qmk_firmware/quantum/action_tapping.c

782 lines
31 KiB
C
Raw Normal View History

#include <stdint.h>
#include <stdbool.h>
2022-03-09 18:43:12 +00:00
#include "action.h"
#include "action_layer.h"
#include "action_tapping.h"
#include "keycode.h"
#include "timer.h"
#ifndef NO_ACTION_TAPPING
# if defined(IGNORE_MOD_TAP_INTERRUPT_PER_KEY)
# error "IGNORE_MOD_TAP_INTERRUPT_PER_KEY has been removed; the code needs to be ported to use HOLD_ON_OTHER_KEY_PRESS_PER_KEY instead."
# elif defined(IGNORE_MOD_TAP_INTERRUPT)
# error "IGNORE_MOD_TAP_INTERRUPT is no longer necessary as it is now the default behavior of mod-tap keys. Please remove it from your config."
# endif
2021-11-01 19:18:33 +00:00
# ifndef COMBO_ENABLE
# define IS_TAPPING_RECORD(r) (KEYEQ(tapping_key.event.key, (r->event.key)))
2021-11-01 19:18:33 +00:00
# else
# define IS_TAPPING_RECORD(r) (KEYEQ(tapping_key.event.key, (r->event.key)) && tapping_key.keycode == r->keycode)
2021-11-01 19:18:33 +00:00
# endif
# define WITHIN_TAPPING_TERM(e) (TIMER_DIFF_16(e.time, tapping_key.event.time) < GET_TAPPING_TERM(get_record_keycode(&tapping_key, false), &tapping_key))
# define WITHIN_QUICK_TAP_TERM(e) (TIMER_DIFF_16(e.time, tapping_key.event.time) < GET_QUICK_TAP_TERM(get_record_keycode(&tapping_key, false), &tapping_key))
# ifdef DYNAMIC_TAPPING_TERM_ENABLE
uint16_t g_tapping_term = TAPPING_TERM;
# endif
# ifdef TAPPING_TERM_PER_KEY
__attribute__((weak)) uint16_t get_tapping_term(uint16_t keycode, keyrecord_t *record) {
# ifdef DYNAMIC_TAPPING_TERM_ENABLE
return g_tapping_term;
# else
return TAPPING_TERM;
# endif
}
2019-08-30 18:19:03 +00:00
# endif
# ifdef QUICK_TAP_TERM_PER_KEY
__attribute__((weak)) uint16_t get_quick_tap_term(uint16_t keycode, keyrecord_t *record) {
return QUICK_TAP_TERM;
}
# endif
# ifdef PERMISSIVE_HOLD_PER_KEY
__attribute__((weak)) bool get_permissive_hold(uint16_t keycode, keyrecord_t *record) {
return false;
}
# endif
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
extern const char chordal_hold_layout[MATRIX_ROWS][MATRIX_COLS] PROGMEM;
# define REGISTERED_TAPS_SIZE 8
// Array of tap-hold keys that have been settled as tapped but not yet released.
static keypos_t registered_taps[REGISTERED_TAPS_SIZE] = {};
static uint8_t num_registered_taps = 0;
/** Adds `key` to the registered_taps array. */
static void registered_taps_add(keypos_t key);
/** Returns the index of `key` in registered_taps, or -1 if not found. */
static int8_t registered_tap_find(keypos_t key);
/** Removes index `i` from the registered_taps array. */
static void registered_taps_del_index(uint8_t i);
/** Logs the registered_taps array for debugging. */
static void debug_registered_taps(void);
/** \brief Finds which queued events should be held according to Chordal Hold.
*
* In a situation with multiple unsettled tap-hold key presses, scan the queue
* up until the first release, non-tap-hold, or one-shot event and find the
* latest event in the queue that settles as held according to
* get_chordal_hold().
*
* \return Index of the first tap, or equivalently, one past the latest hold.
