Re: completion match ordering

[dana <dana@xxxxxxx>]

On 25 Nov 2018, at 19:25, Oliver Kiddle <okiddle@xxxxxxxxxxx> wrote:
>Any thoughts on the interface? Is something more terse like -on
>preferable?

Experimenting with it in (what would be) 'real-world' applications, i do find it
kind of wordy, and if/when additional ordering options (like reverse) are added
it'll be even more so. Maybe a compromise would be possible, where the ordering
options can be abbreviated like -onum,rev? I could be over-complicating it,
though.

On 25 Nov 2018, at 19:25, Oliver Kiddle <okiddle@xxxxxxxxxxx> wrote:
>Besides match, nosort and numeric, should any other orderings be
>supported? Reverse or case-sensitive perhaps?

I like the idea of reverse and case modifiers as i mentioned in our other
conversation. But i suppose in a lot of cases it'll be simple enough to just do
`compadd -o nosort - ${(Oi)foo}` or whatever, so... it'd be cool, but not
necessarily critical.

Other than that, i didn't test *super* extensively, but it seems to work well
for the sort of thing that inspired my original question (keeping numbers in the
right order, as in _lz4).

dana