How to detect UTF8 with BOM encoding in Perl - perl

Question

I have simple Perl script that does comparison of two files.
Result I write in different files with UTF8 BOM encoding.
To save text in BOM file I do printing chr(65279) into the beginning of the result file. Sometimes input text already contains BOM char in the begging of the text and my script prints one more.
The question is: How I can workaround it to do not print this BOM char twice.
See below text of my Perl's code:
use strict;
use warnings;
use List::Compare;
use Cwd 'abs_path';
use open ':encoding(utf8)';
use open IO => ':encoding(utf8)';
open F, "<$ARGV[0]" or die $!;
open S, "<$ARGV[1]" or die $!;
my #a=<F>;
my #b=<S>;
close F;
close S;
my $lc = List::Compare->new(\#a, \#b);
my #intersection = $lc->get_intersection;
my #missing = $lc->get_unique;
my #extra = $lc->get_complement;
open EXTRA, ">".$ARGV[2]."file_extra.txt" or die("Unable to open the file");
open MISSING, ">".$ARGV[2]."file_missing.txt" or die("Unable to open the file");
open SUBTRACTED, ">".$ARGV[2]."file_subtr.txt" or die("Unable to open the file");
#Turn on UTF-8 BOM support
print EXTRA chr(65279);
print MISSING chr(65279);
print SUBTRACTED chr(65279);
print MISSING #missing;
print EXTRA #extra;
print SUBTRACTED #intersection;
close MISSING;
close EXTRA;
close SUBTRACTED;

Answer 1

Strip it while reading file content (in your example apply s/^\x{FEFF}// to $a[0] and $b[0]) and then either add it in front of output when you print results, if you really need it, but better yet - don't print it back at all, as it is useless for UTF-8.

Answer 2

If you have double BOM, this is probably because one BOM comes from your input. So you should clean up your input before processing it:
s/^\x{FEFF}/ for $a[0], $b[0];