Re-encode submission source in UTF-8

Try to re-encode the submission source if it currently is not valid in
UTF-8. Add some extra safeguard by forcing another sanity check re-encode
from UTF-8 to the original encoding, which filters out binary blobs and
verifies that the re-encoding is non-destructive.

Author: Kevinjil

webapp/src/Controller/Jury/SubmissionController.php

@@ -920,7 +920,7 @@ public function sourceAction(
             $files[$f->getFilename()][$submitId] = [
                 'rank' => $f->getRank(),
                 'filename' => $f->getFilename(),
-                'source'   => mb_check_encoding($f->getSourcecode(), 'UTF-8') ? $f->getSourcecode() : "Could not display file as UTF-8, is it binary?",
+                'source'   => Utils::reencodeUtf8($f->getSourcecode()),
             ];
 
             // Keep track of the single filename within a submission for handling renaming.

webapp/src/Twig/TwigExtension.php

@@ -901,7 +901,7 @@ public function codeEditor(
 HTML;
         $rank   = $index;
         $id     = sprintf('editor%s', $rank);
-        $source = mb_check_encoding($code, 'UTF-8') ? $code : "Could not display file as UTF-8, is it binary?";
+        $source = Utils::reencodeUtf8($code);
         if ($elementToUpdate) {
             $extraForEdit = <<<JS
 editor.getModel().onDidChangeContent(() => {
@@ -951,7 +951,7 @@ public function getMonacoModel(SubmissionFile $file): string
         }
         $this->renderedSources[$file->getSubmitfileid()] = true;
 
-        $source = mb_check_encoding($file->getSourcecode(), 'UTF-8') ? $file->getSourcecode() : "Could not display file as UTF-8, is it binary?";
+        $source = Utils::reencodeUtf8($file->getSourcecode());
         return sprintf(
             <<<JS
 monaco.editor.createModel(

webapp/src/Utils/Utils.php

@@ -1056,6 +1056,23 @@ public static function jsonEncode(mixed $data): string
         return json_encode($data, JSON_PRESERVE_ZERO_FRACTION | JSON_UNESCAPED_SLASHES | JSON_THROW_ON_ERROR);
     }
 
+    public static function reencodeUtf8(string $source): string {
+        $detectOrder = array_unique(array_merge(mb_detect_order(), mb_list_encodings()));
+        $encoding = mb_detect_encoding($source, $detectOrder, strict: true);
+        if ($encoding !== false) {
+            $encoded = $encoding !== 'UTF-8' ? mb_convert_encoding($source, 'UTF-8', $encoding) : $source;
+            // Some binary files are strictly valid in an encoding but fail to re-encoding correctly.
+            // A successive strict call to mb_detect_encoding still says it is validly encoded, but rendering the string fails.
+            // This will filter these files.
+            $sanity = mb_convert_encoding($encoded, $encoding, 'UTF-8');
+            if ($source === $sanity) {
+                return $encoded;
+            }
+        }
+        return "Could not display file as UTF-8.\n"
+             . "Check the supported PHP mbstring encodings on the `Config checker` page if you did not expect this file to be binary.";
+    }
+
     /**
      * @return array<string, string>
      */