X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fmixcloud.py;h=bcac13ec5edfdeb83216e732b46ae34a36656205;hb=2f483bc1c389709623117079439708783122b5ec;hp=d4a4963eeb959d7d45f5faa8246d8089d5f82e46;hpb=dd91dfcd67a0d1db25836f734579742ff73b0c66;p=youtube-dl diff --git a/youtube_dl/extractor/mixcloud.py b/youtube_dl/extractor/mixcloud.py index d4a4963ee..bcac13ec5 100644 --- a/youtube_dl/extractor/mixcloud.py +++ b/youtube_dl/extractor/mixcloud.py @@ -1,27 +1,32 @@ 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, - parse_count, str_to_int, + try_get, + urljoin, ) class MixcloudIE(InfoExtractor): - _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/([^/]+)/(?!stream|uploads|favorites|listens|playlists)([^/]+)' + _VALID_URL = r'https?://(?:(?:www|beta|m)\.)?mixcloud\.com/([^/]+)/(?!stream|uploads|favorites|listens|playlists)([^/]+)' IE_NAME = 'mixcloud' _TESTS = [{ @@ -33,9 +38,8 @@ class MixcloudIE(InfoExtractor): 'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.', 'uploader': 'Daniel Holbach', 'uploader_id': 'dholbach', - 'thumbnail': 're:https?://.*\.jpg', + 'thumbnail': r're:https?://.*\.jpg', 'view_count': int, - 'like_count': int, }, }, { 'url': 'http://www.mixcloud.com/gillespeterson/caribou-7-inch-vinyl-mix-chat/', @@ -48,20 +52,18 @@ class MixcloudIE(InfoExtractor): 'uploader_id': 'gillespeterson', 'thumbnail': 're:https?://.*', 'view_count': int, - 'like_count': int, }, + }, { + 'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/', + '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) @@ -71,79 +73,169 @@ 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'] - - PREFIX = ( - r'm-play-on-spacebar[^>]+' - r'(?:\s+[a-zA-Z0-9-]+(?:="[^"]+")?)*?\s+') - title = self._html_search_regex( - PREFIX + r'm-title="([^"]+)"', webpage, 'title') - thumbnail = self._proto_relative_url(self._html_search_regex( - PREFIX + r'm-thumbnail-url="([^"]+)"', webpage, 'thumbnail', - fatal=False)) - uploader = self._html_search_regex( - PREFIX + 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) - like_count = parse_count(self._search_regex( - r'\bbutton-favorite[^>]+>.*?]+class=["\']toggle-number[^>]+>\s*([^<]+)', - webpage, 'like count', fatal=False)) - view_count = str_to_int(self._search_regex( - [r'([0-9,.]+)'], - webpage, 'play count', fatal=False)) + 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, 'uploader_id': uploader_id, 'view_count': view_count, - 'like_count': like_count, } class MixcloudPlaylistBaseIE(InfoExtractor): _PAGE_SIZE = 24 - def _fetch_tracks_page(self, path, video_id, page_name, current_page): - resp = self._download_webpage( + def _find_urls_in_page(self, page): + for url in re.findall(r'm-play-button m-url="(?P[^"]+)"', page): + 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): + real_page_number = real_page_number or current_page + 1 + return self._download_webpage( 'https://www.mixcloud.com/%s/' % path, video_id, note='Download %s (page %d)' % (page_name, current_page + 1), errnote='Unable to download %s' % page_name, - query={'page': (current_page + 1), 'list': 'main', '_ajax': '1'}, + query={'page': real_page_number, 'list': 'main', '_ajax': '1'}, headers={'X-Requested-With': 'XMLHttpRequest'}) - for url in re.findall(r'm-play-button m-url="(?P[^"]+)"', resp): - yield self.url_result( - compat_urlparse.urljoin('https://www.mixcloud.com', clean_html(url)), - MixcloudIE.ie_key()) + def _tracks_page_func(self, page, video_id, page_name, current_page): + resp = self._fetch_tracks_page(page, video_id, page_name, current_page) + + for item in self._find_urls_in_page(resp): + yield item def _get_user_description(self, page_content): return self._html_search_regex( - r']+class="description-text"[^>]*>(.+?)', + r']+class="profile-bio"[^>]*>(.+?)', page_content, 'user description', fatal=False) class MixcloudUserIE(MixcloudPlaylistBaseIE): - _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/(?P[^/]+)/(?Puploads|favorites|listens)?/?