*/
static uint8_t waiting_buffer_find_chordal_hold_tap(void);
/** Processes queued events up to and including `key` as tapped. */
static void waiting_buffer_chordal_hold_taps_until(keypos_t key);
/** \brief Processes and pops buffered events until the first tap-hold event. */
static void waiting_buffer_process_regular(void);
static bool is_mt_or_lt(uint16_t keycode) {
return IS_QK_MOD_TAP(keycode) || IS_QK_LAYER_TAP(keycode);
}
# endif // CHORDAL_HOLD
# ifdef HOLD_ON_OTHER_KEY_PRESS_PER_KEY
__attribute__((weak)) bool get_hold_on_other_key_press(uint16_t keycode, keyrecord_t *record) {
return false;
}
# endif
# if defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT)
# include "process_auto_shift.h"
# endif
2019-08-30 18:19:03 +00:00
static keyrecord_t tapping_key = {};
static keyrecord_t waiting_buffer[WAITING_BUFFER_SIZE] = {};
2019-08-30 18:19:03 +00:00
static uint8_t waiting_buffer_head = 0;
static uint8_t waiting_buffer_tail = 0;
static bool process_tapping(keyrecord_t *record);
static bool waiting_buffer_enq(keyrecord_t record);
static void waiting_buffer_clear(void);
static bool waiting_buffer_typed(keyevent_t event);
static bool waiting_buffer_has_anykey_pressed(void);
static void waiting_buffer_scan_tap(void);
static void debug_tapping_key(void);
static void debug_waiting_buffer(void);
/** \brief Action Tapping Process
*
* FIXME: Needs doc
*/
2019-08-30 18:19:03 +00:00
void action_tapping_process(keyrecord_t record) {
if (process_tapping(&record)) {
if (IS_EVENT(record.event)) {
ac_dprintf("processed: ");
2019-08-30 18:19:03 +00:00
debug_record(record);
ac_dprintf("\n");
}
} else {
if (!waiting_buffer_enq(record)) {
// clear all in case of overflow.
ac_dprintf("OVERFLOW: CLEAR ALL STATES\n");
clear_keyboard();
waiting_buffer_clear();
tapping_key = (keyrecord_t){0};
}
}
// process waiting_buffer
if (IS_EVENT(record.event) && waiting_buffer_head != waiting_buffer_tail) {
ac_dprintf("---- action_exec: process waiting_buffer -----\n");
}
for (; waiting_buffer_tail != waiting_buffer_head; waiting_buffer_tail = (waiting_buffer_tail + 1) % WAITING_BUFFER_SIZE) {
if (process_tapping(&waiting_buffer[waiting_buffer_tail])) {
ac_dprintf("processed: waiting_buffer[%u] =", waiting_buffer_tail);
2019-08-30 18:19:03 +00:00
debug_record(waiting_buffer[waiting_buffer_tail]);
ac_dprintf("\n\n");
} else {
break;
}
}
if (IS_EVENT(record.event)) {
ac_dprintf("\n");
}
}
/* Some conditionally defined helper macros to keep process_tapping more
* readable. The conditional definition of tapping_keycode and all the
* conditional uses of it are hidden inside macros named TAP_...
*/
# define TAP_DEFINE_KEYCODE const uint16_t tapping_keycode = get_record_keycode(&tapping_key, false)
# if defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT)
# ifdef RETRO_TAPPING_PER_KEY
# define TAP_GET_RETRO_TAPPING(keyp) get_auto_shifted_key(tapping_keycode, keyp) && get_retro_tapping(tapping_keycode, &tapping_key)
# else
# define TAP_GET_RETRO_TAPPING(keyp) get_auto_shifted_key(tapping_keycode, keyp)
# endif
/* Used to extend TAPPING_TERM:
* indefinitely if RETRO_SHIFT does not have a value
* to RETRO_SHIFT if RETRO_SHIFT is set
* for possibly retro shifted keys.
*/
# define MAYBE_RETRO_SHIFTING(ev, keyp) (get_auto_shifted_key(tapping_keycode, keyp) && TAP_GET_RETRO_TAPPING(keyp) && ((RETRO_SHIFT + 0) == 0 || TIMER_DIFF_16((ev).time, tapping_key.event.time) < (RETRO_SHIFT + 0)))
# define TAP_IS_LT IS_QK_LAYER_TAP(tapping_keycode)
# define TAP_IS_MT IS_QK_MOD_TAP(tapping_keycode)
# define TAP_IS_RETRO IS_RETRO(tapping_keycode)
# else
# define TAP_GET_RETRO_TAPPING(keyp) false
# define MAYBE_RETRO_SHIFTING(ev, kp) false
# define TAP_IS_LT false
# define TAP_IS_MT false
# define TAP_IS_RETRO false
# endif
# ifdef PERMISSIVE_HOLD_PER_KEY
# define TAP_GET_PERMISSIVE_HOLD get_permissive_hold(tapping_keycode, &tapping_key)
# elif defined(PERMISSIVE_HOLD)
# define TAP_GET_PERMISSIVE_HOLD true
# else
# define TAP_GET_PERMISSIVE_HOLD false
# endif
# ifdef HOLD_ON_OTHER_KEY_PRESS_PER_KEY
# define TAP_GET_HOLD_ON_OTHER_KEY_PRESS get_hold_on_other_key_press(tapping_keycode, &tapping_key)
# elif defined(HOLD_ON_OTHER_KEY_PRESS)
# define TAP_GET_HOLD_ON_OTHER_KEY_PRESS true
# else
# define TAP_GET_HOLD_ON_OTHER_KEY_PRESS false
# endif
/** \brief Tapping
*
* Rule: Tap key is typed(pressed and released) within TAPPING_TERM.
