From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-return-6525-mason-zsh=primenet.com.au@sunsite.auc.dk>
Received: (qmail 4984 invoked from network); 8 Jun 1999 12:35:39 -0000
Received: from sunsite.auc.dk (130.225.51.30)
  by ns1.primenet.com.au with SMTP; 8 Jun 1999 12:35:39 -0000
Received: (qmail 556 invoked by alias); 8 Jun 1999 12:35:21 -0000
Mailing-List: contact zsh-workers-help@sunsite.auc.dk; run by ezmlm
Precedence: bulk
X-No-Archive: yes
X-Seq: 6525
Received: (qmail 549 invoked from network); 8 Jun 1999 12:35:17 -0000
Message-Id: <9906081206.AA21204@ibmth.df.unipi.it>
To: zsh-workers@sunsite.auc.dk (Zsh hackers list)
Subject: PATCH: pws-21: parentheses in command word
Date: Tue, 08 Jun 1999 14:06:39 +0200
From: Peter Stephenson <pws@ibmth.df.unipi.it>

I mentioned this a while ago, but can't find that in the archive, because I
can never find things.

For some reason, parentheses are treated specially in the command word in
zsh, not just at the start (where, of course, they start a subshell).  I
just found a concrete example of where this causes problems:

% ./*(/om[1])<TAB>

should complete to the most recently modified directory.  Instead, the
command word is taken to be `./*' and the `(/om[1])' an argument.

I thought maybe it was for compatility with other shells, but it seems to
have the opposite effect.  Here's a stupid (but perfectly valid) example:

In ksh:
$ touch print
$ @(print|noprint) foo
foo
$ mv print noprint
$ @(print|noprint) foo
ksh: noprint: permission denied

With zsh running as ksh, the @ is taken to be the command.  This is bad.
(Presumably this is another reason why ksh doesn't use naked parentheses
for globbing.)

I suggest trying with this `feature' disabled, unless someone can come up
with a good reason for it.

*** Src/lex.c.icp	Mon Jan 25 14:59:49 1999
--- Src/lex.c.orig	Tue Jun  8 12:28:02 1999
***************
*** 898,906 ****
--- 898,917 ----
  		    e = hgetc();
  		    hungetc(e);
  		    lexstop = 0;
+ #if 1
+ 		    /* For command words, parentheses are only
+ 		     * special at the start.  But now we're tokenising
+ 		     * the remaining string.  So I don't see what
+ 		     * the old incmdpos test here is for.
+ 		     *   pws 1999/6/8
+ 		     */
+ 		    if (e == ')')
+ 			goto brk;
+ #else
  		    if (e == ')' ||
  			(incmdpos && !brct && peek != ENVSTRING))
  			goto brk;
+ #endif
  		}
  		pct++;
  	    }

-- 
Peter Stephenson <pws@ibmth.df.unipi.it>       Tel: +39 050 844536
WWW:  http://www.ifh.de/~pws/
Dipartimento di Fisica, Via Buonarroti 2, 56127 Pisa, Italy