Then why are you asking Perl to read until the next newline?

Besides, your data has two newlines.

84 47 00 0c 00 00 11 0a 03 50 35 04 00 64 0a 04
                     ^^                   ^^
16 12 00 1e 00 1d 00 00 65 09 08 54 52 03 2f 82
05 10 00 02 6e 06 0b a8 53 11 67 00 00 00 7a 7f
00 69 4a 42 47 48 4a 41 0c 00 6a 53 54 4d 44 54
4f 1c 00 66 04 00 00
[download]

If you want to read the whole file at once, why do you have a loop? Undefine $/ and get rid of the loop.

local $/;
my $data = <DATA>;
[download]

If you're not trying to read the entire file at once, what are you trying to do?

Thank you very much for your help. This is my last coding after some modification based on your advise. Please guide me if I am wrong.


#!/usr/bin/perl -w

use Cwd;
use warnings;
use strict;
use Getopt::Long;

use constant FILEHDR => 4;
use constant CDRLEN  => 286;

my ($trace, $help, $infile);
my $swap = '';
my $indir  = getcwd;
my $outdir = getcwd;

GetOptions (
    "h|help"    => \$help,
    "filename|f=s"    => \$infile,
    "swap|s"    => \$swap,
    "input|i=s"    => \$indir,
    "output|o=s"    => \$outdir,
    "trace|t"    => \$trace
) or usage();


sub usage {
    exit;
}

my $outfile = $infile;
my $data;

if ($infile) {
    #open (OUTPUT, ">$outdir/$outfile");
    open (DATA, "$indir/$infile");
    binmode DATA;
    
    until (eof DATA) {
        read (DATA, $data, 2);
        my $tag    = unpack "H2", substr $data,0,1,'';
        my $length = unpack "C", substr $data,0,1,'';
        $length -= 2;

        if ($length == "81") {
            $length = unpack "C", substr $data,0,1,'';
            $length -= 1;
        }

        if ($length > 0) {
            read (DATA, $data, $length);

            my $rec      = unpack "H*", substr $data,0,$length,'';

            printf ("RECORD TAG    : %s\n", $tag);
            printf ("RECORD LENGTH : %s\n", $length);
            printf ("RECORD        : %s\n", $rec);
        }

    }

    close(DATA);
    #close(OUTPUT);
}
[download]

Without actually testing your code, that seems to be OK. A couple of things to improve it though: you should always test the return value from open,

open (DATA, "$indir/$infile") or die "Unable to open $indir/$infile: $
+!";
[download]

would be a conventional way of doing that. Notice also that read returns the number of bytes read, which you can use to check that you get back the number you expect - if data can be wrong it will be!