X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fyandexmusic.py;h=456f95f698662d03bef59429ccc671b73b6a8913;hb=067aa17edf5a46a8cbc4d6b90864eddf051fa2bc;hp=b47aecb15a4def8543934c9fa2df495fbe3e3bc3;hpb=602814adab97d605d1fe84d6439c33325dba15c7;p=youtube-dl

diff --git a/youtube_dl/extractor/yandexmusic.py b/youtube_dl/extractor/yandexmusic.py
index b47aecb15..456f95f69 100644
--- a/youtube_dl/extractor/yandexmusic.py
+++ b/youtube_dl/extractor/yandexmusic.py
@@ -1,4 +1,4 @@
-# coding=utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
@@ -7,18 +7,80 @@ import hashlib
 from .common import InfoExtractor
 from ..compat import compat_str
 from ..utils import (
+    ExtractorError,
     int_or_none,
     float_or_none,
 )
 
 
 class YandexMusicBaseIE(InfoExtractor):
+    @staticmethod
+    def _handle_error(response):
+        if isinstance(response, dict):
+            error = response.get('error')
+            if error:
+                raise ExtractorError(error, expected=True)
+            if response.get('type') == 'captcha' or 'captcha' in response:
+                YandexMusicBaseIE._raise_captcha()
+
+    @staticmethod
+    def _raise_captcha():
+        raise ExtractorError(
+            'YandexMusic has considered youtube-dl requests automated and '
+            'asks you to solve a CAPTCHA. You can either wait for some '
+            'time until unblocked and optionally use --sleep-interval '
+            'in future or alternatively you can go to https://music.yandex.ru/ '
+            'solve CAPTCHA, then export cookies and pass cookie file to '
+            'youtube-dl with --cookies',
+            expected=True)
+
+    def _download_webpage_handle(self, *args, **kwargs):
+        webpage = super(YandexMusicBaseIE, self)._download_webpage_handle(*args, **kwargs)
+        if 'Нам очень жаль, но Ð·Ð°Ð¿Ñ€Ð¾ÑÑ‹, поступившие с Ð²Ð°ÑˆÐµÐ³Ð¾ IP-адреса, похожи на Ð°Ð²Ñ‚оматические.' in webpage:
+            self._raise_captcha()
+        return webpage
+
+    def _download_json(self, *args, **kwargs):
+        response = super(YandexMusicBaseIE, self)._download_json(*args, **kwargs)
+        self._handle_error(response)
+        return response
+
+
+class YandexMusicTrackIE(YandexMusicBaseIE):
+    IE_NAME = 'yandexmusic:track'
+    IE_DESC = 'Яндекс.Музыка - Трек'
+    _VALID_URL = r'https?://music\.yandex\.(?:ru|kz|ua|by)/album/(?P<album_id>\d+)/track/(?P<id>\d+)'
+
+    _TEST = {
+        'url': 'http://music.yandex.ru/album/540508/track/4878838',
+        'md5': 'f496818aa2f60b6c0062980d2e00dc20',
+        'info_dict': {
+            'id': '4878838',
+            'ext': 'mp3',
+            'title': 'Carlo Ambrosio, Carlo Ambrosio & Fabio Di Bari - Gypsy Eyes 1',
+            'filesize': 4628061,
+            'duration': 193.04,
+            'track': 'Gypsy Eyes 1',
+            'album': 'Gypsy Soul',
+            'album_artist': 'Carlo Ambrosio',
+            'artist': 'Carlo Ambrosio, Carlo Ambrosio & Fabio Di Bari',
+            'release_year': 2009,
+        },
+        'skip': 'Travis CI servers blocked by YandexMusic',
+    }
+
     def _get_track_url(self, storage_dir, track_id):
         data = self._download_json(
             'http://music.yandex.ru/api/v1.5/handlers/api-jsonp.jsx?action=getTrackSrc&p=download-info/%s'
             % storage_dir,
             track_id, 'Downloading track location JSON')
 
+        # Each string is now wrapped in a list, this is probably only temporarily thus
+        # supporting both scenarios (see https://github.com/ytdl-org/youtube-dl/issues/10193)
+        for k, v in data.items():
+            if v and isinstance(v, list):
+                data[k] = v[0]
+
         key = hashlib.md5(('XGRlBW9FXlekgbPrRHuSiA' + data['path'][1:] + data['s']).encode('utf-8')).hexdigest()
         storage = storage_dir.split('.')
 
