diff options
author | Michael F. Lamb <mike@datagrok.org> | 2017-11-03 08:56:06 -0700 |
---|---|---|
committer | Jack Humbert <jack.humb@gmail.com> | 2017-11-03 11:56:06 -0400 |
commit | e899cb8940da04fa2610604f0aab417db7fac119 (patch) | |
tree | 315c8b4b4ed73bc582af672740a64ccf2baa09fe /keyboards/mitosis/keymaps/datagrok/keymap.c | |
parent | 80e489e122962abbfa63ff3383ff303144bfe6f2 (diff) |
several improvements for mitosis:datagrok (#1960)
* mitosis/datagrok: reduce features from rules.mk
* mitosis/datagrok: make both layer keys neighbor shift
* mitosis/datagrok: (no-op) tweak some comments
* mitosis/datagrok: set baudrate to 250k
This requires a corresponding change to the mitosis wireless firmware:
https://github.com/reversebias/mitosis/pull/10
* mitosis/datagrok: move design description from code comment to a readme
* mitosis/datagrok: new layout, new shifted keys, efficient LED code
This is experimental, but compiles and seems to work correctly.
* mitosis/datagrok: whoops, move readme.md
* mitosis/datagrok: a minor layout improvement simplifies custom-shifted code
instead of [, .] [? !], using [, ?] [. !] greatly simplifies the code
needed to perform the shifted-key switching. (And keeps , and . on the
same keys that they are under qwerty.)
also: layout improvements for symbols
* mitosis/datagrok: make my code conform to QMK style guidelines
* mitosis/datagrok: TODO note for layout table in README
* mitosis/datagrok: remove led_set_user until i figure out other changes
need to see if the corresponding changes needed in the keyboard-level code
is okay.
* mitosis/datagrok: simpler layer indicator
* mitosis/datagrok: undo change to keyboard baud; make it in my layout dir.
* mitosis/datagrok: apply same punctuation hack to qwerty layer
* mitosis/datagrok: enable qwerty layer toggle
* mitosis/datagrok: update readme
Diffstat (limited to 'keyboards/mitosis/keymaps/datagrok/keymap.c')
-rw-r--r-- | keyboards/mitosis/keymaps/datagrok/keymap.c | 200 |
1 files changed, 118 insertions, 82 deletions
diff --git a/keyboards/mitosis/keymaps/datagrok/keymap.c b/keyboards/mitosis/keymaps/datagrok/keymap.c index e4460a9b69..36683fbc5d 100644 --- a/keyboards/mitosis/keymaps/datagrok/keymap.c +++ b/keyboards/mitosis/keymaps/datagrok/keymap.c @@ -1,34 +1,12 @@ #include "mitosis.h" -// Arrows in a layer in the home position. -// Fn+Arrows = PgUp PgDn Home End, which is intuitive for me -// -// Since QWERTY and WORKMAN keep angle brackets together, I wanted to place -// other enclosing symbols on the same keys. So, 9 and 0 and [ and ] land there -// in other layers. That informed the numbers placement, which informed the -// function-key placement. -// -// To do: -// - Improve LED indications (may require modding bluetooth firmware): -// - Is any board nonresponsive (which one?) -// - Does either board have a low battery? -// - Use "shifted keys" hack to make programming symbols easier to type -// - Add "media" keysyms, Insert, PrintScr, Pause/Break -// - Dynamically toggle QWERTY or other layouts -// - See if the henkans placement is at all useful for Japanese speakers, or -// abuse different keysyms -// - Overlay a 10key numpad somewhere -// - Mod a speaker onto my receiver and enable tones -// - Mod more indicator LEDs onto my receiver -// - Do something with Num/Caps/Scroll lock? -// - Improve tri-layer behavior - enum mitosis_layers { - _WORKMAN, - _FUNC, - _NUMS, - _NMFN + _xW, // workman + _xQ, // qwerty + _xS, // symbols + _xN, // numbers + _xF // functions }; // Fillers to make layering more clear @@ -38,67 +16,125 @@ enum mitosis_layers #define KC_LMTA KC_LALT // For fun, name the mods like the space cadet keyboard does #define KC_RMTA KC_RALT // META #define KC_LSUP KC_LGUI // SUPER -#define KC_RSUP KC_RGUI // +#define KC_RSUP KC_RGUI // #define KC_RHYP KC_INT4 // HYPER (actually muhenkan 無変換 and