Why doesn't gedit recognize the encoding of my output file created from perl program?

Question

#!/usr/bin/perl -w
use strict;

open (EVENTLOGFILE, "<eventlog.txt") || die("Could not open file eventlog file");
open (EVENTLOGFILE_NODATETIME, ">eventlog_nodatetime.txt") || die("Could not open new event log file");


my($line) = "";

while ($line = <EVENTLOGFILE>) {
 my @fields = split /[ \t]/, $line;
 my($newline) = "";
 my($i) = 1;

 foreach( @fields )
 {
  my($field) = $_;
  if( $i ne 3 )
  {
   $newline = $newline . $field;
  }

  $i++;
 }

 print EVENTLOGFILE_NODATETIME "$newline";
}

close(EVENTLOGFILE);
close(EVENTLOGFILE_NODATETIME); 

If I print out $line each time instead of $newline it can detect the encoding no problem. It's only when I try to modify the lines that it gets messed up.

Answer 1

I guess it isn't encoding (as in say ISO 8859-1 vs UTF-8) but line-endings (CR, LF vs LF).

If you used chomp and printed "\n" you'd probably get line endings converted to platform native.

I think your script might be better written something like this (Untested):

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

open ($old, '<', 'eventlog.txt') or die 'Could not open eventlog.txt';
open ($new, '>', 'eventlog_nodatetime.txt') 
  or die 'Could not open eventlog.nodatetime.txt');

$\ = "\n";

while (<$old>) {
  chomp;
  s/^(\S+\s+\S+\s+)\S+\s+(.*)/$1$2/;
  print $new;
}

close $old;
close $new; 

Or

perl -pe 's/^(\S+\s+\S+\s+)\S+\s+(.*)/$1$2/' eventlog.txt >eventlog.nodatetime.txt

Or use a splice on a split? Or ...