Shell Programming and Scripting

I realize this is an old post, but since I was so frustrated when I was trying to figure out how to do this, I thought it might be valuable to someone else to see this answer to the problem.

This may be pretty ugly to sophisticated programmers. I'm very "wet behind the ears" in Bash, or anything else for that matter. By the way, this is done in a Debian etch environment.

The idea is to save the variable to a file, and then open the file with an editor, make the desired changes, save the result, and exit the editor.

There must be some shell gymnastics going on here that I don't understand, because until I specified the whole path (with pwd), the edit did not take place.

If you want the new variable to exist as a permanent part of your script, you'll have to figure that out yourself. In my (DAR backup) program, I'm saving my editable variables to external files.

Here's my example script. The variable begins with 3 errors to be edited. So you can test it 3 times before you get it right. At the end the temporary file is deleted, because by that time you will have placed the variable where you want it.:

Code:

#!/bin/bash

## To test the editing of a variable from inside a script
if [ -f `pwd`/editvarfile.txt ] ; then echo "" ; else echo "This is a variible ti be editied." > `pwd`/editvarfile.txt ; fi
editvar=$(cat editvarfile.txt)
pickans="n"
while [ "$pickans" = "n" ] ; do
echo -n "The present variable is "$editvar".  Is this correct?  (y/n) "
read pickans
if [ "$pickans" = "n" -o "$pickans" = "N" ] ; then
echo ${editvar} > editvarfile.txt
kwrite editvarfile.txt
editvar=$(cat editvarfile.txt)
fi
done
rm editvarfile.txt