From c557d4d96d36b29041437ad347d611b020f95267 Mon Sep 17 00:00:00 2001
From: Pierre-Olivier Mercier <nemunaire@nemunai.re>
Date: Sat, 1 Nov 2025 18:08:06 +0700
Subject: [PATCH 1/2] Fix calculateTextPlainConsistency algorithm

---
 pkg/analyzer/content.go | 45 ++++++++++++++++++++++++++++-------------
 1 file changed, 31 insertions(+), 14 deletions(-)

diff --git a/pkg/analyzer/content.go b/pkg/analyzer/content.go
index 87c423f..8f5d059 100644
--- a/pkg/analyzer/content.go
+++ b/pkg/analyzer/content.go
@@ -659,30 +659,47 @@ func (c *ContentAnalyzer) calculateTextPlainConsistency(plainText, htmlText stri
 		return 0.0
 	}
 
-	// Count common words
-	commonWords := 0
-	plainWordSet := make(map[string]bool)
+	// Count common words by building sets
+	plainWordSet := make(map[string]int)
 	for _, word := range plainWords {
-		plainWordSet[word] = true
+		plainWordSet[word]++
 	}
 
+	htmlWordSet := make(map[string]int)
 	for _, word := range htmlWords {
-		if plainWordSet[word] {
-			commonWords++
+		htmlWordSet[word]++
+	}
+
+	// Count matches: for each unique word, count minimum occurrences in both texts
+	commonWords := 0
+	for word, plainCount := range plainWordSet {
+		if htmlCount, exists := htmlWordSet[word]; exists {
+			// Count the minimum occurrences between both texts
+			if plainCount < htmlCount {
+				commonWords += plainCount
+			} else {
+				commonWords += htmlCount
+			}
 		}
 	}
 
-	// Calculate ratio (Jaccard similarity approximation)
-	maxWords := len(plainWords)
-	if len(htmlWords) > maxWords {
-		maxWords = len(htmlWords)
-	}
-
-	if maxWords == 0 {
+	// Calculate ratio using total words from both texts (union approach)
+	// This provides a balanced measure: perfect match = 1.0, partial overlap = 0.3-0.8
+	totalWords := len(plainWords) + len(htmlWords)
+	if totalWords == 0 {
 		return 0.0
 	}
 
-	return float32(commonWords) / float32(maxWords)
+	// Divide by average word count for better scoring
+	avgWords := float32(totalWords) / 2.0
+	ratio := float32(commonWords) / avgWords
+
+	// Cap at 1.0 for perfect matches
+	if ratio > 1.0 {
+		ratio = 1.0
+	}
+
+	return ratio
 }
 
 // normalizeText normalizes text for comparison
-- 
2.52.0


From 5395d1535c5f2748f967bdf17f3660dbcc17d6ab Mon Sep 17 00:00:00 2001
From: Renovate Bot <renovate@pomail.fr>
Date: Sat, 1 Nov 2025 13:10:36 +0000
Subject: [PATCH 2/2] Update dependency @types/node to v24.9.2

---
 web/package-lock.json | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/web/package-lock.json b/web/package-lock.json
index 01d6a6d..5aa6da8 100644
--- a/web/package-lock.json
+++ b/web/package-lock.json
@@ -1352,9 +1352,9 @@
             "license": "MIT"
         },
         "node_modules/@types/node": {
-            "version": "24.9.1",
-            "resolved": "https://registry.npmjs.org/@types/node/-/node-24.9.1.tgz",
-            "integrity": "sha512-QoiaXANRkSXK6p0Duvt56W208du4P9Uye9hWLWgGMDTEoKPhuenzNcC4vGUmrNkiOKTlIrBoyNQYNpSwfEZXSg==",
+            "version": "24.9.2",
+            "resolved": "https://registry.npmjs.org/@types/node/-/node-24.9.2.tgz",
+            "integrity": "sha512-uWN8YqxXxqFMX2RqGOrumsKeti4LlmIMIyV0lgut4jx7KQBcBiW6vkDtIBvHnHIquwNfJhk8v2OtmO8zXWHfPA==",
             "dev": true,
             "license": "MIT",
             "peer": true,
-- 
2.52.0