Regex captering repeting group in php without whitespace

I'm trying to parse strings by a regex in PHP that always have this format: FooBar(,[0-9]{7}[0-9A-F]{8})+ or in other words, they have a start value/word followed by 1 or multiple entries, each entry is one comma (,), followed by 7 digits and 8 hexdec characters (digits or uppercase characters A to F).

My Regex to capture this is /^C7(,[0-9]{7}[0-9A-F]{8})+$/ which kind of works. When used in a preg_match_all it returns an array with two entries, the first as expected the input string, however, in the second array there's only one entry, the last matched chunk. (see Example)

I need to capure all the chunks matched by the capturing group. I did some research and found this answer, which seamed to be about the same issue:, So I've adjusted my regex to /(,[0-9]{7}[0-9A-F]{8})+$/, but I still only get one match. This can be tested at I then experimented some more, and found, that if I change the input string, to contain a space (or any not matched character for that matter), between the chunks, like this: C7,22801422CFE0F63 ,2280141C5EF0F63 ,22801402EFD0F63 ,2280138C5ED0F63 ,228024329897530 ,228023829877530 and adjust the regex once again to /(,[0-9]{7}[0-9A-F]{8})+/ it does exactly as it is intended to do!

Question: Is there a way to achieve this, matching all the chunks in this recurring group without adding whitespaces in between? If so, how?


To illustrate the problem:
No Whitespace No Whitespace

enter image description here Whitespace/random chars

Goal: Behaviour of second one, the one with whitespaces, but not adding the whitespaces (respectively the not matched characters).

EDIT 2 (hacky solution)

I kind of found a solution, considering this Answer. The Regex /(?:,)([0-9]{7}[0-9A-F]{8})/ works for me. I'd still like a way, to match the initial FooBar. as that indicates the incoming string should be matched with this regex at all.
(I know I could simply check the string in a second regex for this, I however would love to have it in one regex)

Input: 'C7,22801422CFE0F63,2280141C5EF0F63,22801402EFD0F63,2280138C5ED0F63,228024329897530,228023829877530'

4 answers

  • answered 2018-01-06 15:38 Toto

    Is that what you want?

    $in = 'C7,22801422CFE0F63 ,2280141C5EF0F63 ,22801402EFD0F63 ,2280138C5ED0F63 ,228024329897530 ,228023829877530';
    preg_match_all('/(^\w+|\G)\h*(,[0-9]{7}[0-9A-F]{8})/', $in, $m);


        [0] => Array
                [0] => C7,22801422CFE0F63
                [1] =>  ,2280141C5EF0F63
                [2] =>  ,22801402EFD0F63
                [3] =>  ,2280138C5ED0F63
                [4] =>  ,228024329897530
                [5] =>  ,228023829877530
        [1] => Array
                [0] => C7
                [1] => 
                [2] => 
                [3] => 
                [4] => 
                [5] => 
        [2] => Array
                [0] => ,22801422CFE0F63
                [1] => ,2280141C5EF0F63
                [2] => ,22801402EFD0F63
                [3] => ,2280138C5ED0F63
                [4] => ,228024329897530
                [5] => ,228023829877530


    (               : start group 1
      ^\w+          : beginning of line, 1 or more word characters
      |             : O
      \G            : match form this point
    )               : end group 1
    \h*             : 0 or more horizontal spaces
    (               : start group 2
      ,             : a comma
      [0-9]{7}      : 7 digits
      [0-9A-F]{8}   : 8 hexa
    )               : end group 2

  • answered 2018-01-06 15:38 The fourth bird

    To capture all chucks including the first part, you could try:



    • A non capturing group (?:
    • Match FooBar
    • Or |
    • You format in a on capturing group repeated one or more times (?:[0-9]{7}[0-9A-F]{8})+
    • Close non capturing group


  • answered 2018-01-06 15:38 Casimir et Hippolyte

    You can build a pattern to get contiguous matches using the A flag (that means Anchored). The main interest is that you can extract your values and check the format of the line at the same time using a lookahead:

    $pattern = '~
        (?!^)  # fails at the start of the string
        ( \h*,\h* (?<value>[0-9]{7}[A-F0-9]{8}) )
        # the first capture group is useful to shorten the 
        # the lookahead in the second branch.
    if ( preg_match_all($pattern, $yourstring, $matches) ) {
        echo $matches['first'][0], PHP_EOL;


    The A flag forces each match to start at the beginning of the string or at the end of the previous match.

    The first branch describes a comma separated value and the second branch the start of the line.

    The lookahead (?=(?1)*$) checks forward the structure of the line. If this one fails, no match is possible.

  • answered 2018-01-06 15:38 user6307854

    ehmmm... maybe i can't understand the problem but your regex will work for the first scenario removing the trailing +