Question

Hi I have a problem with my regex pattern:

preg_match_all('/!!\d{3}/', '!!333!!333 !!333 test', $result);

I want this to match !!333 but not !!333!333. How can I modify this regex to match only a max length of 5 characters – two ! and three numbers.

Answer 1

I find the easiest and most descriptive way to do this is with negative lookaheads and lookbehinds.

See:

preg_match_all('/(?<![^\s])!!\d{3}(?![^\s])/', '!!333 !!333!!333 !!333 test !!333', result);

This says: match anything of the form !![0-9][0-9][0-9] which doesn’t have anything other than a space in front or behind it. Note that these lookaheads/lookbehinds aren’t matched themselves, they are “zero-width assertions”, they are thrown away and so you only get “!!333″ etc in your match, not ” !!333″ etc.

It returns

[0] => Array
        (
            [0] => !!333
            [1] => !!333
            [2] => !!333
        )

)

Also

preg_match_all(
    '/(?<![^\s])!!\d{3}(?![^\s])/', 
    '!!333 !!555 !!333 !!123 !!555 !!456 !!333 !!333 !!444 !!444 !!123 !!123 !!123!!123',
    $result));

returns

[0] => Array
    (
        [0] => !!333
        [1] => !!555
        [2] => !!333
        [3] => !!123
        [4] => !!555
        [5] => !!456
        [6] => !!333
        [7] => !!333
        [8] => !!444
        [9] => !!444
        [10] => !!123
        [11] => !!123
    )

That is, all but the last two which are too long.

See Lookahead tutorial.