$' + _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P[^/]+)/(?Puploads|favorites|listens)?/?$' IE_NAME = 'mixcloud:user' _TESTS = [{ @@ -151,7 +243,7 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE): 'info_dict': { 'id': 'dholbach_uploads', 'title': 'Daniel Holbach (uploads)', - 'description': 'md5:327af72d1efeb404a8216c27240d1370', + 'description': 'md5:def36060ac8747b3aabca54924897e47', }, 'playlist_mincount': 11, }, { @@ -159,7 +251,7 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE): 'info_dict': { 'id': 'dholbach_uploads', 'title': 'Daniel Holbach (uploads)', - 'description': 'md5:327af72d1efeb404a8216c27240d1370', + 'description': 'md5:def36060ac8747b3aabca54924897e47', }, 'playlist_mincount': 11, }, { @@ -167,7 +259,7 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE): 'info_dict': { 'id': 'dholbach_favorites', 'title': 'Daniel Holbach (favorites)', - 'description': 'md5:327af72d1efeb404a8216c27240d1370', + 'description': 'md5:def36060ac8747b3aabca54924897e47', }, 'params': { 'playlist_items': '1-100', @@ -178,7 +270,7 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE): 'info_dict': { 'id': 'dholbach_listens', 'title': 'Daniel Holbach (listens)', - 'description': 'md5:327af72d1efeb404a8216c27240d1370', + 'description': 'md5:def36060ac8747b3aabca54924897e47', }, 'params': { 'playlist_items': '1-100', @@ -207,16 +299,16 @@ class MixcloudUserIE(MixcloudPlaylistBaseIE): entries = OnDemandPagedList( functools.partial( - self._fetch_tracks_page, + self._tracks_page_func, '%s/%s' % (user_id, list_type), video_id, 'list of %s' % list_type), - self._PAGE_SIZE, use_cache=True) + self._PAGE_SIZE) return self.playlist_result( entries, video_id, '%s (%s)' % (username, list_type), description) class MixcloudPlaylistIE(MixcloudPlaylistBaseIE): - _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/(?P[^/]+)/playlists/(?P[^/]+)/?$' + _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P[^/]+)/playlists/(?P[^/]+)/?$' IE_NAME = 'mixcloud:playlist' _TESTS = [{ @@ -229,12 +321,7 @@ class MixcloudPlaylistIE(MixcloudPlaylistBaseIE): 'playlist_mincount': 16, }, { 'url': 'https://www.mixcloud.com/maxvibes/playlists/jazzcat-on-ness-radio/', - 'info_dict': { - 'id': 'maxvibes_jazzcat-on-ness-radio', - 'title': 'Jazzcat on Ness Radio', - 'description': 'md5:7bbbf0d6359a0b8cda85224be0f8f263', - }, - 'playlist_mincount': 23 + 'only_matching': True, }] def _real_extract(self, url): @@ -243,20 +330,69 @@ class MixcloudPlaylistIE(MixcloudPlaylistBaseIE): playlist_id = mobj.group('playlist') video_id = '%s_%s' % (user_id, playlist_id) - profile = self._download_webpage( + webpage = self._download_webpage( url, user_id, note='Downloading playlist page', errnote='Unable to download playlist page') - description = self._get_user_description(profile) - playlist_title = self._html_search_regex( - r']+class="[^"]*list-playlist-title[^"]*"[^>]*>(.*?)', - profile, 'playlist title') + title = self._html_search_regex( + r']+class="parent active"[^>]*>\d+]*>([^<]+)', + webpage, 'playlist title', + default=None) or self._og_search_title(webpage, fatal=False) + description = self._get_user_description(webpage) entries = OnDemandPagedList( functools.partial( - self._fetch_tracks_page, + self._tracks_page_func, '%s/playlists/%s' % (user_id, playlist_id), video_id, 'tracklist'), self._PAGE_SIZE) - return self.playlist_result(entries, video_id, playlist_title, description) + return self.playlist_result(entries, video_id, title, description) + + +class MixcloudStreamIE(MixcloudPlaylistBaseIE): + _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P[^/]+)/stream/?$' + IE_NAME = 'mixcloud:stream' + + _TEST = { + 'url': 'https://www.mixcloud.com/FirstEar/stream/', + 'info_dict': { + 'id': 'FirstEar', + 'title': 'First Ear', + 'description': 'Curators of good music\nfirstearmusic.com', + }, + 'playlist_mincount': 192, + } + + def _real_extract(self, url): + user_id = self._match_id(url) + + 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)