Shell Programming and Scripting

Man...no one answered my last two posts...Oh well, so hey guys i have a script that i'm wanting to put a # on a certain line, and then execute the script but not write it to the script. So far I'm doing it the dirty way with:

Code:

sed -i -e "s/^rm/\#rm/" -e "s/^tar/\#tar/" /path/to/file
                    /path/to/file
                    sed -i -e "s/^\#rm/rm/" -e "s/^\#tar/tar/" /path/to/file

Just seems like there's gotta be a better way. I don't know how to get the changed file to execute. Sed just wants to list it. Help. Thanks

Pipe the output from sed to a shell.

Code:

sed '....' orig-script | ksh

As the file is a shellscript of some kind make the script do the work:

Code:

myscript [switches] args...        # does the rm/tar
myscript -n [switches] args ...    # do not do rm/tar

At the start of the script:

Code:

noop=""

In the getopts section (normally a case statement in a getopts loop):

Code:

n) noop=":" ;;

If you are not using a getopts loop you may change this to suit your program.
Prefix the rm/tar statements with $noop:

Code:

$noop rm ....

The colon acts as a "no-operation" operator. Watch out for side-effects, eg

Code:

${some:=thing}

will be evaluated ($some will be set to thing if null).

Andrew

definitely using a getopts loop. I think piping this to a shell is the way to go for me. Thanks you guys