Re: PATCH: 3.1.6-pws-6: compdef -K

[Adam Spiers <adam@xxxxxxxxxxxxxxxxxxxxxxx>]

Peter Stephenson (pws@xxxxxxxxxxxxxxxxx) wrote:
> This implements compdef -K, which is like compdef -k but you can define
> multiple widgets to use the same function.  One use of this is to create
> completion and listing widgets simultaneously, but I've used it to make
> _history_complete_word work better: it now defines two widgets,
> _history-complete-older and _history-complete-newer, which it can use to
> decide which way it's going.

Nice; thanks Peter.  I'm a bit snowed under at the moment, but I will
fix _history_complete_word ASAP (I hadn't twigged that you need to
deal with when compstate[insert] is set to something other than a
number).

> for some other reason even
> with history_stop set it won't stop cycling through them.

Does it stop in one direction only, or neither?  If only one, it's
because of compstate[nmatches] returning all matches before duplicate
removal.  If neither, my code is even worse than I thought, which
wouldn't surprise me much :-)