1 files changed, 764 insertions, 0 deletions
diff --git a/addons/gdUnit4/src/core/parse/GdScriptParser.gd b/addons/gdUnit4/src/core/parse/GdScriptParser.gd
new file mode 100644
index 0000000..a102508
--- /dev/null
+++ b/addons/gdUnit4/src/core/parse/GdScriptParser.gd
@@ -0,0 +1,764 @@
+class_name GdScriptParser
+extends RefCounted
+
+const GdUnitTools := preload("res://addons/gdUnit4/src/core/GdUnitTools.gd")
+
+const TYPE_VOID = GdObjects.TYPE_VOID
+const TYPE_VARIANT = GdObjects.TYPE_VARIANT
+const TYPE_VARARG = GdObjects.TYPE_VARARG
+const TYPE_FUNC = GdObjects.TYPE_FUNC
+const TYPE_FUZZER = GdObjects.TYPE_FUZZER
+const TYPE_ENUM = GdObjects.TYPE_ENUM
+
+
+var TOKEN_NOT_MATCH := Token.new("")
+var TOKEN_SPACE := SkippableToken.new(" ")
+var TOKEN_TABULATOR := SkippableToken.new("\t")
+var TOKEN_NEW_LINE := SkippableToken.new("\n")
+var TOKEN_COMMENT := SkippableToken.new("#")
+var TOKEN_CLASS_NAME := RegExToken.new("class_name", GdUnitTools.to_regex("(class_name)\\s+([\\w\\p{L}\\p{N}_]+) (extends[a-zA-Z]+:)|(class_name)\\s+([\\w\\p{L}\\p{N}_]+)"), 5)
+var TOKEN_INNER_CLASS := TokenInnerClass.new("class", GdUnitTools.to_regex("(class)\\s+(\\w\\p{L}\\p{N}_]+) (extends[a-zA-Z]+:)|(class)\\s+([\\w\\p{L}\\p{N}_]+)"), 5)
+var TOKEN_EXTENDS := RegExToken.new("extends", GdUnitTools.to_regex("extends\\s+"))
+var TOKEN_ENUM := RegExToken.new("enum", GdUnitTools.to_regex("enum\\s+"))
+var TOKEN_FUNCTION_STATIC_DECLARATION := RegExToken.new("static func", GdUnitTools.to_regex("^static\\s+func\\s+([\\w\\p{L}\\p{N}_]+)"), 1)
+var TOKEN_FUNCTION_DECLARATION := RegExToken.new("func", GdUnitTools.to_regex("^func\\s+([\\w\\p{L}\\p{N}_]+)"), 1)
+var TOKEN_FUNCTION := Token.new(".")
+var TOKEN_FUNCTION_RETURN_TYPE := Token.new("->")
+var TOKEN_FUNCTION_END := Token.new("):")
+var TOKEN_ARGUMENT_ASIGNMENT := Token.new("=")
+var TOKEN_ARGUMENT_TYPE_ASIGNMENT := Token.new(":=")
+var TOKEN_ARGUMENT_FUZZER := FuzzerToken.new(GdUnitTools.to_regex("((?!(fuzzer_(seed|iterations)))fuzzer?\\w+)( ?+= ?+| ?+:= ?+| ?+:Fuzzer ?+= ?+|)"))
+var TOKEN_ARGUMENT_TYPE := Token.new(":")
+var TOKEN_ARGUMENT_VARIADIC := Token.new("...")
