Definitely TIMTOWTDI, but so far everyone agrees it's better (at least easier) to generate the permutations first and then sort the list, rather than trying to do both at once. Take a look at the Perl Cookbook section 4.15 (if you have it) for info on sorting lists based on a comparison function; it also has some effiency hints. I didn't do it by strings like the other Monks (which I realize differs from the form of your question), but this might be more efficient, and you can always regenerate the strings at the end. Here's what I came up:
#!/usr/bin/perl -w
use strict;
my (@a, @b, @c, @d, @e, $i1, $i2, $i3, $i4, $i5);
@a = @b = @c = @d = @e = (0, 1, 2); # or whatever
my (@unsorted, @sorted);
for ($i1 = 0; $i1 <= $#a; $i1++) {
for ($i2 = 0; $i2 <= $#b; $i2++) {
for ($i3 = 0; $i3 <= $#c; $i3++) {
for ($i4 = 0; $i4 <= $#d; $i4++) {
for ($i5 = 0; $i5 <= $#e; $i5++) {
push @unsorted, [$i1, $i2, $i3, $i4, $i5];
}
}
}
}
}
@sorted = sort { non_zeros($b) <=> non_zeros($a) ||
${$b}[0] <=> ${$a}[0] ||
${$b}[1] <=> ${$a}[1] ||
${$b}[2] <=> ${$a}[2] ||
${$b}[3] <=> ${$a}[3] ||
${$b}[4] <=> ${$a}[4] } @unsorted;
print map "@$_\n", @sorted;
sub non_zeros {
my @arr = @{$_[0]};
scalar grep { $_ != 0 } @arr;
}
I'm not a big fan of any of the permutation generators (including mine) listed so far; I'll try to think of something cleaner and more general.