Question

I am stuck in comparing two unicode strings in PHP which both contain the special char ‘ö’. One string comes from $_GET, the other one is a filesystem’s folder name (scandir()). Both strings seem to be equal to me, making a

var_dump($filter);
var_dump($tail . '/' . $k);

on them also shows their equality but with different string lenghts (?!):

string '/blöb' (length=7)
string '/blöb' (length=6)

My snippet comparing them looks as follows:

if($filter == ($tail . '/' . $k)) {
    /* ... */
}

What’s going on here?

Additional information: $tail is an empty string:

string '' (length=0)

Answer 1

See here: http://en.wikipedia.org/wiki/Unicode_equivalence and use this: http://www.php.net/manual/en/class.normalizer.php

You probably have a decomposed character in the longer string, meaning an o and then a umlaut combining character which overlays the previous character.

The normalizer function will fix things like that.

As a side note you should always normalize your input if you are using it for equivalence (for example a username – you want to make sure two people don’t choose the same username, even if the binary representation of the string happens to be different).