Question

I want to replace all non-numeric characters in a string except a +.

I’ve tried this question, but it’s not working…

Here is what I have at the moment:

$nn = preg_replace("/([^0-9\/+]+)/", "", $string );

It works 100%, except for removing the + in any way…

EDIT

My input will always contain only 1 +, and should there be more, they should be removed.

Basically, if a user enters a phone number as [phone] it should be returned as [phone] (South African Country Code – I add the +27 at a later stage) But if [phone] is entered, [phone] should be returned.

Answer 1

You should be able to do it with the following regex:

[^0-9+]

In preg_replace():

$nn = preg_replace('/[^0-9+]/', '', $string);

Your current regex also keeps a forward-slash, so to keep that functionality:

$nn = preg_replace('/[^0-9\/+]/', '', $string);

Sample code with output:

<?php
$string = '[phone]';
$nn = preg_replace("/[^0-9+]/", "", $string );
echo $nn . "\n";
?>

Results in:

[phone]

UPDATE (keep only first matching +)
Per your latest question-update, you also only want to keep the first + symbol found. To do this, since there may not be a “rule” regarding the location of the first symbol (such as “it has to be the first character in the string), I would suggest using additional methods other than just preg_replace():

$nn = preg_replace("/[^0-9+]/", "", $string);
if (substr_count($nn, '+') > 1) {
    $firstPlus = strpos($nn, '+') + 1;
    $nn = substr($nn, 0, $firstPlus) . str_replace('+', '', substr($nn, $firstPlus));
}

This code will perform the original preg_replace() as normal and then, if there are more than 1 + symbols in the result, it will get a sub-string of the result up to the first +, then perform a string-replacement to replace all remaining + symbols. You could always use a second preg_replace() here too, but to remove only a + symbol it would be overkill.

Here’s a codepad entry for the sample.