[PATCH 2/2] generalize keyboard detection code

[David Zeuthen]

On Tue, 2007-04-03 at 02:59 +0300, Guillem Jover wrote:
> Some devices do not have a proper keyboard, still they have keys, those
> should be handled by the input addon, but should not be considered a
> keyboard. This patch thus adds two new input capabilities, input.keys
> and input.keypad. A device with more than two keys will have the
> input.keys set and it might have input.keyboard, input.keypad or
> neither depending on the supported keys.

Sounds like a good idea.

> +static gboolean
> +input_test_keyboard (long *bitmask)
> +{
> +       int i;
> +
> +       for (i = KEY_Q; i <= KEY_P; i++) {
> +               if (!test_bit (i, bitmask))
> +                       return FALSE;

So here you are testing for Q, W, E, R, T, Y, U, I, P and if these keys
are present we're a keyboard? I guess that heuristic may fit the bill...

I committed your patch and then this patch on top of it

 http://gitweb.freedesktop.org/?p=hal.git;a=commitdiff;h=0391cb5e4e9a26d1c9494727999d38714cb3af18

to explain the various input.* capabilities. Thanks!

     David