Consider it done. The open-source AI agent that works out of the box · 想到,就能做到。开源、开箱即用的 AI Agent。
TypeScript agentai-agentai-assistantandroidclaude-codecodexelectroniosllmmacosreact-nativetypescriptwindows
test(media): align keyboard coverage with inline native controls
Chris092fc93b4570
diff --git a/apps/desktop/src/renderer/__tests__/chatVideoView.test.ts b/apps/desktop/src/renderer/__tests__/chatVideoView.test.tsindex af26c4542f..4d4853607c 100644--- a/apps/desktop/src/renderer/__tests__/chatVideoView.test.ts+++ b/apps/desktop/src/renderer/__tests__/chatVideoView.test.ts@@ -1,77 +1,58 @@// @vitest-environment jsdom-/**- * ChatVideoView.test.ts — 视频预览键盘交互契约。- *- * 与 chatImageView.test.ts 的图片键盘用例对位,但不 mock VideoLightbox:- * 真实组件的 onClose 排在 200ms fade-out 的 setTimeout 之后(见- * VideoLightbox.handleClose),同步调用 onClose 的 mock 覆盖不到这条时序。- * 用例覆盖:键盘打开、焦点进入 lightbox(FocusScope)并圈禁 Tab、Esc 延迟- * 关闭后焦点回到预览触发器——锁住整条无障碍路径,防止弹窗关闭时序或触发- * 器挂载逻辑调整时静默退化。- */+/** Tier 1: the message player now stays inline. Preserve keyboard access and+ * delegate native playback keys without reopening the superseded lightbox.+ * Native decoder/transport behavior is covered by the actual desktop acceptance;+ * this suite covers focus traversal and event ownership through the real view. */import React from 'react';-import { afterEach, describe, expect, it, vi } from 'vitest';-import { fireEvent, render, screen, waitFor } from '@testing-library/react';+import { afterEach, beforeEach, describe, expect, it, vi } from 'vitest';+import { cleanup, fireEvent, render, screen } from '@testing-library/react';+import userEvent from '@testing-library/user-event';vi.mock('react-i18next', () => ({useTranslation: () => ({ t: (key: string, fallback?: string) => fallback ?? key }),}));--vi.mock('@/hooks/useRemoteMediaUrl', () => ({- useRemoteMediaUrl: (url: string) => url,-}));+vi.mock('@/hooks/useRemoteMediaUrl', () => ({ useRemoteMediaUrl: (url: string) => url }));import { ChatVideoView } from '../components/chat/ChatVideoView';import { __resetMediaBusForTests } from '@/lib/mediaPlaybackBus';-describe('ChatVideoView 键盘交互(真实 VideoLightbox)', () => {+describe('ChatVideoView inline keyboard controls', () => {+ beforeEach(() => {+ vi.stubGlobal('IntersectionObserver', class { observe() {} disconnect() {} });+ vi.spyOn(HTMLMediaElement.prototype, 'pause').mockImplementation(() => {});+ });afterEach(() => {- // VideoLightbox 的 effect cleanup 会自行注销媒体元素,这里兜底清空- // 互斥总线,避免用例中途失败时把注册表泄漏给同 worker 的后续用例。+ cleanup();__resetMediaBusForTests();+ vi.restoreAllMocks();+ vi.unstubAllGlobals();});- it.each(['Enter', ' '])('键盘 %s 打开视频预览:焦点进入 lightbox 并圈禁,Esc 延迟关闭后回到触发器', async (key) => {- render(- React.createElement(ChatVideoView, {- src: 'xdt-video://control/clip.mp4',- filename: 'clip.mp4',- variant: 'tool-output',- }),- );- const trigger = screen.getByRole('button', { name: 'clip.mp4' });- trigger.focus();-- // 键盘路径:Enter/Space → currentTarget.click() → VideoLightbox。- fireEvent.keyDown(trigger, { key });- // 真实 VideoLightbox 经 portal 渲染到 document.body:backdrop 关闭按钮- // 在场,且带 controls 的大图 <video> 挂的是同一个 src。- const backdrop = screen.getByRole('button', { name: 'chat.lightbox.close' });- const lightboxVideo = document.body.querySelector('video[controls]');- expect(lightboxVideo).not.toBeNull();- const videoEl = lightboxVideo as HTMLVideoElement;- expect(videoEl.getAttribute('src')).toBe('xdt-video://control/clip.mp4');-- // 焦点进入 lightbox:打开后 FocusScope 把焦点给 <video>(Space 暂停 /- // 方向键 seek 立即可用),不再留在被遮罩挡住的预览触发器上。- expect(document.activeElement).toBe(videoEl);- expect(document.activeElement).not.toBe(trigger);-- // Tab 圈禁:焦点在 lightbox 内首尾循环(overlay DOM 顺序 backdrop 在前、- // video 在后),不漏到背后的聊天控件。- fireEvent.keyDown(videoEl, { key: 'Tab' });- expect(document.activeElement).toBe(backdrop);- fireEvent.keyDown(backdrop, { key: 'Tab', shiftKey: true });- expect(document.activeElement).toBe(videoEl);-- // Esc 关闭(onClose 排在 200ms 渐隐后,此刻焦点仍在 lightbox 内——同步- // onClose 的 mock 走不到这一步)。- fireEvent.keyDown(document, { key: 'Escape' });- expect(document.activeElement).not.toBe(trigger);-- await waitFor(() => expect(document.activeElement).toBe(trigger));- // 关闭完成后 lightbox 卸载,预览触发器仍在原位。+ it.each(['Enter', ' '])('keeps %s with the native player and lets Tab leave the message', async (key) => {+ const user = userEvent.setup();+ const view = render(React.createElement(React.Fragment, null,+ React.createElement('button', null, 'Before'),+ React.createElement(ChatVideoView, { src: 'xdt-video://control/clip.mp4', filename: 'clip.mp4', variant: 'tool-output' }),+ React.createElement('button', null, 'After'),+ ));+ screen.getByRole('button', { name: 'Before' }).focus();+ await user.tab();+ const video = view.container.querySelector('video')!;+ expect(document.activeElement).toBe(video);+ expect(video.controls).toBe(true);+ expect(video.getAttribute('src')).toBe('xdt-video://control/clip.mp4');++ const event = new KeyboardEvent('keydown', { key, bubbles: true, cancelable: true });+ fireEvent(video, event);+ expect(event.defaultPrevented).toBe(false);+ expect(document.activeElement).toBe(video);+ expect(view.container.querySelectorAll('video')).toHaveLength(1);+ expect(screen.queryByRole('dialog')).toBeNull();expect(screen.queryByRole('button', { name: 'chat.lightbox.close' })).toBeNull();- expect(screen.getByRole('button', { name: 'clip.mp4' })).toBeTruthy();++ await user.tab();+ expect(document.activeElement).toBe(screen.getByRole('button', { name: 'After' }));+ await user.tab({ shift: true });+ expect(document.activeElement).toBe(video);});});diff --git a/apps/desktop/src/renderer/components/chat/ChatVideoView.tsx b/apps/desktop/src/renderer/components/chat/ChatVideoView.tsxindex c1f2a0dd16..3634f905be 100644--- a/apps/desktop/src/renderer/components/chat/ChatVideoView.tsx+++ b/apps/desktop/src/renderer/components/chat/ChatVideoView.tsx@@ -122,6 +122,7 @@ export function ChatVideoView({ src, filename, variant, sessionId }: ChatVideoViref={videoRef}src={displaySrc}aria-label={filename}+ tabIndex={0}preload="metadata"controlsplaysInline