* (without interfering by typing other key)
*/
/* return true when key event is processed or consumed. */
2019-08-30 18:19:03 +00:00
bool process_tapping(keyrecord_t *keyp) {
const keyevent_t event = keyp->event;
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
if (!event.pressed) {
const int8_t i = registered_tap_find(event.key);
if (i != -1) {
// If a tap-hold key was previously settled as tapped, set its
// tap.count correspondingly on release.
keyp->tap.count = 1;
registered_taps_del_index(i);
ac_dprintf("Found tap release for [%d]\n", i);
debug_registered_taps();
}
}
# endif // CHORDAL_HOLD
// state machine is in the "reset" state, no tapping key is to be
// processed
if (IS_NOEVENT(tapping_key.event)) {
if (!IS_EVENT(event)) {
// early return for tick events
} else if (event.pressed && is_tap_record(keyp)) {
// the currently pressed key is a tapping key, therefore transition
// into the "pressed" tapping key state
ac_dprintf("Tapping: Start(Press tap key).\n");
tapping_key = *keyp;
process_record_tap_hint(&tapping_key);
waiting_buffer_scan_tap();
debug_tapping_key();
} else {
// the current key is just a regular key, pass it on for regular
// processing
process_record(keyp);
}
return true;
}
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if (defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT)) || defined(PERMISSIVE_HOLD_PER_KEY) || defined(CHORDAL_HOLD) || defined(HOLD_ON_OTHER_KEY_PRESS_PER_KEY)
TAP_DEFINE_KEYCODE;
# endif
// process "pressed" tapping key state
if (tapping_key.event.pressed) {
if (WITHIN_TAPPING_TERM(event) || MAYBE_RETRO_SHIFTING(event, keyp)) {
if (IS_NOEVENT(event)) {
// early return for tick events
return true;
}
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
if (tapping_key.tap.count == 0) {
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
if (IS_TAPPING_RECORD(keyp) && !event.pressed) {
// first tap!
ac_dprintf("Tapping: First tap(0->1).\n");
tapping_key.tap.count = 1;
debug_tapping_key();
process_record(&tapping_key);
// copy tapping state
keyp->tap = tapping_key.tap;
// enqueue
return false;
}
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
else if (is_mt_or_lt(tapping_keycode) && !event.pressed && waiting_buffer_typed(event) && !get_chordal_hold(tapping_keycode, &tapping_key, get_record_keycode(keyp, false), keyp)) {
// Key release that is not a chord with the tapping key.
// Settle the tapping key and any other pending tap-hold
// keys preceding the press of this key as tapped.
ac_dprintf("Tapping: End. Chord considered a tap\n");
tapping_key.tap.count = 1;
registered_taps_add(tapping_key.event.key);
process_record(&tapping_key);
tapping_key = (keyrecord_t){0};
waiting_buffer_chordal_hold_taps_until(event.key);
debug_registered_taps();
debug_waiting_buffer();
// enqueue
return false;
}
# endif // CHORDAL_HOLD
/* Process a key typed within TAPPING_TERM
* This can register the key before settlement of tapping,
* useful for long TAPPING_TERM but may prevent fast typing.
*/
// clang-format off
else if (
!event.pressed && waiting_buffer_typed(event) &&
(
TAP_GET_PERMISSIVE_HOLD ||
// Causes nested taps to not wait past TAPPING_TERM/RETRO_SHIFT
// unnecessarily and fixes them for Layer Taps.
