... how to parse a string which has repetitive data ...
As choroba pointed out, every (pattern) pair of parentheses in a regex captures something (even undef possibly) to its corresponding capture variable. One way to parse a string using nested regexes is avoid using a gazillion capturing groups. Use the non-capturing (?:pattern) instead for grouping. See perlre, perlrequick, perlretut. In the IP example (but this should generalize to any repetitive data you wish to extract):
>perl -wMstrict -le
"my $decimal_octet =
qr{ 2 (?: [0-4] \d | 5 [0-5]) | [01]? \d? \d }xms;
my $ip =
qr{ (?<! \d) $decimal_octet (?: \. $decimal_octet){3} (?! \d) }xms;
print $ip;
;;
my $s = '123.45.6.234 xx yyy zz 000.12.34.255';
my @ips = $s =~ m{ $ip }xmsg;
printf qq{'$_' } for @ips;
"
(?^msx: (?<! \d) (?^msx: 2 (?: [0-4] \d | 5 [0-5]) | [01]? \d? \d ) (?
+: \. (?^msx: 2 (?: [0-4] \d |
5 [0-5]) | [01]? \d? \d )){3} (?! \d) )
'123.45.6.234' '000.12.34.255'
Note that neither (?:pattern) nor the (?<!pattern) (?!pattern) look-around assertions capture. Indeed, nothing captures (to a capture variable) since data is extracted in list context directly to an array.
