Re: How to include a variable in output for derived subnets

[poj]

"My difficulty is I don't really understand well enough the code.."

Perhaps this demo program will help you

use strict;
use Socket;
my $len = '/29';
print "IP                   IPN        Mask      Subnet\n";
for my $i (1...31){
  print_subnet('192.168.1.'.$i ,$len);
}

sub print_subnet { 
  my ( $ip, $len ) = @_; 
  my $ipn = unpack "N", Socket::inet_aton $ip;
  $len =~ s/^\///; 
  my $maskn = 0xFFFFFFFF << ( 32 - $len ); 
  my $subnet_ipn = $ipn & $maskn; 
  my $subnet_ip = Socket::inet_ntoa pack "N", $subnet_ipn; 
  printf "%-20s %s %x %s\n","$ip/$len",$ipn,$maskn,$subnet_ip;
}

It should produce a table like this. Change len values to see effect.

IP                   IPN        Mask      Subnet
192.168.1.1/30       3232235777 3fffffffc 192.168.1.0
192.168.1.2/30       3232235778 3fffffffc 192.168.1.0
192.168.1.3/30       3232235779 3fffffffc 192.168.1.0
192.168.1.4/30       3232235780 3fffffffc 192.168.1.4
192.168.1.5/30       3232235781 3fffffffc 192.168.1.4
192.168.1.6/30       3232235782 3fffffffc 192.168.1.4
192.168.1.7/30       3232235783 3fffffffc 192.168.1.4
192.168.1.8/30       3232235784 3fffffffc 192.168.1.8
192.168.1.9/30       3232235785 3fffffffc 192.168.1.8
etc

You should be able to add the first comment like this

$subnets{"$ip,$len"}{ipn}   = $ipn;
$subnets{"$ip,$len"}{maskn} = $maskn;
if ( $subnets{"$ip,$len"}{'comment'} eq '' ){
  $subnets{"$ip,$len"}{'comment'} = $comment;
}

Note : If you did this the comment would be erased by a later blank record

$subnets{"$ip,$len"} = { ipn => $ipn, maskn => $maskn };

To print it out use this

print "subnet,prefix,name\n";
print "$_,$subnet{$_}{'comment'}\n"
   for sort { $subnets{$a}{ipn} <=> $subnets{$b}{ipn} };

poj

[stroke]

Hi poj,

Thanks - that pretty much did the trick. I had to tweak the print code block a little as I got a compilation error, but it works a treat. Appreciate your assistance.

As is the way - this has thrown up the next thing to fix - i found some subnet definitions don't end up with a name (the IPs have no comment), I guess I can iterate over the subnets hash and substitute any undef comments with a value. Initially, I did a find/replace in the output CSV file, but it would be nice to resolve it in the code. I'll have a look into this.

The demo program is really useful, thanks. I need to have a proper look at it, but my initial guess as to how the original code is working out the subnets is that rather than doing any de-duplication of subets or similar, it is just populating a subnet address and prefix into the hash and if the same one already exists, it just get overwritten.

Thanks again, appreciate your help.