TAP_GET_RETRO_TAPPING(keyp)
)
) {
// clang-format on
ac_dprintf("Tapping: End. No tap. Interfered by typing key\n");
process_record(&tapping_key);
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
uint8_t first_tap = waiting_buffer_find_chordal_hold_tap();
ac_dprintf("first_tap = %u\n", first_tap);
if (first_tap < WAITING_BUFFER_SIZE) {
for (; waiting_buffer_tail != first_tap; waiting_buffer_tail = (waiting_buffer_tail + 1) % WAITING_BUFFER_SIZE) {
ac_dprintf("Processing [%u]\n", waiting_buffer_tail);
process_record(&waiting_buffer[waiting_buffer_tail]);
}
}
waiting_buffer_chordal_hold_taps_until(event.key);
debug_registered_taps();
debug_waiting_buffer();
# endif // CHORDAL_HOLD
tapping_key = (keyrecord_t){0};
debug_tapping_key();
// enqueue
return false;
}
/* Process release event of a key pressed before tapping starts
* Without this unexpected repeating will occur with having fast repeating setting
* https://github.com/tmk/tmk_keyboard/issues/60
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
*
* NOTE: This workaround causes events to process out of order,
* e.g. in a rolled press of three tap-hold keys like
*
* "A down, B down, C down, A up, B up, C up"
*
* events are processed as
*
* "A down, B down, A up, B up, C down, C up"
*
* It seems incorrect to process keyp before the tapping key.
* This workaround is old, from 2013. This might no longer
* be needed for the original problem it was meant to address.
*/
else if (!event.pressed && !waiting_buffer_typed(event)) {
// Modifier/Layer should be retained till end of this tapping.
action_t action = layer_switch_get_action(event.key);
switch (action.kind.id) {
case ACT_LMODS:
case ACT_RMODS:
if (action.key.mods && !action.key.code) return false;
if (IS_MODIFIER_KEYCODE(action.key.code)) return false;
break;
case ACT_LMODS_TAP:
case ACT_RMODS_TAP:
if (action.key.mods && keyp->tap.count == 0) return false;
if (IS_MODIFIER_KEYCODE(action.key.code)) return false;
break;
case ACT_LAYER_TAP:
case ACT_LAYER_TAP_EXT:
switch (action.layer_tap.code) {
case 0 ...(OP_TAP_TOGGLE - 1):
case OP_ON_OFF:
case OP_OFF_ON:
case OP_SET_CLEAR:
return false;
}
break;
}
// Release of key should be process immediately.
ac_dprintf("Tapping: release event of a key pressed before tapping\n");
process_record(keyp);
return true;
2019-08-30 18:19:03 +00:00
} else {
// set interrupted flag when other key pressed during tapping
if (event.pressed) {
tapping_key.tap.interrupted = true;
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
if (is_mt_or_lt(tapping_keycode) && !get_chordal_hold(tapping_keycode, &tapping_key, get_record_keycode(keyp, false), keyp)) {
// In process_action(), HOLD_ON_OTHER_KEY_PRESS
// will revert interrupted events to holds, so
// this needs to be set false.
tapping_key.tap.interrupted = false;
if (!is_tap_record(keyp)) {
ac_dprintf("Tapping: End. Chord considered a tap\n");
tapping_key.tap.count = 1;
registered_taps_add(tapping_key.event.key);
debug_registered_taps();
process_record(&tapping_key);
tapping_key = (keyrecord_t){0};
}
} else
# endif // CHORDAL_HOLD
if (TAP_GET_HOLD_ON_OTHER_KEY_PRESS
# if defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT)
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
// Auto Shift cannot evaluate this early
// Retro Shift uses the hold action for all nested taps even without HOLD_ON_OTHER_KEY_PRESS, so this is fine to skip
&& !(MAYBE_RETRO_SHIFTING(event, keyp) && get_auto_shifted_key(get_record_keycode(keyp, false), keyp))
# endif
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
) {
// Settle the tapping key as *held*, since
// HOLD_ON_OTHER_KEY_PRESS is enabled for this key.
ac_dprintf("Tapping: End. No tap. Interfered by pressed key\n");
process_record(&tapping_key);
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
if (waiting_buffer_tail != waiting_buffer_head && is_tap_record(&waiting_buffer[waiting_buffer_tail])) {
tapping_key = waiting_buffer[waiting_buffer_tail];
// Pop tail from the queue.
