Re: PATCH: zsh-3.1.5-pws-11: _bindkey

[Peter Stephenson <pws@xxxxxxxxxxxxxxxxx>]

Sven Wischnowsky wrote:
> > +  compgen -b -M 'r:|-=* r:|=*'
> 
> I'd like to ask anyone if we should really do things like this.
> Personally I would prefer to let the user use the global match
> specs for things like this as otherwise the behavor might be somewhat
> unexpected.

Seems to me perfectly OK if it's documented.  It's much more natural here
than it is in file names, since the parts of widget names are often similar
to one another, so much so that it's unlikely to produce any completion you
weren't expecting, unless you mistyped.

--- Completion/Builtins/_bindkey.doc	Tue Mar  9 18:05:37 1999
+++ Completion/Builtins/_bindkey	Wed Mar 10 11:06:07 1999
@@ -1,5 +1,12 @@
 #defcomp bindkey
 
+# Normally, this completes names of zle widgets, whether the builtin ones
+# or ones defined by the user.  Note that a - allows a wildcard before it,
+# so h-b-s-b will complete to history-beginning-search-backward.  You
+# can alter this by removing the -M ... from the second compgen.
+#
+# Where appropriate, will complete keymaps instead of widgets.
+
 if [[ "$words[2]" = -*[DAN]* || "$words[CURRENT-1]" = -*M ]]; then
   compgen -s '$(bindkey -l)'
 else

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