@@ -26,32 +88,50 @@ class YandexMusicBaseIE(InfoExtractor):
                 % (data['host'], key, data['ts'] + data['path'], storage[1]))
 
     def _get_track_info(self, track):
-        return {
+        thumbnail = None
+        cover_uri = track.get('albums', [{}])[0].get('coverUri')
+        if cover_uri:
+            thumbnail = cover_uri.replace('%%', 'orig')
+            if not thumbnail.startswith('http'):
+                thumbnail = 'http://' + thumbnail
+
+        track_title = track['title']
+        track_info = {
             'id': track['id'],
             'ext': 'mp3',
             'url': self._get_track_url(track['storageDir'], track['id']),
-            'title': '%s - %s' % (track['artists'][0]['name'], track['title']),
             'filesize': int_or_none(track.get('fileSize')),
             'duration': float_or_none(track.get('durationMs'), 1000),
+            'thumbnail': thumbnail,
+            'track': track_title,
         }
 
+        def extract_artist(artist_list):
+            if artist_list and isinstance(artist_list, list):
+                artists_names = [a['name'] for a in artist_list if a.get('name')]
+                if artists_names:
+                    return ', '.join(artists_names)
 
-class YandexMusicTrackIE(YandexMusicBaseIE):
-    IE_NAME = 'yandexmusic:track'
-    IE_DESC = 'Яндекс.Музыка - Трек'
-    _VALID_URL = r'https?://music\.yandex\.ru/album/(?P<album_id>\d+)/track/(?P<id>\d+)'
+        albums = track.get('albums')
+        if albums and isinstance(albums, list):
+            album = albums[0]
+            if isinstance(album, dict):
+                year = album.get('year')
+                track_info.update({
+                    'album': album.get('title'),
+                    'album_artist': extract_artist(album.get('artists')),
+                    'release_year': int_or_none(year),
+                })
 
-    _TEST = {
-        'url': 'http://music.yandex.ru/album/540508/track/4878838',
-        'md5': 'f496818aa2f60b6c0062980d2e00dc20',
-        'info_dict': {
-            'id': '4878838',
-            'ext': 'mp3',
-            'title': 'Carlo Ambrosio - Gypsy Eyes 1',
-            'filesize': 4628061,
-            'duration': 193.04,
-        }
-    }
+        track_artist = extract_artist(track.get('artists'))
+        if track_artist:
+            track_info.update({
+                'artist': track_artist,
+                'title': '%s - %s' % (track_artist, track_title),
+            })
+        else:
+            track_info['title'] = track_title
+        return track_info
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
@@ -64,10 +144,18 @@ class YandexMusicTrackIE(YandexMusicBaseIE):
         return self._get_track_info(track)
 
 
-class YandexMusicAlbumIE(YandexMusicBaseIE):
+class YandexMusicPlaylistBaseIE(YandexMusicBaseIE):
+    def _build_playlist(self, tracks):
+        return [
+            self.url_result(
+                'http://music.yandex.ru/album/%s/track/%s' % (track['albums'][0]['id'], track['id']))
+            for track in tracks if track.get('albums') and isinstance(track.get('albums'), list)]
+
+
+class YandexMusicAlbumIE(YandexMusicPlaylistBaseIE):
     IE_NAME = 'yandexmusic:album'
     IE_DESC = 'Яндекс.Музыка - Альбом'
-    _VALID_URL = r'https?://music\.yandex\.ru/album/(?P<id>\d+)/?(\?|$)'
+    _VALID_URL = r'https?://music\.yandex\.(?:ru|kz|ua|by)/album/(?P<id>\d+)/?(\?|$)'
 
     _TEST = {
         'url': 'http://music.yandex.ru/album/540508',
@@ -76,6 +164,7 @@ class YandexMusicAlbumIE(YandexMusicBaseIE):
             'title': 'Carlo Ambrosio - Gypsy Soul (2009)',
         },
         'playlist_count': 50,
+        'skip': 'Travis CI servers blocked by YandexMusic',
     }
 
     def _real_extract(self, url):
@@ -85,7 +174,7 @@ class YandexMusicAlbumIE(YandexMusicBaseIE):
             'http://music.yandex.ru/handlers/album.jsx?album=%s' % album_id,
             album_id, 'Downloading album JSON')
 
