For this, you'll want to use sort and probably split. Beware the distinction between "==" and "eq" (documented in perlop). You might also be interested in join, chomp, and How do I post a question effectively?

sort can take a block which dictates the sort criteria. Consider:

use strict;
use warnings;
use Text::xSV;

my $data = <<'END_DATA';
02/28/2009|02062672169||4.99|9801|1.0|
02/28/2009|78513836100||4.96|9639|1.0|
02/28/2009|71171973992||19.99|9749|1.0|
02/28/2009|01463314931||19.99|9802|1.0|
END_DATA

open my $inFile, '<', \$data or die "Failed to open data file: $!";

my $csv = Text::xSV->new (fh => $inFile, sep => '|');
my @rows;

push @rows, $_ while $_ = $csv->get_row;
@rows = sort {$a->[1] <=> $b->[1]} @rows;
print join "\n", map {join '|', map {$_ || ''} @$_} @rows;
[download]

Prints:

02/28/2009|01463314931||19.99|9802|1.0|
02/28/2009|02062672169||4.99|9801|1.0|
02/28/2009|71171973992||19.99|9749|1.0|
02/28/2009|78513836100||4.96|9639|1.0|
[download]

As you will see in your research, the built-in sort facility (as well as various CPAN modules) provide the means to supply your own “comparison function,” whose job is to determine the relative ordering of two records. This comparison function will of course be called thousands of times.

Having said that, I'd also point out that the sort commands of many operating systems (including Linux/Unix) provide many options ... so that you may well find that you have no need for “your own comparison function.” If the standard Unix sort command is told that:

1. I want to sort by the second field...
2. I want to do a numeric comparison...
3. The field-separator is “a vertical bar” ...

Your research will quickly confirm that this is a very routine and straightforward requirement. It is well worth the time to acquaint yourself with the plethora of existing text-handling tools that are available.

You have a table. Treat it as such.

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

use DBI;
my $file = 'f1';

my $dbh = DBI->connect("dbi:CSV:csv_eol=\n;csv_sep_char=|");
$dbh->{csv_tables}->{FOO}{file} = $file;
$dbh->{csv_tables}->{FOO}{col_names} = [ qw/date id unk1 price unk2 qu
+antity/ ];

my $r = $dbh->selectall_arrayref('SELECT * FROM FOO ORDER BY ID');

use Data::Dumper; print Dumper $r;
[download]

I emphasise the tabular nature of your data because you're probably doing other things that SQL makes easy. And, you probably should also put it in a real db anyway, even if that real db is SQLite.

have a look at this link: http://www.stonehenge.com/merlyn/UnixReview/col64.html it explains exactly what you need.

Cheers,
R.