feat(docs): re-write test gen doc flow with pytest plugin

[danceratopz]

🗒️ Description

This is a rewrite of the automatic test case doc generation as a pytest plugin (as apposed to a standalone script). A preview is available at https://ethereum.github.io/execution-spec-tests/pr-801-preview.

Doc Improvements

1. Test module pages: Added a table that provides and overview of its test functions, e.g., tests/prague/eip2537_bls_12_381_precompiles/test_bls12_g1add.
2. Added test function pages which include a table which tries to give an overview of its parametrized test cases, e.g., tests/prague/eip2537_bls_12_381_precompiles/test_bls12_g1add::test_valid.
3. Added static html files for the tables of the function parameters overview, as they often don't play well within mkdocs styling.
4. Markdown pages: Added a header with a path and ref link for the markdown file, e.g., tests/prague/eip7692_eof_v1/tracker.md

Technical Improvements

1. Doc build fails if no function docstring is available or the test type (e.g. state_test) can not be determined (both used in the function overview tables on test module pages).
2. Using a pytest plugin allows us to lever all the usual pytest flags for test item filtering and fork specification (great for debugging) and to introspect test items to easily grab test metadata.
3. jinja2 is used for templating, which helps separate the content from its formatting. Jinja is a very powerful templating library which helps save a lot of boilerplate code.

🔗 Related Issues

Perhaps solved, plz check whether the styling is acceptable, e.g., test_bls12_g1add/test_valid.html.

• feat(docs): add styling for standalone html table pages #804.

Current issues:

• feat(docs,fw): enable better representations of test parameters in parameter overview tables #803.

Ideas for future improvements:

• feat(forks, docs): allow specification of multiple development forks #802 (not a problem now, but we can only gen docs using one dev fork).
• feat(docs): make it more transparent when a test function is a class method #805.
• fix(docs): move initialization js from jinja2 template to site.js #813.
• feat(docs): make test functions discoverable by full test node id and test function parameter ids #814.
• feat(docs, fw): add a library function that maps a test function id to a (versioned) documentation html page #815.
• docs(bug): function parameter datatables don't switch in dark mode #816.

✅ Checklist

• All: Set appropriate labels for the changes.
• All: Considered squashing commits to improve commit history.
• All: Added an entry to CHANGELOG.md.
• All: Considered updating the online docs in the ./docs/ directory.
• Tests: All converted JSON/YML tests from ethereum/tests have been added to converted-ethereum-tests.txt.
• Tests: A PR with removal of converted JSON/YML tests from ethereum/tests have been opened.
• Tests: Included the type and version of evm t8n tool used to locally execute test cases: e.g., ref with commit hash or geth 1.13.1-stable-3f40e65.
• Tests: Ran mkdocs serve locally and verified the auto-generated docs for new tests in the Test Case Reference are correctly formatted.

[marioevz]

This is absolutely amazing, great PR!
Just some minor comments. Thanks!

[marioevz]

We can import SPEC_TYPES from ethereum_test_specs then do:

Suggested change

	test_types: List[str] = [
	"state_test",
	"state_test_only",
	"blockchain_test",
	"eof_test",
	"eof_state_test",
	]
	test_types: List[str] = [spec_type.pytest_parameter_name() for spec_type in SPEC_TYPES]

[marioevz]

Suggested change

	self.skip_params = ["blockchain_test", "state_test", "eof_test", "fork"]
	self.skip_params = ["fork"] + [
	spec_type.pytest_parameter_name() for spec_type in SPEC_TYPES
	]

[marioevz]

I realized on 7702 tests that I copy-pasted the docstrings of some of the tests.

I think this would be a good time to check for duplicate docstrings in different test functions:

Suggested change

	lines = docstring.splitlines()
	test_function_id = get_test_function_id(item)
	if (
	docstring in docstring_test_function_history
	and docstring_test_function_history[docstring] != test_function_id
	):
	logger.warning(
	f"Duplicate docstring for {test_function_id}: "
	f"{docstring_test_function_history[docstring]} and {test_function_id}"
	)
	else:
	docstring_test_function_history[docstring] = test_function_id
	lines = docstring.splitlines()

docstring_test_function_history in this case is a simple global dictionary to keep track of the docstring texts of all test functions visited so far.

[marioevz]

Suggested change

	def get_template(self) -> str:
	@property
	@abstractmethod
	def template(self) -> str:
	pass

And then define the attribute in each subclass:

@dataclass
class FunctionPageProps(PagePropsBase):
  ...
  template: str = "function.md.j2"