X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/d9d7cd0e85dc712461d9185db9df9d6c900a573b..68c3607ca2d7bdfc72afc94ca5dc6034b41e7bfa:/youtube_dl/extractor/mixcloud.py?ds=inline
diff --git a/youtube_dl/extractor/mixcloud.py b/youtube_dl/extractor/mixcloud.py
index 0efbe66..bcac13e 100644
--- a/youtube_dl/extractor/mixcloud.py
+++ b/youtube_dl/extractor/mixcloud.py
@@ -1,22 +1,27 @@
from __future__ import unicode_literals
-import base64
import functools
import itertools
import re
from .common import InfoExtractor
from ..compat import (
+ compat_b64decode,
compat_chr,
compat_ord,
+ compat_str,
compat_urllib_parse_unquote,
compat_urlparse,
+ compat_zip
)
from ..utils import (
clean_html,
ExtractorError,
+ int_or_none,
OnDemandPagedList,
str_to_int,
+ try_get,
+ urljoin,
)
@@ -53,16 +58,12 @@ class MixcloudIE(InfoExtractor):
'only_matching': True,
}]
- # See https://www.mixcloud.com/media/js2/www_js_2.9e23256562c080482435196ca3975ab5.js
@staticmethod
- def _decrypt_play_info(play_info):
- KEY = 'pleasedontdownloadourmusictheartistswontgetpaid'
-
- play_info = base64.b64decode(play_info.encode('ascii'))
-
+ def _decrypt_xor_cipher(key, ciphertext):
+ """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR."""
return ''.join([
- compat_chr(compat_ord(ch) ^ compat_ord(KEY[idx % len(KEY)]))
- for idx, ch in enumerate(play_info)])
+ compat_chr(compat_ord(ch) ^ compat_ord(k))
+ for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
@@ -72,38 +73,129 @@ class MixcloudIE(InfoExtractor):
webpage = self._download_webpage(url, track_id)
+ # Legacy path
+ encrypted_play_info = self._search_regex(
+ r'm-play-info="([^"]+)"', webpage, 'play info', default=None)
+
+ if encrypted_play_info is not None:
+ # Decode
+ encrypted_play_info = compat_b64decode(encrypted_play_info)
+ else:
+ # New path
+ full_info_json = self._parse_json(self._html_search_regex(
+ r'',
+ webpage, 'play info'), 'play info')
+ for item in full_info_json:
+ item_data = try_get(
+ item, lambda x: x['cloudcast']['data']['cloudcastLookup'],
+ dict)
+ if try_get(item_data, lambda x: x['streamInfo']['url']):
+ info_json = item_data
+ break
+ else:
+ raise ExtractorError('Failed to extract matching stream info')
+
message = self._html_search_regex(
r'(?s)]+class="global-message cloudcast-disabled-notice-light"[^>]*>(.+?)<(?:a|/div)',
webpage, 'error message', default=None)
- encrypted_play_info = self._search_regex(
- r'm-play-info="([^"]+)"', webpage, 'play info')
- play_info = self._parse_json(
- self._decrypt_play_info(encrypted_play_info), track_id)
-
- if message and 'stream_url' not in play_info:
- raise ExtractorError('%s said: %s' % (self.IE_NAME, message), expected=True)
-
- song_url = play_info['stream_url']
-
- title = self._html_search_regex(r'm-title="([^"]+)"', webpage, 'title')
- thumbnail = self._proto_relative_url(self._html_search_regex(
- r'm-thumbnail-url="([^"]+)"', webpage, 'thumbnail', fatal=False))
- uploader = self._html_search_regex(
- r'm-owner-name="([^"]+)"', webpage, 'uploader', fatal=False)
- uploader_id = self._search_regex(
- r'\s+"profile": "([^"]+)",', webpage, 'uploader id', fatal=False)
- description = self._og_search_description(webpage)
- view_count = str_to_int(self._search_regex(
- [r'([0-9,.]+)',
- r'(?:m|data)-tooltip=["\']([\d,.]+) plays'],
- webpage, 'play count', default=None))
+ js_url = self._search_regex(
+ r']+\bsrc=["\"](https://(?:www\.)?mixcloud\.com/media/(?:js2/www_js_4|js/www)\.[^>]+\.js)',
+ webpage, 'js url')
+ js = self._download_webpage(js_url, track_id, 'Downloading JS')
