Merge branch 'MDL-84907-401' of https://github.com/HuongNV13/moodle into MOODLE_401_STABLE

backup/moodle2/restore_qtype_plugin.class.php

@@ -168,8 +168,7 @@ abstract class restore_qtype_plugin extends restore_plugin {
                 $this->questionanswercacheid = $newquestionid;
                 // Cache all cleaned answers for a simple text match.
                 foreach ($answers as $answer) {
-                    // MDL-30018: Clean in the same way as {@link xml_writer::xml_safe_utf8()}.
-                    $clean = preg_replace('/[\x-\x8\xb-\xc\xe-\x1f\x7f]/is','', $answer->answer); // Clean CTRL chars.
+                    $clean = core_text::trim_ctrl_chars($answer->answer); // Clean CTRL chars.
                     $clean = preg_replace("/\r\n|\r/", "\n", $clean); // Normalize line ending.
                     $this->questionanswercache[$clean] = $answer->id;
                 }

backup/moodle2/restore_stepslib.php

@@ -5154,8 +5154,7 @@ class restore_create_categories_and_questions extends restore_structure_step {
                 $potentialhints = $DB->get_records('question_hints',
                         array('questionid' => $newquestionid), '', 'id, hint');
                 foreach ($potentialhints as $potentialhint) {
-                    // Clean in the same way than {@link xml_writer::xml_safe_utf8()}.
-                    $cleanhint = preg_replace('/[\x-\x8\xb-\xc\xe-\x1f\x7f]/is','', $potentialhint->hint); // Clean CTRL chars.
+                    $cleanhint = core_text::trim_ctrl_chars($potentialhint->hint); // Clean CTRL chars.
                     $cleanhint = preg_replace("/\r\n|\r/", "\n", $cleanhint); // Normalize line ending.
                     if ($cleanhint === $data->hint) {
                         $newitemid = $data->id;

backup/util/xml/xml_writer.class.php

@@ -253,14 +253,14 @@ class xml_writer {
     }
 
     /**
-     * Perform some UTF-8 cleaning, stripping the control chars (\x0-\x1f)
-     * but tabs (\x9), newlines (\xa) and returns (\xd). The delete control
+     * Perform some UTF-8 cleaning, stripping the control chars (\x00-\x1f)
+     * but tabs (\x09), newlines (\xa) and returns (\xd). The delete control
      * char (\x7f) is also included. All them are forbiden in XML 1.0 specs.
      * The expression below seems to be UTF-8 safe too because it simply
      * ignores the rest of characters. Also normalize linefeeds and return chars.
      */
     protected function xml_safe_utf8($content) {
-        $content = preg_replace('/[\x-\x8\xb-\xc\xe-\x1f\x7f]/is', '', $content ?? ''); // clean CTRL chars.
+        $content = core_text::trim_ctrl_chars($content ?? '');
         $content = preg_replace("/\r\n|\r/", "\n", $content); // Normalize line&return=>line
         return fix_utf8($content);
     }

lib/classes/text.php

@@ -676,4 +676,16 @@ class core_text {
 
         return mb_convert_case($text, MB_CASE_TITLE, 'UTF-8');
     }
+
+    /**
+     * Trims control characters out of a string.
+     * Example: (\x00-\x1f) and (\x7f)
+     *
+     * @param string $text Input string
+     * @return string Cleaned string value
+     */
+    public static function trim_ctrl_chars(string $text): string {
+        // Remove control characters text.
+        return preg_replace('/[\x00-\x08\x0b-\x0c\x0e-\x1f\x7f]/i', '', $text);
+    }
 }

lib/upgrade.txt

@@ -1,6 +1,10 @@
 This files describes API changes in core libraries and APIs,
 information provided here is intended especially for developers.
 
+=== 4.1.18 ===
+* A new method, `core_text::trim_ctrl_chars()`, has been introduced to clean control characters from text.
+  This ensures cleaner input handling and prevents issues caused by invisible or non-printable characters
+
 === 4.1.12 ===
 
 * Added the ability for unit tests to autoload classes in the `\[component]\tests\` namespace from the `[path/to/component]/tests/classes` directory.

question/type/match/backup/moodle2/restore_qtype_match_plugin.class.php

@@ -157,13 +157,10 @@ class restore_qtype_match_plugin extends restore_qtype_plugin {
                 $this->questionsubcacheid = $newquestionid;
                 // Cache all cleaned answers and questiontext.
                 foreach ($potentialsubs as $potentialsub) {
-                    // Clean in the same way than {@link xml_writer::xml_safe_utf8()}.
-                    $cleanquestion = preg_replace('/[\x-\x8\xb-\xc\xe-\x1f\x7f]/is',
-                            '', $potentialsub->questiontext); // Clean CTRL chars.
+                    $cleanquestion = core_text::trim_ctrl_chars($potentialsub->questiontext); // Clean CTRL chars.
                     $cleanquestion = preg_replace("/\r\n|\r/", "\n", $cleanquestion); // Normalize line ending.
 
-                    $cleananswer = preg_replace('/[\x-\x8\xb-\xc\xe-\x1f\x7f]/is',
-                            '', $potentialsub->answertext); // Clean CTRL chars.
+                    $cleananswer = core_text::trim_ctrl_chars($potentialsub->answertext); // Clean CTRL chars.
                     $cleananswer = preg_replace("/\r\n|\r/", "\n", $cleananswer); // Normalize line ending.
 
                     $this->questionsubcache[$cleanquestion][$cleananswer] = $potentialsub->id;