+var TOKEN_ARGUMENT_SEPARATOR := Token.new(",")
+var TOKEN_BRACKET_ROUND_OPEN := Token.new("(")
+var TOKEN_BRACKET_ROUND_CLOSE := Token.new(")")
+var TOKEN_BRACKET_SQUARE_OPEN := Token.new("[")
+var TOKEN_BRACKET_SQUARE_CLOSE := Token.new("]")
+var TOKEN_BRACKET_CURLY_OPEN := Token.new("{")
+var TOKEN_BRACKET_CURLY_CLOSE := Token.new("}")
+
+
+var OPERATOR_ADD := Operator.new("+")
+var OPERATOR_SUB := Operator.new("-")
+var OPERATOR_MUL := Operator.new("*")
+var OPERATOR_DIV := Operator.new("/")
+var OPERATOR_REMAINDER := Operator.new("%")
+
+var TOKENS :Array[Token] = [
+	TOKEN_SPACE,
+	TOKEN_TABULATOR,
+	TOKEN_NEW_LINE,
+	TOKEN_COMMENT,
+	TOKEN_BRACKET_ROUND_OPEN,
+	TOKEN_BRACKET_ROUND_CLOSE,
+	TOKEN_BRACKET_SQUARE_OPEN,
+	TOKEN_BRACKET_SQUARE_CLOSE,
+	TOKEN_BRACKET_CURLY_OPEN,
+	TOKEN_BRACKET_CURLY_CLOSE,
+	TOKEN_CLASS_NAME,
+	TOKEN_INNER_CLASS,
+	TOKEN_EXTENDS,
+	TOKEN_ENUM,
+	TOKEN_FUNCTION_STATIC_DECLARATION,
+	TOKEN_FUNCTION_DECLARATION,
+	TOKEN_ARGUMENT_FUZZER,
+	TOKEN_ARGUMENT_TYPE_ASIGNMENT,
+	TOKEN_ARGUMENT_ASIGNMENT,
+	TOKEN_ARGUMENT_TYPE,
+	TOKEN_ARGUMENT_VARIADIC,
+	TOKEN_FUNCTION,
+	TOKEN_ARGUMENT_SEPARATOR,
+	TOKEN_FUNCTION_RETURN_TYPE,
+	OPERATOR_ADD,
+	OPERATOR_SUB,
+	OPERATOR_MUL,
+	OPERATOR_DIV,
+	OPERATOR_REMAINDER,
+]
+
+var _regex_strip_comments := GdUnitTools.to_regex("^([^#\"']|'[^']*'|\"[^\"]*\")*\\K#.*")
+var _scanned_inner_classes := PackedStringArray()
+var _script_constants := {}
+var _is_awaiting := GdUnitTools.to_regex("\\bawait\\s+(?![^\"]*\"[^\"]*$)(?!.*#.*await)")
+
+
+static func to_unix_format(input :String) -> String:
+	return input.replace("\r\n", "\n")
+
+
+class Token extends RefCounted:
+	var _token: String
+	var _consumed: int
+	var _is_operator: bool
+
+	func _init(p_token: String, p_is_operator := false) -> void:
+		_token = p_token
+		_is_operator = p_is_operator
+		_consumed = p_token.length()
+
+	func match(input: String, pos: int) -> bool:
+		return input.findn(_token, pos) == pos
+
+	func value() -> Variant:
+		return _token
+
+	func is_operator() -> bool:
+		return _is_operator
+
+	func is_inner_class() -> bool:
+		return _token == "class"
+
+	func is_variable() -> bool:
+		return false
+
+	func is_token(token_name :String) -> bool:
+		return _token == token_name
+
+	func is_skippable() -> bool:
+		return false
+
+	func _to_string() -> String:
+		return "Token{" + _token + "}"
+
+
+class Operator extends Token:
+	func _init(p_value: String) -> void:
+		super(p_value, true)
+
+	func _to_string() -> String:
+		return "OperatorToken{%s}" % [_token]
+
+
+# A skippable token, is just a placeholder like space or tabs
+class SkippableToken extends Token:
+
+	func _init(p_token: String) -> void:
+		super(p_token)
+
+	func is_skippable() -> bool:
+		return true
+
+
+# Token to parse function arguments
+class Variable extends Token:
+	var _plain_value :String
+	var _typed_value :Variant
+	var _type :int = TYPE_NIL
+
+
+	func _init(p_value: String) -> void:
+		super(p_value)
+		_type = _scan_type(p_value)
+		_plain_value = p_value
+		_typed_value = _cast_to_type(p_value, _type)
+
+
+	func _scan_type(p_value: String) -> int:
+		if p_value.begins_with("\"") and p_value.ends_with("\""):
+			return TYPE_STRING
+		var type_ := GdObjects.string_to_type(p_value)
+		if type_ != TYPE_NIL:
+			return type_
+		if p_value.is_valid_int():
+			return TYPE_INT
+		if p_value.is_valid_float():
+			return TYPE_FLOAT
+		if p_value.is_valid_hex_number():
+			return TYPE_INT
+		return TYPE_OBJECT
+
+
+	func _cast_to_type(p_value :String, p_type: int) -> Variant:
+		match p_type:
+			TYPE_STRING:
+				return p_value#.substr(1, p_value.length() - 2)
+			TYPE_INT:
+				return p_value.to_int()
+			TYPE_FLOAT:
+				return p_value.to_float()
+		return p_value
+
+
+	func is_variable() -> bool:
+		return true
+
+
+	func type() -> int:
+		return _type
+
+
