[CORE-7942] 2.4.y.z/description-wysiwyg-fix-trunk

docs/fixes/CORE-7942-whitespace-fix.md

@@ -0,0 +1,128 @@
+# CORE-7942: Whitespace Stripping Fix in Language Detection
+
+## Problem Summary
+
+When displaying HTML content with the `detectLanguage` pipe in description components, spaces between inline elements (like `<a>` tags) and adjacent text were being stripped, causing text to run together.
+
+### Example
+**API Response:**
+```html
+<a href="...">Privacy Policy</a> click here
+```
+
+**Rendered Output (BEFORE FIX):**
+```
+Privacy Policyclick here
+```
+(no space between "Policy" and "click")
+
+**Expected Output (AFTER FIX):**
+```
+Privacy Policy click here
+```
+(space preserved)
+
+## Root Cause Analysis
+
+The issue was in the `addLanguageAttributes()` method in [utils.service.ts](projects/v3/src/app/services/utils.service.ts#L902-L945), which is called by the `detectLanguage` pipe for WCAG 3.1.2 Language of Parts compliance.
+
+### Technical Details
+
+When processing HTML content:
+1. The DOM parser creates separate text nodes for text outside and between elements
+2. For `<a>Privacy Policy</a> click here`, there are 2 text nodes:
+   - Text node 1 (inside `<a>`): `"Privacy Policy"`
+   - Text node 2 (after `<a>`): `" click here"` (with leading space)
+
+### The Bug (Lines 919-925)
+
+```typescript
+const text = node.textContent?.trim() || '';  // ❌ Strips whitespace
+if (text.length >= 10) {
+  const detectedLang = this.detectLanguage(text, baseLang);
+  if (detectedLang) {
+    const span = this.document.createElement('span');
+    span.setAttribute('lang', detectedLang);
+    span.textContent = text;  // ❌ Uses trimmed text, losing leading/trailing spaces
+    node.parentNode?.replaceChild(span, node);
+  }
+}
+```
+
+**What Happened:**
+- Line 919: `node.textContent?.trim()` converted `" click here"` → `"click here"`
+- Line 921: Length check: `"click here".length = 10` ✅ passes threshold
+- Line 924: Assigned trimmed text to new span element
+- Line 925: Replaced original node (with space) with span (without space)
+
+**Result:** Space between "Policy" and "click" disappeared
+
+## Solution Implemented
+
+### 1. Preserve Original Whitespace
+
+Changed line 924 to use `node.textContent` (original) instead of `text` (trimmed):
+
+```typescript
+const text = node.textContent?.trim() || '';  // ✅ Use for validation only
+if (text.length >= 20) {
+  const detectedLang = this.detectLanguage(text, baseLang);
+  if (detectedLang) {
+    const span = this.document.createElement('span');
+    span.setAttribute('lang', detectedLang);
+    span.textContent = node.textContent;  // ✅ Preserve original whitespace
+    node.parentNode?.replaceChild(span, node);
+  }
+}
+```
+
+**Strategy:**
+- Use `text` (trimmed) for validation and language detection logic
+- Use `node.textContent` (original with whitespace) for rendering
+
+### 2. Increase Minimum Length Threshold
+
+Changed minimum length from **10** to **20** characters:
+
+```typescript
+// minimum text length for reliable detection (increased from 10 to 20 for better accuracy)
+const minLength = 20;
+```
+
+**Benefits:**
+1. **Better Accuracy**: Language detection algorithms are more reliable with longer text
+2. **Avoid False Positives**: Short phrases like "click here" (10 chars) won't trigger detection
+3. **Performance**: Reduces unnecessary processing of short text nodes
+4. **WCAG Intent**: WCAG 3.1.2 is meant for substantial foreign language content, not individual words
+
+## Impact Assessment
+
+### Fixed Components
+All components using the `detectLanguage` pipe will benefit:
+- [description.component.html](projects/v3/src/app/components/description/description.component.html)
+- [activity-desktop.component.html](projects/v3/src/app/desktop/activity-desktop/activity-desktop.component.html)
+- [review-desktop.component.html](projects/v3/src/app/desktop/review-desktop/review-desktop.component.html)
+
+### Edge Cases Considered
+- **Text < 20 chars**: Not processed (preserves original spacing by default)
+- **Multiple spaces**: Preserved exactly as in original HTML
+- **Non-breaking spaces (`&nbsp;`)**: Preserved as HTML entities
+- **Mixed content**: Works correctly with inline elements
+
+## Testing Recommendations
+
+1. **Visual Test**: Verify spacing appears correctly in description content with links
+2. **Language Detection**: Confirm foreign language passages (>20 chars) still get `lang` attributes
+3. **Short Text**: Verify short phrases (<20 chars) don't get incorrectly wrapped in `<span lang="...">`
+4. **Regression**: Test existing descriptions with mixed English/foreign content
+
+## Files Modified
+
+- `/projects/v3/src/app/services/utils.service.ts`
+  - Line 849: Increased `minLength` from 10 to 20
+  - Lines 918-926: Added comments and fixed whitespace preservation in `addLanguageAttributes()`
+
+## Related Tickets
+
+- CORE-7942: Description WYSIWYG spacing fix
+- Original accessibility implementation for WCAG 3.1.2 compliance

projects/v3/src/app/services/utils.service.ts

@@ -847,8 +847,8 @@ export class UtilsService {
       return null;
     }
 
-    // Minimum text length for reliable detection
-    const minLength = 10;
+    // minimum text length for reliable detection (increased from 10 to 20 for better accuracy)
+    const minLength = 20;
     const cleanText = text.trim().replace(/<[^>]*>/g, ' ').replace(/\s+/g, ' ');
 
     if (cleanText.length < minLength) {
@@ -915,13 +915,15 @@ export class UtilsService {
 
       const processNode = (node: Node): void => {
         if (node.nodeType === Node.TEXT_NODE) {
+          // use trimmed text for validation only, preserve original whitespace for rendering
           const text = node.textContent?.trim() || '';
-          if (text.length >= 10) {
+          if (text.length >= 20) {
             const detectedLang = this.detectLanguage(text, baseLang);
             if (detectedLang) {
               const span = this.document.createElement('span');
               span.setAttribute('lang', detectedLang);
-              span.textContent = text;
+              // preserve original whitespace from node.textContent (not trimmed)
+              span.textContent = node.textContent;
               node.parentNode?.replaceChild(span, node);
             }
           }