From 1722c55400ff30bb5aee5dd7a262f0b7e9ce2f0e Mon Sep 17 00:00:00 2001
From: doe1080 <98906116+doe1080@users.noreply.github.com>
Date: Fri, 13 Jun 2025 08:25:08 +0900
Subject: [PATCH] [ie/hypergryph] Improve metadata extraction (#13415)

Closes #13384
Authored by: doe1080, eason1478

Co-authored-by: eason1478 <134664337+eason1478@users.noreply.github.com>
---
 yt_dlp/extractor/hypergryph.py | 56 +++++++++++++++++++++++++++-------
 1 file changed, 45 insertions(+), 11 deletions(-)

diff --git a/yt_dlp/extractor/hypergryph.py b/yt_dlp/extractor/hypergryph.py
index 1fb2e9a98..f405d14b5 100644
--- a/yt_dlp/extractor/hypergryph.py
+++ b/yt_dlp/extractor/hypergryph.py
@@ -1,32 +1,66 @@
 from .common import InfoExtractor
-from ..utils import js_to_json, traverse_obj
+from ..utils import (
+    ExtractorError,
+    clean_html,
+    url_or_none,
+)
+from ..utils.traversal import subs_list_to_dict, traverse_obj
 
 
 class MonsterSirenHypergryphMusicIE(InfoExtractor):
+    IE_NAME = 'monstersiren'
+    IE_DESC = '塞壬唱片'
+    _API_BASE = 'https://monster-siren.hypergryph.com/api'
     _VALID_URL = r'https?://monster-siren\.hypergryph\.com/music/(?P<id>\d+)'
     _TESTS = [{
         'url': 'https://monster-siren.hypergryph.com/music/514562',
         'info_dict': {
             'id': '514562',
             'ext': 'wav',
-            'artists': ['塞壬唱片-MSR'],
-            'album': 'Flame Shadow',
             'title': 'Flame Shadow',
+            'album': 'Flame Shadow',
+            'artists': ['塞壬唱片-MSR'],
+            'description': 'md5:19e2acfcd1b65b41b29e8079ab948053',
+            'thumbnail': r're:https?://web\.hycdn\.cn/siren/pic/.+\.jpg',
+        },
+    }, {
+        'url': 'https://monster-siren.hypergryph.com/music/514518',
+        'info_dict': {
+            'id': '514518',
+            'ext': 'wav',
+            'title': 'Heavenly Me (Instrumental)',
+            'album': 'Heavenly Me',
+            'artists': ['塞壬唱片-MSR', 'AIYUE blessed : 理名'],
+            'description': 'md5:ce790b41c932d1ad72eb791d1d8ae598',
+            'thumbnail': r're:https?://web\.hycdn\.cn/siren/pic/.+\.jpg',
         },
     }]
 
     def _real_extract(self, url):
         audio_id = self._match_id(url)
-        webpage = self._download_webpage(url, audio_id)
-        json_data = self._search_json(
-            r'window\.g_initialProps\s*=', webpage, 'data', audio_id, transform_source=js_to_json)
+        song = self._download_json(f'{self._API_BASE}/song/{audio_id}', audio_id)
+        if traverse_obj(song, 'code') != 0:
+            msg = traverse_obj(song, ('msg', {str}, filter))
+            raise ExtractorError(
+                msg or 'API returned an error response', expected=bool(msg))
+
+        album = None
+        if album_id := traverse_obj(song, ('data', 'albumCid', {str})):
+            album = self._download_json(
+                f'{self._API_BASE}/album/{album_id}/detail', album_id, fatal=False)
 
         return {
             'id': audio_id,
-            'title': traverse_obj(json_data, ('player', 'songDetail', 'name')),
-            'url': traverse_obj(json_data, ('player', 'songDetail', 'sourceUrl')),
-            'ext': 'wav',
             'vcodec': 'none',
-            'artists': traverse_obj(json_data, ('player', 'songDetail', 'artists', ...)),
-            'album': traverse_obj(json_data, ('musicPlay', 'albumDetail', 'name')),
+            **traverse_obj(song, ('data', {
+                'title': ('name', {str}),
+                'artists': ('artists', ..., {str}),
+                'subtitles': ({'url': 'lyricUrl'}, all, {subs_list_to_dict(lang='en')}),
+                'url': ('sourceUrl', {url_or_none}),
+            })),
+            **traverse_obj(album, ('data', {
+                'album': ('name', {str}),
+                'description': ('intro', {clean_html}),
+                'thumbnail': ('coverUrl', {url_or_none}),
+            })),
         }