Shell Programming and Scripting

I have a script called script.sh that has the following in it.

cat /home/auto/script.sh

Code:

#!/bin/bash

ARGA=$1

if [ "${ARGA}" = "jsmith" ] ; then
        echo "${ARGA}:Confirmed"
else
        echo "${ARGA}:Unconfirmed"
fi

I need a way to run this script, and then, after script has confirmed the argument given to it, i'd like to comment out the section of the script that confirmed the argument (lines 2-9). in other words, i basically want to add a "#" to the beginning of lines 2 to 9. Then, after those lines are commented out, id like to append the word "date" to the end of the script. then run the script again.

I'm trying to do this all in one go. But i'm not sure how to go about it.

The logic should be something like this:

Code:

/home/auto/script.sh jsmith | sed -n 2-9,s~^*~#^*~g /home/auto/script.sh > /home/auto/script.sh | echo "date" >> /home/auto/script.sh ; /home/auto/script.sh

I'm aware this command is 100% wrong, but the idea of it is what i'm after. i'm hoping someone can re-write it for me. I want to be able to do this with a one-liner. or at least something close to it.

i would prefer not to have to create several little temp scripts which are to be deleted later (unless they can be created in memory, not hard disk). i want all this to be done within the "script.sh" script. but if there are no other solutions, i wouldn't mind it creating temps.

That's a very interesting idea, having shell scripts that more or less edit themselves. I do see a possible practical problem here. If the script comments out the test lines and adds 'date' at the end, what use is the script in the future? Just commenting those lines and adding 'date' doesn't seem useful, but there might be some other variation that is perhaps quite useful.

there is a use. but explaining the use will take forever. so i just want to get this small (seemingly small), section out of the way first before even attempting to go into deeper details. but if someone can accomplish the situation i laid out above here, everything else is simple.

i'm sure (at least i hope) someone on here knows of a way to do this.

Code:

$ /tmp/script.sh jsmith; sed "2,9 s/^/#/" /tmp/script.sh > /tmp/temp.x; mv /tmp/temp.x /tmp/script.sh; chmod 755 /tmp/script.sh; echo "date" >> /tmp/script.sh; /tmp/script.sh
jsmith:Confirmed
Sun Apr 21 13:35:52 PDT 2013

interesting!
is there a way to avoid the part i bolded in red? basically i want to (if i can), avoid creating these temp files.

If your sed allows sed -i then here is a simpler way:

Code:

$ /tmp/script.sh jsmith; sed -i "2,9 s/^/#/" /tmp/script.sh; echo "date" >> /tmp/script.sh; /tmp/script.sh
jsmith:Confirmed
Sun Apr 21 15:34:28 PDT 2013

sed (and perl) -i still create temp files. It just saves you the trouble of having to do it yourself.

A different approach:

Code:

sed '2,9 s/^/#/; $a\
date' script.sh | bash

Regards,
Alister