[mixcloud] improve extraction

- improve metadata extraction(closes #11721)
- fix playlist extraction(closes #22378)
- fix user mixes extraction(closes #15197)(closes #17865)
This commit is contained in:
Remita Amine 2019-11-06 20:41:49 +01:00
parent 55adb63e54
commit 5d92b407e0
2 changed files with 225 additions and 274 deletions

View File

@ -639,7 +639,6 @@ from .mixcloud import (
MixcloudIE, MixcloudIE,
MixcloudUserIE, MixcloudUserIE,
MixcloudPlaylistIE, MixcloudPlaylistIE,
MixcloudStreamIE,
) )
from .mlb import MLBIE from .mlb import MLBIE
from .mnet import MnetIE from .mnet import MnetIE

View File

@ -1,6 +1,5 @@
from __future__ import unicode_literals from __future__ import unicode_literals
import functools
import itertools import itertools
import re import re
@ -11,28 +10,37 @@ from ..compat import (
compat_ord, compat_ord,
compat_str, compat_str,
compat_urllib_parse_unquote, compat_urllib_parse_unquote,
compat_urlparse,
compat_zip compat_zip
) )
from ..utils import ( from ..utils import (
clean_html,
ExtractorError,
int_or_none, int_or_none,
OnDemandPagedList, parse_iso8601,
str_to_int, strip_or_none,
try_get, try_get,
urljoin,
) )
class MixcloudIE(InfoExtractor): class MixcloudBaseIE(InfoExtractor):
def _call_api(self, object_type, object_fields, display_id, username, slug=None):
lookup_key = object_type + 'Lookup'
return self._download_json(
'https://www.mixcloud.com/graphql', display_id, query={
'query': '''{
%s(lookup: {username: "%s"%s}) {
%s
}
}''' % (lookup_key, username, ', slug: "%s"' % slug if slug else '', object_fields)
})['data'][lookup_key]
class MixcloudIE(MixcloudBaseIE):
_VALID_URL = r'https?://(?:(?:www|beta|m)\.)?mixcloud\.com/([^/]+)/(?!stream|uploads|favorites|listens|playlists)([^/]+)' _VALID_URL = r'https?://(?:(?:www|beta|m)\.)?mixcloud\.com/([^/]+)/(?!stream|uploads|favorites|listens|playlists)([^/]+)'
IE_NAME = 'mixcloud' IE_NAME = 'mixcloud'
_TESTS = [{ _TESTS = [{
'url': 'http://www.mixcloud.com/dholbach/cryptkeeper/', 'url': 'http://www.mixcloud.com/dholbach/cryptkeeper/',
'info_dict': { 'info_dict': {
'id': 'dholbach-cryptkeeper', 'id': 'dholbach_cryptkeeper',
'ext': 'm4a', 'ext': 'm4a',
'title': 'Cryptkeeper', 'title': 'Cryptkeeper',
'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.', 'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
@ -40,11 +48,13 @@ class MixcloudIE(InfoExtractor):
'uploader_id': 'dholbach', 'uploader_id': 'dholbach',
'thumbnail': r're:https?://.*\.jpg', 'thumbnail': r're:https?://.*\.jpg',
'view_count': int, 'view_count': int,
'timestamp': 1321359578,
'upload_date': '20111115',
}, },
}, { }, {
'url': 'http://www.mixcloud.com/gillespeterson/caribou-7-inch-vinyl-mix-chat/', 'url': 'http://www.mixcloud.com/gillespeterson/caribou-7-inch-vinyl-mix-chat/',
'info_dict': { 'info_dict': {
'id': 'gillespeterson-caribou-7-inch-vinyl-mix-chat', 'id': 'gillespeterson_caribou-7-inch-vinyl-mix-chat',
'ext': 'mp3', 'ext': 'mp3',
'title': 'Caribou 7 inch Vinyl Mix & Chat', 'title': 'Caribou 7 inch Vinyl Mix & Chat',
'description': 'md5:2b8aec6adce69f9d41724647c65875e8', 'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
@ -52,11 +62,14 @@ class MixcloudIE(InfoExtractor):
'uploader_id': 'gillespeterson', 'uploader_id': 'gillespeterson',
'thumbnail': 're:https?://.*', 'thumbnail': 're:https?://.*',
'view_count': int, 'view_count': int,
'timestamp': 1422987057,
'upload_date': '20150203',
}, },
}, { }, {
'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/', 'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/',
'only_matching': True, 'only_matching': True,
