Root directory had accumulated audit/exploration markdown files cluttering the project root. Moved all audit-related files to docs/audits/ with a README.md index, and updated cross-references in K6_LOAD_TESTING_GUIDE.md and README_FRONTEND_DOCS.md. Co-Authored-By: Paperclip <noreply@paperclip.ing>
8.9 KiB
Accessibility Code Fixes - Detailed Implementation Guide
Fix #1: File Upload Input aria-label
File: apps/web/components/listings/image-upload.tsx
Line: 118
Before
<input
ref={inputRef}
type="file"
accept="image/jpeg,image/png,image/webp"
multiple
className="hidden"
onChange={(e) => {
if (e.target.files) addFiles(e.target.files);
e.target.value = '';
}}
/>
After
<input
ref={inputRef}
type="file"
accept="image/jpeg,image/png,image/webp"
multiple
className="hidden"
aria-label="Chọn ảnh để tải lên"
onChange={(e) => {
if (e.target.files) addFiles(e.target.files);
e.target.value = '';
}}
/>
Why: Hidden inputs need aria-label so screen readers can announce their purpose when focused.
Fix #2: Search Dialog Input aria-label
File: apps/web/app/[locale]/(public)/search/page.tsx
Line: 189
Before
<input
type="text"
value={saveName}
onChange={(e) => setSaveName(e.target.value)}
placeholder="Tên tìm kiếm (VD: Chung cư Q7 dưới 3 tỷ)"
className="mb-3 w-full rounded-md border bg-background px-3 py-2 text-sm outline-none focus:ring-2 focus:ring-primary"
maxLength={100}
onKeyDown={(e) => e.key === 'Enter' && handleSaveSearch()}
/>
After
<input
type="text"
value={saveName}
onChange={(e) => setSaveName(e.target.value)}
placeholder="Tên tìm kiếm (VD: Chung cư Q7 dưới 3 tỷ)"
aria-label="Tên bộ lọc tìm kiếm"
className="mb-3 w-full rounded-md border bg-background px-3 py-2 text-sm outline-none focus:ring-2 focus:ring-primary"
maxLength={100}
onKeyDown={(e) => e.key === 'Enter' && handleSaveSearch()}
/>
Why: Text inputs need aria-label when no associated label element exists. Placeholder is not a substitute for aria-label.
Fix #3: Admin Moderation - Select All Checkbox
File: apps/web/app/[locale]/(admin)/admin/moderation/page.tsx
Line: 222
Before
<TableHead className="w-10">
<input
type="checkbox"
checked={selected.size === result.data.length && result.data.length > 0}
onChange={toggleSelectAll}
className="rounded border-input"
/>
</TableHead>
After
<TableHead className="w-10">
<input
type="checkbox"
aria-label="Chọn tất cả tin đăng"
checked={selected.size === result.data.length && result.data.length > 0}
onChange={toggleSelectAll}
className="rounded border-input"
/>
</TableHead>
Why: Checkbox in table headers need aria-label to distinguish them from row checkboxes.
Fix #4: Admin Moderation - Row Checkboxes
File: apps/web/app/[locale]/(admin)/admin/moderation/page.tsx
Line: 242
Before
<TableCell>
<input
type="checkbox"
checked={selected.has(item.listingId)}
onChange={() => toggleSelect(item.listingId)}
className="rounded border-input"
/>
</TableCell>
After (Option 1 - Simple)
<TableCell>
<input
type="checkbox"
aria-label={`Chọn tin đăng: ${item.listingId}`}
checked={selected.has(item.listingId)}
onChange={() => toggleSelect(item.listingId)}
className="rounded border-input"
/>
</TableCell>
After (Option 2 - Better with title)
<TableCell>
<input
type="checkbox"
aria-label={`Chọn tin đăng: ${item.title || item.listingId}`}
checked={selected.has(item.listingId)}
onChange={() => toggleSelect(item.listingId)}
className="rounded border-input"
/>
</TableCell>
Why: Each checkbox needs unique aria-label that includes context about what listing it represents.
Fix #5: Test Mock Image Component
File: apps/web/app/[locale]/(public)/search/__tests__/search.spec.tsx
Line: 46
Before
vi.mock('next/image', () => ({
default: (props: Record<string, unknown>) => <img {...props} />,
}));
After (Option 1 - Simple)
vi.mock('next/image', () => ({
default: (props: Record<string, unknown>) => <img {...props} alt={props.alt || ''} />,
}));
After (Option 2 - With Warning)
vi.mock('next/image', () => ({
default: (props: Record<string, unknown>) => {
if (!props.alt) {
console.warn('Image mock: Missing alt attribute', props);
}
return <img {...props} alt={props.alt || 'image'} />;
},
}));
Why: Mock should enforce alt attribute to catch missing alts in tests before production.
