The default `onSetup` will call `.begin`, to support initializing plugins using
the V1 plugins while using `KALEIDOSCOPE_INIT_PLUGINS`. However, plugins that
implement a compatibility layer so that they can be used with both the new API,
and with `Kaleidoscope.use()` will have a `.begin` method too. Which the default
`onSetup` will call, and we'll register the compatibility layer too, in addition
to the new-style event handlers. This results in many things running twice,
which leads to all kinds of problems.
For this reason, override `onSetup`, so that it does not call `begin`. When used
with `Kaleidoscope.use()`, the plugin will still work, so compatibility is
maintained. But the bug is now gone.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Unfortunately, the way we reported the use of an old-style API also triggered
when not using it. Change that to only trigger when we DO use the old API, by
marking the `combo_t` constructor private.
Unfortunately, this does not allow us to use a custom error message.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of manually counting set bits, use gcc's `__builtin_popcount()` for a -
supposedly - more efficient count.
Also initialize the counter to zero, while there.
Fixes#11.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of iterating through all the bits, use `__builtin_popcountl()`, provided
by gcc, which should be considerably more efficient.
Fixes#27.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
For a while now, Kaleidoscope does nothing if keys are idle. This example relied
on events being fired in the idle case too. So instead of relying on that, move
the holding logic to `loop()`.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
`Consumer_SNapshot` really should have been `Consumer_Snapshot`, so lets fix it.
However, the typo'd name is left in place as an alias, in order to not break any
existing user code, however unlikely the use of this key is.
Sadly, we can't easily add a deprecation warning, because key_defs.h, which
defines `Key`, depends on `key_defs_consumerctl.h`, so we'd end up with a
circular dependency if we tried to add a deprecation.
Fixes#339.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
As per UPGRADE.md, remove `Kaleidoscope.setup(KEYMAP_SIZE)`,
`event_handler_hook_use`, `loop_hook_use`, `USE_PLUGINS`, `MOMENTARY_OFFSET`,
`key_was_pressed`, `key_is_pressed`, `key_toggled_on`, and `key_toggled_off`.
These were deprecated between July and October 2017, and have been marked for
deletion for over a month.
Also updated UPGRADE.md, moving the section about these to a new, "deprecated
and removed" section.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
When using the old API, fail with a helpful error message that points to
`UPGRADING.md`, which explains in simple detail the migration process.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
The old `RxCy` macros were recently changed to be key indexes instead of
per-hand bit indexes, and `KeyboardHardware.isKeyswitchPressed()` and
`KeyboardHardware.pressedKeyswitchCount()` were introduced as a way to peek into
the keyswitch state. This little change migrates TestMode to use them.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of using a list of left/right-hand states and an overrideable global
callback, use a map of action and key-list pairs. This makes the plugin much
more portable, does not require any hardware-specific knowledge within the
plugin, and does not require us to treat the hands separately.
This in turn, results in a friendlier user interface, at the cost of limiting
the maximum length of a combination to five keys. A small price to pay.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of `getKeyswitchStateAtPosition`, which is long, unintuitive and feels
wrong too, introduce `isKeyswitchPressed`, shorter, better, more
reasonable (because it returns a bool - we support only two states anyway!).
Additionally, add `pressedKeyswitchCount()`, which returns the number of key
switches pressed.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of `getKeyswitchStateAtPosition`, which is long, unintuitive and feels
wrong too, introduce `isKeyswitchPressed`, shorter, better, more
reasonable (because it returns a bool - we support only two states anyway!).
Additionally, add `pressedKeyswitchCount()`, which returns the number of key
switches pressed.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
This is sufficiently low-level that it is OK to use `KeyboardHardware` for it.
It's not a HID thing, either.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Move `KEY_INDEX` here, and turn it into a `constexpr` function, `keyIndex`, and
convert the `RxCy` macros to constexpr values instead.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
To allow the hardware plugin to use a more efficient way of representing the
index (if need be), and to be able to turn it into a constexpr, move it to the
hardware plugin.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
To be used by the hardware implementations, `KEY_INDEX` tells us the index of a
key, from which we can figure out the row and column as needed. The index starts
at one, so that plugins that work with a list of key indexes can use zero as a
sentinel. This is important, because when we initialize arrays with fewer
elements than the declared array size, the remaining elements will be zero. We
can use this to avoid having to explicitly add a sentinel in user-facing code.
Additionally, we add `getKeyswitchStateAtPosition` to the HID facade. See its
documentation in `Kaleidoscope-Hardware`.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Implement `getKeyswitchStateAtPosition`, a hardware-agnostic way to peek into
the keyswitch state. Also transition the `RxCy` macros to `KEY_INDEX`, to make
it easier for hardware with more keys than 64 to implement them, and to make
their values unique across a keyboard, not just across a keyboard half.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Instead of only having an ifdef in the body of deprecated methods, and leaving
it up to the compiler to optimize out the empty & unused, explicitly wrap the
declaration of them within an ifdef too. This will make it easier to remove
everything V1 at a later point, and we're not at the mercy of the compiler,
either.
Fixes#327.
Signed-off-by: Gergely Nagy <algernon@keyboard.io>
Gergely Nagy
Jesse Vincent
Jack Zhou
Gergely Nagy