🧼 split layout and input entrypoints

[natemoo-re]

Building on #50 (brotli+z85 encoding), this PR splits the WASM compilation into two scoped modules and exposes them as separate export subpaths, enabling per-consumer dead-code elimination and per-module opt tuning.

• WASM compilation now based on two scoped modules: module-layout.c and module-input.c
• new ./layout and ./input export subpaths exposed via build-npm.ts, allows downstream bundlers to tree-shake
• removed unified wasm.ts; updated .gitignore and Makefile clean targets accordingly

Per-consumer DCE

Consumer that import only one subpath no longer need to load the other module's WASM.

consumer	unified (main)	split	Δ
clayterm/input	43.0 kB	3.7 kB	−91%
clayterm/layout	43.0 kB	39.2 kB	−9%

Full bundle (JS + WASM), gzipped: input-only 47.8 kB → 8.5 kB (−82%); layout-only 48.8 kB → 45.4 kB (−7%).

Honest size note

Now that brotli+z85 (#50) is on main, the split does not reduce total install size. mem/utf8/wcwidth are duplicated across both modules, so the combined WASM is ~equal (42.1 kB split vs 42.4 kB unified, brotli) and the full package is marginally larger (115.4 kB → 121.5 kB unpacked, +5.3%). The wins here are scoping and per-module tunability, not aggregate size. Narrower compilation units also let --gc-sections apply more aggressive DCE per module.

[pkg-pr-new]

Open in StackBlitz

npm i https://pkg.pr.new/clayterm@37

commit: 51f4d26

[ghostdevv]

There are quite a few references to both wasm.ts and clayterm.wasm still

[cowboyd]

Looks good to me. So tiny! 🤩

[natemoo-re]

@cowboyd ah I need to rerun those numbers after rolling back the more aggressive stuff! still tiny but not quite this tiny (yet)

[codspeed-hq]

Merging this PR will degrade performance by 30.86%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
❌ 3 regressed benchmarks
✅ 16 untouched benchmarks

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	UTF-8 3-byte character	85 µs	176.6 µs	-51.86%
❌	Simulation	UTF-8 4-byte emoji	85.5 µs	173.2 µs	-50.62%
❌	Simulation	Kitty protocol (CSI u with modifiers)	97.6 µs	185.5 µs	-47.38%
⚡	Simulation	long input burst (200 bytes)	979.3 µs	536.1 µs	+82.66%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing ref/opt-split (51f4d26) with main (ebaf397)}