Hit the semicolon key once, send a semicolon. Hit it twice, rapidly -- send a colon. Hit it three times, and your keyboard's LEDs do a wild dance. That's just one example of what Tap Dance can do. It's one of the nicest community-contributed features in the firmware, conceived and created by [algernon](https://github.com/algernon) in [#451](https://github.com/qmk/qmk_firmware/pull/451). Here's how algernon describes the feature:
With this feature one can specify keys that behave differently, based on the amount of times they have been tapped, and when interrupted, they get handled before the interrupter.
First, you will need `TAP_DANCE_ENABLE = yes` in your `rules.mk`, because the feature is disabled by default. This adds a little less than 1k to the firmware size.
The `TAPPING_TERM` time is the maximum time allowed between taps of your Tap Dance key, and is measured in milliseconds. For example, if you used the above `#define` statement and set up a Tap Dance key that sends `Space` on single-tap and `Enter` on double-tap, then this key will send `ENT` only if you tap this key twice in less than 175ms. If you tap the key, wait more than 175ms, and tap the key again you'll end up sending `SPC SPC` instead. The `TAPPING_TERM_PER_KEY` definition is only needed if you control the tapping term through a [custom `get_tapping_term` function](tap_hold.md#tapping_term), which may be needed because `TAPPING_TERM` affects not just tap-dance keys.
Next, you will want to define some tap-dance keys, which is easiest to do with the `TD()` macro. That macro takes a number which will later be used as an index into the `tap_dance_actions` array and turns it into a tap-dance keycode.
After this, you'll want to use the `tap_dance_actions` array to specify what actions shall be taken when a tap-dance key is in action. Currently, there are five possible options:
*`ACTION_TAP_DANCE_DOUBLE(kc1, kc2)`: Sends the `kc1` keycode when tapped once, `kc2` otherwise. When the key is held, the appropriate keycode is registered: `kc1` when pressed and held, `kc2` when tapped once, then pressed and held.
*`ACTION_TAP_DANCE_LAYER_MOVE(kc, layer)`: Sends the `kc` keycode when tapped once, or moves to `layer`. (this functions like the `TO` layer keycode).
*`ACTION_TAP_DANCE_LAYER_TOGGLE(kc, layer)`: Sends the `kc` keycode when tapped once, or toggles the state of `layer`. (this functions like the `TG` layer keycode).
*`ACTION_TAP_DANCE_FN_ADVANCED(on_each_tap_fn, on_dance_finished_fn, on_dance_reset_fn)`: Calls the first specified function - defined in the user keymap - on every tap, the second function when the dance action finishes (like the previous option), and the last function when the tap dance action resets.
*`ACTION_TAP_DANCE_FN_ADVANCED_WITH_RELEASE(on_each_tap_fn, on_each_release_fn, on_dance_finished_fn, on_dance_reset_fn)`: This macro is identical to `ACTION_TAP_DANCE_FN_ADVANCED` with the addition of `on_each_release_fn` which is invoked every time the key for the tap dance is released. It is worth noting that `on_each_release_fn` will still be called even when the key is released after the dance finishes (e.g. if the key is released after being pressed and held for longer than the `TAPPING_TERM`).
The first option is enough for a lot of cases, that just want dual roles. For example, `ACTION_TAP_DANCE_DOUBLE(KC_SPC, KC_ENT)` will result in `Space` being sent on single-tap, `Enter` otherwise.
For more complicated cases, like blink the LEDs, fiddle with the backlighting, and so on, use the fourth or fifth option. Examples of each are listed below.
Well, that's the bulk of it! You should now be able to work through the examples below, and to develop your own Tap Dance functionality. But if you want a deeper understanding of what's going on behind the scenes, then read on for the explanation of how it all works!
