refactor(execution): NativeExecutor refactor

[universalmind303]

Small refactor to NativeExecutor to make it a bit more flexible for spark connect.

some changes made

• make the following configurable:
  • runtime
  • progress_bar_manager
  • enable_explain_analyze
• refactor the engine to take in the plan when performing .run not at creation.
  • this is useful as we can now reuse an engine instead of having to create a new one for every plan.
  • Note: we could probably follow up with a refactor the python code to have a single "native" engine instead of creating it every time

The runtime configuration is especially important for spark connect as previously we were using spawn_blocking when running a plan, but now we can leverage the existing compute runtime instead of creating a new one.

[codspeed-hq]

CodSpeed Performance Report

Merging #3689 will improve performances by 55.47%

_{Comparing universalmind303:execution-refactor (6f02729) with main (34d2036)}

Summary

⚡ 1 improvements
✅ 26 untouched benchmarks

Benchmarks breakdown

	Benchmark	main	universalmind303:execution-refactor	Change
⚡	test_iter_rows_first_row[100 Small Files]	197.1 ms	126.8 ms	+55.47%