+	func value() -> Variant:
+		return _typed_value
+
+
+	func plain_value() -> String:
+		return _plain_value
+
+
+	func _to_string() -> String:
+		return "Variable{%s: %s : '%s'}" % [_plain_value, GdObjects.type_as_string(_type), _token]
+
+
+class RegExToken extends Token:
+	var _regex: RegEx
+	var _extract_group_index: int
+	var _value := ""
+
+
+	func _init(token: String, regex: RegEx, extract_group_index: int = -1) -> void:
+		super(token, false)
+		_regex = regex
+		_extract_group_index = extract_group_index
+
+
+	func match(input: String, pos: int) -> bool:
+		var matching := _regex.search(input, pos)
+		if matching == null or pos != matching.get_start():
+			return false
+		if _extract_group_index != -1:
+			_value = matching.get_string(_extract_group_index)
+		_consumed = matching.get_end() - matching.get_start()
+		return true
+
+
+	func value() -> String:
+		return _value
+
+
+# Token to parse Fuzzers
+class FuzzerToken extends RegExToken:
+
+
+	func _init(regex: RegEx) -> void:
+		super("fuzzer", regex, 1)
+
+
+	func name() -> String:
+		return value()
+
+
+	func type() -> int:
+		return GdObjects.TYPE_FUZZER
+
+
+	func _to_string() -> String:
+		return "FuzzerToken{%s: '%s'}" % [value(), _token]
+
+
+class TokenInnerClass extends RegExToken:
+	var _content := PackedStringArray()
+
+
+	static func _strip_leading_spaces(input: String) -> String:
+		var characters := input.to_utf8_buffer()
+		while not characters.is_empty():
+			if characters[0] != 0x20:
+				break
+			characters.remove_at(0)
+		return characters.get_string_from_utf8()
+
+
+	static func _consumed_bytes(row: String) -> int:
+		return row.replace(" ", "").replace("	", "").length()
+
+
+	func _init(token: String, p_regex: RegEx, extract_group_index: int = -1) -> void:
+		super(token, p_regex, extract_group_index)
+
+
+	func is_class_name(clazz_name: String) -> bool:
+		return value() == clazz_name
+
+
+	func content() -> PackedStringArray:
+		return _content
+
+
+	@warning_ignore_start("return_value_discarded")
+	func parse(source_rows: PackedStringArray, offset: int) -> void:
+		# add class signature
+		_content.clear()
+		_content.append(source_rows[offset])
+		# parse class content
+		for row_index in range(offset+1, source_rows.size()):
+			# scan until next non tab
+			var source_row := source_rows[row_index]
+			var row := TokenInnerClass._strip_leading_spaces(source_row)
+			if row.is_empty() or row.begins_with("\t") or row.begins_with("#"):
+				# fold all line to left by removing leading tabs and spaces
+				if source_row.begins_with("\t"):
+					source_row = source_row.trim_prefix("\t")
+				# refomat invalid empty lines
+				if source_row.dedent().is_empty():
+					_content.append("")
+				else:
+					_content.append(source_row)
+				continue
+			break
+		_consumed += TokenInnerClass._consumed_bytes("".join(_content))
+	@warning_ignore_restore("return_value_discarded")
+
+
+	func _to_string() -> String:
+		return "TokenInnerClass{%s}" % [value()]
+
+
+
+func get_token(input: String, current_index: int) -> Token:
+	for t in TOKENS:
+		if t.match(input, current_index):
+			return t
+	return TOKEN_NOT_MATCH
+
+
+func next_token(input: String, current_index: int, ignore_tokens :Array[Token] = []) -> Token:
+	var token := TOKEN_NOT_MATCH
+	for t :Token in TOKENS.filter(func(t :Token) -> bool: return not ignore_tokens.has(t)):
+
+		if t.match(input, current_index):
+			token = t
+			break
+	if token == OPERATOR_SUB:
+		token = tokenize_value(input, current_index, token)
+	if token == TOKEN_NOT_MATCH:
+		return tokenize_value(input, current_index, token, ignore_tokens.has(TOKEN_FUNCTION))
+	return token
+
+
+func tokenize_value(input: String, current: int, token: Token, ignore_dots := false) -> Token:
+	var next := 0
+	var current_token := ""
+	# test for '--', '+-', '*-', '/-', '%-', or at least '-x'
