Merge pull request cypht-org#1133 from kambereBr/unicode-domdocument-…

…html [ENH] Improve Unicode Handling in DOMDocument HTML Loading
Yannick243 · Aug 7, 2024 · 7ee1ef4 · 7ee1ef4
2 parents a6ea5b6 + 904727e
commit 7ee1ef4
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/modules/core/message_functions.php b/modules/core/message_functions.php
@@ -449,7 +449,8 @@ class HTMLToText {
 
     function __construct($html) {
         $doc = new DOMDocument();
-        $doc->loadHTML(html_entity_decode($html, ENT_QUOTES | ENT_HTML5, 'UTF-8'));
+        $html = mb_convert_encoding($html, 'HTML-ENTITIES', 'UTF-8');
+        $doc->loadHTML(htmlentities($html, ENT_QUOTES, 'UTF-8'));
         if (trim($html) && $doc->hasChildNodes()) {
             $this->parse_nodes($doc->childNodes);
         }