Relax allowed characters in file names

Allows all unicode letters and all unicode numbers instead of only a-z, A-Z, 0-9.

Signed-off-by: Maximilian Krög <maxi_kroeg@web.de>

Author: MoonE

libraries/classes/Sanitize.php

@@ -247,14 +247,8 @@ static function (array $match): string {
      */
     public static function sanitizeFilename($filename, $replaceDots = false)
     {
-        $pattern = '/[^A-Za-z0-9_';
-        // if we don't have to replace dots
-        if (! $replaceDots) {
-            // then add the dot to the list of legit characters
-            $pattern .= '.';
-        }
-
-        $pattern .= '-]/';
+        // Keep only numbers (N), letters (L), dash, underbar, and maybe dot
+        $pattern = '/[^\p{N}\p{L}' . ($replaceDots ? '' : '.') . '_-]/u';
         $filename = preg_replace($pattern, '_', $filename);
 
         return $filename;

test/classes/SanitizeTest.php

@@ -6,6 +6,10 @@
 
 use PhpMyAdmin\Sanitize;
 
+use function implode;
+use function range;
+use function str_repeat;
+
 /**
  * @covers \PhpMyAdmin\Sanitize
  */
@@ -158,10 +162,30 @@ public function testEscape(): void
 
     /**
      * Test for Sanitize::sanitizeFilename
+     *
+     * @dataProvider providerTestSanitizeFileName
      */
-    public function testSanitizeFilename(): void
+    public function testSanitizeFilename(string $expected, string $input, bool $replaceDot): void
     {
-        self::assertSame('File_name_123', Sanitize::sanitizeFilename('File_name 123'));
+        self::assertSame($expected, Sanitize::sanitizeFilename($input, $replaceDot));
+    }
+
+    /** @psalm-return array{string,string,bool} */
+    public static function providerTestSanitizeFileName(): array
+    {
+        return [
+            ['Hello123', 'Hello123', false],
+            ['宮保雞丁', '宮保雞丁', false],
+            ['Україна', 'Україна', false],
+            ['-_-', '-.-', true],
+            ['-.-', '-.-', false],
+            ['___', '"\'"', false],
+            ['_test_', '<test>', false],
+            ['Hello__World_', "Hello\r\nWorld!", false],
+            ['_', "\u{fffd}", false],
+            ['_', '🚀', false],
+            [str_repeat('_', 32), implode('', range("\0", "\x1f")), false],
+        ];
     }
 
     /**