}] }]
_DECRYPTION_KEY = 'IFYOUWANTTHEARTISTSTOGETPAIDDONOTDOWNLOADFROMMIXCLOUD'
@staticmethod @staticmethod
def _decrypt_xor_cipher(key, ciphertext): def _decrypt_xor_cipher(key, ciphertext):
@ -66,177 +79,193 @@ class MixcloudIE(InfoExtractor):
for ch, k in compat_zip(ciphertext, itertools.cycle(key))]) for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
def _real_extract(self, url): def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url) username, slug = re.match(self._VALID_URL, url).groups()
uploader = mobj.group(1) username, slug = compat_urllib_parse_unquote(username), compat_urllib_parse_unquote(slug)
cloudcast_name = mobj.group(2) track_id = '%s_%s' % (username, slug)
track_id = compat_urllib_parse_unquote('-'.join((uploader, cloudcast_name)))
webpage = self._download_webpage(url, track_id) cloudcast = self._call_api('cloudcast', '''audioLength
comments(first: 100) {
edges {
node {
comment
created
user {
displayName
username
}
}
}
totalCount
}
description
favorites {
totalCount
}
featuringArtistList
isExclusive
name
owner {
displayName
url
username
}
picture(width: 1024, height: 1024) {
url
}
plays
publishDate
reposts {
totalCount
}
streamInfo {
dashUrl
hlsUrl
url
}
tags {
tag {
name
}
}''', track_id, username, slug)
# Legacy path title = cloudcast['name']
encrypted_play_info = self._search_regex(
r'm-play-info="([^"]+)"', webpage, 'play info', default=None)
if encrypted_play_info is not None: stream_info = cloudcast['streamInfo']
# Decode formats = []
encrypted_play_info = compat_b64decode(encrypted_play_info)
else:
# New path
full_info_json = self._parse_json(self._html_search_regex(
r'<script id="relay-data" type="text/x-mixcloud">([^<]+)</script>',
webpage, 'play info'), 'play info')
for item in full_info_json:
item_data = try_get(item, [
lambda x: x['cloudcast']['data']['cloudcastLookup'],
lambda x: x['cloudcastLookup']['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( for url_key in ('url', 'hlsUrl', 'dashUrl'):
r'(?s)<div[^>]+class="global-message cloudcast-disabled-notice-light"[^>]*>(.+?)<(?:a|/div)', format_url = stream_info.get(url_key)
webpage, 'error message', default=None) if not format_url:
js_url = self._search_regex(
r'<script[^>]+\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 continue
break decrypted = self._decrypt_xor_cipher(
else: self._DECRYPTION_KEY, compat_b64decode(format_url))
raise ExtractorError('Failed to extract encryption key') 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,
},
})
if encrypted_play_info is not None: if not formats and cloudcast.get('isExclusive'):
play_info = self._parse_json(self._decrypt_xor_cipher(key, encrypted_play_info), 'play info') self.raise_login_required()
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') self._sort_formats(formats)
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'<meta itemprop="interactionCount" content="UserPlays:([0-9]+)"',
r'/listeners/?">([0-9,.]+)</a>',
r'(?:m|data)-tooltip=["\']([\d,.]+) plays'],
webpage, 'play count', default=None))
else: comments = []
title = info_json['name'] for edge in (try_get(cloudcast, lambda x: x['comments']['edges']) or []):
thumbnail = urljoin( node = edge.get('node') or {}
'https://thumbnailer.mixcloud.com/unsafe/600x600/', text = strip_or_none(node.get('comment'))
try_get(info_json, lambda x: x['picture']['urlRoot'], compat_str)) if not text:
uploader = try_get(info_json, lambda x: x['owner']['displayName']) continue
uploader_id = try_get(info_json, lambda x: x['owner']['username']) user = node.get('user') or {}
description = try_get(info_json, lambda x: x['description']) comments.append({
