Sorry to bother, I feel permanently lost when it comes to regex...
I have to match a string which occurs in a longer sequence of hex-values. My test-string is this:
BF1301020302000017BF1301030101010300FF6ABF130201010300FFC0BF1303010303030100FF98
Pattern is this:
I tried BF13(\w\w)+?00(\w\w){1} but it's obviously wrong.
The test-string is supposed to match and output these values:
Thanks, guys!
This one will do the job :
BF13(?:0[123])+00[A-Z0-9]{4}
Explanation
BF13 BF13 literally
(?:...)+ Followed by something (non capturing group) at least one time (+)
0[123] a zero followed by 1, 2 or 3
00 Followed by 00
[A-Z0-9]{4} Followed by uppercase char or a digit 4 times
Sample PHP code Test online
$re = '/BF13(?:0[123])+00[A-Z0-9]{4}/';
$str = 'BF1301020302000017BF1301030101010300FF6ABF130201010300FFC0BF1303010303030100FF98';
preg_match_all($re, $str, $matches, PREG_SET_ORDER, 0);
foreach ($matches as $val) {
echo "matched: " . $val[0] . "
";
}
You have a couple of options:
Input:
$in='BF1301020302000017BF1301030101010300FF6ABF130201010300FFC0BF1303010303030100FF98';
Method #1 - preg_match_all() (Regex Pattern Explanation/Demo):
var_export(preg_match_all('/BF13(?:0[123])+0{2}[A-F0-9]{4}/',$in,$out)?$out[0]:[]);
// *my pattern is a couple of steps faster than stej4n's
// and doesn't make the mistake of putting commas in the character class
Method #2: - preg_split() (Regex Pattern Explanation/Demo):
var_export(preg_split('/0{2}[A-F0-9]{4}\K/',$in,null,PREG_SPLIT_NO_EMPTY));
// K moves the match starting point -- preserving all characters when splitting
// I prefer this method because it requires a small pattern and
// it returns an array, as opposed to true/false with a variable declaration
// Another pattern for preg_split() is just slightly slower, but needs less parameters:
// preg_split('/0{2}[A-F0-9]{4}\K(?!$)/',$in)
Output (either way):
array (
0 => 'BF1301020302000017',
1 => 'BF1301030101010300FF6A',
2 => 'BF130201010300FFC0',
3 => 'BF1303010303030100FF98',
)