Question

I have the following possible strings:

NL DE 
NL,DE
nl DE
nl de
NL/DE
NL,mismatch,DE

I’m looking for the preg_match that produces the following output given the inputs above.

array(
  [0]=>"NL",
  [1]=>"DE"
);

I’ve tried the following code:

preg_match_all('/(\w{2,2})/ui', $info["country"], $m);

but that seems to also cut up the word mismatch, which is undesired.

The regex should only match two letter country codes, everything else should be ignored.

How can I do this using preg_match in PHP?

Answer 1

// @claudrian Variant
function SplitCountries($string){
    // Sanity check
    if(!is_string($string)) return false;
    // Split string by non-letters (case insensitive)
    $slices = preg_split('~[^a-z]+~i', trim($string));
    // Keep only 2-letter words
    $slices = preg_grep('~^[a-z]{2}$~i', $slices);
    // Keep uniques
    $slices = array_unique($slices);
    // Done
    return $slices;
}

// @Wiseguy Variant
function SplitCountries($string){
    // Sanity check
    if(!is_string($string)) return false;
    // Capture only two letter packs
    if(!preg_match_all('~\\b[a-z]{2}\\b~i', trim($string), $slices)){
        return false;
    }
    // Keep uniques
    $slices = array_unique($slices[0]);
    // Done
    return $slices;
}

Hope it helps.