Re: [PATCH] Trapping exit in tests, using return for errors

From: Junio C Hamano <junkio@cox.net>
Date: Aug 11, 2005, 06:06 UTC
Message-ID: <7virydnh1o.fsf@assigned-by-dhcp.cox.net>
In-Reply-To: <1123732581.18644.37.camel@dv>

Pavel Roskin <proski@gnu.org> writes:

> This patch does following:
>
> All instances of "exit", "exit 1" and "(exit 1)" in tests have been
> replaced with "return 1".  In fact, "(exit 1)" had no effect.

Are you sure about all of the above?

You are right about "... || exit" in the expect_success tests; they are broken. But '(exit 1)' or just saying 'false' at the end should have done the right thing. Worse yet, return does not seem to really work as expected, except if all you want to do was to tell the test driver "I failed", in which case "bloopl" would work just as well.

prompt$ cat k.sh
what="$1"
eval '
        echo foo
        case '$what' in
        false)
                (exit 1) ;;
        exit)
                exit 1 ;;
        return)
                return 1 ;;
        esac
'
echo "status $?"
prompt$ bash k.sh exit
foo
prompt$ bash k.sh false
foo
status 1
prompt$ bash k.sh return
foo
k.sh: line 20: return: can only `return' from a function or sourced script
status 1
prompt$