#!/usr/bin/perl
use strict;
use warnings;

#### untested ####

open my $fh_transfile, '<', "filename" 
    or die "unable to open translation file $!";

my %table;

while my $line (<$fh_transfile>)
{
   my($nat_ip, $real_ip) = split ' ', $line;
   $table{$nat_ip} = $real_ip;
}
close $fh_transfile;


open my $fh_bigfile , '<', "filename" or die
    "unable to open big input file $!";
open my $fh_out, '>', "filenameout" or die
    "unable to open big output file $!"

while my $line (<$fh_bigfile>)
{
   chomp $line;
   my @tokens = split ',',$line;
   
   if ($table{$tokens[8]})
   {
       $tokens[8] = $table{$tokens[8]};
   }

   print $fh_out join(",",@tokens),"\n";
}
[download]

Update: I got a msg with a question about why is there a "chomp $line;" in the second while loop and not in the first while loop? I'll put the answer here as others may have the same question...

The purpose of chomp() is to remove the line ending, represented as "\n" in Perl. If there was no chomp() in the second while loop, then the last element of @tokens would have the line ending included in that last element after the split on ','.

In the first while loop, split ' ',$line; does NOT mean split on the "space" character. This is a special case coded into Perl and is translated into: split on any sequence of whitespace characters, (space,tab,form feed,end of line). So in the first while loop, the split ' ',$line; removes the line ending because it is included in the set of things to split upon. A chomp() before that split would not hurt, but it is not necessary.

The difference between split /\s+/, $line; and split ' ', $line; is that in the second version, any whitespace at the front of the line is removed while in the first version, leading whitespace would cause the first element of @tokens to be a null field. Easier demo'ed than further attempts at english explanations:

use strict;
use warnings;
use Data::Dumper;

my $line = "     X  Y    \tZ   A  \n";

my @tokens = split ' ', $line;
print Dumper \@tokens;

@tokens = split /\s+/, $line;
print Dumper \@tokens;

__END__
$VAR1 = [              #split ' ' version
          'X',         #note ending removed
          'Y',
          'Z',
          'A'
        ];
$VAR1 = [              #split /\s+/ version
          '',          #note ending removed
          'X',
          'Y',
          'Z',
          'A'
        ];
[download]

It's hard to give an example of the hash unless you give us examples of your two files...

The better your examples match your real data, the better the example program will work.

G'day EntropyDF,

Welcome to the Monastery.

"File 1 is a CSV seperated log file. For this need the IP address is always in split 8 of the file."

Do not try to write your own CSV parsing code. This is fraught with all sorts of problems and is one wheel that definitely does not need to be reinvented. Use Text::CSV for this task.

"However if I could not rely on that it would be good as there are other log types this coudl apply to."

Abstract your code into a subroutine and pass it the wanted field number. As you've chosen to show no code, I'm in no position to indicate how you'd modify such code to achieve this. The field number could come from any number of places, such as command line argument, config file, database, etc.; it would presumably be associated with whatever "other log types" refers to.