fix: add UTF-8 guard to PatternDriver and use original text in PhoneticDriver matches

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: deemonic

src/Drivers/PatternDriver.php

@@ -18,6 +18,10 @@ public function detect(string $text, Dictionary $dictionary, MaskStrategyInterfa
             return new Result($text ?? '', $text ?? '', [], 0);
         }
 
+        if (!mb_check_encoding($text, 'UTF-8')) {
+            $text = mb_convert_encoding($text, 'UTF-8', 'UTF-8');
+        }
+
         $matchedWords = [];
         $lowerText = mb_strtolower($text, 'UTF-8');
         $profanities = $dictionary->getProfanities();

src/Drivers/PhoneticDriver.php

@@ -90,8 +90,10 @@ public function detect(string $text, Dictionary $dictionary, MaskStrategyInterfa
                 continue;
             }
 
+            $originalWord = mb_substr($text, $start, $length);
+
             $matchedWords[] = new MatchedWord(
-                text: $word,
+                text: $originalWord,
                 base: $baseWord,
                 severity: $dictionary->getSeverity($baseWord),
                 position: $start,