Sort hash of hashes by values

[BrianAtWork]

Maybe I'm going about this the wrong way, but here's what I'm trying to do:

I have a hash that is storing the output from the unix "df" command, to get the disk used and disk free for each filesystem. Here is how I build the hash:

sub get_df {
    my %hash=();
    my @data=qx(df -t -k);
    chomp(@data);
    foreach (@data){
        next if (/^Filesystem/);
        my @parts=split(' ',$_);
        next unless ($parts[2] =~/\d+/ && $parts[3] =~/\d+/);
        $hash{$parts[5]}{"used"}=$parts[2];
        $hash{$parts[5]}{"free"}=$parts[3];
        $hash{$parts[5]}{"pused"}=$parts[4];
    }#end foreach
    return %hash;
}

#sample output from the df -t -k:
#/dev/hd3  131072  9804 121268  8% /home

When I process the hash I want to sort by the percentage used in order to show the highest disk usage at the top. I know how to sort a hash by the values instead of the keys, using byval or something like this:

for (sort {$hash{$a} <=> $hash{$b}} keys %hash){

but my hash structure isn't as easy as that (I don't think).
I need to go through each filesystem, and pull out the percentage out of that hash and sort by that value.

And like I said, maybe I shouldn't even be using a hash of hashes - it seemed like a good storage method to me. This will be output on a webpage, and graphs will be created by using the data in the hash as well, so calling the hash by "used", "free", and "pused" seems like the way to go, but I will wait and see if anyone else has something to say.

Thanks in advance!

Brian

 

[BrianAtWork]

I tinkered with my previous method to sort it. I'm not sure if it's right or not, but this is what I did:

for (sort {$hash{$b}{"pused"} <=> $hash{$a}{"pused"}} keys %hash){

I was making it too complicated - I just had to take a step away from it for a second.

I'll still take comments if this should be done differently though.

Thanks!

Brian

 

[MillerH]

for (sort {$hash{$b}{used} <=> $hash{$b}{used}} keys %hash)