waiting_buffer_tail = (waiting_buffer_tail + 1) % WAITING_BUFFER_SIZE;
debug_waiting_buffer();
} else
# endif // CHORDAL_HOLD
{
tapping_key = (keyrecord_t){0};
}
debug_tapping_key();
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# if defined(CHORDAL_HOLD)
waiting_buffer_process_regular();
# endif // CHORDAL_HOLD
}
}
// enqueue
return false;
}
}
// tap_count > 0
else {
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
if (IS_TAPPING_RECORD(keyp) && !event.pressed) {
ac_dprintf("Tapping: Tap release(%u)\n", tapping_key.tap.count);
keyp->tap = tapping_key.tap;
process_record(keyp);
tapping_key = *keyp;
debug_tapping_key();
return true;
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
} else if (is_tap_record(keyp) && event.pressed) {
if (tapping_key.tap.count > 1) {
ac_dprintf("Tapping: Start new tap with releasing last tap(>1).\n");
// unregister key
2021-11-01 19:18:33 +00:00
process_record(&(keyrecord_t){
.tap = tapping_key.tap,
.event.key = tapping_key.event.key,
.event.time = event.time,
.event.pressed = false,
.event.type = tapping_key.event.type,
2021-11-01 19:18:33 +00:00
# ifdef COMBO_ENABLE
.keycode = tapping_key.keycode,
# endif
});
} else {
ac_dprintf("Tapping: Start while last tap(1).\n");
}
tapping_key = *keyp;
waiting_buffer_scan_tap();
debug_tapping_key();
return true;
2019-08-30 18:19:03 +00:00
} else {
ac_dprintf("Tapping: key event while last tap(>0).\n");
# if defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT)
retroshift_swap_times();
# endif
process_record(keyp);
return true;
}
}
}
// after TAPPING_TERM
else {
if (tapping_key.tap.count == 0) {
ac_dprintf("Tapping: End. Timeout. Not tap(0): ");
2019-08-30 18:19:03 +00:00
debug_event(event);
ac_dprintf("\n");
process_record(&tapping_key);
tapping_key = (keyrecord_t){0};
debug_tapping_key();
return false;
2019-08-30 18:19:03 +00:00
} else {
if (IS_NOEVENT(event)) {
return true;
}
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
if (IS_TAPPING_RECORD(keyp) && !event.pressed) {
ac_dprintf("Tapping: End. last timeout tap release(>0).");
keyp->tap = tapping_key.tap;
process_record(keyp);
tapping_key = (keyrecord_t){0};
return true;
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
} else if (is_tap_record(keyp) && event.pressed) {
if (tapping_key.tap.count > 1) {
ac_dprintf("Tapping: Start new tap with releasing last timeout tap(>1).\n");
// unregister key
2021-11-01 19:18:33 +00:00
process_record(&(keyrecord_t){
.tap = tapping_key.tap,
.event.key = tapping_key.event.key,
.event.time = event.time,
.event.pressed = false,
.event.type = tapping_key.event.type,
2021-11-01 19:18:33 +00:00
# ifdef COMBO_ENABLE
.keycode = tapping_key.keycode,
# endif
});
} else {
ac_dprintf("Tapping: Start while last timeout tap(1).\n");
}
tapping_key = *keyp;
waiting_buffer_scan_tap();
debug_tapping_key();
return true;
2019-08-30 18:19:03 +00:00
} else {
ac_dprintf("Tapping: key event while last timeout tap(>0).\n");
process_record(keyp);
return true;
}
}
}
}
// process "released" tapping key state
else {
if (WITHIN_TAPPING_TERM(event) || MAYBE_RETRO_SHIFTING(event, keyp)) {
if (IS_NOEVENT(event)) {
// early return for tick events
return true;
}
if (event.pressed) {
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
if (IS_TAPPING_RECORD(keyp)) {
if (WITHIN_QUICK_TAP_TERM(event) && !tapping_key.tap.interrupted && tapping_key.tap.count > 0) {
// sequential tap.
