The **keymap** is an array composed of one or more layers.
Each **layer** is an array of **keycodes**, defining **actions** for each physical key.
Layers can be activated and deactivated independently. Multiple layers may be active at once, resulting in the currently-active **layer state**. Each layer has an index between 0-31. As active layers are stacked together, higher layers take precedence over lower layers.
The current keymap layer state is determined by two parameters: the *default layer*, and the individual *layer states*. Changing the default layer is useful for switching key layouts completely; for example, switching to Dvorak, Colemak or Workman instead of QWERTY. Individual layer states, on the other hand, can be used to overlay the base layer with other functions such as navigation keys, function keys (F1-F12), media keys or other actions.
Because the default layer is really just a special case affecting the overall layer state, it is important to first understand how the layer state is determined.
The **layer state** indicates the current on/off status of all layers. It is defined in the firmware by a 32-bit integer, `layer_state`, which stores each layer's on/off status in a single bit: 0 for off, 1 for on. As layers are activated and deactivated, their respective bits are flipped, changing the value of `layer_state`.
The **default layer** is the base keymap layer (0-31) which is always active and considered the "bottom" of the stack. When the firmware boots, the default layer is the only active layer. It is set to layer 0 by default, though this can be changed ~~in *config.h*~~ via Boot Magic settings.
Initial state of Keymap Change base layout
----------------------- ------------------
31 31
30 30
29 29
: :
: : ____________
2 ____________ 2 / /
1 / / ,->1 /___________/
,->0 /___________/ | 0
| |
`--- default_layer = 0 `--- default_layer = 1
layer_state = 0x00000001 layer_state = 0x00000002
Note that the `default_layer_state` variable only determines the lowest value to which `layer_state` may be set, and that `default_layer_state` is used by the core firmware when determining the starting value of `layer_state` before applying changes. In other words, the default layer will *always* be set to *on* in `layer_state`.
The default layer is defined in the firmware by the `default_layer_state` variable, which is identical in format to the `layer_state` variable exlpained above. The value may be changed using the following functions:
-`default_layer_state_set(state)` sets the state to the specified 32-bit integer value.
- AND/OR/XOR functions set the state based on a boolean logic comparison between the current state and the specified 32-bit integer value:
-`default_layer_state_and(state)`
-`default_layer_state_or(state)`
-`default_layer_state_xor(state)`
For example, to set layer 3 as the default layer:
```C
// convert 3 to a 32-bit unsigned long value, and set the default layer
Note that ***higher layers have priority in the layer stack***. The firmware starts at the topmost active layer, and works down to the bottom to find the an active keycode. Once the search encounters any keycode other than **`KC_TRNS`** (transparent) on an active layer, the search is halted and the remaining lower layers aren't examined, even if they are active.
**Note:** a layer must be activated before it may be included in the stack search.
`KC_TRNS` is a special placeholder which can be used on overlay layers. This allows for the creation of "partial" layers which fall back on the lower layers, eliminating a good deal of repetition in keymap files.
The keymap is defined in the **`uint8_t keymaps[]`** array, a 2-dimensional array of rows and columns corresponding to positions in the keyboard matrix. But most often the layers are defined using C macros to allow for easier reading and editing of the keymap files. To use complex actions you need to define `Fn` action in the **`action_t fn_actions[]`** array.
`Fn1` is a momentary layer switching key with tapping function--tapping the key as one would normally use it, sends the semicolon **';'** keycode, while holding the key down switches layers.
See [`common/keycode.h`](../common/keycode.h) or keycode table below for the detail. Keycode is internal **8bit code** to indicate action performed on key in keymap. Keycode has `KC_` prefixed symbol respectively. Most of keycodes like `KC_A` have simple action registers key to host on press and unregister on release, while some of other keycodes has some special actions like `Fn` keys, Media control keys, System control keys and Mousekeys.
***In `KEYMAP()` macro you should omit prefix part `KC_` of keycode to keep keymap compact.*** For example, just use `A` instead you place `KC_A` in `KEYMAP()`. Some keycodes has 4-letter **short name** in addition to descriptive name, you'll prefer short one in `KEYMAP()`.
