tests: parameterize

[jkowalleck]

i found it annoying that tests stopped on the first error.

so i changed the tests to be properly parameterized. this way, tests can run in parallel and dont fail fast - so that ALL errors are found, not only the first one ...

[jkowalleck]

@willynilly please review

[Copilot]

Pull request overview

Updates the syntax validation test suite to use pytest parameterization so each JSON test vector runs as an individual test case, improving failure reporting granularity.

Changes:

• Replaces nested loops in test_is_valid_syntax / test_not_is_valid_syntax with @pytest.mark.parametrize cases generated from JSON data.
• Introduces a shared syntax_data_as_params generator to build parameter sets and test ids.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

pytest.param(..., id=f"{term}-{example['value']}") will embed raw example values into the test id. The JSON test data includes control characters and newlines/tabs (e.g., "\t", "\n", "\u001f"), which can produce unreadable or even multi-line nodeids and make CI output/log parsing brittle. Consider escaping/sanitizing and truncating the value (e.g., using a repr/escaped form plus an index) when building the id.

Suggested change

	def syntax_data_as_params(src_cb):
	for term, examples in src_cb().items():
	for example in examples:
	yield pytest.param(term, example, id=f"{term}-{example['value']}")
	def _syntax_param_id(term, value, index, max_length=40):
	safe_value = str(value).encode("unicode_escape").decode("ascii")
	if len(safe_value) > max_length:
	safe_value = f"{safe_value[: max_length - 3]}..."
	return f"{term}-{index}-{safe_value}"
	def syntax_data_as_params(src_cb):
	for term, examples in src_cb().items():
	for index, example in enumerate(examples):
	yield pytest.param(
	term,
	example,
	id=_syntax_param_id(term, example["value"], index),
	)

[Copilot]

The debug output / failure messages interpolate *_example['value'] directly. Since the test vectors include whitespace/control characters, this can make the printed line and assertion message hard to read (or span multiple lines). Consider formatting the value as an escaped/repr form in both the print and the assertion message so failures are diagnosable.

Suggested change

	f"Testing {term} with {valid_example['value']} : {valid_example['reason']}"
	)
	assert (
	actual is True
	), f"Failed term: {term} for '{valid_example['value']}' : {valid_example['reason']}"
	@pytest.mark.parametrize("term,invalid_example", syntax_data_as_params(invalid_syntax_data))
	def test_not_is_valid_syntax(term, invalid_example):
	actual = h.is_valid_syntax(term=term, value=invalid_example["value"])
	print("")
	print(
	f"Testing {term} with {invalid_example['value']} : {invalid_example['reason']}"
	)
	assert (
	actual is False
	), f"Failed term: {term} for '{invalid_example['value']}' : {invalid_example['reason']}"
	f"Testing {term} with {valid_example['value']!r} : {valid_example['reason']}"
	)
	assert (
	actual is True
	), f"Failed term: {term} for {valid_example['value']!r} : {valid_example['reason']}"
	@pytest.mark.parametrize("term,invalid_example", syntax_data_as_params(invalid_syntax_data))
	def test_not_is_valid_syntax(term, invalid_example):
	actual = h.is_valid_syntax(term=term, value=invalid_example["value"])
	print("")
	print(
	f"Testing {term} with {invalid_example['value']!r} : {invalid_example['reason']}"
	)
	assert (
	actual is False
	), f"Failed term: {term} for {invalid_example['value']!r} : {invalid_example['reason']}"

[Copilot]

PR description mentions tests "can run in parallel", but the repo's testing extras currently only declare pytest (no pytest-xdist), so parametrization alone won't make the suite run in parallel by default. Either adjust the description/expectation, or add xdist to the testing extras / CI if parallel execution is a goal.

[Copilot]

Parametrizing every example creates one pytest test case per JSON entry; with invalid_syntax.json being large, this can significantly increase collection time and per-test overhead compared to a single loop-based test. If runtime becomes an issue, consider parametrizing at a coarser granularity (e.g., per term) while still aggregating failures within the test, so you keep the "report all failures" behavior without thousands of separate test nodes.

Suggested change

	def syntax_data_as_params(src_cb):
	for term, examples in src_cb().items():
	for example in examples:
	yield pytest.param(term, example, id=f"{term}-{example['value']}")
	@pytest.mark.parametrize("term,valid_example", syntax_data_as_params(valid_syntax_data))
	def test_is_valid_syntax(term, valid_example):
	actual = h.is_valid_syntax(term=term, value=valid_example["value"])
	print("")
	print(
	f"Testing {term} with {valid_example['value']} : {valid_example['reason']}"
	)
	assert (
	actual is True
	), f"Failed term: {term} for '{valid_example['value']}' : {valid_example['reason']}"
	@pytest.mark.parametrize("term,invalid_example", syntax_data_as_params(invalid_syntax_data))
	def test_not_is_valid_syntax(term, invalid_example):
	actual = h.is_valid_syntax(term=term, value=invalid_example["value"])
	print("")
	print(
	f"Testing {term} with {invalid_example['value']} : {invalid_example['reason']}"
	)
	assert (
	actual is False
	), f"Failed term: {term} for '{invalid_example['value']}' : {invalid_example['reason']}"
	def syntax_data_by_term(src_cb):
	for term, examples in src_cb().items():
	yield pytest.param(term, examples, id=term)
	@pytest.mark.parametrize("term,valid_examples", syntax_data_by_term(valid_syntax_data))
	def test_is_valid_syntax(term, valid_examples):
	failures = []
	for valid_example in valid_examples:
	actual = h.is_valid_syntax(term=term, value=valid_example["value"])
	print("")
	print(
	f"Testing {term} with {valid_example['value']} : {valid_example['reason']}"
	)
	if actual is not True:
	failures.append(
	f"Failed term: {term} for '{valid_example['value']}' : {valid_example['reason']}"
	)
	if failures:
	pytest.fail("\n".join(failures))
	@pytest.mark.parametrize("term,invalid_examples", syntax_data_by_term(invalid_syntax_data))
	def test_not_is_valid_syntax(term, invalid_examples):
	failures = []
	for invalid_example in invalid_examples:
	actual = h.is_valid_syntax(term=term, value=invalid_example["value"])
	print("")
	print(
	f"Testing {term} with {invalid_example['value']} : {invalid_example['reason']}"
	)
	if actual is not False:
	failures.append(
	f"Failed term: {term} for '{invalid_example['value']}' : {invalid_example['reason']}"
	)
	if failures:
	pytest.fail("\n".join(failures))

[jkowalleck]

Hi! Since this repository is no longer actively maintained, I created and now maintain a fork of this package long-term:

• PyPI: https://pypi.org/project/rfc3987-syntax2/
• Repository: https://github.com/jkowalleck/rfc3987-syntax2

Would you mind retargeting (or reopening) this pull request against the maintained fork so your fix/feature can be reviewed and merged there?

Thank you!