in reply to text-file parsing with automatic field ordering

Is this type of referencing okay or insane, or is there mayber a simpler/better/faster way?

With some minor corrections and a pair of things I would do differently, it seems fine to me.

Update: here's one possible way I'd do it (note in particular that I don't use %index):

#!/usr/bin/perl

use strict;
use warnings;

chomp (my $headerline = <DATA>);
$headerline =~ s/^#// or die "Not a correct headerline";
my @keys=split /:/, $headerline;

my @array;
while (<DATA>) {
    chomp;
    my %hash;
    @hash{@keys}=split /:/;
    push @array, \%hash;
}

# examples
print "1st name : $array[0]{'name'}\n";
print "3rd category : $array[2]{'category'}\n";

# traverse entire array
for my $idx (0..$#array) {
    my $row=$idx+1;
    print "Row $row: ",
      (map "$_=$array[$idx]{$_} ", @keys), "\n";
}

__DATA__
#name:surname:category
tiger:woods:golfer
tyler:hamilton:f1
james:stewart:supercross
roger:federer:tennis

[download]

It seems simpler to me.

In Section Seekers of Perl Wisdom