Add execution target that recursively canonicalises as well as canonical with canonical validity

[robert3005]

Should there be an explicit option to execute validity just by one step?

[codspeed-hq]

CodSpeed Performance Report

Merging this PR will degrade performance by 99.99%

_{Comparing rk/recursivecanonical (7689e15) with develop (28acfa4)}

Summary

⚡ 1 improved benchmark
❌ 53 regressed benchmarks
✅ 1084 untouched benchmarks
⏩ 1265 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_bool_into_canonical[(1000, 10)]	66 µs	88.6 µs	-25.57%
❌	Simulation	take_fsl_nullable_random[16, 1000]	6.8 µs	411.2 µs	-98.35%
❌	Simulation	take_fsl_nullable_random[1024, 100]	6.8 µs	1,999.3 µs	-99.66%
❌	Simulation	take_fsl_nullable_random[256, 100]	6.8 µs	537.7 µs	-98.73%
❌	Simulation	take_fsl_nullable_random[1024, 1000]	6.8 µs	22,594.1 µs	-99.97%
❌	Simulation	take_fsl_nullable_random[16, 100]	8.2 µs	137.5 µs	-94.06%
❌	Simulation	take_fsl_nullable_random[256, 1000]	6.8 µs	4,739.2 µs	-99.86%
❌	Simulation	take_fsl_nullable_random[4096, 1000]	6.8 µs	90,778.6 µs	-99.99%
❌	Simulation	take_fsl_nullable_random[4096, 100]	6.8 µs	8,113.2 µs	-99.92%
❌	Simulation	take_fsl_nullable_random[64, 1000]	6.8 µs	1,276.8 µs	-99.46%
❌	Simulation	take_fsl_random[16, 1000]	6.8 µs	352.8 µs	-98.06%
❌	Simulation	take_fsl_random[1024, 1000]	6.8 µs	23,028.9 µs	-99.97%
❌	Simulation	take_fsl_nullable_random[64, 100]	6.8 µs	177.4 µs	-96.14%
❌	Simulation	take_fsl_random[64, 100]	6.8 µs	159.8 µs	-95.71%
❌	Simulation	take_fsl_random[16, 100]	6.9 µs	67.9 µs	-89.88%
❌	Simulation	take_fsl_random[256, 100]	6.9 µs	530.8 µs	-98.7%
❌	Simulation	take_fsl_random[256, 1000]	6.8 µs	4,934 µs	-99.86%
❌	Simulation	take_fsl_random[64, 1000]	6.8 µs	1,263.2 µs	-99.46%
❌	Simulation	take_fsl_random[4096, 1000]	6.8 µs	92,697 µs	-99.99%
❌	Simulation	take_fsl_random[4096, 100]	6.9 µs	8,199.2 µs	-99.92%
...	...	...	...	...	...

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

Footnotes

1. 1265 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