As a general thing providing a focused working stand alone sample script illustrating the problem will get you better answers. Consider:

use strict;
use warnings;

my @lines = split /\n/, <<LINES;
token token token
taken token tiki
LINES

chomp @lines;

for my $line (@lines) {
    print "$line:\n";
    while ($line =~ /token/gi) {
        print "   >$`|$'<\n";
    }
}
[download]

Prints:

token token token:
   >| token token<
   >token | token<
   >token token |<
taken token tiki:
   >taken | tiki<
[download]

while (lc($line) =~ /token/g) { loops until killed because each time through the loop lc($line) is re-evaluated so the regular expression restarts, so don't do that. So long as $line isn't changed the /g makes the regular expression carry on from where it left off last time through the loop, which is what you want.