From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <zsh-workers-return-27391-mason-zsh=primenet.com.au@zsh.org>
Received: (qmail 11587 invoked by alias); 12 Nov 2009 02:57:18 -0000
Mailing-List: contact zsh-workers-help@zsh.org; run by ezmlm
Precedence: bulk
X-No-Archive: yes
List-Id: Zsh Workers List <zsh-workers.zsh.org>
List-Post: <mailto:zsh-workers@zsh.org>
List-Help: <mailto:zsh-workers-help@zsh.org>
X-Seq: 27391
Received: (qmail 28298 invoked from network); 12 Nov 2009 02:57:15 -0000
X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on f.primenet.com.au
X-Spam-Level: 
X-Spam-Status: No, score=-2.6 required=5.0 tests=AWL,BAYES_00 autolearn=ham
	version=3.2.5
Received-SPF: none (ns1.primenet.com.au: domain at klanderman.net does not designate permitted sender hosts)
From: Greg Klanderman <gak@klanderman.net>
To: zsh-workers@zsh.org
Subject: Re: bug in 'rm' completion
Reply-To: gak@klanderman.net
Date: Wed, 11 Nov 2009 21:48:48 -0500
In-Reply-To: <091110075221.ZM27832@torch.brasslantern.com> (Bart Schaefer's message of "Tue, 10 Nov 2009 07:52:21 -0800")
Message-ID: <m3d43omp9r.fsf@klanderman.net>
User-Agent: Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.4.17 (linux)
References: <19191.43212.832827.724369@gargle.gargle.HOWL>
	<m38wef4qqs.fsf@klanderman.net>
	<091109092926.ZM26572@torch.brasslantern.com>
	<m3639j4mcl.fsf@klanderman.net> <m3iqdjmm31.fsf@klanderman.net>
	<091110075221.ZM27832@torch.brasslantern.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii

>>>>> On November 10, 2009 Bart Schaefer <schaefer@brasslantern.com> wrote:

> } +    ignored=(${line[1,CURRENT-1]//(#m)[\[\]()\\*?#<>~\^]/\\$MATCH})
> } +    ignored+=(${line[CURRENT+1,-1]//(#m)[\[\]()\\*?#<>~\^]/\\$MATCH})

> I'd be inclined to wrap that in a check that ((CURRENT > 1))

As long as you mean to wrap just the first line with that check..
otherwise it won't work correctly, for example:

% rm -rf <tab> foo bar baz

> just so
> it's obvious that we don't ignore the only word on the line, rather
> than rely on $line[1,0] to do the right thing, but otherwise this
> seems right to me.

Do you want the analogous check wrapped around the second line?
Can we rely on $line[$#line+1,-1] to do the right thing?  If not
then presumably you want something like the new patch below..

thanks,
Greg


Index: Completion/Unix/Command/_rm
===================================================================
RCS file: /cvsroot/zsh/zsh/Completion/Unix/Command/_rm,v
retrieving revision 1.2
diff -u -r1.2 _rm
--- Completion/Unix/Command/_rm	23 Nov 2008 18:23:29 -0000	1.2
+++ Completion/Unix/Command/_rm	12 Nov 2009 02:47:20 -0000
@@ -5,7 +5,7 @@
   '(-f --force)'{-f,--force}'[ignore nonexistent files, never prompt]'
   '(-I --interactive)-i[prompt before every removal]'
   '(-r -R --recursive)'{-r,-R,--recursive}'[remove directories and their contents recursively]'
-  '*:files:->file'
+  '*::files:->file'
 )
 if _pick_variant gnu=gnu unix --help; then
   opts+=(-S)
@@ -33,7 +33,11 @@
 case $state in
   (file)
     declare -a ignored
-    ignored=(${line//(#m)[\[\]()\\*?#<>~\^]/\\$MATCH})
+    ignored=()
+    (( CURRENT > 1 )) &&
+      ignored+=(${line[1,CURRENT-1]//(#m)[\[\]()\\*?#<>~\^]/\\$MATCH})
+    (( CURRENT < $#line )) &&
+      ignored+=(${line[CURRENT+1,-1]//(#m)[\[\]()\\*?#<>~\^]/\\$MATCH})
     _files -F ignored && ret=0
     ;;
 esac