+ # Known plaintext attack
+ if encrypted_play_info:
+ kps = ['{"stream_url":']
+ kpa_target = encrypted_play_info
+ else:
+ kps = ['https://', 'http://']
+ kpa_target = compat_b64decode(info_json['streamInfo']['url'])
+ for kp in kps:
+ partial_key = self._decrypt_xor_cipher(kpa_target, kp)
+ for quote in ["'", '"']:
+ key = self._search_regex(
+ r'{0}({1}[^{0}]*){0}'.format(quote, re.escape(partial_key)),
+ js, 'encryption key', default=None)
+ if key is not None:
+ break
+ else:
+ continue
+ break
+ else:
+ raise ExtractorError('Failed to extract encryption key')
+
+ if encrypted_play_info is not None:
+ play_info = self._parse_json(self._decrypt_xor_cipher(key, encrypted_play_info), 'play info')
+ if message and 'stream_url' not in play_info:
+ raise ExtractorError('%s said: %s' % (self.IE_NAME, message), expected=True)
+ song_url = play_info['stream_url']
+ formats = [{
+ 'format_id': 'normal',
+ 'url': song_url
+ }]
+
+ title = self._html_search_regex(r'm-title="([^"]+)"', webpage, 'title')
+ thumbnail = self._proto_relative_url(self._html_search_regex(
+ r'm-thumbnail-url="([^"]+)"', webpage, 'thumbnail', fatal=False))
+ uploader = self._html_search_regex(
+ r'm-owner-name="([^"]+)"', webpage, 'uploader', fatal=False)
+ uploader_id = self._search_regex(
+ r'\s+"profile": "([^"]+)",', webpage, 'uploader id', fatal=False)
+ description = self._og_search_description(webpage)
+ view_count = str_to_int(self._search_regex(
+ [r'([0-9,.]+)',
+ r'(?:m|data)-tooltip=["\']([\d,.]+) plays'],
+ webpage, 'play count', default=None))
+
+ else:
+ title = info_json['name']
+ thumbnail = urljoin(
+ 'https://thumbnailer.mixcloud.com/unsafe/600x600/',
+ try_get(info_json, lambda x: x['picture']['urlRoot'], compat_str))
+ uploader = try_get(info_json, lambda x: x['owner']['displayName'])
+ uploader_id = try_get(info_json, lambda x: x['owner']['username'])
+ description = try_get(info_json, lambda x: x['description'])
+ view_count = int_or_none(try_get(info_json, lambda x: x['plays']))
+
+ stream_info = info_json['streamInfo']
+ formats = []
+
+ def decrypt_url(f_url):
+ for k in (key, 'IFYOUWANTTHEARTISTSTOGETPAIDDONOTDOWNLOADFROMMIXCLOUD'):
+ decrypted_url = self._decrypt_xor_cipher(k, f_url)
+ if re.search(r'^https?://[0-9a-z.]+/[0-9A-Za-z/.?=&_-]+$', decrypted_url):
+ return decrypted_url
+
+ for url_key in ('url', 'hlsUrl', 'dashUrl'):
+ format_url = stream_info.get(url_key)
+ if not format_url:
+ continue
+ decrypted = decrypt_url(compat_b64decode(format_url))
+ if not decrypted:
+ continue
+ if url_key == 'hlsUrl':
+ formats.extend(self._extract_m3u8_formats(
+ decrypted, track_id, 'mp4', entry_protocol='m3u8_native',
+ m3u8_id='hls', fatal=False))
+ elif url_key == 'dashUrl':
+ formats.extend(self._extract_mpd_formats(
+ decrypted, track_id, mpd_id='dash', fatal=False))
+ else:
+ formats.append({
+ 'format_id': 'http',
+ 'url': decrypted,
+ 'downloader_options': {
+ # Mixcloud starts throttling at >~5M
+ 'http_chunk_size': 5242880,
+ },
+ })
+ self._sort_formats(formats)
return {
'id': track_id,
'title': title,
- 'url': song_url,
+ 'formats': formats,
'description': description,
'thumbnail': thumbnail,
'uploader': uploader,
@@ -209,7 +301,7 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE):
functools.partial(
self._tracks_page_func,
'%s/%s' % (user_id, list_type), video_id, 'list of %s' % list_type),
- self._PAGE_SIZE, use_cache=True)
+ self._PAGE_SIZE)
return self.playlist_result(
entries, video_id, '%s (%s)' % (username, list_type), description)