Germans have some special characters like ä, ü, ö and ß. They aren't very search engine friendly, so I want to replace them in the file names of uploading catalog images. That sounds not that complicated, I know... but:

Responsible for actions like that is the method getCorrectFileName() in the class Varien_File_Uploader, which looks like this:

 * Correct filename with special chars and spaces
 * @param string $fileName
 * @return string
public function getCorrectFileName($fileName)
    $fileName = preg_replace('/[^a-z0-9_\\-\\.]+/i', '_', $fileName);
    $fileInfo = pathinfo($fileName);

    if (preg_match('/^_+$/', $fileInfo['filename'])) {
        $fileName = 'file.' . $fileInfo['extension'];
    return $fileName;

We could simply extend it with strtr() or str_replace() or even preg_replace() before the original preg_replace() which could looks like this:

$fileName = strtr($fileName, [
    'Ä' => 'Ae',
    'ä' => 'ae',
    'Ö' => 'Oe',
    'Ü' => 'Ue',
    'ü' => 'ue',
    'ß' => 'ss',

But no matter which of these functions I use, it ignores the special characters in the file name. When I'm overriding $fileName at the beginning with a string like "täst-file.jpg" everything works fine..

My first idea was an encoding issue, so I tried to convert the string to UTF-8, but mb_detect_encoding($fileName, 'UTF-8', true) said it was already a valid UTF-8 encoded string and I ran out of ideas..

Does anyone know what I'm doing wrong or what a solution for this problem is?

Thanks in advance. :)


1 Answer 1


Okay, here's a solution...

All common ways to replace chars in a string like str_replace(), preg_replace(), strtr(), ... did NOT work in this special case.

I don't know exactly, which encoding (server, php, etc.) is responsible for not detecting the special chars, but converting the chars to ASCII finally work.

After splitting the filename via str_split(), I recognized that I got a sequence of 3 values for every german special char - "ä" (a, Ì, ˆ) or "Ü" (U, Ì, ˆ), so my solution may looks a bit weird:

public function getCorrectFileName($fileName)

    $fileInfo   = pathinfo($fileName);
    $split      = str_split($fileName);
    $splitCount = count($split);
    $fileName   = '';
    for ($i = 0; $i < $splitCount; $i++) {
        if (isset($split[$i]) == false || $split[$i] == null) {
        } else if (in_array(ord($split[$i]), [65, 79, 85, 97, 111, 117]) && ord($split[$i+1]) == 204 && ord($split[$i+2]) == 136) {
            $fileName .= chr(ord($split[$i])) . 'e';
        } else if (ord($split[$i]) == 195 && ord($split[$i+1]) == 159) {
            $fileName .= 'ss';
        } else {
            $fileName .= chr(ord($split[$i]));
    // @HACK END
    $fileName = preg_replace('/[^a-z0-9_\\-\\.]+/i', '_', $fileName);

    if (preg_match('/^_+$/', $fileInfo['filename'])) {
        $fileName = 'file.' . $fileInfo['extension'];
    return $fileName;

I think this solution could also work for special chars in other languages. Just split the string and see via ord() which ASCII values you get...

I hope that helps.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.