From 070775dfd9e35446ec95da86433f0f28a5e08720 Mon Sep 17 00:00:00 2001
From: Linghong <ink29535@proton.me>
Date: Sat, 8 Aug 2026 16:30:26 +0000
Subject: [PATCH] =?UTF-8?q?=E9=9F=B3=E9=A2=91=E4=BF=A1=E6=81=AF=E9=A1=B5?=
 =?UTF-8?q?=20Task=20=E5=88=A4=E6=8D=AE=E6=94=B9=E7=94=A8=E8=B5=84?=
 =?UTF-8?q?=E6=BA=90=E5=8E=9F=E5=A7=8B=E7=B1=BB=E5=9E=8B?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

layer.assetKind 是 assetKindOverride ?? resourceAssetKind，用户在画布上改一次「素材标签」
就能翻转判据：背景音乐被改标成「音效」会换用完整 id，音效被改标成「背景音乐」则退回
formatTaskIdForDisplay 的截断假值——上一版修复被一次改标就废掉。

- 判据换成 resolveLayerResourceAssetKind；它在资源类型缺省时回退到 layer.assetKind，
  刚生成、尚未注册资源行的音效层同样命中
- soundEffect 那一支保留为补充判据：它和生成 action 一样长在 generationInputs 上，
  元数据整块缺失时会一并消失，不能当主判据
- 补两条回归：背景音乐改标为音效仍截断、音效改标为背景音乐仍显示完整 id

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
---
 .../ImageCanvasMetadataModalView.test.tsx     | 59 +++++++++++++++++++
 .../ImageCanvasMetadataModalView.tsx          | 11 ++--
 2 files changed, 66 insertions(+), 4 deletions(-)

diff --git a/src/components/image-editor/ImageCanvasMetadataModalView.test.tsx b/src/components/image-editor/ImageCanvasMetadataModalView.test.tsx
index cd1aaf265..2ca6814df 100644
--- a/src/components/image-editor/ImageCanvasMetadataModalView.test.tsx
+++ b/src/components/image-editor/ImageCanvasMetadataModalView.test.tsx
@@ -380,6 +380,65 @@ describe('ImageCanvasMetadataModalView', () => {
     expect(within(dialog).queryByText('1234')).toBeNull();
   });
 
+  it('keeps the task id display tied to the resource kind across asset tag overrides', () => {
+    // 中文注释：素材标签只改 assetKindOverride，资源原始类型才是展示口径的依据。背景音乐被
+    // 改标成「音效」不该换用完整 id，音效被改标成「背景音乐」更不能退回截断假值。
+    const { unmount } = render(
+      <ImageCanvasMetadataModalView
+        layer={createLayer({
+          title: '游戏背景音乐',
+          src: '/generated-character-drafts/editor-audios/music.wav',
+          mediaType: 'audio',
+          resourceAssetKind: 'background-music',
+          assetKindOverride: 'sound-effect',
+          assetKind: 'sound-effect',
+          originalWidth: 420,
+          originalHeight: 120,
+          model: 'suno',
+          taskId: 'audio-task-75',
+          generationInputs: {
+            fields: [{ title: '音乐描述', value: '紧张的地下城循环音乐' }],
+            references: [],
+          },
+        })}
+        onClose={vi.fn()}
+      />,
+    );
+
+    const musicDialog = screen.getByRole('dialog', { name: '音频信息' });
+    expect(within(musicDialog).getByText('75')).toBeTruthy();
+    expect(within(musicDialog).queryByText('audio-task-75')).toBeNull();
+    unmount();
+
+    render(
+      <ImageCanvasMetadataModalView
+        layer={createLayer({
+          title: '金币音效',
+          src: '/generated-character-drafts/editor-audios/coin.mp3',
+          mediaType: 'audio',
+          resourceAssetKind: 'sound-effect',
+          assetKindOverride: 'background-music',
+          assetKind: 'background-music',
+          originalWidth: 420,
+          originalHeight: 120,
+          model: 'eleven_text_to_sound_v2',
+          taskId: 'task-sfx-v2-1234-abcd',
+          generationInputs: {
+            fields: [{ title: '用户描述', value: '金币落地' }],
+            references: [],
+          },
+        })}
+        onClose={vi.fn()}
+      />,
+    );
+
+    const soundEffectDialog = screen.getByRole('dialog', { name: '音频信息' });
+    expect(
+      within(soundEffectDialog).getByText('task-sfx-v2-1234-abcd'),
+    ).toBeTruthy();
+    expect(within(soundEffectDialog).queryByText('1234')).toBeNull();
+  });
+
   it('does not render a dialog when no layer is selected', () => {
     render(<ImageCanvasMetadataModalView layer={null} onClose={vi.fn()} />);
 
diff --git a/src/components/image-editor/ImageCanvasMetadataModalView.tsx b/src/components/image-editor/ImageCanvasMetadataModalView.tsx
index 7a9237bd9..c3385dda8 100644
--- a/src/components/image-editor/ImageCanvasMetadataModalView.tsx
+++ b/src/components/image-editor/ImageCanvasMetadataModalView.tsx
@@ -1,4 +1,5 @@
 import { UnifiedModal } from '../common/UnifiedModal';
+import { resolveLayerResourceAssetKind } from './ImageCanvasEditorModel';
 import type { CanvasLayer } from './ImageCanvasEditorTypes';
 import { formatTaskIdForDisplay } from './ImageCanvasExportModel';
 import {
@@ -106,10 +107,12 @@ export function ImageCanvasMetadataModalView({
           <dt>Task</dt>
           <dd>
             {/* 中文注释：音效的 taskId 是 operation id 形态，`formatTaskIdForDisplay` 的
-                「取最后一段数字」会把它截成 `0` 这类误导值。判据取图层自身的 assetKind，
-                不再依赖 soundEffect 元数据是否解析成功——元数据一旦缺失，Task 这一栏不该
-                跟着降级成假值。 */}
-            {layer.assetKind === 'sound-effect' ||
+                「取最后一段数字」会把它截成 `0` 这类误导值。判据必须落在资源原始类型上：
+                `layer.assetKind` 是 `assetKindOverride ?? resourceAssetKind`，用户在画布上
+                改一次「素材标签」就能让音效层显示成截断假值、让背景音乐层改用完整 id。
+                soundEffect 那一支只是补充——它和 action 一样长在 generationInputs 上，元数据
+                整块缺失时会一并消失，而那正是这里要兜住的场景。 */}
+            {resolveLayerResourceAssetKind(layer) === 'sound-effect' ||
             layer.generationInputs?.soundEffect?.schemaVersion === 2
               ? layer.taskId || '-'
               : formatTaskIdForDisplay(layer.taskId)}