X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fmixcloud.py;h=bcac13ec5edfdeb83216e732b46ae34a36656205;hb=56667d622c3f6e7594a04f8cd5f4371875940725;hp=769b68dc36e7813b45d3f6ddb655ef8a8ed28c19;hpb=dcaf00fb3eb716ea146a8e8870c7753b9acf67b8;p=youtube-dl

diff --git a/youtube_dl/extractor/mixcloud.py b/youtube_dl/extractor/mixcloud.py
index 769b68dc3..bcac13ec5 100644
--- a/youtube_dl/extractor/mixcloud.py
+++ b/youtube_dl/extractor/mixcloud.py
@@ -1,24 +1,32 @@
 from __future__ import unicode_literals
 
+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_urllib_request
+    compat_urlparse,
+    compat_zip
 )
 from ..utils import (
+    clean_html,
     ExtractorError,
-    HEADRequest,
-    NO_DEFAULT,
-    parse_count,
+    int_or_none,
+    OnDemandPagedList,
     str_to_int,
-    clean_html
+    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 = [{
@@ -30,9 +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/',
@@ -43,22 +50,20 @@ class MixcloudIE(InfoExtractor):
             'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
             'uploader': 'Gilles Peterson Worldwide',
             'uploader_id': 'gillespeterson',
-            'thumbnail': 're:https?://.*/images/',
+            '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,
     }]
 
-    def _check_url(self, url, track_id, ext):
-        try:
-            # We only want to know if the request succeed
-            # don't download the whole file
-            self._request_webpage(
-                HEADRequest(url), track_id,
-                'Trying %s URL' % ext)
-            return True
-        except ExtractorError:
-            return False
+    @staticmethod
+    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(k))
+            for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
@@ -68,246 +73,326 @@ 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'<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'],
+                    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)<div[^>]+class="global-message cloudcast-disabled-notice-light"[^>]*>(.+?)<(?:a|/div)',
             webpage, 'error message', default=None)
 
-        preview_url = self._search_regex(
-            r'\s(?:data-preview-url|m-preview)="([^"]+)"',
-            webpage, 'preview url', default=None if message else NO_DEFAULT)
-
-        if message:
-            raise ExtractorError('%s said: %s' % (self.IE_NAME, message), expected=True)
-
-        song_url = re.sub(r'audiocdn(\d+)', r'stream\1', preview_url)
-        song_url = song_url.replace('/previews/', '/c/originals/')
-        if not self._check_url(song_url, track_id, 'mp3'):
-            song_url = song_url.replace('.mp3', '.m4a').replace('originals/', 'm4a/64/')
-            if not self._check_url(song_url, track_id, 'm4a'):
-                raise ExtractorError('Unable to extract track 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[^>]+>.*?<span[^>]+class=["\']toggle-number[^>]+>\s*([^<]+)',
-            webpage, 'like count', fatal=False))
-        view_count = str_to_int(self._search_regex(
-            [r'<meta itemprop="interactionCount" content="UserPlays:([0-9]+)"',
-             r'/listeners/?">([0-9,.]+)</a>'],
-            webpage, 'play count', fatal=False))
+        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
+            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'<meta itemprop="interactionCount" content="UserPlays:([0-9]+)"',
+                 r'/listeners/?">([0-9,.]+)</a>',
+                 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 MixcloudUserIE(InfoExtractor):
-    """
-    Information extractor for Mixcloud users.
-    It can retrieve a list of a user's uploads, favorites or listens.
-    """
+class MixcloudPlaylistBaseIE(InfoExtractor):
+    _PAGE_SIZE = 24
+
+    def _find_urls_in_page(self, page):
+        for url in re.findall(r'm-play-button m-url="(?P<url>[^"]+)"', 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': real_page_number, 'list': 'main', '_ajax': '1'},
+            headers={'X-Requested-With': 'XMLHttpRequest'})
+
+    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'<div[^>]+class="profile-bio"[^>]*>(.+?)</div>',
+            page_content, 'user description', fatal=False)
+
 
-    _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/(?P<type>uploads|favorites|listens)?/?$'
+class MixcloudUserIE(MixcloudPlaylistBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/(?P<type>uploads|favorites|listens)?/?$'
     IE_NAME = 'mixcloud:user'
 
