* This is a very simple fix to a rather complicated issue. Essentially,
hold-taps will "release" (raise) their captured keys before actually
telling the event manager they have captured a key. This means the event
manager ends up assigning the `last_listener_index` to the hold-tap
subscription rather than the combo. So when the combo calls
`ZMK_EVENT_RELEASE` it raises after the hold-tap instead of after the
combo as the combo code expects.
* The corresponding test (which fails without this change) has also been added.
* An event can be captured and released in the same event handler, before
the last_listener_index would have been updated. This causes some handlers
to be triggered multiple times.
* The solution is to update the last_listener_index before calling the next
event handler, so capturing and releasing within an event handler is harmless.
* Also see discussion at https://github.com/zmkfirmware/zmk/pull/1401
* If our handler dedides our undedided hold-tap,
return early before continuing.
* Fix incorrect pointer logic, resulting in combo
candidate filtering leaving incorrect timeout details.
Co-authored-by: Andrew Rae <ajrae.nv@gmail.com>
Co-authored-by: okke <okke@formsma.nl>
* Allows removing multilib from docker images
* Run properly in aarch64 host docker
containers for testing on Rasberry Pi.
* Small sticky-keys fix to initialize w/ correct
constant for max uin32_t value.
The current combo completion check only makes sure the last key in the
combo is set. This works when the combo is typed correctly initially, or
when reraising events in a combo of length two. However, it fails for
longer combos since the last event in pressed_keys might be set, but the
first (or subsequent) event in pressed_keys can be NULL thanks to
release_pressed_keys.
Also added a regression test.
This test doesn't test a long combo that completes, but rather one
that's incomplete (so the combo doesn't trigger). Renaming to avoid
confusion when we add more long combo tests.
To properly retrigger hold-taps when a combo is not activated, some
position down events are reraised instead of released. The corresponding
position up events were never reraised, causing a potential stuck key.
this makes LS(LEFT_CONTROL) work as if shift and control were both
pressed explicitly. Previously, the left shift would have been released
as soon as another key was pressed. The implicit behavior is useful in
case of LS(NUMBER_1) when rolling over to other keys.
Also see #361.
