From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 14871 invoked from network); 25 Jul 2000 07:23:44 -0000 Received: from sunsite.auc.dk (130.225.51.30) by ns1.primenet.com.au with SMTP; 25 Jul 2000 07:23:44 -0000 Received: (qmail 14738 invoked by alias); 25 Jul 2000 07:23:35 -0000 Mailing-List: contact zsh-workers-help@sunsite.auc.dk; run by ezmlm Precedence: bulk X-No-Archive: yes X-Seq: 12363 Received: (qmail 14719 invoked from network); 25 Jul 2000 07:23:34 -0000 Date: Tue, 25 Jul 2000 09:23:31 +0200 (MET DST) Message-Id: <200007250723.JAA06371@beta.informatik.hu-berlin.de> From: Sven Wischnowsky To: zsh-workers@sunsite.auc.dk In-reply-to: "Andrej Borsenkow"'s message of Tue, 25 Jul 2000 11:09:58 +0400 Subject: RE: noglob and _match Andrej Borsenkow wrote: > > Andrej Borsenkow wrote: > > > > > bor@itsrm2% noglob ls *TAB > > > > > > bor@itsrm2% noglob ls Brodsky/ > > > Completing file > > > Brodsky/ addrbook.d/ dikz/ observe/ test/ > > > ..... > > > > > > Is it correct in this case? I suspect, the same aplies to _expand as > > > well. > > > > Why not? I mean: you explicitly requested completion, so the > > completion system *must* think that you know what you're doing. > > > > In this case I want to list all files that start with `*' :-))) Ah, right, sorry, stupid me. Ok, first, with a completer with _match after _complete (where it belongs) and files starting with a `*', `ls *' *does* give me only the files matching `\**' as possible completions. If I have _expand in the completer (before _complete, of course), that takes precedence and I get the possible expansions in a menu completion. But even for _expand it can be fixed, of course: zstyle -e :completion::::: completer \ 'if [[ $words[1] = noglob ]]; then reply=(_complete) else reply=(_expand _complete _match) fi' And because of this possibility and because being able to selectively expand glob patterns for noglob I wouldn't like to have that changed. Especially because it would then probably be harder or even impossible to get the current behaviour by setting some styles. Except, of course, we add yet another style that says if it should behave as it does now or in the way you want it. Bye Sven -- Sven Wischnowsky wischnow@informatik.hu-berlin.de