Re: PATCH: completion or'ing and grouping

[Sven Wischnowsky <wischnow@informatik.hu-berlin.de>]

Peter Stephenson wrote:

> 
> Excellent, everything seems to be working at the moment with the new
> patches (though I expect you'll be hearing from me again...).
> For some reason the mail system screwed up that last patch and I
> get to get it from the archive.  Strange.  (Is there a way of
> retrieving a patch as plain text, i.e. before HTMLifying?  Otherwise
> back-conversion is liable to screw up whitespace.)
> 

Interesting, I got your message 4518 screwed up (in two parts).

> One thing worried me a bit: this hunk from zle_tricky.c
> 
> *** 4791,4796 ****
> --- 4800,4808 ----
>   	int ics = cs, ocs, pl = 0, l, lp, ls;
>   	char *ps;
>   	Cline lc;
> + 
> + 	if (!ainfo)
> + 	    return;
>   
>   	fixsuffix();
> 
> didn't agree with what I had before, which came from the
> patch-match.gz patch (natch).  Namely, the fixsuffix() wasn't there in
> what that patch introduced.  I just added it by hand.
> 

It's from the patch in message 4528.

> Secondly, after I applied patch-match.gz, completion did nothing
> whatsoever.  After panicking I remembered compctl -T.  But according
> to my initialisation files, I never had a compctl -T before, and when
> zsh started up, compctl -T -q was in effect (why?)  If this is
> intentional, hadn't that better now be compctl -T -q -tc?  Otherwise
> we're going to have a very large number of irate users.

I don't understand this one. In my patched version the -T completion
is correctly set to `compctl -T -tc' by default.

Bye
 Sven


--
Sven Wischnowsky                         wischnow@informatik.hu-berlin.de