In the CB I was ineffectively trying to suggest that instead of @arrayA and @arrayB, you could more conveniently use a hash of arrays like this:
use strict;
use warnings;
my $str="A: Gold, Black, Purple, Blue, Red B: Black, Neon Pink, Ne
+on Yellow, Neon Green, Neon Purple, Red, White, Neon Orange, Navy";
my %arrays;
my( @substrings ) = $str =~ m/(\w:\s*.+?)(?=\w:|$)/g;
foreach my $substring ( @substrings ) {
my( $key, $colors ) = split /:\s*/, $substring;
my( @colors ) = split /,\s*/, $colors;
$arrays{ $key } = [ @colors ];
}
foreach my $key ( sort keys %arrays ) {
print "Array $key: @{$arrays{$key}}\n";
}
With this approach you're not stuck trying to treat diversely named variables as if they're a set (which is what you're trying to do when you deal with @arrayA and @arrayB). The hash approach holds both arrays as anonymous arrays referred to by the reference held in key A and key B. Hash key manipulation is much easier to do than contortions with variable names. The road down which you were traveling would eventually lead you to a desire to use symbolic references. The road my approach takes you down leads to proper references; much easier and safer to work with.
I'd have a look at perlref, perlreftut, and perllol, and finally, perldsc for more insight. I know it's a lot of reading, but spend a couple hours with it, and you'll be miles ahead on the learning curve.