`_check_method_and_attr_name` in invalid name checker allows names with three leading underscores

[a1-su]

Currently, in the _check_method_and_attr_name function in invalid_name_checker.py, it allows three leading underscores instead of only two. While double leading underscores invoke Python's name-mangling rules, triple underscores have no special semantic meaning and should still be flagged by the invalid_name_checker.py.

pyta/packages/python-ta/src/python_ta/checkers/invalid_name_checker.py

Lines 234 to 251 in cbf34b3

	def _check_method_and_attr_name(node_type: str, name: str) -> list[str]:
	"""Returns a list of strings, each detailing how `name` violates Python naming conventions for
	method and instance or class attribute names.
	Returns an empty list if `name` is a valid method, instance, or attribute name."""
	error_msgs = []
	# Also consider the case of invoking Python's name mangling rules with leading dunderscores.
	if not (_is_in_snake_case(name) or (name.startswith("__") and _is_in_snake_case(name[2:]))):
	error_msgs.append(
	f'{node_type.capitalize()} name "{name}" should be in snake_case format. '
	f"{node_type.capitalize()} names should be lowercase, with words "
	f"separated by underscores. A single leading underscore can be used to "
	f"denote a private {node_type} while a double leading underscore invokes "
	f"Python's name-mangling rules."
	)
	return error_msgs

When name.startswith("__") is checked and the first two underscores are sliced out with _is_in_snake_case(name[2:]), it doesn't account for the _is_in_snake_case method allowing an optional leading underscore.

pyta/packages/python-ta/src/python_ta/checkers/invalid_name_checker.py

Lines 45 to 56 in cbf34b3

	def _is_in_snake_case(name: str) -> bool:
	"""Returns whether `name` is in snake_case.
	`name` is in snake_case if:
	- `name` starts with a lowercase letter or an underscore (to denote private fields) followed
	by a lowercase letter,
	- each word is separated by an underscore, and
	- each word is in lowercase.
	"""
	pattern = "(_?[a-z][a-z0-9_]*)$"
	return re.match(pattern, name) is not None