+	var test_for_sign := (token == null or token.is_operator()) and input[current] == "-"
+	while current + next < len(input):
+		var character := input[current + next] as String
+		# if first charater a sign
+		# or allowend charset
+		# or is a float value
+		if (test_for_sign and next==0) \
+			or is_allowed_character(character) \
+			or (character == "." and (ignore_dots or current_token.is_valid_int())):
+			current_token += character
+			next += 1
+			continue
+		break
+	if current_token != "":
+		return Variable.new(current_token)
+	return TOKEN_NOT_MATCH
+
+
+# const ALLOWED_CHARACTERS := "0123456789_abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ\""
+func is_allowed_character(input: String) -> bool:
+	var code_point := input.unicode_at(0)
+	# Unicode
+	if code_point > 127:
+		# This is a Unicode character (Chinese, Japanese, etc.)
+		return true
+	# ASCII digit 0-9
+	if code_point >= 48 and code_point <= 57:
+		return true
+	# ASCII lowercase a-z
+	if code_point >= 97 and code_point <= 122:
+		return true
+	# ASCII uppercase A-Z
+	if code_point >= 65 and code_point <= 90:
+		return true
+	# underscore _
+	if code_point == 95:
+		return true
+	# quotes '"
+	if code_point == 34 or code_point == 39:
+		return true
+	return false
+
+
+func parse_return_token(input: String) -> Variable:
+	var index := input.rfind(TOKEN_FUNCTION_RETURN_TYPE._token)
+	if index == -1:
+		return TOKEN_NOT_MATCH
+	index += TOKEN_FUNCTION_RETURN_TYPE._consumed
+	# We scan for the return value exclusive '.' token because it could be referenced to a
+	# external or internal class e.g.  'func foo() -> InnerClass.Bar:'
+	var token := next_token(input, index, [TOKEN_FUNCTION])
+	while !token.is_variable() and token != TOKEN_NOT_MATCH:
+		index += token._consumed
+		token = next_token(input, index, [TOKEN_FUNCTION])
+	return token
+
+
+func get_function_descriptors(script: GDScript, included_functions: PackedStringArray = []) -> Array[GdFunctionDescriptor]:
+	var fds: Array[GdFunctionDescriptor] = []
+	for method_descriptor in script.get_script_method_list():
+		var func_name: String = method_descriptor["name"]
+		if included_functions.is_empty() or func_name in included_functions:
+			# exclude type set/geters
+			if is_getter_or_setter(func_name):
+				continue
+			if not fds.any(func(fd: GdFunctionDescriptor) -> bool: return fd.name() == func_name):
+				fds.append(GdFunctionDescriptor.extract_from(method_descriptor, false))
+
+	# we need to enrich it by default arguments and line number by parsing the script
+	# the engine core functions has no valid methods to get this info
+	_prescan_script(script)
+	_enrich_function_descriptor(script, fds)
+	return fds
+
+
+func is_getter_or_setter(func_name: String) -> bool:
+	return func_name.begins_with("@") and (func_name.ends_with("getter") or func_name.ends_with("setter"))
+
+
+func _parse_function_arguments(input: String) -> Array[Dictionary]:
+	var arguments: Array[Dictionary] = []
+	var current_index := 0
+	var token: Token = null
+	var bracket := 0
+	var in_function := false
+
+
+	while current_index < len(input):
+		token = next_token(input, current_index)
+		# fallback to not end in a endless loop
+		if token == TOKEN_NOT_MATCH:
+			var error : = """
+				Parsing Error: Invalid token at pos %d found.
+				Please report this error!
+				source_code:
+				--------------------------------------------------------------
+				%s
+				--------------------------------------------------------------
+			""".dedent() % [current_index, input]
+			push_error(error)
+			current_index += 1
+			continue
+		current_index += token._consumed
+		if token.is_skippable():
+			continue
+		if token == TOKEN_BRACKET_ROUND_OPEN :
+			in_function = true
+			bracket += 1
+		if token == TOKEN_BRACKET_ROUND_CLOSE:
+			bracket -= 1
+		# if function end?