     _TESTS = [{
         'url': 'http://www.mixcloud.com/dholbach/',
         'info_dict': {
-            'id': 'dholbach/uploads',
+            'id': 'dholbach_uploads',
             'title': 'Daniel Holbach (uploads)',
-            'description': 'md5:327af72d1efeb404a8216c27240d1370',
+            'description': 'md5:def36060ac8747b3aabca54924897e47',
         },
-        'playlist_mincount': 11
+        'playlist_mincount': 11,
     }, {
         'url': 'http://www.mixcloud.com/dholbach/uploads/',
         'info_dict': {
-            'id': 'dholbach/uploads',
+            'id': 'dholbach_uploads',
             'title': 'Daniel Holbach (uploads)',
-            'description': 'md5:327af72d1efeb404a8216c27240d1370',
+            'description': 'md5:def36060ac8747b3aabca54924897e47',
         },
-        'playlist_mincount': 11
+        'playlist_mincount': 11,
     }, {
         'url': 'http://www.mixcloud.com/dholbach/favorites/',
         'info_dict': {
-            'id': 'dholbach/favorites',
+            'id': 'dholbach_favorites',
             'title': 'Daniel Holbach (favorites)',
-            'description': 'md5:327af72d1efeb404a8216c27240d1370',
+            'description': 'md5:def36060ac8747b3aabca54924897e47',
         },
-        'playlist_mincount': 244
+        'params': {
+            'playlist_items': '1-100',
+        },
+        'playlist_mincount': 100,
     }, {
         'url': 'http://www.mixcloud.com/dholbach/listens/',
         'info_dict': {
-            'id': 'dholbach/listens',
+            'id': 'dholbach_listens',
             'title': 'Daniel Holbach (listens)',
-            'description': 'md5:327af72d1efeb404a8216c27240d1370',
+            'description': 'md5:def36060ac8747b3aabca54924897e47',
+        },
+        'params': {
+            'playlist_items': '1-100',
         },
-        'playlist_mincount': 846
+        'playlist_mincount': 100,
     }]
 
-    def _fetch_tracks(self, base_url, video_id, dl_note=None, dl_errnote=None):
-        # retrieve all fragments of a list of tracks with fake AJAX calls
-        track_urls = []
-        current_page = 1
-        while True:
-            # fake a AJAX request to retrieve a list fragment
-            page_url = base_url + "?page=%d&list=main&_ajax=1" % current_page
-            req = compat_urllib_request.Request(page_url, headers={"X-Requested-With": "XMLHttpRequest"})
-            resp = self._download_webpage(req, video_id, note=dl_note + " (page %d)" % current_page, errnote=dl_errnote)
-
-            # extract all track URLs from fragment
-            urls = re.findall(r'm-play-button m-url="(?P<url>[^"]+)"', resp)
-            # clean up URLs
-            urls = map(clean_html, urls)
-            # create absolute URLs
-            urls = map(lambda u: "https://www.mixcloud.com" + u, urls)
-            track_urls.extend(urls)
-
-            # advance to next fragment, if any
-            if " m-next-page-url=" in resp:
-                current_page += 1
-            else:
-                break
-
-        return track_urls
-
-    def _handle_track_urls(self, urls):
-        return map(lambda u: self.url_result(u, "Mixcloud"), urls)
-
-    def _get_user_description(self, page_content):
-        return self._html_search_regex(
-            r'<div class="description-text">.*?<p>(?P<description>.*?)</p></div></div></div>',
-            page_content,
-            "user description",
-            group="description",
-            fatal=False,
-            default="")
-
-    def _get_username(self, page_content):
-        return self._og_search_title(page_content)
-
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
-        user_id = mobj.group("user")
-        list_type = mobj.group("type")
+        user_id = mobj.group('user')
+        list_type = mobj.group('type')
 
         # if only a profile URL was supplied, default to download all uploads
         if list_type is None:
-            list_type = "uploads"
+            list_type = 'uploads'
 
-        video_id = "%s/%s" % (user_id, list_type)
+        video_id = '%s_%s' % (user_id, list_type)
 