### 1.0 Other key
-`KC_NO` for no action
-`KC_TRNS` for layer transparency (See above)
### 1.1 Normal key
-`KC_A` to `KC_Z`, `KC_1` to `KC_0` for alpha numeric key
`KC_FNnn` are keycodes for `Fn` key which not given any actions at the beginning unlike most of keycodes has its own inborn action. To use these keycodes in `KEYMAP()` you need to assign action you want at first. Action of `Fn` key is defined in `action_t fn_actions[]` and its index of the array is identical with number part of `KC_FNnn`. Thus `KC_FN0` keycode indicates the action defined in first element of the array. ***32 `Fn` keys can be defined at most.***
See keycode table in [`doc/keycode.txt`](./keycode.txt) for description of keycodes.
In regard to implementation side most of keycodes are identical with [HID usage][HID_usage](pdf) sent to host for real and some virtual keycodes are defined to support special actions.
See [`common/action_code.h`](../common/action_code.h). Action is a **16bit code** and defines function to perform on events of a key like press, release, holding and tapping.
Most of keys just register 8bit scancode to host, but to support other complex features needs 16bit extended action codes internally. However, using 16bit action codes in keymap results in double size in memory compared to using just keycodes. To avoid this waste 8bit keycodes are used in `KEYMAP()` instead of action codes.
***You can just use keycodes of `Normal key`, `Modifier`, `Mousekey` and `System & Media key` in keymap*** to indicate corresponding actions instead of using action codes. While ***to use other special actions you should use keycode of `Fn` key defined in `fn_actions[]`.***
### 2.1 Key Action
This is a simple action that registers scancodes(HID usage in fact) to host on press event of key and unregister on release.
***This action usually won't be used expressly in keymap*** because you can just use keycodes in `KEYMAP()` instead.
You can define these actions on *'A'* key and *'left shift'* modifier with:
ACTION_KEY(KC_A)
ACTION_KEY(KC_LSFT)
#### 2.1.2 Modified key
This action is comprised of strokes of modifiers and a key. `Macro` action is needed if you want more complex key strokes.
Say you want to assign a key to `Shift + 1` to get character *'!'* or `Alt + Tab` to switch application windows.
ACTION_MODS_KEY(MOD_LSFT, KC_1)
ACTION_MODS_KEY(MOD_LALT, KC_TAB)
Or `Alt,Shift + Tab` can be defined. `ACTION_MODS_KEY(mods, key)` requires **4-bit modifier state** and a **keycode** as arguments. See `keycode.h` for `MOD_BIT()` macro.
ACTION_MODS_KEY(MOD_LALT | MOD_LSFT, KC_TAB)
#### 2.1.3 Multiple Modifiers
Registers multiple modifiers with pressing a key. To specify multiple modifiers use `|`.
ACTION_MODS(MOD_ALT | MOD_LSFT)
#### 2.1.3 Modifier with Tap key([Dual role][dual_role])
Works as a modifier key while holding, but registers a key on tap(press and release quickly).
ACTION_MODS_TAP_KEY(MOD_RCTL, KC_ENT)
### 2.2 Layer Action
These actions operate layers of keymap.
#### Parameters
You can specify a **target layer** of action and **when the action is executed**. Some actions take a **bit value** for bitwise operation.
The function must always return a valid macro, and default implementation of `action_get_macro` always returns `MACRO_NONE` which has no effect.
#### 2.3.1.1 Limitations
Similar to the Function Action system, the selector functions is passed a `keyrecord_t` object, so it can inspect the key state (e.g. different macros on key press or release), and key itself.
Unlike the Function Action system,`macros` are pre-recorded key sequences, so you can only select from a list. If you want to use dynamic macros then you should look at the more complex function action system.
This C function is called every time key is operated, argument `id` selects action to be performed and `opt` can be used for option. Function `id` can be 0-255 and `opt` can be 0-15.
`keyrecord_t` is comprised of key event and tap count. `keyevent_t` indicates which and when key is pressed or released. From `tap_count` you can know tap state, 0 means no tap. These information will be used in user function to decide how action of key is performed.
typedef struct {
keyevent_t event;
uint8_t tap_count;
} keyrecord_t;
typedef struct {
key_t key;
bool pressed;
uint16_t time;
} keyevent_t;
typedef struct {
uint8_t col;
uint8_t row;
} key_t;
***TODO: sample implementation***
See `keyboard/hhkb/keymap.c` for sample.
### 2.5 Backlight Action
These actions control the backlight.
#### 2.5.1 Change backlight level
Increase backlight level.
ACTION_BACKLIGHT_INCREASE()
Decrease backlight level.