+		if in_function and bracket == 0:
+			return arguments
+		# is function
+		if token == TOKEN_FUNCTION_DECLARATION:
+			continue
+
+		# is value argument
+		if in_function:
+			var arg_value := ""
+			var current_argument := {
+				"name" : "",
+				"value" : GdFunctionArgument.UNDEFINED,
+				"type" : TYPE_VARIANT
+			}
+
+			# parse type and default value
+			while current_index < len(input):
+				token = next_token(input, current_index)
+				current_index += token._consumed
+				if token.is_skippable():
+					continue
+
+				if token.is_variable() && current_argument["name"] == "":
+					arguments.append(current_argument)
+					current_argument["name"] = (token as Variable).plain_value()
+					continue
+
+				match token:
+							# is fuzzer argument
+					TOKEN_ARGUMENT_FUZZER:
+						arg_value = _parse_end_function(input.substr(current_index), true)
+						current_index += arg_value.length()
+						current_argument["name"] = (token as FuzzerToken).name()
+						current_argument["value"] = arg_value.lstrip(" ")
+						current_argument["type"] = TYPE_FUZZER
+						arguments.append(current_argument)
+						continue
+
+					TOKEN_ARGUMENT_VARIADIC:
+						current_argument["type"] = TYPE_VARARG
+
+					TOKEN_ARGUMENT_TYPE:
+						token = next_token(input, current_index)
+						if token == TOKEN_SPACE:
+							current_index += token._consumed
+							token = next_token(input, current_index)
+							current_index += token._consumed
+						if current_argument["type"] != TYPE_VARARG:
+							current_argument["type"] = GdObjects.string_to_type((token as Variable).plain_value())
+
+					TOKEN_ARGUMENT_TYPE_ASIGNMENT:
+						arg_value = _parse_end_function(input.substr(current_index), true)
+						current_index += arg_value.length()
+						current_argument["value"] = arg_value.lstrip(" ")
+					TOKEN_ARGUMENT_ASIGNMENT:
+						token = next_token(input, current_index)
+						arg_value = _parse_end_function(input.substr(current_index), true)
+						current_index += arg_value.length()
+						current_argument["value"] = arg_value.lstrip(" ")
+
+					TOKEN_BRACKET_SQUARE_OPEN:
+						bracket += 1
+					TOKEN_BRACKET_CURLY_OPEN:
+						bracket += 1
+					TOKEN_BRACKET_ROUND_OPEN :
+						bracket += 1
+						# if value a function?
+						if bracket > 1:
+							# complete the argument value
+							var func_begin := input.substr(current_index-TOKEN_BRACKET_ROUND_OPEN ._consumed)
+							var func_body := _parse_end_function(func_begin)
+							arg_value += func_body
+							# fix parse index to end of value
+							current_index += func_body.length() - TOKEN_BRACKET_ROUND_OPEN ._consumed - TOKEN_BRACKET_ROUND_CLOSE._consumed
+					TOKEN_BRACKET_SQUARE_CLOSE:
+						bracket -= 1
+					TOKEN_BRACKET_CURLY_CLOSE:
+						bracket -= 1
+					TOKEN_BRACKET_ROUND_CLOSE:
+						bracket -= 1
+						# end of function
+						if bracket == 0:
+							break
+					TOKEN_ARGUMENT_SEPARATOR:
+						if bracket <= 1:
+							# next argument
+							current_argument = {
+								"name" : "",
+								"value" : GdFunctionArgument.UNDEFINED,
+								"type" : GdObjects.TYPE_VARIANT
+							}
+							continue
+	return arguments
+
+
+func _parse_end_function(input: String, remove_trailing_char := false) -> String:
+	# find end of function
+	var current_index := 0
+	var bracket_count := 0
+	var in_array := 0
+	var in_dict := 0
+	var end_of_func := false
+
+	while current_index < len(input) and not end_of_func:
+		var character := input[current_index]
+		# step over strings
+		if character == "'" :
+			current_index = input.find("'", current_index+1) + 1
+			if current_index == 0:
+				push_error("Parsing error on '%s', can't evaluate end of string." % input)
+				return ""
+			continue
+		if character == '"' :
+			# test for string blocks
+			if input.find('"""', current_index) == current_index:
+				current_index = input.find('"""', current_index+3) + 3
+			else:
+				current_index = input.find('"', current_index+1) + 1
+			if current_index == 0:
+				push_error("Parsing error on '%s', can't evaluate end of string." % input)
+				return ""
+			continue
+
+		match character:
+			# count if inside an array
+			"[": in_array += 1
+			"]": in_array -= 1
+			# count if inside an dictionary
+			"{": in_dict += 1
+			"}": in_dict -= 1
+			# count if inside a function
+			"(": bracket_count += 1
+			")":
+				bracket_count -= 1
+				if bracket_count < 0 and in_array <= 0 and in_dict <= 0:
+					end_of_func = true
+			",":
+				if bracket_count == 0 and in_array == 0 and in_dict <= 0:
+					end_of_func = true
+		current_index += 1
+	if remove_trailing_char:
+		# check if the parsed value ends with comma or end of doubled breaked
+		# `<value>,` or `<function>())`
+		var trailing_char := input[current_index-1]
+		if trailing_char == ',' or (bracket_count < 0 and trailing_char == ')'):
+			return input.substr(0, current_index-1)
+	return input.substr(0, current_index)
+
+
+func extract_inner_class(source_rows: PackedStringArray, clazz_name :String) -> PackedStringArray:
+	for row_index in source_rows.size():
+		var input := source_rows[row_index]
+		var token := next_token(input, 0)
+		if token.is_inner_class():
+			@warning_ignore("unsafe_method_access")
+			if token.is_class_name(clazz_name):
+				@warning_ignore("unsafe_method_access")
+				token.parse(source_rows, row_index)
+				@warning_ignore("unsafe_method_access")
+				return token.content()
+	return PackedStringArray()
+
+
+func extract_func_signature(rows: PackedStringArray, index: int) -> String:
+	var signature := ""
+
+	for rowIndex in range(index, rows.size()):
+		var row := rows[rowIndex]
+		row = _regex_strip_comments.sub(row, "").strip_edges(false)
+		if row.is_empty():
+			continue
+		signature += row + "\n"
+		if is_func_end(row):
+			return signature.strip_edges()
+	push_error("Can't fully extract function signature of '%s'" % rows[index])
+	return ""
+
+
+func get_class_name(script :GDScript) -> String:
+	var source_code := GdScriptParser.to_unix_format(script.source_code)
+	var source_rows := source_code.split("\n")
+
+	for index :int in min(10, source_rows.size()):
+		var input := source_rows[index]
+		var token := next_token(input, 0)
+		if token == TOKEN_CLASS_NAME:
+			return token.value()
+	# if no class_name found extract from file name
+	return GdObjects.to_pascal_case(script.resource_path.get_basename().get_file())
+
+
+func parse_func_name(input: String) -> String:
+	if TOKEN_FUNCTION_DECLARATION.match(input, 0):
+		return TOKEN_FUNCTION_DECLARATION.value()
+	if TOKEN_FUNCTION_STATIC_DECLARATION.match(input, 0):
+		return TOKEN_FUNCTION_STATIC_DECLARATION.value()
+	push_error("Can't extract function name from '%s'" % input)
+	return ""
+
+
+## Enriches the function descriptor by line number and argument default values
+## - enrich all function descriptors form current script up to all inherited scrips
+func _enrich_function_descriptor(script: GDScript, fds: Array[GdFunctionDescriptor]) -> void:
+	var enriched_functions := {}  # Use Dictionary for O(1) lookup instead of PackedStringArray
+	var script_to_scan := script
+	while script_to_scan != null:
+		# do not scan the test suite base class itself
+		if script_to_scan.resource_path == "res://addons/gdUnit4/src/GdUnitTestSuite.gd":
+			break
+
+		var rows := script_to_scan.source_code.split("\n")
+		for rowIndex in rows.size():
+			var input := rows[rowIndex]
+			# step over inner class functions
+			if input.begins_with("\t"):
+				continue
+			# skip comments and empty lines
+			if input.begins_with("#") or input.length() == 0:
+				continue
+			var token := next_token(input, 0)
+			if token != TOKEN_FUNCTION_STATIC_DECLARATION and token != TOKEN_FUNCTION_DECLARATION:
