perf: Only memoize revisitable parameters

[CathalMullan]

No description provided.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/builder.rs	100.00% <100.00%> (ø)
src/compiler.rs	100.00% <100.00%> (ø)
src/needle.rs	100.00% <100.00%> (+3.44%)	⬆️
src/node.rs	96.55% <100.00%> (-0.04%)	⬇️
src/state.rs	100.00% <100.00%> (ø)

[codspeed-hq]

Merging this PR will not alter performance

⚡ 20 improved benchmarks
❌ 14 (👁 14) regressed benchmarks
✅ 170 untouched benchmarks

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	dynamic_depth[1000]	2.8 µs	2.5 µs	+15.03%
⚡	Simulation	dynamic_depth[100]	2.8 µs	2.5 µs	+12.51%
⚡	Simulation	dynamic_inline[1]	2.1 µs	1.9 µs	+9.31%
⚡	Simulation	wildcard_depth[100]	2.5 µs	2.3 µs	+8.17%
⚡	Simulation	wildcard_suffixes[100]	2.7 µs	2.5 µs	+7.72%
⚡	Simulation	wildcard_suffixes[1000]	2.7 µs	2.5 µs	+7.71%
⚡	Simulation	wildcard_chain[1000]	963.7 µs	909.2 µs	+6%
⚡	Simulation	wildcard_depth[1000]	2.5 µs	2.4 µs	+5.6%
⚡	Simulation	wildcard_chain[100]	103.7 µs	98.3 µs	+5.46%
⚡	Simulation	dynamic_depth[10]	2.6 µs	2.5 µs	+5.36%
⚡	Simulation	wildcard_endings[100]	2.6 µs	2.5 µs	+5.23%
⚡	Simulation	wildcard_endings[1000]	2.7 µs	2.6 µs	+5.11%
⚡	Simulation	wildcard_open[1]	2.9 µs	2.8 µs	+4.89%
⚡	Simulation	wildcard_open[10]	3.1 µs	2.9 µs	+4.69%
⚡	Simulation	wildcard_competing[1]	2.5 µs	2.4 µs	+4.33%
⚡	Simulation	dynamic_competing[1]	2.5 µs	2.4 µs	+4.17%
⚡	Simulation	wildcard_suffixes[1]	2.5 µs	2.4 µs	+4.14%
⚡	Simulation	wildcard_suffixes[10]	2.6 µs	2.5 µs	+4.09%
⚡	Simulation	wildcard_nested_inline[10]	3.3 µs	3.2 µs	+3.91%
⚡	Simulation	wildcard_backtrack[10]	3.1 µs	2.9 µs	+3.68%
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing revisit (74f875d) with main (4ff295e)}