henkan 変換) #define KC_LHYP KC_INT5 // or NFER/XFER. -// I didn't want to mess about with new keymappings and custom logic etc. to -// enable tri-state layers like mitosis default does. This layout accomplishes -// it with a small quirk that triggering both layers then releasing one -// out-of-order will leave the tri-state triggered. Which doesn't bother me. +// Momentary tri-state layers. Mitosis default keymap does this too but employs +// new keymappings and a bunch of conditional code. This simpler keymap +// accomplishes it, but with a small quirk: triggering both layers then +// releasing one out-of-order will leave the tri-state triggered until the +// other is released. Which doesn't bother me. const uint16_t PROGMEM keymaps[][MATRIX_ROWS][MATRIX_COLS] = { - [_WORKMAN] = { - {KC_Q, KC_D, KC_R, KC_W, KC_B, KC_J, KC_F, KC_U, KC_P, KC_SCLN}, - {KC_A, KC_S, KC_H, KC_T, KC_G, KC_Y, KC_N, KC_E, KC_O, KC_I}, - {KC_Z, KC_X, KC_M, KC_C, KC_V, KC_K, KC_L, KC_COMM, KC_DOT, KC_SLSH}, - {XXXXXXX, MO(_NUMS), KC_LCTL, KC_LSFT, KC_BSPC, KC_SPC, KC_RSFT, KC_RCTL, MO(_NUMS), XXXXXXX}, - {XXXXXXX, KC_LHYP, KC_LSUP, KC_LMTA, MO(_FUNC), MO(_FUNC), KC_RMTA, KC_RSUP, KC_RHYP, XXXXXXX} - }, - [_FUNC] = { - {KC_ESC, _______, KC_UP, _______, _______, _______, _______, _______, _______, KC_QUOT}, - {KC_TAB, KC_LEFT, KC_DOWN, KC_RGHT, _______, _______, _______, _______, _______, KC_GRV}, - {_______, _______, _______, _______, _______, _______, _______, KC_LBRC, KC_RBRC, KC_BSLS}, - {XXXXXXX, MO(_NMFN), _______, _______, KC_DEL, KC_ENT, _______, _______, MO(_NMFN), XXXXXXX}, - {XXXXXXX, _______, _______, _______, ___M___, ___M___, _______, _______, _______, XXXXXXX}, - }, - [_NUMS] = { - {_______, _______, _______, _______, _______, _______, KC_1, KC_2, KC_3, KC_4}, - {_______, _______, _______, _______, _______, _______, KC_5, KC_6, KC_7, KC_8}, - {_______, _______, _______, _______, _______, _______, KC_MINS, KC_9, KC_0, KC_EQL}, - {XXXXXXX, ___M___, _______, _______, _______, _______, _______, _______, ___M___, XXXXXXX}, - {XXXXXXX, _______, _______, _______, MO(_NMFN), MO(_NMFN), _______, _______, _______, XXXXXXX}, - }, - [_NMFN] = { - {_______, _______, KC_PGUP, _______, _______, _______, KC_F1, KC_F2, KC_F3, KC_F4}, - {_______, KC_HOME, KC_PGDN, KC_END, _______, _______, KC_F5, KC_F6, KC_F7, KC_F8}, - {_______, _______, _______, _______, _______, _______, KC_F8, KC_F9, KC_F10, KC_F12}, - {XXXXXXX, ___M___, _______, _______, _______, _______, _______, _______, ___M___, XXXXXXX}, - {XXXXXXX, _______, _______, _______, ___M___, ___M___, _______, _______, _______, XXXXXXX}, - }, + [_xW] = { + {KC_Q, KC_D, KC_R, KC_W, KC_B, KC_J, KC_F, KC_U, KC_P, KC_SCLN}, + {KC_A, KC_S, KC_H, KC_T, KC_G, KC_Y, KC_N, KC_E, KC_O, KC_I}, + {KC_Z, KC_X, KC_M, KC_C, KC_V, KC_K, KC_L, KC_COMM, KC_DOT, KC_QUOT}, + {XXXXXXX, KC_LSUP, KC_LCTL, MO(_xN), KC_LSFT, KC_RSFT, MO(_xN), KC_RCTL, KC_RSUP, XXXXXXX}, + {XXXXXXX, KC_LHYP, KC_LMTA, MO(_xS), KC_BSPC, KC_SPC, MO(_xS), KC_RMTA, KC_RHYP, XXXXXXX} + }, + [_xQ] = { + {KC_Q, KC_W, KC_E, KC_R, KC_T, KC_Y, KC_U, KC_I, KC_O, KC_P}, + {KC_A, KC_S, KC_D, KC_F, KC_G, KC_H, KC_J, KC_K, KC_L, KC_SCLN}, + {KC_Z, KC_X, KC_C, KC_V, KC_B, KC_N, KC_M, KC_COMM, KC_DOT, KC_QUOT}, + {XXXXXXX, _______, _______, _______, _______, _______, _______, _______, _______, XXXXXXX}, + {XXXXXXX, _______, _______, _______, _______, _______, _______, _______, _______, XXXXXXX} + }, + [_xS] = { + {KC_ESC, _______, KC_UP, _______, _______, KC_AT, KC_HASH, KC_DLR, KC_PERC, KC_TILD}, + {KC_TAB, KC_LEFT, KC_DOWN, KC_RGHT, _______, KC_CIRC, KC_AMPR, KC_PIPE, KC_GRV, KC_UNDS}, + {KC_BSLS, KC_RPRN, KC_RCBR, KC_RBRC, KC_RABK, KC_LABK, KC_LBRC, KC_LCBR, KC_LPRN, KC_SLSH}, + {XXXXXXX, _______, _______, MO(_xF), _______, _______, MO(_xF), _______, _______, XXXXXXX}, + {XXXXXXX, _______, _______, ___M___, KC_DEL, KC_ENT, ___M