-        entries = [self._get_track_info(track) for track in album['volumes'][0]]
+        entries = self._build_playlist(album['volumes'][0])
 
         title = '%s - %s' % (album['artists'][0]['name'], album['title'])
         year = album.get('year')
@@ -95,12 +184,12 @@ class YandexMusicAlbumIE(YandexMusicBaseIE):
         return self.playlist_result(entries, compat_str(album['id']), title)
 
 
-class YandexMusicPlaylistIE(YandexMusicBaseIE):
+class YandexMusicPlaylistIE(YandexMusicPlaylistBaseIE):
     IE_NAME = 'yandexmusic:playlist'
     IE_DESC = 'Яндекс.Музыка - Плейлист'
-    _VALID_URL = r'https?://music\.yandex\.ru/users/[^/]+/playlists/(?P<id>\d+)'
+    _VALID_URL = r'https?://music\.yandex\.(?P<tld>ru|kz|ua|by)/users/(?P<user>[^/]+)/playlists/(?P<id>\d+)'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://music.yandex.ru/users/music.partners/playlists/1245',
         'info_dict': {
             'id': '1245',
@@ -108,20 +197,74 @@ class YandexMusicPlaylistIE(YandexMusicBaseIE):
             'description': 'md5:3b9f27b0efbe53f2ee1e844d07155cc9',
         },
         'playlist_count': 6,
-    }
+        'skip': 'Travis CI servers blocked by YandexMusic',
+    }, {
+        # playlist exceeding the limit of 150 tracks shipped with webpage (see
+        # https://github.com/ytdl-org/youtube-dl/issues/6666)
+        'url': 'https://music.yandex.ru/users/ya.playlist/playlists/1036',
+        'info_dict': {
+            'id': '1036',
+            'title': 'Музыка 90-х',
+        },
+        'playlist_mincount': 300,
+        'skip': 'Travis CI servers blocked by YandexMusic',
+    }]
 
     def _real_extract(self, url):
-        playlist_id = self._match_id(url)
+        mobj = re.match(self._VALID_URL, url)
+        tld = mobj.group('tld')
+        user = mobj.group('user')
+        playlist_id = mobj.group('id')
 
-        webpage = self._download_webpage(url, playlist_id)
+        playlist = self._download_json(
+            'https://music.yandex.%s/handlers/playlist.jsx' % tld,
+            playlist_id, 'Downloading missing tracks JSON',
+            fatal=False,
+            headers={
+                'Referer': url,
+                'X-Requested-With': 'XMLHttpRequest',
+                'X-Retpath-Y': url,
+            },
+            query={
+                'owner': user,
+                'kinds': playlist_id,
+                'light': 'true',
+                'lang': tld,
+                'external-domain': 'music.yandex.%s' % tld,
+                'overembed': 'false',
+            })['playlist']
 
-        playlist = self._parse_json(
-            self._search_regex(
-                r'var\s+Mu\s*=\s*({.+?});\s*</script>', webpage, 'player'),
-            playlist_id)['pageData']['playlist']
+        tracks = playlist['tracks']
+        track_ids = [compat_str(track_id) for track_id in playlist['trackIds']]
 
-        entries = [self._get_track_info(track) for track in playlist['tracks']]
+        # tracks dictionary shipped with playlist.jsx API is limited to 150 tracks,
+        # missing tracks should be retrieved manually.
+        if len(tracks) < len(track_ids):
+            present_track_ids = set([
+                compat_str(track['id'])
+                for track in tracks if track.get('id')])
+            missing_track_ids = [
+                track_id for track_id in track_ids
+                if track_id not in present_track_ids]
+            missing_tracks = self._download_json(
+                'https://music.yandex.%s/handlers/track-entries.jsx' % tld,
+                playlist_id, 'Downloading missing tracks JSON',
+                fatal=False,
+                headers={
+                    'Referer': url,
+                    'X-Requested-With': 'XMLHttpRequest',
+                },
+                query={
+                    'entries': ','.join(missing_track_ids),
+                    'lang': tld,
+                    'external-domain': 'music.yandex.%s' % tld,
+                    'overembed': 'false',
+                    'strict': 'true',
+                })
+            if missing_tracks:
+                tracks.extend(missing_tracks)
 
         return self.playlist_result(
-            entries, compat_str(playlist_id),
-            playlist['title'], playlist.get('description'))
+            self._build_playlist(tracks),
+            compat_str(playlist_id),
+            playlist.get('title'), playlist.get('description'))