Same code ( what is this, Lisp? ), reformatted to be more readable and do a sanity check for the zero-columns case. Note that it doesn't seem to do exactly what the questioner asked - see the 22/7 case:

#!/usr/bin/perl -w

use strict;

while (<DATA> ) {
    my ( $items, $cols ) = split;
    print "Spread for $items, $cols = ", 
            spread( $items, $cols ), 
            "\n";
}
    
sub spread {
    my ($a, $cols) = @_;
    $a ||= 0;
    return unless $cols > 0; # avoid nasty surprises
    return $a if $cols == 1; #trivial case 
    
    if ( $a % $cols) {
         return join '-', (  (int( $a / ($cols - 1) ))  x ( $cols - 1)
+, ( $a % ($cols - 1)));
    } else {
        return join '-', (int ( $a / $cols)) x $cols;
    }
}

__DATA__
19 3
20 5
22 7
2 47
0 9
899 0
[download]

Results:

Spread for 19, 3 = 9-9-1
Spread for 20, 5 = 4-4-4-4-4
Spread for 22, 7 = 3-3-3-3-3-3-4
Spread for 2, 47 = 0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0
+-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-0-2
Spread for 0, 9 = 0-0-0-0-0-0-0-0-0
Spread for 899, 0 =
[download]