feat(input): raw device layer — multi-key held state, analog, mouse, gamepad, touch, full-state replay (#50)

The raw device layer the input proposal sketched, over the action maps +
record/replay of #7. Beyond one key per frame, gameplay can read:

- Multiple simultaneous held keys: Input.key_down / key_pressed / key_released,
  with clean rising/falling edges (hold left AND jump).
- Analog from keys: Input.axis(neg, pos) and a normalized Input.vector(l,r,u,d)
  (diagonals scaled by 1/sqrt(2)), plus Input.strength(action).
- Mouse: Input.mouse_x/y, mouse_dx/dy (per-frame delta), mouse_down(btn), wheel.
- Gamepads: Input.pad_connected/pad_button/pad_axis (SDL-order buttons, -1..1
  sticks); touch: Input.touch_count/touch_x/touch_y.

The held set is fed by the platform when windowed — cocoa.ll now tracks
keyDown/keyUp into a 256-bit held-key bitset (win_held) and the mouse
buttons/wheel (win_mouse), gated so headless builds DCE the native calls — and
by the Input.press / Input.set_mouse / Input.set_pad / Input.set_touch injection
on every target (Godot-style action injection: replays, AI, network-fed input).
Input.record / replay now snapshot the full per-frame device state (held set +
mouse), extending #7's single-key tape.

Everything is integer and deterministic, so the same inputs reproduce the same
frame on every run and headless. The gamepad/touch native hardware bindings
(GameController.framework / NSTouch) feed the same injected state and are the one
remaining platform-glue follow-up; the software layer, semantics and replay are
complete and driven deterministically today.

Worked example + regression: examples/library/input_device.ludic
(1 1 0 1 0 1 71 -71 5 1 3 1 2 1 0 0 1, injection-driven headless). 23 new
docs/language/input pages. Full suite 78 passed, self-host C-free fixpoint
intact, no golden drift; cocoa.ll assembles and a windowed build links.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

selfhost/frontend/parse.ludic

@@ -183,6 +183,9 @@ function p_postfix() -> Node {
       # Input.* action-map / record-replay methods (#7) -> splice input.ludic.
       # Input.key stays bare (no runtime), so gate on the new methods only.
       if e.a.kind == E_ID and e.a.s == "Input" and (e.s == "bind" or e.s == "rebind" or e.s == "poll" or e.s == "down" or e.s == "pressed" or e.s == "record" or e.s == "replay") { g_uses_input = true }
+      # #50 device layer — any of the multi-key / analog / mouse / gamepad / touch
+      # methods also splices input.ludic (Input.key stays bare, no runtime).
+      if e.a.kind == E_ID and e.a.s == "Input" and (e.s == "key_down" or e.s == "key_pressed" or e.s == "key_released" or e.s == "press" or e.s == "release" or e.s == "axis" or e.s == "vector" or e.s == "strength" or e.s == "mouse_x" or e.s == "mouse_y" or e.s == "mouse_dx" or e.s == "mouse_dy" or e.s == "mouse_down" or e.s == "wheel" or e.s == "set_mouse" or e.s == "pad_connected" or e.s == "pad_button" or e.s == "pad_axis" or e.s == "set_pad" or e.s == "touch_count" or e.s == "touch_x" or e.s == "touch_y" or e.s == "set_touch") { g_uses_input = true }
       # Anim.play/clip/on_frame/fired + Motion.to (#48): the ergonomic writes over
       # the SpriteAnim/Motion components live in systems.ludic and use the world
       # table, so splice it and force the reflection ABI even if the game leaves