                 'format_id': 'normal',
 137                 'url': song_url
 138             }]
 139
 140             title = self._html_search_regex(r'm-title="([^"]+)"', webpage, 'title')
 141             thumbnail = self._proto_relative_url(self._html_search_regex(
 142                 r'm-thumbnail-url="([^"]+)"', webpage, 'thumbnail', fatal=False))
 143             uploader = self._html_search_regex(
 144                 r'm-owner-name="([^"]+)"', webpage, 'uploader', fatal=False)
 145             uploader_id = self._search_regex(
 146                 r'\s+"profile": "([^"]+)",', webpage, 'uploader id', fatal=False)
 147             description = self._og_search_description(webpage)
 148             view_count = str_to_int(self._search_regex(
 149                 [r'<meta itemprop="interactionCount" content="UserPlays:([0-9]+)"',
 150                  r'/listeners/?">([0-9,.]+)</a>',
 151                  r'(?:m|data)-tooltip=["\']([\d,.]+) plays'],
 152                 webpage, 'play count', default=None))
 153
 154         else:
 155             title = info_json['name']
 156             thumbnail = try_get(info_json,
 157                                 lambda x: 'https://thumbnailer.mixcloud.com/unsafe/600x600/' + x['picture']['urlRoot'])
 158             uploader = try_get(info_json, lambda x: x['owner']['displayName'])
 159             uploader_id = try_get(info_json, lambda x: x['owner']['username'])
 160             description = try_get(info_json, lambda x: x['description'])
 161             view_count = try_get(info_json, lambda x: x['plays'])
 162
 163             stream_info = info_json['streamInfo']
 164             formats = []
 165             self._decrypt_and_extend(stream_info, 'url', lambda x: [{
 166                 'format_id': 'normal',
 167                 'url': x
 168             }], key, formats)
 169             self._decrypt_and_extend(stream_info, 'hlsUrl', lambda x: self._extract_m3u8_formats(x, title), key,
 170                                      formats)
 171             self._decrypt_and_extend(stream_info, 'dashUrl', lambda x: self._extract_mpd_formats(x, title), key,
 172                                      formats)
 173
 174         return {
 175             'id': track_id,
 176             'title': title,
 177             'formats': formats,
 178             'description': description,
 179             'thumbnail': thumbnail,
 180             'uploader': uploader,
 181             'uploader_id': uploader_id,
 182             'view_count': view_count,
 183         }
 184
 185
 186 class MixcloudPlaylistBaseIE(InfoExtractor):
 187     _PAGE_SIZE = 24
 188
 189     def _find_urls_in_page(self, page):
 190         for url in re.findall(r'm-play-button m-url="(?P<url>[^"]+)"', page):
 191             yield self.url_result(
 192                 compat_urlparse.urljoin('https://www.mixcloud.com', clean_html(url)),
 193                 MixcloudIE.ie_key())
 194
 195     def _fetch_tracks_page(self, path, video_id, page_name, current_page, real_page_number=None):
 196         real_page_number = real_page_number or current_page + 1
 197         return self._download_webpage(
 198             'https://www.mixcloud.com/%s/' % path, video_id,
 199             note='Download %s (page %d)' % (page_name, current_page + 1),
 200             errnote='Unable to download %s' % page_name,
 201             query={'page': real_page_number, 'list': 'main', '_ajax': '1'},
 202             headers={'X-Requested-With': 'XMLHttpRequest'})
 203
 204     def _tracks_page_func(self, page, video_id, page_name, current_page):
 205         resp = self._fetch_tracks_page(page, video_id, page_name, current_page)
 206
 207         for item in self._find_urls_in_page(resp):
 208             yield item
 209
 210     def _get_user_description(self, page_content):
 211         return self._html_search_regex(
 212             r'<div[^>]+class="profile-bio"[^>]*>(.+?)</div>',
 213             page_content, 'user description', fatal=False)
 214
 215
 216 class MixcloudUserIE(MixcloudPlaylistBaseIE):
 217     _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/(?P<type>uploads|favorites|listens)?/?$'
 218     IE_NAME = 'mixcloud:user'
 219
 220     _TESTS = [{
 221         'url': 'http://www.mixcloud.com/dholbach/',
 222         'info_dict': {
 223             'id': 'dholbach_uploads',
 224             'title': 'Daniel Holbach (uploads)',
 225             'description': 'md5:def36060ac8747b3aabca54924897e47',
 226         },
 227         'playlist_mincount': 11,
 228     }, {
 229         'url': 'http://www.mixcloud.com/dholbach/uploads/',
 230         'info_dict': {
 231             'id': 'dholbach_uploads',
 232             'title': 'Daniel Holbach (uploads)',
 233             'description': 'md5:def36060ac8747b3aabca54924897e47',
 234         },
 235         'playlist_mincount': 11,
 236     }, {
 237         'url': 'http://www.mixcloud.com/dholbach/favorites/',
 238         'info_dict': {
 239             'id': 'dholbach_favorites',
 240             'title': 'Daniel Holbach (favorites)',
 241             'description': 'md5:def36060ac8747b3aabca54924897e47',
 242         },
 243         'params': {
 244             'playlist_items': '1-100',
 245         },
