Shell Programming and Scripting

HI all
I am working on a script, few details are as follows. I have one properties File and one script. The property file contains the JOBID which are to be executed and the Script runs these jobs ONE by ONE. After completing the JOB I need to comment that job in the property File. This is the step where I am getting stucked up.

Also I need to uncomment the Jobs after completion of the script

Note:: The point is that once a job is completed its commented and if say the script fails then it will start from the point where it ended.

Contents are as follows
1. Properties file

EODJOB=6000
EODJOB=6001
EODJOB=6002
EODJOB=1010
EODJOB=1070
EODJOB=1071
EODJOB=6003
EODJOB=6004
EODJOB=6005

2. Script

#! /usr/bin/ksh
# A test to automate the whole EOD

echo "Starting FX Front Office EOD..."

brn_num=$1
set -e on
error.txt 2>&1

rootdir=$PWD
propfile=$rootdir/FX_EOD_JOB.properties

if [ `grep -c "^EODJOB=" $propfile` -gt 0 ]
then
echo "Running the EOD Process"
for i in `grep "^EODJOB=" $propfile | cut -f2 -d"="`
do
#. ./runEod.sh $i $brn_num>> error.txt 2>&1
if [ $# -eq 0 ] #ok execution
then
#############################
Code to comment the Line in Properies file.
Say fist job picked is 6000 so after execution
the Corrsponding job should be commented. Like
#EODJOB=6000

#############################

fi

echo $i Process completed for Branch No:=$brn_num
done
#############################
Code to uncomment all the Jobs in .properties file
#############################
fi




Kindly help with ur valuable information

I would suggest you keep two files instead; one with the original queue and another which lists all completed jobs. Then you can find which ones are still not run by removing all the completed ones from the queue, and if any remain, you run them. If the files are sorted, some things are easier. If you always proceed from the first line to the next, some things are easier still. Commenting out stuff in the queue file seems error-prone and subject to race conditions.

Many thanks for the reply
Will try out that option as well
However Just for my Knowledge.
How can I update a line in UNIX?

Yes, but generally it means re-writing the entire file. Most tools (sed, perl) create a temporary file, and after processing is complete, moves the temp file to the original one.

It's also possible to open files for reading and writing, but in this case, you'll still have to re-write the lines after the change (because you inserted a character). If your lines began with, let's say, a single space, you could replace the space with a # using this technique. Then you could do a read-write in-place, changing the space for the # and without re-writing the file. You can do this in perl or C.