___, _______, _______, XXXXXXX}, + }, + [_xN] = { + {_______, _______, _______, _______, _______, KC_PSLS, KC_P7, KC_P8, KC_P9, KC_P0}, + {_______, _______, _______, _______, _______, KC_PAST, KC_P4, KC_P5, KC_P6, KC_PPLS}, + {_______, _______, _______, _______, _______, KC_PMNS, KC_P1, KC_P2, KC_P3, KC_PEQL}, + {XXXXXXX, _______, _______, ___M___, _______, _______, ___M___, _______, _______, XXXXXXX}, + {XXXXXXX, _______, _______, MO(_xF), _______, KC_PENT, MO(_xF), _______, _______, XXXXXXX}, + }, + [_xF] = { + {_______, _______, KC_PGUP, _______, KC_VOLU, KC_F13, KC_F7, KC_F8, KC_F9, KC_F10}, + {_______, KC_HOME, KC_PGDN, KC_END, KC_VOLD, KC_F14, KC_F4, KC_F5, KC_F6, KC_F11}, + {TG(_xQ), KC_MPRV, KC_MPLY, KC_MNXT, KC_MUTE, KC_F15, KC_F1, KC_F2, KC_F3, KC_F12}, + {XXXXXXX, _______, _______, ___M___, _______, _______, ___M___, _______, _______, XXXXXXX}, + {XXXXXXX, _______, _______, _______, ___M___, ___M___, _______, _______, _______, XXXXXXX}, + }, }; +// This is a hack to place <question mark> on <shift-comma> and <exclaimation +// mark> on <shift-period>, when using an operating system configured for a +// US/qwerty layout. +bool comm_shifted = false; +bool ques_shifted = false; +bool process_record_user(uint16_t keycode, keyrecord_t *record) { + uint8_t shifted; + uint16_t s_keycode; + bool *k_shifted; + + switch (keycode) { + case KC_COMM: + s_keycode = KC_SLSH; + k_shifted = &comm_shifted; + break; + case KC_DOT: + s_keycode = KC_1; + k_shifted = &ques_shifted; + break; + default: + return true; + } + + shifted = get_mods() & (MOD_BIT(KC_LSHIFT)|MOD_BIT(KC_RSHIFT)); + + // Keydown. If shift is currently pressed, register its alternate keycode. + if (record->event.pressed && shifted) { + *k_shifted = true; + register_code(s_keycode); + return false; + // Keyup. If shift was pressed back when the key was pressed, unregister + // its alternate keycode. + } else if (!(record->event.pressed) && *k_shifted) { + *k_shifted = false; + unregister_code(s_keycode); + return false; + // Otherwise, behave as normal. + } else { + return true; + } +} + +// Set the bits of A selected by MASK to the corresponding bits of B +#define setbits(A, B, MASK) A = (A & (B | ~MASK)) | (B & MASK) void matrix_scan_user(void) { - uint8_t layer = biton32(layer_state); + // + // Bit # 7 6 5 4 3 2 1 0 + // layer_state: [ | | | _xF | _xN | _xS | _xQ | _xW ] + // usb_led [ | | |kana |cmps |scrl |caps | num ] + // PORTB: [ NC | 10 | 9 | 8 | 14 | 16 | 15 |rxled] + // PORTC: [ NC | 5 | | | | | | ] + // PORTD: [ 6 | NC |txled| 4 | tx* | rx* | grn | p29 ] + // PORTE: [ | 7 | | | | | | ] + // PORTF: [ a0 | a1 | red | blu | | | NC | NC ] + // + // PD0 is connected to the pairing switch and p29 on the wireless module. + // PF0,PF1,PB7,PC7,PD6 are not broken out by the pro micro board. I don't understand why. + // PB1-PB6,PD4,PD5,PD6,PF6,PF7 are not connected to the Mitosis receiver + // board. Each may be connected to an LED by way of a resistor (4.7k to + // match the others) for a total of 14 additional indicators. - switch (layer) { - case _WORKMAN: - red_led_off; - grn_led_off; - break; - case _FUNC: - red_led_off; - grn_led_on; - break; - case _NUMS: - red_led_on; - grn_led_off; - break; - case _NMFN: - red_led_on; - grn_led_on; - break; - default: - break; - } -}; + // A simple (but technically inaccurate) model of the momentary layer state: + // Fn1 key makes _xS active; indicator = red + // Fn2 key makes _xN active; indicator = blue + // Both keys make _xF active; indicator = purple + // Toggling QWERTY mode makes indicator include green, so (red/blue/purple becomes yellow/cyan/white) + + // negated because for ports 0=LED on. + uint32_t portf_bits = ~(layer_state|layer_state<<1|(layer_state&0b100)<<3); + setbits(PORTF, portf_bits, 0b00110000); + setbits(PORTD, ~layer_state, 0b00000010); +} + +// vim: set sw=2 et: |