Ok I've tried different versions of this program

#!/usr/bin/perl

use strict;

print get_time() ."\n";

    my $count = 0;
    
    my $pr_regex= "program.jsp?id=1";
    $pr_regex = qr/\Q$pr_regex\E/oi;
    #open(LOGFILE,"file.txt");
    @ARGV = qw(file.txt);
    close ARGV;

    #while (<LOGFILE>) {
    while (<>) {
            $count ++ if m/$pr_regex/oi;
    }
        
print qq|$count\n|;    
print "\n" .get_time() ."\n";
exit;
    
    
sub get_time {
    my ($sec,$min,$hour,@junk) = localtime(time);
    $min = '0' . $min if ($min<10);
    $sec = '0' . $sec if ($sec<10);
    return qq|$hour:$min:$sec|;
}
[download]

and the output is :

bash-2.03$ perl -w agrsel_mark3.cgi
14:27:05
203

14:27:26

so around 20 seconds to find one string. That's after a little tweaking to get it down from 26 seconds.
Here's a version of my original (just looking for one string though):

#!/usr/bin/perl

use strict;

print get_time() ."\n";

    my $count = 0;
    
    my $pr_regex= "program.jsp?id=1";
    $count = `grep -c '$pr_regex' file.txt`;
        
        
print qq|$count\n|;    
print "\n" .get_time() ."\n";
exit;
    
    
sub get_time {
    my ($sec,$min,$hour,@junk) = localtime(time);
    $min = '0' . $min if ($min<10);
    $sec = '0' . $sec if ($sec<10);
    return qq|$hour:$min:$sec|;
}
[download]

and the output:
bash-2.03$ perl -w agrsel_mark4.cgi
14:27:34
203


14:27:40

about 6 seconds.
actually running the full program it takes about 8 seconds a string over the first 68 strings, not quite 9 minutes.
And the regex version takes about 26 minutes to run the first 68 strings.
a little quick math tells me I'm looking at 2 hours versus 6 hours when I start really using the program.
I've tried the reg_ex version a few different ways but the time doesn't get any better then 20 seconds.
Any ideas on how to jump this up a little?
Thanks again
John