G'day rruser,

Here's a solution using the builtin modules Time::Piece and Time::Seconds. It only requires a single script (instead of your three) and doesn't create an additional file concatenating all the data from the original three files.

With the data you posted, "COMPANY      BCYX 156095 L  2013,11,30   2013,12,08" has the greatest time spread (8 days). For my testing, I duplicated that record in file2 and file3 and added a unique record to file1 which also had a spread of 8 days. Here's the data I worked with:

$ cat pm_1071030_file1.txt
COMPANY      ABCD 764200 E  2013,12,13   2013,12,19
COMPANY      BCDX 156167 L  2013,11,29   2013,12,03
COMPANY      BCYX 165230 L  2013,12,13   2013,12,19
COMPANY      BCYX 156095 L  2013,11,29   2013,12,07

$ cat pm_1071030_file2.txt
COMPANY      ABCD 764200 E  2013,12,13   2013,12,19
COMPANY      BCDX 156167 L  2013,12,28   2013,12,31
COMPANY      BCYX 156095 L  2013,11,30   2013,12,08

$ cat pm_1071030_file3.txt
COMPANY      ABCD 764200 E  2013,12,13   2013,12,17
COMPANY      BCDX 156167 L  2013,11,30   2013,12,03
COMPANY      BCYX 165230 L  2013,12,13   2013,12,17
COMPANY      BCYX 156095 L  2013,11,30   2013,12,08
COMPANY      BCYX 156095 L  2013,11,30   2013,12,08
[download]

Here's the script:

#!/usr/bin/env perl

use strict;
use warnings;
use autodie;

use Time::Piece;
use Time::Seconds;

my $format = '%Y,%m,%d';
my $free_days = 4;
my @wanted_record_data = (0, []);

for my $file (map { "pm_1071030_file${_}.txt" } 1 .. 3) {
    open my $fh, '<', $file;

    while (<$fh>) {
        my ($start, $end) = (split)[4, 5];
        my $t0 = Time::Piece->strptime($start, $format);
        my $t1 = Time::Piece->strptime($end, $format);
        my $diff = ($t1 - $t0)->days;
        next if $diff <= $free_days;
        push @{$wanted_record_data[1]}, $_ if $diff == $wanted_record_
+data[0];
        @wanted_record_data = ($diff, [$_]) if $diff > $wanted_record_
+data[0];
    }
}

print "Greatest time spread = $wanted_record_data[0] days.\n";
print "Records with this time spread:\n";
my %seen;
print for grep { ! $seen{$_}++ } @{$wanted_record_data[1]};
[download]

Here's the output:

Greatest time spread = 8 days.
Records with this time spread:
COMPANY      BCYX 156095 L  2013,11,29   2013,12,07
COMPANY      BCYX 156095 L  2013,11,30   2013,12,08
[download]