Hi Everyone,

I need some help on my bash script, I'm trying to rename a certain line in a file which might be 1 or more.

IDXCOUNT=`grep -c 'index .* on ' $FILENAME`;

for n in $(eval echo {1..$IDXCOUNT});
do
timestamp=$(date +"%s");
echo "Renaming index idx_$timestamp..";
if [ $n -eq 1 ]; then
sed -i "0,/^index [^)]* on /s/index idx_$timestamp on /" $FILENAME;

My problem is, if the sed target is 2 or more it will generate an idx_$timestamp that will cause duplicate index value when the script finish. My goal is to have the script recognize that when there are multiple index in the file, it will try to rename it one by one.

I'm new to bash so I'm not sure if I explained my issue well but I will appreciate any help!!

Thanks!

You mean in more than one place, the time stamp is identical? You're using a time stamp in seconds, and if it processes more than one such line in less than a second, you're going to have as many identical lines as the script can do within a second.

Your first problem is thinking you have only one ... I would suggest restating what it is you are trying to do, as the current script will only lead you to more bad scripting habits.

Please also use [code][/code] tags around code and data to make them more legible

grail has some good points.
I'd simplify the first line
You should also consider using 'set -vx' at the top of your program for debugging

Ok ... you went from asking a question to SOLVED. Good to see you worked it out, would you like to share so others may learn from you?