Let's go over the three functions mentioned in `ACTION_TAP_DANCE_FN_ADVANCED` in a little more detail. They all receive the same two arguments: a pointer to a structure that holds all dance related state information, and a pointer to a use case specific state variable. The three functions differ in when they are called. The first, `on_each_tap_fn()`, is called every time the tap dance key is *pressed*. Before it is called, the counter is incremented and the timer is reset. The second function, `on_dance_finished_fn()`, is called when the tap dance is interrupted or ends because `TAPPING_TERM` milliseconds have passed since the last tap. When the `finished` field of the dance state structure is set to `true`, the `on_dance_finished_fn()` is skipped. After `on_dance_finished_fn()` was called or would have been called, but no sooner than when the tap dance key is *released*, `on_dance_reset_fn()` is called. It is possible to end a tap dance immediately, skipping `on_dance_finished_fn()`, but not `on_dance_reset_fn`, by calling `reset_tap_dance(state)`.
To accomplish this logic, the tap dance mechanics use three entry points. The main entry point is `process_tap_dance()`, called from `process_record_quantum()`*after*`process_record_kb()` and `process_record_user()`. This function is responsible for calling `on_each_tap_fn()` and `on_dance_reset_fn()`. In order to handle interruptions of a tap dance, another entry point, `preprocess_tap_dance()` is run right at the beginning of `process_record_quantum()`. This function checks whether the key pressed is a tap-dance key. If it is not, and a tap-dance was in action, we handle that first, and enqueue the newly pressed key. If it is a tap-dance key, then we check if it is the same as the already active one (if there's one active, that is). If it is not, we fire off the old one first, then register the new one. Finally, `tap_dance_task()` periodically checks whether `TAPPING_TERM` has passed since the last key press and finishes a tap dance if that is the case.
This means that you have `TAPPING_TERM` time to tap the key again; you do not have to input all the taps within a single `TAPPING_TERM` timeframe. This allows for longer tap counts, with minimal impact on responsiveness.
#### Example 3: Send `:` on Tap, `;` on Hold :id=example-3
With a little effort, powerful tap-hold configurations can be implemented as tap dances. To emit taps as early as possible, we need to act on releases of the tap dance key. There is no callback for this in the tap dance framework, so we use `process_record_user()`.
> In this configuration "hold" takes place **after** tap dance timeout. To achieve instant hold, remove `state->interrupted` checks in conditions. As a result you may use comfortable longer tapping periods to have more time for taps and not to wait too long for holds (try starting with doubled `TAPPING_TERM`).
Tap dance can be used to emulate `MT()` and `LT()` behavior when the tapped code is not a basic keycode. This is useful to send tapped keycodes that normally require `Shift`, such as parentheses or curly braces—or other modified keycodes, such as `Control + X`.
Below your layers and custom keycodes, add the following:
Tap Dance can be used to mimic MO(layer) and TG(layer) functionality. For this example, we will set up a key to function as `KC_QUOT` on single-tap, as `MO(_MY_LAYER)` on single-hold, and `TG(_MY_LAYER)` on double-tap.
The first step is to include the following code towards the beginning of your `keymap.c`:
The above code is similar to that used in previous examples. The one point to note is that we need to be able to check which layers are active at any time so we can toggle them if needed. To do this we use the `layer_state_is(layer)` function which returns `true` if the given `layer` is active.
The `case: TD_SINGLE_TAP` in `ql_finished` is similar to the above examples. The `TD_SINGLE_HOLD` case works in conjunction with `ql_reset()` to switch to `_MY_LAYER` while the tap dance key is held, and to switch away from `_MY_LAYER` when the key is released. This mirrors the use of `MO(_MY_LAYER)`. The `TD_DOUBLE_TAP` case works by checking whether `_MY_LAYER` is the active layer, and toggling it on or off accordingly. This mirrors the use of `TG(_MY_LAYER)`.
`tap_dance_actions[]` works similar to the above examples. Note that, additionally, I set a longer tapping term for the tap dance keys. This is because I like my `TAPPING_TERM` to be short (\~175ms) for my non-tap-dance keys but find that this is too quick for me to reliably complete tap dance actions - thus the increased time of 275ms here. In order for the per-key tapping terms to take effect, `TAPPING_TERM_PER_KEY` must be defined in your `config.h`.