Question

I have the following regex

((\$|(\\\[)).*?(\$|(\\\])))

which should capture everything between $$ and \[\] and I tested it on http://gskinner.com/RegExr/ and it’s working.

PHP variant is (doubled backslashes)

((\$|(\\\\\[)).*?(\$|(\\\\\])))

and I would like to split my text based on that regex. How can I tell that it uses just the first (and largest group) and not these small ones?

preg_split('/((\$|(\\\\\[)).*?(\$|(\\\\\])))/', $text, -1, PREG_SPLIT_DELIM_CAPTURE);

So for text This is my $test$ for something. I should get an array

[0] => This is my 
[1] => $test$
[2] =>  for something.

But I get

[0] => This is my 
[1] => $test$
[2] => $
[3] => 
[4] => $
[5] =>  for something.

Answer 1

You would need something like this:

$text = 'This is my $test$ for \[something\] new!';
print_r(preg_split('/(\$.*?\$|\\\\\[.*?\\\\\])/', $text, -1, PREG_SPLIT_DELIM_CAPTURE));

Output:

Array
(
    [0] => This is my 
    [1] => $test$
    [2] =>  for 
    [3] => \[something\]
    [4] =>  new!
)

---

IMHO, your regex is (probably) wrong. It would fail for texts like Hello $there\]. If you need to capture texts between two $s and a pair of \[ and \], then you need the regexp like:

          <-------------> Match text between \[ and \]
/(\$.*?\$|\\\\\[.*?\\\\\])/
  <----->   Match text between dollars