From 22ebbff7554126742e5934404e202f2a09bfeef9 Mon Sep 17 00:00:00 2001
From: Gigi <dergigi@pm.me>
Date: Tue, 14 Oct 2025 00:25:18 +0200
Subject: [PATCH] fix(highlights): add text validation before applying
 highlights

- Validate extracted range text matches search text before highlighting
- Check single-node matches are not empty or whitespace-only
- Compare both exact and normalized text to handle whitespace variations
- Prevent broken/corrupted highlights from being applied to DOM
- Add detailed logging for validation failures to aid debugging
---
 src/utils/highlightMatching/domUtils.ts | 25 +++++++++++++++++++++++++
 1 file changed, 25 insertions(+)

diff --git a/src/utils/highlightMatching/domUtils.ts b/src/utils/highlightMatching/domUtils.ts
index 00fe3454..b5755c29 100644
--- a/src/utils/highlightMatching/domUtils.ts
+++ b/src/utils/highlightMatching/domUtils.ts
@@ -74,6 +74,13 @@ export function tryMarkInTextNodes(
     const before = text.substring(0, actualIndex)
     const match = text.substring(actualIndex, actualIndex + searchText.length)
     const after = text.substring(actualIndex + searchText.length)
+    
+    // Validate the match makes sense (not just whitespace or empty)
+    if (!match || match.trim().length === 0) {
+      console.warn('Invalid match (empty or whitespace only)')
+      continue
+    }
+    
     const mark = createMarkElement(highlight, match, highlightStyle)
     
     replaceTextWithMark(textNode, before, after, mark)
@@ -202,6 +209,24 @@ function tryMultiNodeMatch(
       return false
     }
     
+    // Get the text content before extraction to verify it matches
+    const rangeText = range.toString()
+    const normalizedRangeText = normalizeWhitespace(rangeText)
+    const normalizedSearchText = normalizeWhitespace(searchText)
+    
+    // Validate that the extracted text matches what we're searching for
+    if (!rangeText.includes(searchText) && 
+        !normalizedRangeText.includes(normalizedSearchText) &&
+        normalizedRangeText !== normalizedSearchText) {
+      console.warn('Range text does not match search text:', {
+        rangeText: rangeText.substring(0, 100),
+        searchText: searchText.substring(0, 100),
+        rangeLength: rangeText.length,
+        searchLength: searchText.length
+      })
+      return false
+    }
+    
     // Extract the content from the range
     const extractedContent = range.extractContents()