refactor(core): address audit findings (bugfixes, dedup, dead-code, comment hygiene)

[BANANASJIM]

Addresses a batch of codebase-audit findings across src/core/.

Changes

• gyro (correctness): selectTiltAxis returns null for .yaw instead of 0.0. Previously axis_y = "yaw" in tilt mode produced a constant-zero joy_y (non-null), which spuriously set suppress_right_stick_gyro and zeroed the physical stick Y axis. Now the axis is correctly absent. + regression test.
• gyro (comment hygiene): removed the vader5 origin annotation; renumbered the process-step comments to a contiguous [1..5].
• macro_player (correctness): emitPendingReleases now walks executed steps in reverse to compute net-held keys, removing the fixed 32-entry held[] cap that silently dropped releases for macros with >32 outstanding .down steps (stuck keys on cancel / layer-switch). + regression test (40 distinct held keys all released).
• macro_player (comment hygiene): stripped resolved issue #99 references from doc comments.
• mapper (dead-code): removed self.injected_buttons &= ~mask in the pending-tap-release flush — injected_buttons is unconditionally zeroed before output, so the masking had no effect; fixed the misleading comment.
• mapper (perf): hoisted effectiveStickConfig(.left/.right) so each is computed once per apply() frame instead of twice.
• mapper (comment hygiene): stripped resolved issue #99 reference from the axis-floor comment.
• mapper (test-gap): added integration tests for gesture tap and gesture hold (gamepad bit persists across frames then clears on release) through apply(), and for pause_for_release macro drain on layer deactivation.

Skipped (verified, not changed)

• timer_queue.cancel tombstone rewrite — micro-perf on a <32-entry heap; tombstone-by-token risks suppressing re-armed tokens. Current rebuild is correct and simpler.
• gyro/stick mode []const u8 → enum — wide cross-file refactor (runtime structs, resolve fns, ~11 call sites) for poll-rate string compares the finding itself calls harmless.
• layer.processLayerTriggers ResolvedLayerConfig — same perf-only tradeoff, requires a parallel resolved array kept in sync.
• currentMappedGamepadFrame dedup into buildEmitState — ~60-line refactor of intentionally-divergent logic; high behavior-change risk for a simplification.

Test plan

• scripts/padctl-docker test → EXIT=0 (full suite green).
• Each new regression test proven falsifiable by reverting its production fix and confirming only that test fails for the right reason.

refs: codebase audit

Summary by CodeRabbit

Bug Fixes

• Corrected tilt mode behavior where yaw axis was incorrectly converted to joystick output
• Expanded macro support to handle more than 32 distinct keys held simultaneously, eliminating previous tracking limitations
• Enhanced gesture tap/hold remapping and macro/layer deactivation to ensure proper key release handling across frames

[qodo-code-review]

Qodo reviews are paused for this user.

Troubleshooting steps vary by plan Learn more →

On a Teams plan?
Reviews resume once this user has a paid seat and their Git account is linked in Qodo.
Link Git account →

Using GitHub Enterprise Server, GitLab Self-Managed, or Bitbucket Data Center?
These require an Enterprise plan - Contact us
Contact us →

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 4420b732-4697-4bed-8d86-a15e766b89ca

📥 Commits

Reviewing files that changed from the base of the PR and between 8e1c4d6 and 78e3ddf.

📒 Files selected for processing (3)

• src/core/gyro.zig
• src/core/macro_player.zig
• src/core/mapper.zig

---

📝 Walkthrough

Walkthrough

Three core input-processing improvements: tilt mode now skips yaw-to-stick conversion by returning null from selectTiltAxis; macro pending-release handling removes the 32-entry tracking limit via reverse-scan logic; mapper.apply optimizes tap-release clearing and reuses stick-config computations while expanding gesture and macro/layer test coverage.

Changes

Input Processing Improvements

Layer / File(s)	Summary
Tilt axis null handling src/core/gyro.zig	selectTiltAxis returns null for yaw instead of 0.0, preventing tilt-mode yaw-to-joystick conversion; processMotion step comments clarified; test added verifying joy_y == null when configured with axis_y = .yaw.
Macro pending release refactor src/core/macro_player.zig	emitPendingReleases changed from forward-simulation with fixed 32-entry buffer to reverse-scan of executed steps, determining which .down targets remain net-held and emitting releases for uncovered targets; test added validating release of 40 distinct held keys.
Tap release and stick config optimization src/core/mapper.zig	Mapper.apply now nulls pending_tap_release without explicit injected_buttons bit masking; effectiveStickConfig calls moved earlier to share left_cfg/right_cfg across gyro override and stick suppression logic, eliminating duplicate config computations.
Gesture remap and macro/layer interaction tests src/core/mapper.zig	Added gesture tap/hold remap tests verifying tap key emission and gamepad bit persistence; added macro/layer interaction test confirming pause_for_release macros drain and emit releases correctly on layer deactivation.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• BANANASJIM/padctl#313: Both PRs modify selectTiltAxis tilt-axis handling in src/core/gyro.zig and update related tilt joystick tests.
• BANANASJIM/padctl#339: Main PR's MacroPlayer.emitPendingReleases refactor is directly related to the retrieved PR's use of the same method for clearing macro-held analog floors.
• BANANASJIM/padctl#198: Both PRs modify Mapper.apply flow for tap emission and pending-release handling across frames.

Poem

🐰 A hop through input flows so clean,
Tilt axes null where yaw has been,
Macros release all forty keys,
Configs reused with utmost ease—
Padctl's precision, fresh and keen!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately reflects the PR's primary focus: addressing multiple audit findings across the core codebase with a mix of bugfixes, code deduplication, dead code removal, and comment improvements.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/audit-core

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}