How to convert this script to perl

[rchilson11]

DT=`echo $1`
NOTIFY=rchilson@hotmail.com
PROD_DIR=/apps*
if [ "$DT" = '' ] || [ "$DT" = ' ' ] ; then
DTPULL=`date -d yesterday '+%Y-%m-%d'`
else
DTPULL=`date --date "$DT" '+%Y-%m-%d'` || exit -1
fi
(
TOT=`dsh -g prod "gzcat /apps*/servlet-timer.log.2010-12-13* | grep 2009" | wc -l`
echo "Transaction Report: $DTPULL $HOSTNAME"
echo ""
echo " 2009............................................:$TOT
echo ""
) | mailx -s "Prod Transaction Stats for: $DTPULL" $NOTIFY

 

[rchilson11]

I want to learn perl and figured I could learn best by examples. I have been reading tech articles and I would like to know the easiest way to convert this bash script to perl. Thanks!!

 

[rchilson11]

this is how far I got..

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

# Specify a date
my $date = shift || undef;

#Search string as the script argument
my $2009 = shift || die "No Cardholder2010 found!\n";

# Get the gzipped log files as an array
my @files = glob '/apps/servlet-timer.log*' .
(defined $date ? $date : '') . '.gz';


# Loop on the log files, open and read them, and
# print out any lines with given Search string in them
foreach my $file (@files) {
open FILE, "gunzip -c $file|" or die $!;

while (my $line = <FILE>) {
print $line if index($line, '"' . $2009 . '"') >= 0;
}

close FILE;
}

 

[MikeLacey]

Nicely done, so far as I can see. Does it behave as you expect so far? I particularly like the clear layout of the script, v cool and easy to read.

Two things.

1 - the variable $2009, I don't know if that (a variable name made all of digits) will work or not, change to something starting with a letter if it's a problem, like this: $yr2009

2 - I'd suggest one change which might make it a bit more Perl ish, and it will run a little faster as well.

print $line if $line =~ /$2009/;

So the next step then - to store and email the results to someone?

Mike

http://www.myspace.com/micahhowzat