-        # download the user's profile to retrieve some metadata
-        profile = self._download_webpage("https://www.mixcloud.com/%s/" % user_id,
-                                         video_id,
-                                         note="Downloading user profile",
-                                         errnote="Unable to download user profile")
+        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._get_username(profile)
+        username = self._og_search_title(profile)
         description = self._get_user_description(profile)
 
-        # retrieve all page fragments of uploads, favorites or listens
-        track_urls = self._fetch_tracks(
-            "https://www.mixcloud.com/%s/%s/" % (user_id, list_type),
-            video_id,
-            dl_note="Downloading list of %s" % list_type,
-            dl_errnote="Unable to download list of %s" % list_type)
+        entries = OnDemandPagedList(
+            functools.partial(
+                self._tracks_page_func,
+                '%s/%s' % (user_id, list_type), video_id, 'list of %s' % list_type),
+            self._PAGE_SIZE)
 
-        # let MixcloudIE handle each track URL
-        entries = self._handle_track_urls(track_urls)
-
-        return {
-            '_type': 'playlist',
-            'entries': entries,
-            'title': "%s (%s)" % (username, list_type),
-            'id': video_id,
-            "description": description
-        }
+        return self.playlist_result(
+            entries, video_id, '%s (%s)' % (username, list_type), description)
 
 
-class MixcloudPlaylistIE(MixcloudUserIE):
-    """
-    Information extractor for Mixcloud playlists.
-    """
-
-    _VALID_URL = r'^(?:https?://)?(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/playlists/(?P<playlist>[^/]+)/?$'
+class MixcloudPlaylistIE(MixcloudPlaylistBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?mixcloud\.com/(?P<user>[^/]+)/playlists/(?P<playlist>[^/]+)/?$'
     IE_NAME = 'mixcloud:playlist'
 
     _TESTS = [{
         'url': 'https://www.mixcloud.com/RedBullThre3style/playlists/tokyo-finalists-2015/',
         'info_dict': {
-            'id': 'RedBullThre3style/playlists/tokyo-finalists-2015',
+            'id': 'RedBullThre3style_tokyo-finalists-2015',
             'title': 'National Champions 2015',
             'description': 'md5:6ff5fb01ac76a31abc9b3939c16243a3',
         },
-        'playlist_mincount': 16
+        'playlist_mincount': 16,
     }, {
         'url': 'https://www.mixcloud.com/maxvibes/playlists/jazzcat-on-ness-radio/',
-        'info_dict': {
-            'id': 'maxvibes/playlists/jazzcat-on-ness-radio',
-            'title': 'Jazzcat on Ness Radio',
-            'description': 'md5:7bbbf0d6359a0b8cda85224be0f8f263',
-        },
-        'playlist_mincount': 23
+        'only_matching': True,
     }]
 
-    def _get_playlist_title(self, page_content):
-        return self._html_search_regex(
-            r'<span class="main-list-title list-playlist-title ">(?P<title>.*?)</span>',
-            page_content,
-            "playlist title",
-            group="title",
-            fatal=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/playlists/%s" % (user_id, playlist_id)
+        user_id = mobj.group('user')
+        playlist_id = mobj.group('playlist')
+        video_id = '%s_%s' % (user_id, playlist_id)
 
-        # download the playlist page to retrieve some metadata
-        profile = self._download_webpage(url,
-                                         user_id,
-                                         note="Downloading playlist page",
-                                         errnote="Unable to download playlist page")
+        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._get_playlist_title(profile)
+        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)
 
-        # retrieve all page fragments of playlist
-        track_urls = self._fetch_tracks(
-            "https://www.mixcloud.com/%s/playlists/%s/" % (user_id, playlist_id),
-            video_id,
-            dl_note="Downloading tracklist of %s" % playlist_title,
-            dl_errnote="Unable to tracklist of %s" % playlist_title)
+        entries = OnDemandPagedList(
+            functools.partial(
+                self._tracks_page_func,
+                '%s/playlists/%s' % (user_id, playlist_id), video_id, 'tracklist'),
+            self._PAGE_SIZE)
 
-        # let MixcloudIE handle each track
-        entries = self._handle_track_urls(track_urls)
+        return self.playlist_result(entries, video_id, title, description)
 
-        return {
-            '_type': 'playlist',
-            'entries': entries,
-            'title': playlist_title,
-            'id': video_id,
-            "description": 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': {
+            '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)