Here is an equality function for arrays of strings, takes references to two arrays as argument:

sub aeq {
    my ($first, $second, @comp) = @_;
    return 0 if @{$first} != @{$second};
    @comp = grep {$first->[$_] ne $second->[$_]} 0..$#$first;
    return not @comp;
}
[download]

We check to see that the lengths are the same, return false ( not equal) if they are not. Then we use grep over the indexes to scan for unequal elements, putting the index in another array for each mismatch. Finally we return the negation of the mismath array's length as result; no mismatches gave a zero length array.

Update: Thanks to gmax for spotting my blunder in obtaining the last index. Fixed. For kappa's concern, I was thinking more of avoiding a run off the end of $second if it was shorter. For efficiency, I'd write this in terms of a flipflop, .. in scalar context. On reflection, I'd either do that, or in list context return an array of indexes that match:

sub aeq {
    my ($first, $second, @comp) = @_;
    return 0 if @{$first} != @{$second};
    @comp = grep {$first->[$_] eq $second->[$_]} 0..$#$first;
    wantarray ?
        @comp :
        @comp == @$first;
}
[download]

The choice is, as you say, between efficiency and general utility. Pick efficiency if you never need to filter for matches.

U² Bah, I'm suffering from a little brain damage, repaired the newer code.

After Compline,
Zaxo