use 5.010;
use strict;

# A few definitions...
use constant {
    ITERATIONS   => 20,   # number of random strings to generate
    LENGTH       => 72,   # length of each string
    X_COUNT      => 4,    # number of X characters in each string
    };

# We need a uniq() function. List::MoreUtils provides a fast XS based
# one, but if they don't have it installed, fall back to a pure Perl
# alternative.
BEGIN {
    *uniq = eval { require List::MoreUtils; 1 }
        ? \&List::MoreUtils::uniq
        : sub (@) { my %seen = (); grep { not $seen{$_}++ } @_ };
}

# We want to generate more than one random string.
foreach my $iter (1 .. ITERATIONS)
{
    # Make a pool of random numbers. uniq ensures no duplicates,
    # but means that we might end up with fewer than we need,
    # hence a loop to top up the pool if it's below the number
    # we need.
    my @positions;
    while (@positions < X_COUNT)
    {
        push @positions, map { int rand LENGTH } 0 .. X_COUNT;
        @positions = uniq @positions;
    }
    
    # Take just the first X_COUNT items from our pool, then
    # map it into a hash structure.
    my %exes =
        map { $_ => 1 }
        @positions[0 .. X_COUNT-1];
    
    # Create our array of length LENGTH. Indexes found in the
    # %exes hash get an "X"; other indexes get a hyphen.
    my @array =
        map { $exes{$_} ? 'X' : '-' }
        1 .. LENGTH;
    
    # Print out the array as a string.
    say join q{}, @array;
}
[download]