+				continue
+
+			var function_name: String = token.value()
+			# Skip if already enriched (from parent class scan)
+			if enriched_functions.has(function_name):
+				continue
+
+			# Find matching function descriptor
+			var fd: GdFunctionDescriptor = null
+			for candidate in fds:
+				if candidate.name() == function_name:
+					fd = candidate
+					break
+			if fd == null:
+				continue
+			# Mark as enriched
+			enriched_functions[function_name] = true
+			var func_signature := extract_func_signature(rows, rowIndex)
+			var func_arguments := _parse_function_arguments(func_signature)
+			# enrich missing default values
+			fd.enrich_arguments(func_arguments)
+			fd.enrich_file_info(script_to_scan.resource_path, rowIndex + 1)
+			fd._is_coroutine = is_func_coroutine(rows, rowIndex)
+			# enrich return class name if not set
+			if fd.return_type() == TYPE_OBJECT and fd._return_class in ["", "Resource", "RefCounted"]:
+				var var_token := parse_return_token(func_signature)
+				if var_token != TOKEN_NOT_MATCH and var_token.type() == TYPE_OBJECT:
+					fd._return_class = _patch_inner_class_names(var_token.plain_value(), "")
+		# if the script ihnerits we need to scan this also
+		script_to_scan = script_to_scan.get_base_script()
+
+
+func is_func_coroutine(rows :PackedStringArray, index :int) -> bool:
+	var is_coroutine := false
+	for rowIndex in range(index+1, rows.size()):
+		var input := rows[rowIndex].strip_edges()
+		if input.begins_with("#") or input.is_empty():
+			continue
+		var token := next_token(input, 0)
+		# scan until next function
+		if token == TOKEN_FUNCTION_STATIC_DECLARATION or token == TOKEN_FUNCTION_DECLARATION:
+			break
+
+		if _is_awaiting.search(input):
+			return true
+	return is_coroutine
+
+
+func is_inner_class(clazz_path :PackedStringArray) -> bool:
+	return clazz_path.size() > 1
+
+
+func is_func_end(row :String) -> bool:
+	return row.strip_edges(false, true).ends_with(":")
+
+
+func _patch_inner_class_names(clazz :String, clazz_name :String = "") -> String:
+	var inner_clazz_name := clazz.split(".")[0]
+	if _scanned_inner_classes.has(inner_clazz_name):
+		return inner_clazz_name
+		#var base_clazz := clazz_name.split(".")[0]
+		#return base_clazz + "." + clazz
+	if _script_constants.has(clazz):
+		return clazz_name + "." + clazz
+	return clazz
+
+
+func _prescan_script(script: GDScript) -> void:
+	_script_constants = script.get_script_constant_map()
+	for key :String in _script_constants.keys():
+		var value :Variant = _script_constants.get(key)
+		if value is GDScript:
+			@warning_ignore("return_value_discarded")
+			_scanned_inner_classes.append(key)
+
+
+func parse(clazz_name :String, clazz_path :PackedStringArray) -> GdUnitResult:
+	if clazz_path.is_empty():
+		return GdUnitResult.error("Invalid script path '%s'" % clazz_path)
+	var is_inner_class_ := is_inner_class(clazz_path)
+	var script :GDScript = load(clazz_path[0])
+	_prescan_script(script)
+
+	if is_inner_class_:
+		var inner_class_name := clazz_path[1]
+		if _scanned_inner_classes.has(inner_class_name):
+			# do load only on inner class source code and enrich the stored script instance
+			var source_code := _load_inner_class(script, inner_class_name)
+			script = _script_constants.get(inner_class_name)
+			script.source_code = source_code
+	var function_descriptors := get_function_descriptors(script)
+	var gd_class := GdClassDescriptor.new(clazz_name, is_inner_class_, function_descriptors)
+	return GdUnitResult.success(gd_class)
+
+
+func _load_inner_class(script: GDScript, inner_clazz: String) -> String:
+	var source_rows := GdScriptParser.to_unix_format(script.source_code).split("\n")
+	# extract all inner class names
+	var inner_class_code := extract_inner_class(source_rows, inner_clazz)
+	return "\n".join(inner_class_code)