Fix #6 (Enhancement): Image Upload Drag-Drop Accessibility
File: apps/web/components/listings/image-upload.tsx
Lines: 86-128
Current Code
<div
onDragOver={handleDragOver}
onDragLeave={handleDragLeave}
onDrop={handleDrop}
onClick={() => inputRef.current?.click()}
className={cn(
'flex cursor-pointer flex-col items-center justify-center rounded-lg border-2 border-dashed p-8 transition-colors',
isDragging
? 'border-primary bg-primary/5'
: 'border-muted-foreground/25 hover:border-primary/50',
)}
>
<svg>...</svg>
<p className="text-sm font-medium">Kéo thả ảnh vào đây hoặc nhấp để chọn</p>
<p className="mt-1 text-xs text-muted-foreground">
JPG, PNG, WebP - Tối đa {maxFiles} ảnh, mỗi ảnh 10MB
</p>
<input
ref={inputRef}
type="file"
accept="image/jpeg,image/png,image/webp"
multiple
className="hidden"
onChange={(e) => {
if (e.target.files) addFiles(e.target.files);
e.target.value = '';
}}
/>
</div>
Enhanced Code
<div
onDragOver={handleDragOver}
onDragLeave={handleDragLeave}
onDrop={handleDrop}
onClick={() => inputRef.current?.click()}
role="button"
tabIndex={0}
aria-label="Khu vực kéo thả hoặc nhấp để tải ảnh lên"
onKeyDown={(e) => {
if (e.key === 'Enter' || e.key === ' ') {
e.preventDefault();
inputRef.current?.click();
}
}}
className={cn(
'flex cursor-pointer flex-col items-center justify-center rounded-lg border-2 border-dashed p-8 transition-colors focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-primary focus-visible:ring-offset-2',
isDragging
? 'border-primary bg-primary/5'
: 'border-muted-foreground/25 hover:border-primary/50',
)}
>
<svg>...</svg>
<p className="text-sm font-medium">Kéo thả ảnh vào đây hoặc nhấp để chọn</p>
<p className="mt-1 text-xs text-muted-foreground">
JPG, PNG, WebP - Tối đa {maxFiles} ảnh, mỗi ảnh 10MB
</p>
<input
ref={inputRef}
type="file"
accept="image/jpeg,image/png,image/webp"
multiple
className="hidden"
aria-label="Chọn ảnh để tải lên"
onChange={(e) => {
if (e.target.files) addFiles(e.target.files);
e.target.value = '';
}}
/>
</div>
Changes:
role="button"- Identifies div as an interactive buttontabIndex={0}- Makes div keyboard accessiblearia-label- Describes the purpose to screen readersonKeyDownhandler - Allows Enter/Space to activatefocus-visiblestyles - Shows focus indicator for keyboard navigation
Why: Makes drag-drop area fully keyboard accessible for users who can't use a mouse.
Testing After Implementation
1. Screen Reader Testing
# Use VoiceOver (Mac), NVDA (Windows), or JAWS
# Navigate to each fixed element and verify:
# - Input is announced with its aria-label
# - Checkbox is announced with its aria-label
# - Purpose is clear from screen reader announcement
2. Keyboard Navigation Testing
# Tab through the page
# Verify:
# - All interactive elements are reachable via Tab
# - Focus is visible on all elements
# - Enter/Space activates buttons and checkboxes
# - Image upload drag area is focused and can be activated with keyboard
3. Automated Testing
# Run axe
npm run test:a11y
# Or use Lighthouse
npx lighthouse https://localhost:3000 --view
# ESLint JSX Accessibility Plugin should catch these issues:
npm run lint
4. Visual Testing
# Verify with browser dev tools:
# - Inspect each input to confirm aria-label attribute exists
# - Check for proper focus styles
# - Verify focus ring colors meet contrast requirements
Summary of Changes
| Issue | File | Line | Type | Severity |
|---|---|---|---|---|
| File input missing aria-label | image-upload.tsx | 118 | aria-label | HIGH |
| Search input missing aria-label | search/page.tsx | 189 | aria-label | HIGH |
| Header checkbox missing aria-label | moderation/page.tsx | 222 | aria-label | HIGH |
| Row checkboxes missing aria-label | moderation/page.tsx | 242 | aria-label | HIGH |
| Mock Image missing alt | search.spec.tsx | 46 | alt attribute | MEDIUM |
| Drag-drop area not keyboard accessible | image-upload.tsx | 86-128 | enhancement | MEDIUM |
Estimated Implementation Time
- Fix #1: 2 minutes
- Fix #2: 2 minutes
- Fix #3: 2 minutes
- Fix #4: 3 minutes (need to find item.title in context)
- Fix #5: 2 minutes
- Fix #6: 10 minutes
- Testing: 15-20 minutes
Total: ~35-45 minutes