ACTION_BACKLIGHT_DECREASE()
Step through backlight levels.
ACTION_BACKLIGHT_STEP()
Turn a specific backlight level on or off.
ACTION_BACKLIGHT_LEVEL(1)
#### 2.5.2 Turn on / off backlight
Turn the backlight on and off without changing level.
ACTION_BACKLIGHT_TOGGLE()
## 3. Layer switching Example
There are some ways to switch layer with 'Layer' actions.
### 3.1 Momentary switching
Momentary switching changes layer only while holding Fn key.
This action makes 'Layer 1' active(valid) on key press event and inactive on release event. Namely you can overlay a layer on lower layers or default layer temporarily with this action.
ACTION_LAYER_MOMENTARY(1)
Note that after switching on press the actions on destination layer(Layer 1) are performed.
***Thus you shall need to place an action to go back on destination layer***, or you will be stuck in destination layer without way to get back. Usually you need to place same action or 'KC_TRNS` on destination layer to get back.
### 3.2 Toggle switching
Toggle switching performed after releasing a key. With this action you can keep staying on the destination layer until you type the key again to return.
This performs toggle switching action of 'Layer 2'.
ACTION_LAYER_TOGGLE(2)
### 3.3 Momentary switching with Tap key
These actions switch a layer only while holding a key but register the key on tap. **Tap** means to press and release a key quickly.
ACTION_LAYER_TAP_KEY(2, KC_SCLN)
With this you can place a layer switching action on normal key like ';' without losing its original key register function. This action allows you to have layer switching action without necessity of a dedicated key. It means you can have it even on home row of keyboard.
### 3.4 Momentary switching with Tap Toggle
This switches layer only while holding a key but toggle layer with several taps. **Tap** means to press and release key quickly.
ACTION_LAYER_TAP_TOGGLE(1)
Number of taps can be configured with `TAPPING_TOGGLE` in `config.h`, `5` by default.
### 3.5 Momentary switching with Modifiers
This registers modifier key(s) simultaneously with layer switching.
Tapping is to press and release a key quickly. Tapping speed is determined with setting of `TAPPING_TERM`, which can be defined in `config.h`, 200ms by default.
### 4.1 Tap Key
This is a feature to assign normal key action and modifier including layer switching to just same one physical key. This is a kind of [Dual role key][dual_role]. It works as modifier when holding the key but registers normal key when tapping.
This is a feature to assign both toggle layer and momentary switch layer action to just same one physical key. It works as momentary layer switch when holding a key but toggle switch with several taps.
ACTION_LAYER_TAP_TOGGLE(1)
### 4.3 Oneshot Modifier
This runs onetime effects which modify only on just one following key. It works as normal modifier key when holding down while oneshot modifier when tapping.
ACTION_MODS_ONESHOT(MOD_LSFT)
Say you want to type 'The', you have to push and hold Shift key before type 't' then release it before type 'h' and 'e', otherwise you'll get 'THe' or 'the' unintentionally. With Oneshot Modifier you can tap Shift then type 't', 'h' and 'e' normally, you don't need to holding Shift key properly here. This mean you can release Shift before 't' is pressed down.
Oneshot effect is cancel unless following key is pressed down within `ONESHOT_TIMEOUT` of `config.h`. No timeout when it is `0` or not defined.
### 4.4 Tap Toggle Mods
Similar to layer tap toggle, this works as a momentary modifier when holding, but toggles on with several taps. A single tap will 'unstick' the modifier again.
ACTION_MODS_TAP_TOGGLE(MOD_LSFT)
## 5. Legacy Keymap
This was used in prior version and still works due to legacy support code in `common/keymap.c`. Legacy keymap doesn't support many of features that new keymap offers. ***It is not recommended to use Legacy Keymap for new project.***
To enable Legacy Keymap support define this macro in `config.h`.
#define USE_LEGACY_KEYMAP
Legacy Keymap uses two arrays `fn_layer[]` and `fn_keycode[]` to define Fn key. The index of arrays corresponds with postfix number of `Fn` key. Array `fn_layer[]` indicates destination layer to switch and `fn_keycode[]` has keycodes to send when tapping `Fn` key.
In following setting example, `Fn0`, `Fn1` and `Fn2` switch layer to 1, 2 and 2 respectively. `Fn2` registers `Space` key when tapping while `Fn0` and `Fn1` doesn't send any key.