keyp->tap = tapping_key.tap;
if (keyp->tap.count < 15) keyp->tap.count += 1;
ac_dprintf("Tapping: Tap press(%u)\n", keyp->tap.count);
process_record(keyp);
tapping_key = *keyp;
debug_tapping_key();
return true;
}
// FIX: start new tap again
tapping_key = *keyp;
return true;
Process combos earlier & overlapping combos (#8591) * Combo processing improvements. Now it is possible to use ModTap and LayerTap keys as part of combos. Overlapping combos also don't trigger all the combos, just exactly the one that you press. New settings: - COMBO_MUST_HOLD_MODS - COMBO_MOD_TERM - COMBO_TERM_PER_COMBO - COMBO_MUST_HOLD_PER_COMBO - COMBO_STRICT_TIMER - COMBO_NO_TIMER * Remove the size flags from combo_t struct boolean members. This in the end actually saves space as the members are accessed so many times. The amount of operations needed to access the bits uses more memory than setting the size saves. * Fix `process_combo_key_release` not called correctly with tap-only combos * Fix not passing a pointer when NO_ACTION_TAPPING is defined. * Docs for `COMBO_ONLY_FROM_LAYER` * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update quantum/process_keycode/process_combo.c Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Add `EXTRA_SHORT_COMBOS` option. Stuff combo's `disabled` and `active` flags into `state`. Possibly can save some space. * Add more examples and clarify things with dict management system. - Simple examples now has a combo that has modifiers included. - The slightly more advanced examples now are actually more advanced instead of just `tap_code16(<modded-keycode>)`. - Added a note that `COMBO_ACTION`s are not needed anymore as you can just use custom keycodes. - Added a note that the `g/keymap_combo.h` macros use the `process_combo_event` function and that it is not usable in one's keymap afterwards. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Change "the" combo action example to "email" example. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Fix sneaky infinite loop with `combo_disable()` No need to call `dump_key_buffer` when disabling combos because the buffer is either being dumped if a combo-key was pressed, or the buffer is empty if a non-combo-key is pressed. * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> * Update docs/feature_combo.md Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: precondition <57645186+precondition@users.noreply.github.com> Co-authored-by: Drashna Jaelre <drashna@live.com>
2021-08-05 23:44:57 +00:00
} else if (is_tap_record(keyp)) {
// Sequential tap can be interfered with other tap key.
ac_dprintf("Tapping: Start with interfering other tap.\n");
tapping_key = *keyp;
waiting_buffer_scan_tap();
debug_tapping_key();
return true;
} else {
// should none in buffer
// FIX: interrupted when other key is pressed
tapping_key.tap.interrupted = true;
process_record(keyp);
return true;
}
} else {
ac_dprintf("Tapping: other key just after tap.\n");
process_record(keyp);
return true;
}
} else {
// Timeout - reset state machine.
ac_dprintf("Tapping: End(Timeout after releasing last tap): ");
2019-08-30 18:19:03 +00:00
debug_event(event);
ac_dprintf("\n");
tapping_key = (keyrecord_t){0};
debug_tapping_key();
return false;
}
}
}
/** \brief Waiting buffer enq
*
* FIXME: Needs docs
*/
2019-08-30 18:19:03 +00:00
bool waiting_buffer_enq(keyrecord_t record) {
if (IS_NOEVENT(record.event)) {
return true;
}
if ((waiting_buffer_head + 1) % WAITING_BUFFER_SIZE == waiting_buffer_tail) {
ac_dprintf("waiting_buffer_enq: Over flow.\n");
return false;
}
waiting_buffer[waiting_buffer_head] = record;
2019-08-30 18:19:03 +00:00
waiting_buffer_head = (waiting_buffer_head + 1) % WAITING_BUFFER_SIZE;
ac_dprintf("waiting_buffer_enq: ");
2019-08-30 18:19:03 +00:00
debug_waiting_buffer();
return true;
}
/** \brief Waiting buffer clear
*
* FIXME: Needs docs
*/
2019-08-30 18:19:03 +00:00
void waiting_buffer_clear(void) {
waiting_buffer_head = 0;
waiting_buffer_tail = 0;
}
/** \brief Waiting buffer typed
*
* FIXME: Needs docs
*/
2019-08-30 18:19:03 +00:00
bool waiting_buffer_typed(keyevent_t event) {
for (uint8_t i = waiting_buffer_tail; i != waiting_buffer_head; i = (i + 1) % WAITING_BUFFER_SIZE) {
2019-08-30 18:19:03 +00:00
if (KEYEQ(event.key, waiting_buffer[i].event.key) && event.pressed != waiting_buffer[i].event.pressed) {
return true;
}
}
return false;
}
/** \brief Waiting buffer has anykey pressed
*
* FIXME: Needs docs
*/
2019-08-30 18:19:03 +00:00
