NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/comp/unix/shell/5342

Internet Message Format | 1993-01-11 | 1.5 KB

Path: sparky!uunet!spool.mu.edu!agate!usenet.ins.cwru.edu!odin!chet
From: [e-mail] (Chet Ramey)
Newsgroups: comp.unix.shell
Subject: Re: Bourne shell functions; what of 'exit'?
Date: 11 Jan 1993 19:00:35 GMT
Organization: Case Western Reserve University, Cleveland OH (USA)
Lines: 38
Message-ID: <[e-mail]>
References: <[e-mail]>
NNTP-Posting-Host: odin.ins.cwru.edu

In article <[e-mail]> [e-mail] writes:
$ Bourne shell functions seem to be inconsistent in the way they handle an 'exit'
$ statement.  For example, the following:
$
$     func() {
$         .
$         .
$         exit 1
$     }
$
$     func
$
$ causes the whole shell script to exit, while this:
$
$     func() {
$         .
$         .
$         exit 1
$     }
$
$     junk=`func`
$     echo $?
$
$ doesn't.  At first I thought this meant func was executed in a subshell, but
$ echoing $$ both places gives the same PID.  Is this context-sensitive treatment
$ of 'exit' a feature?  Can it be counted on in portable scripts?

The command substitution is indeed carried out in a subshell (what Posix.2
terms a `subshell environment').  $$ is the same in both shells.  This is
traditional sh behavior and is specified in Posix.2, so I think it can be
counted on for portable shell scripts.

Chet
--
``The use of history as therapy means the corruption of history as history.''
    -- Arthur Schlesinger

Chet Ramey, Case Western Reserve University    Internet: [e-mail]