find /path/to/dir -type f -atime +10 -exec rm {} \;
[download]

use File::Find::Rule;
File::Find::Rule->file() # only files
                ->atime("+10") # that haven't been accessed in 10 days
                ->exec( sub { unlink( $_[2] ) } ) # delete them
                ->in('/path/to/dir'); # starting in this path
[download]

automatically delete these files after 10 days they are last accessed.

To answer the second half of your question first, you could use a very nifty tool like so:

find2perl ${dir_I_am_watching} -atime +10 \
    -exec rm {} \; > myNewScript.pl
[download]

As to the first half of your question I'm not sure what you are asking. If what you are asking is to look in a particular directory and act upon any file in that directory then the strategy I've mapped out for you above should work just fine. Just replace the shell variable I've cited (${dir_I_am_watching}) with the actual name of the directory and it will be hardcoded into your generated script.

Of course one of the immediate tweaks I'd make to the script is to eliminate hardcoding and add logic surrounding GetOpt::Long and provide a command line switch to the script that sets what directory it is examining as well as a command line switch that sets the age of the file I want to eliminate. Part of that logic would set defaults for sure, but by having command line switches there to modify those variables you've increased the usability of the resultant script a bit.

Lastly, you failed to say waht platform you are doing this all on. This set of strategies will work fine on *nix platforms but on others YMMV. I'm not 100% sure it will work on *doze or Mac and I'm too lazy to go find out.

HTH

Here is some code, more or less lifted straight from the horse's mouth

{
    my $now;
    sub days_ago {
        # days as number of seconds
        $now ||= time;
        return $now - (86400 * shift);
    }
}

for my $file (

    File::Find::Rule->new
        ->mtime( '<' . days_ago(2) )
        ->name( qr/^CGItemp\d+$/ )
        ->file()
        ->maxdepth(1)
        ->in('tmp'),

    File::Find::Rule->new
        ->mtime( '<' . days_ago(14) )
        ->file()
        ->maxdepth(1)
        ->in('/tmp/ep'),

    File::Find::Rule->new
        ->mtime( '<' . days_ago(30) )
        ->file()
        ->in(qw(
            /var/db/ep/epdata/reports
            /var/db/ep/cmd/sync-log
            /var/db/ep/cmd/analyse-oracle/log
        )),

) {
    $body = "Old temporary files purged:\n" unless length($body);
    $body .= "$file "
        . (unlink($file) ? 'ok' : "not ok: $!")
        . "\n";
    ++$files;
}
[download]

At the end of this, if anything was processed ($file > 0), it is sent off as an e-mail message to the interested parties (me). This is run daily at around 5:00 in the morning.

What you want is if the file has not been accessed for 10+ days, delete. (-atime is last accessed, -mtime last mod, -ctime creationtime)

Make sure however your contraption reads the file, is actually changing atime, likely it is. Whenever the file is accessed, *ix changes the atime.

Login to your shell

Open your crontab # crontab -e

If you want it to run every hour:
59 * * * * /usr/bin/find /path/to/dir -type f -atime +10 -exec rm {} \;

that's an example, look up # man crontab

Re: removing old dirs

This package provides a program that can be used to clean out temporary-file directories. It recursively searches the directory, refusing to chdir() across symlinks, and removes files that haven't been accessed in a user-specified amount of time. You can specify a set of files to protect from deletion with a shell pattern. It will not remove files owned by the process EUID that have the `w' bit clear, unless you ask it to, much like `rm -f'. `tmpreaper' will not remove symlinks, sockets, fifos, or special files unless given a command line option enabling it to.

well the best way i think , is using stat function

# this will return last access time in seconds and you can compare tim
+e then 
my $last_access = (stat($filename))[8];
[download]

HTH