__attribute__((unused)) bool waiting_buffer_has_anykey_pressed(void) {
for (uint8_t i = waiting_buffer_tail; i != waiting_buffer_head; i = (i + 1) % WAITING_BUFFER_SIZE) {
if (waiting_buffer[i].event.pressed) return true;
}
return false;
}
/** \brief Scan buffer for tapping
*
* FIXME: Needs docs
*/
2019-08-30 18:19:03 +00:00
void waiting_buffer_scan_tap(void) {
// early return if:
// - tapping already is settled
// - invalid state: tapping_key released && tap.count == 0
if ((tapping_key.tap.count > 0) || !tapping_key.event.pressed) {
return;
}
# if (defined(AUTO_SHIFT_ENABLE) && defined(RETRO_SHIFT))
TAP_DEFINE_KEYCODE;
# endif
for (uint8_t i = waiting_buffer_tail; i != waiting_buffer_head; i = (i + 1) % WAITING_BUFFER_SIZE) {
keyrecord_t *candidate = &waiting_buffer[i];
// clang-format off
if (IS_EVENT(candidate->event) && KEYEQ(candidate->event.key, tapping_key.event.key) && !candidate->event.pressed && (
WITHIN_TAPPING_TERM(waiting_buffer[i].event) || MAYBE_RETRO_SHIFTING(waiting_buffer[i].event, &tapping_key)
)) {
// clang-format on
tapping_key.tap.count = 1;
candidate->tap.count = 1;
process_record(&tapping_key);
ac_dprintf("waiting_buffer_scan_tap: found at [%u]\n", i);
debug_waiting_buffer();
return;
}
}
}
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
# ifdef CHORDAL_HOLD
__attribute__((weak)) bool get_chordal_hold(uint16_t tap_hold_keycode, keyrecord_t *tap_hold_record, uint16_t other_keycode, keyrecord_t *other_record) {
return get_chordal_hold_default(tap_hold_record, other_record);
}
bool get_chordal_hold_default(keyrecord_t *tap_hold_record, keyrecord_t *other_record) {
if (tap_hold_record->event.type != KEY_EVENT || other_record->event.type != KEY_EVENT) {
return true; // Return true on combos or other non-key events.
}
char tap_hold_hand = chordal_hold_handedness(tap_hold_record->event.key);
if (tap_hold_hand == '*') {
return true;
}
char other_hand = chordal_hold_handedness(other_record->event.key);
return other_hand == '*' || tap_hold_hand != other_hand;
}
__attribute__((weak)) char chordal_hold_handedness(keypos_t key) {
return (char)pgm_read_byte(&chordal_hold_layout[key.row][key.col]);
}
static void registered_taps_add(keypos_t key) {
if (num_registered_taps >= REGISTERED_TAPS_SIZE) {
ac_dprintf("TAPS OVERFLOW: CLEAR ALL STATES\n");
clear_keyboard();
num_registered_taps = 0;
}
registered_taps[num_registered_taps] = key;
++num_registered_taps;
}
static int8_t registered_tap_find(keypos_t key) {
for (int8_t i = 0; i < num_registered_taps; ++i) {
if (KEYEQ(registered_taps[i], key)) {
return i;
}
}
return -1;
}
static void registered_taps_del_index(uint8_t i) {
if (i < num_registered_taps) {
--num_registered_taps;
if (i < num_registered_taps) {
registered_taps[i] = registered_taps[num_registered_taps];
}
}
}
static void debug_registered_taps(void) {
ac_dprintf("registered_taps = { ");
for (int8_t i = 0; i < num_registered_taps; ++i) {
ac_dprintf("%02X%02X ", registered_taps[i].row, registered_taps[i].col);
}
ac_dprintf("}\n");
}
static uint8_t waiting_buffer_find_chordal_hold_tap(void) {
keyrecord_t *prev = &tapping_key;
uint16_t prev_keycode = get_record_keycode(&tapping_key, false);
uint8_t first_tap = WAITING_BUFFER_SIZE;
for (uint8_t i = waiting_buffer_tail; i != waiting_buffer_head; i = (i + 1) % WAITING_BUFFER_SIZE) {
keyrecord_t * cur = &waiting_buffer[i];
const uint16_t cur_keycode = get_record_keycode(cur, false);
if (!cur->event.pressed || !is_mt_or_lt(prev_keycode)) {
break;
} else if (get_chordal_hold(prev_keycode, prev, cur_keycode, cur)) {
first_tap = i; // Track one index past the latest hold.
}
prev = cur;
prev_keycode = cur_keycode;
}
return first_tap;
}
static void waiting_buffer_chordal_hold_taps_until(keypos_t key) {
while (waiting_buffer_tail != waiting_buffer_head) {
keyrecord_t *record = &waiting_buffer[waiting_buffer_tail];
ac_dprintf("waiting_buffer_chordal_hold_taps_until: processing [%u]\n", waiting_buffer_tail);
if (record->event.pressed && is_tap_record(record)) {
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
record->tap.count = 1;
registered_taps_add(record->event.key);
}
process_record(record);
waiting_buffer_tail = (waiting_buffer_tail + 1) % WAITING_BUFFER_SIZE;
if (KEYEQ(key, record->event.key) && record->event.pressed) {
break;
}
}
}
static void waiting_buffer_process_regular(void) {
for (; waiting_buffer_tail != waiting_buffer_head; waiting_buffer_tail = (waiting_buffer_tail + 1) % WAITING_BUFFER_SIZE) {
if (is_tap_record(&waiting_buffer[waiting_buffer_tail])) {
break; // Stop once a tap-hold key event is reached.
