With commit
9ca87ab312
and
c62b0e5316
the winit event loop sends our KeyPress for
winit::event::WindowEvent::KeyboardInput only when the virtual key code
matches our specially encoded keys. For regular keys we rely on
winit::event::WindowEvent::ReceivedCharacter, for which we sent only
KeyRelease (which TextInput listens for), but not KeyPressed. With
this patch we simulate press & release.
* Rename logo to the more generic meta as keyboard modifier.
* Use control as the real modifier and map command key to it with
winit and keep Qt semantics as-is.
We rely on ReceivedCharacter to send the event to the item for
combinations like Ctrl+C, instead of the single KeyEvent from winit. We
use the latter only for special key codes.
This will give a nicer API to expose to .60. If the struct weren't
repr(C) then the booleans would be nicely packed, but alas that's not
happening. On the other hand we're not keeping many instances of them
around.
Fold CharacterInput into KeyPressed/KeyReleased and store the "key" as a string.
Also, instead of exposing the KeyCode we're encoding special characters
into the string.
Reduce the dependency of the GLRenderer to a new trait that exposes the
EventLoopTarget and EventLoopProxy. Those are provided by either an
winit::event_loop::EventLoop or, once the loop is started using the
self-consuming run(), by the event loop target provided to the run_fn.
This way renderers can be created from either within run or before.
The initial event loop instance is kept in TLS. When starting the loop,
it's taken out and instead the event loop target is placed into a scoped
tls variable.