 246         'playlist_mincount': 100,
 247     }, {
 248         'url': 'http://www.mixcloud.com/dholbach/listens/',
 249         'info_dict': {
 250             'id': 'dholbach_listens',
 251             'title': 'Daniel Holbach (listens)',
 252             'description': 'md5:def36060ac8747b3aabca54924897e47',
 253         },
 254         'params': {
 255             'playlist_items': '1-100',
 256         },
 257         'playlist_mincount': 100,
 258     }]
 259
 260     def _real_extract(self, url):
 261         mobj = re.match(self._VALID_URL, url)
 262         user_id = mobj.group('user')
 263         list_type = mobj.group('type')
 264
 265         # if only a profile URL was supplied, default to download all uploads
 266         if list_type is None:
 267             list_type = 'uploads'
 268
 269         video_id = '%s_%s' % (user_id, list_type)
 270
 271         profile = self._download_webpage(
 272             'https://www.mixcloud.com/%s/' % user_id, video_id,
 273             note='Downloading user profile',
 274             errnote='Unable to download user profile')
 275
 276         username = self._og_search_title(profile)
 277         description = self._get_user_description(profile)
 278
 279         entries = OnDemandPagedList(
 280             functools.partial(
 281                 self._tracks_page_func,
 282                 '%s/%s' % (user_id, list_type), video_id, 'list of %s' % list_type),
 283             self._PAGE_SIZE, use_cache=True)
 284
 285         return self.playlist_result(
 286             entries, video_id, '%s (%s)' % (username, list_type), description)
 287
 288
 289 class MixcloudPlaylistIE(MixcloudPlaylistBaseIE):
 290     _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/playlists/(?P<playlist>[^/]+)/?$'
 291     IE_NAME = 'mixcloud:playlist'
 292
 293     _TESTS = [{
 294         'url': 'https://www.mixcloud.com/RedBullThre3style/playlists/tokyo-finalists-2015/',
 295         'info_dict': {
 296             'id': 'RedBullThre3style_tokyo-finalists-2015',
 297             'title': 'National Champions 2015',
 298             'description': 'md5:6ff5fb01ac76a31abc9b3939c16243a3',
 299         },
 300         'playlist_mincount': 16,
 301     }, {
 302         'url': 'https://www.mixcloud.com/maxvibes/playlists/jazzcat-on-ness-radio/',
 303         'only_matching': True,
 304     }]
 305
 306     def _real_extract(self, url):
 307         mobj = re.match(self._VALID_URL, url)
 308         user_id = mobj.group('user')
 309         playlist_id = mobj.group('playlist')
 310         video_id = '%s_%s' % (user_id, playlist_id)
 311
 312         webpage = self._download_webpage(
 313             url, user_id,
 314             note='Downloading playlist page',
 315             errnote='Unable to download playlist page')
 316
 317         title = self._html_search_regex(
 318             r'<a[^>]+class="parent active"[^>]*><b>\d+</b><span[^>]*>([^<]+)',
 319             webpage, 'playlist title',
 320             default=None) or self._og_search_title(webpage, fatal=False)
 321         description = self._get_user_description(webpage)
 322
 323         entries = OnDemandPagedList(
 324             functools.partial(
 325                 self._tracks_page_func,
 326                 '%s/playlists/%s' % (user_id, playlist_id), video_id, 'tracklist'),
 327             self._PAGE_SIZE)
 328
 329         return self.playlist_result(entries, video_id, title, description)
 330
 331
 332 class MixcloudStreamIE(MixcloudPlaylistBaseIE):
 333     _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<id>[^/]+)/stream/?$'
 334     IE_NAME = 'mixcloud:stream'
 335
 336     _TEST = {
 337         'url': 'https://www.mixcloud.com/FirstEar/stream/',
 338         'info_dict': {
 339             'id': 'FirstEar',
 340             'title': 'First Ear',
 341             'description': 'Curators of good music\nfirstearmusic.com',
 342         },
 343         'playlist_mincount': 192,
 344     }
 345
 346     def _real_extract(self, url):
 347         user_id = self._match_id(url)
 348
 349         webpage = self._download_webpage(url, user_id)
 350
 351         entries = []
 352         prev_page_url = None
 353
 354         def _handle_page(page):
 355             entries.extend(self._find_urls_in_page(page))
 356             return self._search_regex(
 357                 r'm-next-page-url="([^"]+)"', page,
 358                 'next page URL', default=None)
 359
 360         next_page_url = _handle_page(webpage)
 361
 362         for idx in itertools.count(0):
 363             if not next_page_url or prev_page_url == next_page_url:
 364                 break
 365
 366             prev_page_url = next_page_url
 367             current_page = int(self._search_regex(
 368                 r'\?page=(\d+)', next_page_url, 'next page number'))
 369
 370             next_page_url = _handle_page(self._fetch_tracks_page(
 371                 '%s/stream' % user_id, user_id, 'stream', idx,
 372                 real_page_number=current_page))
 373
 374         username = self._og_search_title(webpage)
 375         description = self._get_user_description(webpage)
 376
 377         return self.playlist_result(entries, user_id, username, description)