}
ac_dprintf("waiting_buffer_process_regular: processing [%u]\n", waiting_buffer_tail);
process_record(&waiting_buffer[waiting_buffer_tail]);
}
debug_waiting_buffer();
}
# endif // CHORDAL_HOLD
/** \brief Logs tapping key if ACTION_DEBUG is enabled. */
2019-08-30 18:19:03 +00:00
static void debug_tapping_key(void) {
ac_dprintf("TAPPING_KEY=");
2019-08-30 18:19:03 +00:00
debug_record(tapping_key);
ac_dprintf("\n");
}
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
/** \brief Logs waiting buffer if ACTION_DEBUG is enabled. */
2019-08-30 18:19:03 +00:00
static void debug_waiting_buffer(void) {
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
ac_dprintf("{");
for (uint8_t i = waiting_buffer_tail; i != waiting_buffer_head; i = (i + 1) % WAITING_BUFFER_SIZE) {
[Core] Add Chordal Hold, an "opposite hands rule" tap-hold option similar to Achordion, Bilateral Combinations. (#24560) * Chordal Hold: restrict what chords settle as hold * Chordal Hold: docs and further improvements * Fix formatting. * Doc rewording and minor edit. * Support Chordal Hold of multiple tap-hold keys. * Fix formatting. * Simplification and additional test. * Fix formatting. * Tighten tests. * Add test two_mod_taps_same_hand_hold_til_timeout. * Revise handing of pairs of tap-hold keys. * Generate a default chordal_hold_layout. * Document chordal_hold_handedness(). * Add license notice to new and branched files in PR. * Add `tapping.chordal_hold` property for info.json. * Update docs/reference_info_json.md * Revise "hand" jsonschema. * Chordal Hold: Improved layout handedness heuristic. This commit improves the heuristic used in generate-keyboard-c for inferring key handedness from keyboard.json geometry data. Heuristic summary: 1. If the layout is symmetric (e.g. most split keyboards), guess the handedness based on the sign of (x - layout_x_midpoint). 2. Otherwise, if the layout has a key of >=6u width, it is probably the spacebar. Form a dividing line through the spacebar, nearly vertical but with a slight angle to follow typical row stagger. 3. Otherwise, assume handedness based on the widest horizontal separation. I have tested this strategy on a couple dozen keyboards and found it to work reliably. * Use Optional instead of `| None`. * Refactor to avoid lambdas. * Remove trailing comma in chordal_hold_layout. * Minor docs edits. * Revise to allow combining multiple same-hand mods. This commit revises Chordal Hold as described in discussion in https://github.com/qmk/qmk_firmware/pull/24560#discussion_r1894655238 1. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RCTL_T(KC_A)↑" before the tapping term, RCTL_T(KC_A) is settled as tapped. 2. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, RSFT_T(KC_C)↑", both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 3. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, KC_U↓" (all keys on the same side), both RCTL_T(KC_A) and RSFT_T(KC_C) are settled as tapped. 4. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓, LSFT_T(KC_T)↓", with the third key on the other side, we allow Permissive Hold or Hold On Other Keypress to decide how/when to settle the keys. 5. In "RCTL_T(KC_A)↓, RSFT_T(KC_C)↓" held until the tapping term, the keys are settled as held. 1–3 provide same-hand roll protection. 4–5 are for combining multiple same-hand modifiers. I've updated the unit tests and have been running it on my keyboard, for a few hours so far, and all seems good. I really like this scheme. It allows combining same-side mods, yet it also has roll protection on streaks. For me, this feels like Achordion, but clearly better streak handling and improved responsiveness. * Fix formatting. * Add a couple tests with LT keys. * Remove stale use of CHORDAL_HOLD_LAYOUT. * Fix misspelling lastest -> latest * Handling tweak for LTs and tests. * Fix formatting. * More tests with LT keys. * Fix formatting.
2025-01-27 11:32:23 +00:00
ac_dprintf(" [%u]=", i);
2019-08-30 18:19:03 +00:00
debug_record(waiting_buffer[i]);
}
ac_dprintf("}\n");
}
#endif