I usually do this sort of things with a data structure of my tests and a simple loop with the repetitive test code. E.g.

#!/usr/bin/perl

use strict;
use warnings;
use Test::More; # no plan here -- plan later
use Test::Cmd;

my @cases = (
    # label, args, exit, expected, regex
    [ "plain", "", 0, "Linux", "Linux" ],
    [ "with node", "-n", 0, "Desktop1", "Desktop" ],
);

my $tests_per_case = 3;
plan tests =>  $tests_per_case * @cases;

for my $case ( @cases ) {
    my ($label, $args, $exit, $expected, $regex) = @$case;
    my $cmd = Test::Cmd->new( prog => "/bin/uname $args" );
    is $cmd->run, $exit, '$label\: uname ran okay';
    is $cmd->stdout, $expected, "$label\: Test uname $args";
    like $cmd->stdout, qr/$regex/, "$label\: Found uname parameter '$r
+egex'";
}
[download]