perf[fsst]: like pushdown using a dfa

[joseph-isaacs]

Fsst like execution without decompression.

This uses a DFA over the symbol table and the like expression.

Once this is proved out we could think about putting this in fsst-rs?

[codspeed-hq]

Merging this PR will improve performance by ×4

⚡ 2 improved benchmarks
✅ 1000 untouched benchmarks
🆕 20 new benchmarks
⏩ 1522 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
🆕	Simulation	pure_pack_manual_word	N/A	200.4 µs	N/A
🆕	Simulation	with_offsets_bool_buf_64	N/A	948.2 µs	N/A
🆕	Simulation	with_offsets_collect_bool	N/A	890.5 µs	N/A
🆕	Simulation	with_offsets_manual_word	N/A	866.4 µs	N/A
🆕	Simulation	pure_pack_collect_bool	N/A	200.6 µs	N/A
🆕	Simulation	pure_pack_bool_buf_64	N/A	185.9 µs	N/A
🆕	Simulation	fsst_contains[cb]	N/A	19.3 ms	N/A
🆕	Simulation	fsst_contains[log]	N/A	22.7 ms	N/A
🆕	Simulation	fsst_contains[path]	N/A	5.6 ms	N/A
🆕	Simulation	fsst_contains[email]	N/A	6.7 ms	N/A
🆕	Simulation	fsst_contains[rare]	N/A	21.9 ms	N/A
🆕	Simulation	fsst_prefix[cb]	N/A	5.3 ms	N/A
🆕	Simulation	fsst_contains[json]	N/A	12.5 ms	N/A
🆕	Simulation	fsst_prefix[log]	N/A	5.4 ms	N/A
🆕	Simulation	fsst_prefix[email]	N/A	3.2 ms	N/A
🆕	Simulation	fsst_prefix[rare]	N/A	4.9 ms	N/A
🆕	Simulation	fsst_contains[urls]	N/A	13.1 ms	N/A
🆕	Simulation	fsst_prefix[json]	N/A	4.1 ms	N/A
🆕	Simulation	fsst_prefix[urls]	N/A	3.4 ms	N/A
🆕	Simulation	fsst_prefix[path]	N/A	3.3 ms	N/A
...	...	...	...	...	...

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

---

_{Comparing ji/fsst-like-kernel (e08fb69) with develop (7899299)}

Footnotes

1. 1522 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. ↩