Remove repetitve Consecutive words from string

Question

I want to remove repetitive words from the string (only consecutive).

$str = 'abc,def,fgh,fgh,xna,fgh,xyz,xyz,xyz,tr,tr,xna';

My desired output string is:

abc,def,fgh,xna,fgh,xyz,tr,xna

I can get the result I want in php using this:

$ip = explode(',', $str);
$op = [];$last = null;
for($i=0;$i<count($ip);$i++){
    if ($last == $ip[$i]) {
        continue;
    }
    $op[]=$last=$ip[$i];
}
$ip = implode(',', $op);

But was looking for the regex approach. So far I have got closer with these two regexes:

$after = preg_replace('/(?:^|,)([^,]+)(?=.*,\1(?:,|$))/m', '', $str);
output : abc,def,fgh,xyz,tr,xna

$after = preg_replace('/([^,]+)(,[ ]*\1)+/m', '', $str);
output : abc,degh,fgh,xna,fgh,,,xna

Answer 1

You should use

preg_replace('~(?<![^,])([^,]+)(?:,\1)+(?![^,])~', '$1', $str)

See the regex demo

If there is a need to support any 0 or more whitespace chars between the commas and repetitive values, add \s* (0 or more whitespaces) pattern before \1.

Details

• (?<![^,]) - start of string or any char but a comma
• ([^,]+) - Group 1: any one or more chars other than a comma
• (?:,\1)+ - one or more sequences of a comma and the value in Group 1
• (?![^,]) - end of string or a char other than a comma.

Answer 2

$after = preg_replace('/(?<=^|,)([^,]+)(,\s*\1)+/', '$1', $str);

P.S. You can get rid of \s* from the regexp above if there is no whitespace expecter after ,. I just looked at your [ ]* and figured you may have whitespace.

Answer 3

Iterating with strtok, only glue pieces that are not like the last:

<?php

$str = 'abc,def,fgh,fgh,xna,fgh,xyz,xyz,xyz,tr,tr,xna';

$out = $last = strtok($str, ',');
while($current = strtok(','))
    if($current !== $last)
        $out .= ',' . ($last = $current);

echo $out;

Output:

abc,def,fgh,xna,fgh,xyz,tr,xna