From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from euclid.skiles.gatech.edu (list@euclid.skiles.gatech.edu [130.207.146.50]) by melb.werple.net.au (8.7.5/8.7.3/2) with ESMTP id BAA18835 for <mason@werple.mira.net.au>; Sat, 29 Jun 1996 01:21:15 +1000 (EST)
Received: (from list@localhost) by euclid.skiles.gatech.edu (8.7.3/8.7.3) id LAA08924; Fri, 28 Jun 1996 11:08:51 -0400 (EDT)
Resent-Date: Fri, 28 Jun 1996 11:08:51 -0400 (EDT)
From: Zoltan Hidvegi <hzoli@cs.elte.hu>
Message-Id: <199606281508.RAA01280@bolyai.cs.elte.hu>
Subject: Re: arr=('') broken
To: pws@ifh.de (Peter Stephenson)
Date: Fri, 28 Jun 1996 17:08:26 +0200 (MET DST)
Cc: zsh-workers@math.gatech.edu
In-Reply-To: <199606281452.QAA23567@hydra.ifh.de> from Peter Stephenson at "Jun 28, 96 04:52:14 pm"
Organization: Dept. of Comp. Sci., Eotvos University, Budapest, Hungary
Phone: (36 1)2669833 ext: 2667, home phone: (36 1) 2752368
X-Mailer: ELM [version 2.4ME+ PL16 (25)]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Resent-Message-ID: <"jBpxL1.0.MB2.3M_qn"@euclid>
Resent-From: zsh-workers@math.gatech.edu
X-Mailing-List: <zsh-workers@math.gatech.edu> archive/latest/1460
X-Loop: zsh-workers@math.gatech.edu
Precedence: list
Resent-Sender: zsh-workers-request@math.gatech.edu

> Assigning empty strings to arrays (set -A still works OK) was broken
> by the changes in untokenisation.  This broke my multicomp function.
> 
> % sofar=('')
> % print $#sofar
> 0
> 
> The patch should be something like the following.  I've written it
> this way on the assumption that if we get as far as the array code,
> the list must have passed through globlist and hence has had unwanted
> elements pruned already.  As I understand it, globlist() now always
> returns untokenised text, and one needs to be careful not to
> untokenise twice which can in principal do bad things to
> eight-bit-cleanliness, which is why I added the first hunk and removed
> the untokenize() in the second hunk.

I also discovered this bug and fixed yesterday :-) in a bit different way.
There is no problem calling untokenize twice.  Untokenize does not affect
8-bit cleannes.  8-bit conversion is only done when zsh communicates with
the outer world.

zsh-3.0-pre1 will be out today which will have this fixed.

Zoltan