view_count = int_or_none(try_get(info_json, lambda x: x['plays'])) 'author': user.get('displayName'),
'author_id': user.get('username'),
'text': text,
'timestamp': parse_iso8601(node.get('created')),
})
stream_info = info_json['streamInfo'] tags = []
formats = [] for t in cloudcast.get('tags'):
tag = try_get(t, lambda x: x['tag']['name'], compat_str)
if not tag:
tags.append(tag)
def decrypt_url(f_url): get_count = lambda x: int_or_none(try_get(cloudcast, lambda y: y[x]['totalCount']))
for k in (key, 'IFYOUWANTTHEARTISTSTOGETPAIDDONOTDOWNLOADFROMMIXCLOUD'):
decrypted_url = self._decrypt_xor_cipher(k, f_url)
if re.search(r'^https?://[0-9A-Za-z.]+/[0-9A-Za-z/.?=&_-]+$', decrypted_url):
return decrypted_url
for url_key in ('url', 'hlsUrl', 'dashUrl'): owner = cloudcast.get('owner') or {}
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 { return {
'id': track_id, 'id': track_id,
'title': title, 'title': title,
'formats': formats, 'formats': formats,
'description': description, 'description': cloudcast.get('description'),
'thumbnail': thumbnail, 'thumbnail': try_get(cloudcast, lambda x: x['picture']['url'], compat_str),
'uploader': uploader, 'uploader': owner.get('displayName'),
'uploader_id': uploader_id, 'timestamp': parse_iso8601(cloudcast.get('publishDate')),
'view_count': view_count, 'uploader_id': owner.get('username'),
'uploader_url': owner.get('url'),
'duration': int_or_none(cloudcast.get('audioLength')),
'view_count': int_or_none(cloudcast.get('plays')),
'like_count': get_count('favorites'),
'repost_count': get_count('reposts'),
'comment_count': get_count('comments'),
'comments': comments,
'tags': tags,
'artist': ', '.join(cloudcast.get('featuringArtistList') or []) or None,
} }
class MixcloudPlaylistBaseIE(InfoExtractor): class MixcloudPlaylistBaseIE(MixcloudBaseIE):
_PAGE_SIZE = 24 def _get_cloudcast(self, node):
return node
def _find_urls_in_page(self, page): def _get_playlist_title(self, title, slug):
for url in re.findall(r'm-play-button m-url="(?P<url>[^"]+)"', page): return title
yield self.url_result(
compat_urlparse.urljoin('https://www.mixcloud.com', clean_html(url)),
MixcloudIE.ie_key())
def _fetch_tracks_page(self, path, video_id, page_name, current_page, real_page_number=None): def _real_extract(self, url):
real_page_number = real_page_number or current_page + 1 username, slug = re.match(self._VALID_URL, url).groups()
return self._download_webpage( username = compat_urllib_parse_unquote(username)
'https://www.mixcloud.com/%s/' % path, video_id, if not slug:
note='Download %s (page %d)' % (page_name, current_page + 1), slug = 'uploads'
errnote='Unable to download %s' % page_name, else:
query={'page': real_page_number, 'list': 'main', '_ajax': '1'}, slug = compat_urllib_parse_unquote(slug)
headers={'X-Requested-With': 'XMLHttpRequest'}) playlist_id = '%s_%s' % (username, slug)
def _tracks_page_func(self, page, video_id, page_name, current_page): is_playlist_type = self._ROOT_TYPE == 'playlist'
resp = self._fetch_tracks_page(page, video_id, page_name, current_page) playlist_type = 'items' if is_playlist_type else slug
list_filter = ''
for item in self._find_urls_in_page(resp): has_next_page = True
yield item entries = []
while has_next_page:
playlist = self._call_api(
self._ROOT_TYPE, '''%s
%s
%s(first: 100%s) {
edges {
node {
%s
}
}
pageInfo {
endCursor
hasNextPage
}
}''' % (self._TITLE_KEY, self._DESCRIPTION_KEY, playlist_type, list_filter, self._NODE_TEMPLATE),
playlist_id, username, slug if is_playlist_type else None)
def _get_user_description(self, page_content): items = playlist.get(playlist_type) or {}
