justabunchofdisks.com a technology blog

Wading into the Perl Scripting Pool

This week I wrote my first Perl script.

Quick background – I am working on a project which requires a script to do the following:

• to delete files based on predetermined age of file
• to delete files in multiple locations
• to log the files it has deleted
• has to be able to be scheduled

Pretty simple requirements. Of course my own additional requirement would be – has to run on Linux (no Microsoft fees for this project).

I immediately thought to do this in Perl, because (a) Perl was basically written for file manipulation, and (b) it’s the all purpose programming language.

So my goals are pretty simple. And (hopefully) here is my script (version 1):

#!/usr/bin/perl
###############################################################
# This is a script to delete all files that have file creation
# dates older than X days from 2 directories. Script will output
# separate text files logging which files have been deleted
# along with date/time stamp. Logs will be appended and kept.
###############################################################
$dir1 = "/path/to/first/directory";
$age1 = X; # age in days
$dir2 = "/path/to/second/directory";
$age2 = X; # age in days
my $log_file1="/path/to/first/log/log1.txt";
my $log_file2="/path/to/second/log/log2.txt";
my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday,
$isdst)=localtime(time);
my $date = sprintf ('%4d-%02d-%02d %02d:%02d:%02d',
$year+1900,$mon+1,$mday,$hour,$min,$sec);
###############################################################
# Begin script ....
###############################################################
opendir FH, "$dir1" || die $!;
my @filenames = readdir(FH);closedir FH;
foreach my $file(@filenames){
if(-e "$dir1/$file"){
         if (-M "$dir1/$file" > $age1) {
next if ($file =~ m/^\./);
open OUTPUT, (">>$log_file1") || die "Couldn't open log, $!";
unlink("$dir1/$file");
print OUTPUT "$file - deleted on $date.\n";
close OUTPUT || die "Couldn't close log, $!";}
          }
}
###############################################################
opendir FH, "$dir2" || die $!;
my @filenames = readdir(FH);closedir FH;
foreach my $file(@filenames){
if(-e "$dir2/$file"){
         if (-M "$dir2/$file" > $age2) {
next if ($file =~ m/^\./);
open OUTPUT, (">>$log_file2") || die "Couldn't open log, $!";
unlink("$dir2/$file");
print OUTPUT "$file - deleted on $date.\n";
close OUTPUT || die "Couldn't close log, $!"}
          }
}
###############################################################
# Finish script ....
###############################################################
print "Script Completed!\n";

And it works!

I have been testing on my workstation running Centos 5. I mount two CIFS shares which are hosted on NetApp Filers (one on the East Coast, one on the West Coast). It outputs the 2 text files which log all the deleted files.

I have yet to show this script to the Unix folks. And if they don’t laugh at me, I should be able to use in production. (I should be able to schedule this with cron when I roll this out into production. Also plan to use fstab to automate the CIFS mounts.)

Some future goals I have for this are:

• Output to html so Sys Ops can view the logs through browser instead of text file
• Email text files to Sys Ops so that each time script is run they receive an email