From ae87bea35cc11de1cae34f27e5b2df6bed8ccebf Mon Sep 17 00:00:00 2001
From: "METANEOCORTEX\\Kotti" <rainer.kottenhoff@gmail.com>
Date: Thu, 12 Mar 2026 01:49:14 +0100
Subject: [PATCH 1/2] fix: encoding detection bug

---
 src/EncodingDetection.cpp | 22 +++++++++-------------
 1 file changed, 9 insertions(+), 13 deletions(-)

diff --git a/src/EncodingDetection.cpp b/src/EncodingDetection.cpp
index c59929b18..18e35e3d7 100644
--- a/src/EncodingDetection.cpp
+++ b/src/EncodingDetection.cpp
@@ -1233,7 +1233,7 @@ extern "C" ENC_DET_T Encoding_DetectEncoding(const HPATHL hpath, const char* lpD
             Encoding_AnalyzeText(lpData, cbNbytes4Analysis, &encDetRes, iAnalyzeHint);
             // ---------------------------------------------------------------------------
         }
-        encDetRes.bPureASCII7Bit = (encDetRes.analyzedEncoding == CPI_ASCII_7BIT) || IsPureAscii7Bit(lpData, cbData);
+        encDetRes.bPureASCII7Bit = IsPureAscii7Bit(lpData, cbData);
 
         if (encDetRes.analyzedEncoding == CPI_NONE) {
             encDetRes.analyzedEncoding = iAnalyzeHint;
@@ -1297,22 +1297,17 @@ extern "C" ENC_DET_T Encoding_DetectEncoding(const HPATHL hpath, const char* lpD
     }
     else if (Encoding_IsValid(encDetRes.analyzedEncoding) && (encDetRes.bIsAnalysisReliable || !Settings.UseReliableCEDonly))
     {
-        if (!encDetRes.bIsAnalysisReliable && !Encoding_IsUTF8(encDetRes.analyzedEncoding) && encDetRes.bValidUTF8) {
-            encDetRes.Encoding = CPI_UTF8; // unreliable non-UTF-8 guess, but data is valid UTF-8
-        } else {
-            encDetRes.Encoding = encDetRes.analyzedEncoding;
-        }
+        encDetRes.Encoding = (encDetRes.analyzedEncoding == CPI_ASCII_7BIT) ? CPI_UTF8 : encDetRes.analyzedEncoding;
     }
-    else if (!encDetRes.bIsAnalysisReliable && (Encoding_IsValid(encDetRes.analyzedEncoding) || encDetRes.bPureASCII7Bit))
-    {
-        // UCHARDET below confidence threshold (UseReliableCEDonly is true)
-        encDetRes.Encoding = encDetRes.bValidUTF8 ? CPI_UTF8 : CPI_ANSI_DEFAULT;
-    }
-    else if (Encoding_IsUNICODE(encDetRes.unicodeAnalysis) && (iConfidence > 50))
+    else if (Encoding_IsUNICODE(encDetRes.unicodeAnalysis))
     {
         // unicodeAnalysis (IsTextUnicode) confirms Unicode structure,
         // iConfidence is from UCHARDET analysis — use analyzedEncoding (intentional)
-        encDetRes.Encoding = encDetRes.analyzedEncoding;
+        encDetRes.Encoding = Encoding_IsValid(encDetRes.analyzedEncoding) ? encDetRes.analyzedEncoding : encDetRes.unicodeAnalysis;
+    }
+    else if (encDetRes.bPureASCII7Bit) {
+        // UCHARDET below confidence threshold (UseReliableCEDonly is true)
+        encDetRes.Encoding = encDetRes.bValidUTF8 ? CPI_UTF8 : CPI_ANSI_DEFAULT;
     }
     else if (Encoding_IsValid(Encoding_SrcWeak(CPI_GET)))
     {
@@ -1323,6 +1318,7 @@ extern "C" ENC_DET_T Encoding_DetectEncoding(const HPATHL hpath, const char* lpD
         encDetRes.Encoding = iAnalyzeHint;
     }
 
+    // final check
     if (!Encoding_IsValid(encDetRes.Encoding)) {
         encDetRes.Encoding = CPI_PREFERRED_ENCODING;
     }

From 8d5a289dadcc4f1053667a37739058708f83fa15 Mon Sep 17 00:00:00 2001
From: "METANEOCORTEX\\Kotti" <rainer.kottenhoff@gmail.com>
Date: Thu, 12 Mar 2026 02:18:22 +0100
Subject: [PATCH 2/2] fix: revert ASCII as UTF-8 for unicode detection - false
 positives

---
 src/EncodingDetection.cpp | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/src/EncodingDetection.cpp b/src/EncodingDetection.cpp
index 18e35e3d7..460a50f04 100644
--- a/src/EncodingDetection.cpp
+++ b/src/EncodingDetection.cpp
@@ -1303,9 +1303,17 @@ extern "C" ENC_DET_T Encoding_DetectEncoding(const HPATHL hpath, const char* lpD
     {
         // unicodeAnalysis (IsTextUnicode) confirms Unicode structure,
         // iConfidence is from UCHARDET analysis — use analyzedEncoding (intentional)
-        encDetRes.Encoding = Encoding_IsValid(encDetRes.analyzedEncoding) ? encDetRes.analyzedEncoding : encDetRes.unicodeAnalysis;
+        if (Encoding_IsValid(encDetRes.analyzedEncoding)) {
+            encDetRes.Encoding = encDetRes.analyzedEncoding;
+        } 
+        //~else if ((encDetRes.analyzedEncoding == CPI_ASCII_7BIT) && encDetRes.bValidUTF8) {
+        //~    encDetRes.Encoding = CPI_UTF8;
+        //~}
+        else {
+            encDetRes.Encoding = encDetRes.unicodeAnalysis;
+        }
     }
-    else if (encDetRes.bPureASCII7Bit) {
+    else if (encDetRes.bPureASCII7Bit || (encDetRes.analyzedEncoding == CPI_ASCII_7BIT)) {
         // UCHARDET below confidence threshold (UseReliableCEDonly is true)
         encDetRes.Encoding = encDetRes.bValidUTF8 ? CPI_UTF8 : CPI_ANSI_DEFAULT;
     }