return self._html_search_regex( for edge in items.get('edges', []):
r'<div[^>]+class="profile-bio"[^>]*>(.+?)</div>', cloudcast = self._get_cloudcast(edge.get('node') or {})
page_content, 'user description', fatal=False) cloudcast_url = cloudcast.get('url')
if not cloudcast_url:
continue
entries.append(self.url_result(
cloudcast_url, MixcloudIE.ie_key(), cloudcast.get('slug')))
page_info = items['pageInfo']
has_next_page = page_info['hasNextPage']
list_filter = ', after: "%s"' % page_info['endCursor']
return self.playlist_result(
entries, playlist_id,
self._get_playlist_title(playlist[self._TITLE_KEY], slug),
playlist.get(self._DESCRIPTION_KEY))
class MixcloudUserIE(MixcloudPlaylistBaseIE): class MixcloudUserIE(MixcloudPlaylistBaseIE):
_VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/(?P<type>uploads|favorites|listens)?/?$' _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<id>[^/]+)/(?P<type>uploads|favorites|listens|stream)?/?$'
IE_NAME = 'mixcloud:user' IE_NAME = 'mixcloud:user'
_TESTS = [{ _TESTS = [{
@ -244,68 +273,58 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE):
'info_dict': { 'info_dict': {
'id': 'dholbach_uploads', 'id': 'dholbach_uploads',
'title': 'Daniel Holbach (uploads)', 'title': 'Daniel Holbach (uploads)',
'description': 'md5:def36060ac8747b3aabca54924897e47', 'description': 'md5:b60d776f0bab534c5dabe0a34e47a789',
}, },
'playlist_mincount': 11, 'playlist_mincount': 36,
}, { }, {
'url': 'http://www.mixcloud.com/dholbach/uploads/', 'url': 'http://www.mixcloud.com/dholbach/uploads/',
'info_dict': { 'info_dict': {
'id': 'dholbach_uploads', 'id': 'dholbach_uploads',
'title': 'Daniel Holbach (uploads)', 'title': 'Daniel Holbach (uploads)',
'description': 'md5:def36060ac8747b3aabca54924897e47', 'description': 'md5:b60d776f0bab534c5dabe0a34e47a789',
}, },
'playlist_mincount': 11, 'playlist_mincount': 36,
}, { }, {
'url': 'http://www.mixcloud.com/dholbach/favorites/', 'url': 'http://www.mixcloud.com/dholbach/favorites/',
'info_dict': { 'info_dict': {
'id': 'dholbach_favorites', 'id': 'dholbach_favorites',
'title': 'Daniel Holbach (favorites)', 'title': 'Daniel Holbach (favorites)',
'description': 'md5:def36060ac8747b3aabca54924897e47', 'description': 'md5:b60d776f0bab534c5dabe0a34e47a789',
}, },
'params': { # 'params': {
'playlist_items': '1-100', # 'playlist_items': '1-100',
}, # },
'playlist_mincount': 100, 'playlist_mincount': 396,
}, { }, {
'url': 'http://www.mixcloud.com/dholbach/listens/', 'url': 'http://www.mixcloud.com/dholbach/listens/',
'info_dict': { 'info_dict': {
'id': 'dholbach_listens', 'id': 'dholbach_listens',
'title': 'Daniel Holbach (listens)', 'title': 'Daniel Holbach (listens)',
'description': 'md5:def36060ac8747b3aabca54924897e47', 'description': 'md5:b60d776f0bab534c5dabe0a34e47a789',
}, },
'params': { # 'params': {
'playlist_items': '1-100', # 'playlist_items': '1-100',
# },
'playlist_mincount': 1623,
'skip': 'Large list',
}, {
'url': 'https://www.mixcloud.com/FirstEar/stream/',
'info_dict': {
'id': 'FirstEar_stream',
'title': 'First Ear (stream)',
'description': 'Curators of good music\r\n\r\nfirstearmusic.com',
}, },
'playlist_mincount': 100, 'playlist_mincount': 271,
}] }]
def _real_extract(self, url): _TITLE_KEY = 'displayName'
mobj = re.match(self._VALID_URL, url) _DESCRIPTION_KEY = 'biog'
user_id = mobj.group('user') _ROOT_TYPE = 'user'
list_type = mobj.group('type') _NODE_TEMPLATE = '''slug
url'''
# if only a profile URL was supplied, default to download all uploads def _get_playlist_title(self, title, slug):
if list_type is None: return '%s (%s)' % (title, slug)
