robinxx_map 0.1.0

High-performance, thread-safe open-addressing hash map using Robin Hood displacement & xxHash3.
Documentation
# Task Breakdown & Traceability: robinxx_map
**Role:** Architect Oversight → Senior Engineer Execution
**Methodology:** Spec-Driven Development (SDD) Phased Execution
**MSRV:** Rust 1.95.0

## Phase Traceability Matrix
| ID | Task | Phase | Acceptance Criteria | Links | Status | Owner |
|----|------|-------|---------------------|-------|--------|-------|
| T-01 | Project Init & Config | Setup | `Cargo.toml` valid, MSRV 1.95.0, `no_std` + `alloc` ready | NFR-04 || Senior Eng |
| T-02 | Core Primitives & Hash | Core | `RobinHoodKey` impl, `xxhash-rust` binding, unit tests pass | FR-08 || Senior Eng |
| T-03 | Raw Table & Memory | Core | SoA allocation, 64B align, DfH probing, resize, `unsafe` + `# Safety` | FR-01, NFR-02 || Senior Eng |
| T-04 | Concurrency & Sync | Core | `SpinMutex` impl, `Send + Sync`, thread-safe lock/unlock | NFR-03 || Senior Eng |
| T-05 | Public API & Entry/Iter | Integration | `with_key` closure API, `Entry` (`K: Clone`), `Iter`, `Reserve` | FR-02~07 || Senior Eng |
| T-06 | Testing & Validation | Validation | `proptest` invariants, `miri` clean, coverage ≥90/100% | NFR-02, NFR-05 || Senior Eng |
| T-07 | Benchmarking Suite | Performance | `criterion` vs `std` & `hashbrown`, `--noplot`, stable CI | NFR-01 || Senior Eng |
| T-08 | Docs, CI & Release | Compliance | `rustdoc -D warnings`, `clippy`, `fmt`, MIT 2026, GH Actions | NFR-04, NFR-05 || Senior Eng |

## Validation Sequence (Per Phase)
1. `cargo check` → type/ownership validation
2. `cargo test` → unit/integration/doctest
3. `cargo miri test` → undefined behavior & pointer safety validation (Mandatory)
4. `cargo bench -- --noplot` → performance regression check
5. `RUSTDOCFLAGS="-D warnings" cargo doc` → documentation strictness
6. `cargo clippy -- -D warnings` → lint compliance
7. `cargo fmt --check` → formatting
8. `cargo llvm-cov --open` → coverage verification

## Branching & Merge Policy
- **Branches:** `feat/<scope>`, `test/<module>`, `perf/<benchmark>`, `roadmap/<v0.2>`
- **PR Requirements:** Pass validation sequence. Attach benchmark delta. Miri clean required.
- **Merge:** Squash merge after Architect review & CI green.

## Performance & Regression Guardrails
- Baseline pinned per release: `cargo bench -- --save-baseline vX.X.X`
- Regression >5% → revert, profile with `cargo flamegraph`/`perf`
- Cache contention (`std. dev.` >15%) → adjust prefetch/alignment params.

## Rollout & Publish Checklist
- [x] Metadata valid (`Cargo.toml`, MSRV 1.95.0)
- [x] License match (`MIT 2026`, `Dzulkifli Anwar`)
- [x] Docs built clean (`rustdoc` 0 warnings)
- [x] Tests pass (`proptest` + `criterion` stable)
- [x] Bench stable (vs `std` & `hashbrown`)
- [x] Clippy clean (`-D warnings`)
- [x] `CHANGELOG.md`, `CONTRIBUTING.md`, `CODE_OF_CONDUCT.md` ready
- [x] Tag & push → GitHub/Codeberg
- [x] `cargo publish --dry-run`**Manual** `cargo publish`

## Architect Review Notes
- `SpinMutex` critical sections must remain <50ns average to avoid thread contention.
- `xxhash-rust` feature flag `xxh3` explicitly enabled in `Cargo.toml`.
- Auto-vectorization relies on contiguous SoA layout, avoid pointer chasing in probe loops.
- **v0.2.0 Focus:** API Parity (`get`/`get_mut` refs) is highest priority before concurrency scaling.