If you have problems using count_chars with a multibyte string, you can change the page encoding. Alternatively, you can also use this mb_count_chars version of the function. Basically it is mode "1" of the original function.

<?php
/**
 * Counts character occurences in a multibyte string
 * @param string $input UTF-8 data
 * @return array associative array of characters.
 */
function mb_count_chars($input) {
    $l = mb_strlen($input, 'UTF-8');
    $unique = array();
    for($i = 0; $i < $l; $i++) {
        $char = mb_substr($input, $i, 1, 'UTF-8');
        if(!array_key_exists($char, $unique))
            $unique[$char] = 0;
        $unique[$char]++;
    }
    return $unique;
}

$input = "Let's try some Greek letters: αααααΕεΙιΜμΨψ, Russian: ЙЙЫЫЩН, Czech: ěščřžýáíé";
print_r( mb_count_chars($input) ); 
//returns: Array ( [L] => 1 [e] => 7 [t] => 4 ['] => 1 [s] => 5 [ ] => 9 [r] => 3 [y] => 1 [o] => 1 [m] => 1 [G] => 1 [k] => 1 [l] => 1 [:] => 3 [α] => 5 [Ε] => 1 [ε] => 1 [Ι] => 1 [ι] => 1 [Μ] => 1 [μ] => 1 [Ψ] => 1 [ψ] => 1 [,] => 2 [R] => 1 [u] => 1 [i] => 1 [a] => 1 [n] => 1 [Й] => 2 [Ы] => 2 [Щ] => 1 [Н] => 1 [C] => 1 [z] => 1 [c] => 1 [h] => 1 [ě] => 1 [š] => 1 [č] => 1 [ř] => 1 [ž] => 1 [ý] => 1 [á] => 1 [í] => 1 [é] => 1 )
?>

Checking that two strings are anagram:

<?php

function isAnagram($string1, $string2)
{
  return count_chars($string1, 1) === count_chars($string2, 1);
}

isAnagram('act', 'cat'); // true

?>

<?php

// Require (n) unique characters in a string
// Modification of a function below which ads some flexibility in how many unique characters are required in a given string.

$pass = '123456' ; // true
$pass = '111222' ; // false

req_unique($pass,3);

function req_unique($string,$unique=3) {
    if ( count(count_chars($string,1)) < $unique) {
        echo 'false';
    }else{
        echo 'true';
    }
}

?>

After much trial and error trying to create a function that finds the number of unique characters in a string I same across count_chars() - my 20+ lines of useless code were wiped for this:

<?
function unichar($string) {
$two= strtolower(str_replace(' ', '', $string));
$res = count(count_chars($two, 1));
return $res;
}

/* examples :: */

echo unichar("bob"); // 2
echo unichar("Invisibility"); //8
echo unichar("The quick brown fox slyly jumped over the lazy dog"); //26

?>

I have no idea where this could be used, but it's quite fun

count_chars for multibyte supported.

<?php

function mb_count_chars ($string, $mode = 0) {
    
    $result =  array_fill(0, 256, 0);
    
    for ($i = 0, $size = mb_strlen($string); $i < $size; $i++) {
        $char = mb_substr($string, $i, 1);
        if (strlen($char) > 1) {
            continue;
        }
        
        $code = ord($char);
        if ($code >= 0 && $code <= 255) {
            $result[$code]++;
        }
    }
    
    switch ($mode) {
        case 1: // same as 0 but only byte-values with a frequency greater than zero are listed.
            foreach ($result as $key => $value) {
                if ($value == 0) {
                    unset($result[$key]);
                }
            }
            break;
        case 2: // same as 0 but only byte-values with a frequency equal to zero are listed.
            foreach ($result as $key => $value) {
                if ($value > 0) {
                    unset($result[$key]);
                }
            }
            break;
        case 3: // a string containing all unique characters is returned.
            $buildString = '';
            foreach ($result as $key => $value) {
                if ($value > 0) {
                    $buildString .= chr($key);
                }
            }
            return $buildString;
        case 4: // a string containing all not used characters is returned.
            $buildString = '';
            foreach ($result as $key => $value) {
                if ($value == 0) {
                    $buildString .= chr($key);
                }
            }
            return $buildString;
    }
    
    // change key names...
    foreach ($result as $key => $value) {
        $result[chr($key)] = $value;
        unset($result[$key]);
    }
    
    return $result;
    
}
?>

This function is great for input validation.  I frequently need to check that all characters in a string are 7-bit ASCII (and not null).  This is the fastest function I have found yet:

<?php
function is7bit($string) {
    // empty strings are 7-bit clean
    if (!strlen($string)) {
        return true;
    }
    // count_chars returns the characters in ascending octet order
    $str = count_chars($str, 3);
    // Check for null character
    if (!ord($str[0])) {
        return false;
    }
    // Check for 8-bit character
    if (ord($str[strlen($str)-1]) & 128) {
        return false;
    }
    return true;
}
?>

Here's a function to count number of strings in a string. It can be used as a simple utf8-enabled count_chars (but limited to a single mode)...

<?php
function utf8_count_strings($stringChar)
{
    $num = -1;
    $lenStringChar = strlen($stringChar);
    
    for ($lastPosition = 0; 
        $lastPosition !== false;
        $lastPosition = strpos($textSnippet, $stringChar, $lastPosition + $lenStringChar))
    {
        $num++;
    }
    
    return $num;
}
?>

// Usefulness of the two functions

<?php
 $string="aaabbc";

 // You just want to count the letter a
 $acount=substr_count($string,"a");

 // You want to count both letter a and letter b
 $counts=count_chars($string,0);
 $acount=$counts[ord("a")];
 $bcount=$counts[ord("b")];
?>