list_type = 'uploads'
video_id = '%s_%s' % (user_id, list_type)
profile = self._download_webpage(
'https://www.mixcloud.com/%s/' % user_id, video_id,
note='Downloading user profile',
errnote='Unable to download user profile')
username = self._og_search_title(profile)
description = self._get_user_description(profile)
entries = OnDemandPagedList(
functools.partial(
self._tracks_page_func,
'%s/%s' % (user_id, list_type), video_id, 'list of %s' % list_type),
self._PAGE_SIZE)
return self.playlist_result(
entries, video_id, '%s (%s)' % (username, list_type), description)
class MixcloudPlaylistIE(MixcloudPlaylistBaseIE): class MixcloudPlaylistIE(MixcloudPlaylistBaseIE):
@ -313,87 +332,20 @@ class MixcloudPlaylistIE(MixcloudPlaylistBaseIE):
IE_NAME = 'mixcloud:playlist' IE_NAME = 'mixcloud:playlist'
_TESTS = [{ _TESTS = [{
'url': 'https://www.mixcloud.com/RedBullThre3style/playlists/tokyo-finalists-2015/',
'info_dict': {
'id': 'RedBullThre3style_tokyo-finalists-2015',
'title': 'National Champions 2015',
'description': 'md5:6ff5fb01ac76a31abc9b3939c16243a3',
},
'playlist_mincount': 16,
}, {
'url': 'https://www.mixcloud.com/maxvibes/playlists/jazzcat-on-ness-radio/', 'url': 'https://www.mixcloud.com/maxvibes/playlists/jazzcat-on-ness-radio/',
'only_matching': True,
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
user_id = mobj.group('user')
playlist_id = mobj.group('playlist')
video_id = '%s_%s' % (user_id, playlist_id)
webpage = self._download_webpage(
url, user_id,
note='Downloading playlist page',
errnote='Unable to download playlist page')
title = self._html_search_regex(
r'<a[^>]+class="parent active"[^>]*><b>\d+</b><span[^>]*>([^<]+)',
webpage, 'playlist title',
default=None) or self._og_search_title(webpage, fatal=False)
description = self._get_user_description(webpage)
entries = OnDemandPagedList(
functools.partial(
self._tracks_page_func,
'%s/playlists/%s' % (user_id, playlist_id), video_id, 'tracklist'),
self._PAGE_SIZE)
return self.playlist_result(entries, video_id, title, description)
class MixcloudStreamIE(MixcloudPlaylistBaseIE):
_VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<id>[^/]+)/stream/?$'
IE_NAME = 'mixcloud:stream'
_TEST = {
'url': 'https://www.mixcloud.com/FirstEar/stream/',
'info_dict': { 'info_dict': {
'id': 'FirstEar', 'id': 'maxvibes_jazzcat-on-ness-radio',
'title': 'First Ear', 'title': 'Ness Radio sessions',
'description': 'Curators of good music\nfirstearmusic.com',
}, },
'playlist_mincount': 192, 'playlist_mincount': 59,
} }]
_TITLE_KEY = 'name'
_DESCRIPTION_KEY = 'description'
_ROOT_TYPE = 'playlist'
_NODE_TEMPLATE = '''cloudcast {
slug
url
}'''
def _real_extract(self, url): def _get_cloudcast(self, node):
user_id = self._match_id(url) return node.get('cloudcast') or {}
webpage = self._download_webpage(url, user_id)
entries = []
prev_page_url = None
def _handle_page(page):
entries.extend(self._find_urls_in_page(page))
return self._search_regex(
r'm-next-page-url="([^"]+)"', page,
'next page URL', default=None)
next_page_url = _handle_page(webpage)
for idx in itertools.count(0):
if not next_page_url or prev_page_url == next_page_url:
break
prev_page_url = next_page_url
current_page = int(self._search_regex(
r'\?page=(\d+)', next_page_url, 'next page number'))
next_page_url = _handle_page(self._fetch_tracks_page(
'%s/stream' % user_id, user_id, 'stream', idx,
real_page_number=current_page))
username = self._og_search_title(webpage)
description = self._get_user_description(webpage)
return self.playlist_result(entries, user_id, username, description)