From 3dfd720d098b4d49d69cfc77e6376f22bcd90934 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Xingchen=20Song=28=E5=AE=8B=E6=98=9F=E8=BE=B0=29?=
 <xingchensong1996@163.com>
Date: Sat, 14 Sep 2024 07:16:34 +0800
Subject: [PATCH] [ie/ximalaya] Add VIP support (#10832)

Closes #6928
Authored by: xingchensong, seproDev

Co-authored-by: sepro <4618135+seproDev@users.noreply.github.com>
---
 yt_dlp/extractor/ximalaya.py | 91 +++++++++++++++++++++++++++++++++---
 1 file changed, 85 insertions(+), 6 deletions(-)

diff --git a/yt_dlp/extractor/ximalaya.py b/yt_dlp/extractor/ximalaya.py
index e900a4ad9..d63964a00 100644
--- a/yt_dlp/extractor/ximalaya.py
+++ b/yt_dlp/extractor/ximalaya.py
@@ -1,7 +1,17 @@
+import base64
 import math
+import time
 
 from .common import InfoExtractor
-from ..utils import InAdvancePagedList, str_or_none, traverse_obj, try_call
+from .videa import VideaIE
+from ..utils import (
+    InAdvancePagedList,
+    int_or_none,
+    str_or_none,
+    traverse_obj,
+    try_call,
+    update_url_query,
+)
 
 
 class XimalayaBaseIE(InfoExtractor):
@@ -71,23 +81,92 @@ class XimalayaIE(XimalayaBaseIE):
                 'like_count': int,
             },
         },
+        {
+            # VIP-restricted audio
+            'url': 'https://www.ximalaya.com/sound/562111701',
+            'only_matching': True,
+        },
     ]
 
+    @staticmethod
+    def _decrypt_filename(file_id, seed):
+        cgstr = ''
+        key = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ/\\:._-1234567890'
+        for _ in key:
+            seed = float(int(211 * seed + 30031) % 65536)
+            r = int(seed / 65536 * len(key))
+            cgstr += key[r]
+            key = key.replace(key[r], '')
+        parts = file_id.split('*')
+        filename = ''.join(cgstr[int(part)] for part in parts if part.isdecimal())
+        if not filename.startswith('/'):
+            filename = '/' + filename
+        return filename
+
+    @staticmethod
+    def _decrypt_url_params(encrypted_params):
+        params = VideaIE.rc4(
+            base64.b64decode(encrypted_params), 'xkt3a41psizxrh9l').split('-')
+        # sign, token, timestamp
+        return params[1], params[2], params[3]
+
     def _real_extract(self, url):
         scheme = 'https' if url.startswith('https') else 'http'
 
         audio_id = self._match_id(url)
-        audio_info_file = f'{scheme}://m.ximalaya.com/tracks/{audio_id}.json'
         audio_info = self._download_json(
-            audio_info_file, audio_id,
-            f'Downloading info json {audio_info_file}', 'Unable to download info file')
+            f'{scheme}://m.ximalaya.com/tracks/{audio_id}.json', audio_id,
+            'Downloading info json', 'Unable to download info file')
 
-        formats = [{
+        formats = []
+        # NOTE: VIP-restricted audio
+        if audio_info.get('is_paid'):
+            ts = int(time.time())
+            vip_info = self._download_json(
+                f'{scheme}://mpay.ximalaya.com/mobile/track/pay/{audio_id}/{ts}',
+                audio_id, 'Downloading VIP info json', 'Unable to download VIP info file',
+                query={'device': 'pc', 'isBackend': 'true', '_': ts})
+            filename = self._decrypt_filename(vip_info['fileId'], vip_info['seed'])
+            sign, token, timestamp = self._decrypt_url_params(vip_info['ep'])
+            vip_url = update_url_query(
+                f'{vip_info["domain"]}/download/{vip_info["apiVersion"]}{filename}', {
+                    'sign': sign,
+                    'token': token,
+                    'timestamp': timestamp,
+                    'buy_key': vip_info['buyKey'],
+                    'duration': vip_info['duration'],
+                })
+            fmt = {
+                'format_id': 'vip',
+                'url': vip_url,
+                'vcodec': 'none',
+            }
+            if '_preview_' in vip_url:
+                self.report_warning(
+                    f'This tracks requires a VIP account. Using a sample instead. {self._login_hint()}')
+                fmt.update({
+                    'format_note': 'Sample',
+                    'preference': -10,
+                    **traverse_obj(vip_info, {
+                        'filesize': ('sampleLength', {int_or_none}),
+                        'duration': ('sampleDuration', {int_or_none}),
+                    }),
+                })
+            else:
+                fmt.update(traverse_obj(vip_info, {
+                    'filesize': ('totalLength', {int_or_none}),
+                    'duration': ('duration', {int_or_none}),
+                }))
+
+            fmt['abr'] = try_call(lambda: fmt['filesize'] * 8 / fmt['duration'] / 1024)
+            formats.append(fmt)
+
+        formats.extend([{
             'format_id': f'{bps}k',
             'url': audio_info[k],
             'abr': bps,
             'vcodec': 'none',
-        } for bps, k in ((24, 'play_path_32'), (64, 'play_path_64')) if audio_info.get(k)]
+        } for bps, k in ((24, 'play_path_32'), (64, 'play_path_64')) if audio_info.get(k)])
 
         thumbnails = []
         for k in audio_info: