Guidance on implementing `KeyboardLayout` about pc-keyboard

[jonathanpallant]

probably because it was the first to be implemented?

exactly right.

In which cases is it appropriate to fall back on the Us104Key (or any other) layout?

I would say, whenever the majority of the keys are handled by the 'parent' variant. It didn't seem to make sense to duplicate the entire match for US and UK, and have to fix bugs in both, so UK started as a special case of US.

Then again, at least De105Key and No105Key duplicate certain KeyCode mappings

Yeah that's not ideal.

Does it make sense to create longer (with more than 2 layouts) layout chains?

Sure. I would prioritise code space over execution type for a piece of code that at most executes a couple of times per second.

Thoughts

Maybe one solution is to produce text-format tables of scan-code-to-Unicode-value and run a pre-build pre-processor that converts all of those tables into an optimally packed tree of match statements.

[SPuntte]

Thanks for the swift reply :)

I would say, whenever the majority of the keys are handled by the 'parent' variant. It didn't seem to make sense to duplicate the entire match for US and UK, and have to fix bugs in both, so UK started as a special case of US.

Agreed.

Sure. I would prioritise code space over execution type for a piece of code that at most executes a couple of times per second.

Sorry, I'm not sure if I understood that correctly.. Did you mean that, keyboard code being relatively "cold", you'd prefer smaller binary size over faster execution?

Thoughts

-- pre-build pre-processor --

Yup, sounds intriguing, and something that could be a fruitful exercise in DSL and compiler design. Maybe something like kbd could be considered – though dynamic and maybe a tad complex – a form of "prior art"?

Personally, I currently lack a realistic use for such a system. However, I could see the support for dead keys, commonly used on European layouts, being a natural driver for something like this.

[jonathanpallant]

Sorry, I'm not sure if I understood that correctly

It was a typo. I meant "execution time". So, yes, you interpreted my spelling mistakes correctly anyway.

[SPuntte]

Alright. I think I got what I was after so I'm going to close this. Cheers!