feat: add java tests

codebase_rag/graph_updater.py

@@ -403,7 +403,9 @@ def should_skip_path(path: Path) -> bool:
 
     def _process_function_calls(self) -> None:
         """Third pass: Process function calls using the cached ASTs."""
-        for file_path, (root_node, language) in self.ast_cache.items():
+        # Create a copy of items to prevent "OrderedDict mutated during iteration" errors
+        ast_cache_items = list(self.ast_cache.items())
+        for file_path, (root_node, language) in ast_cache_items:
             self.factory.call_processor.process_calls_in_file(
                 file_path, root_node, language, self.queries
             )

codebase_rag/language_config.py

@@ -220,12 +220,41 @@ class LanguageConfig:
             "interface_declaration",
             "enum_declaration",
             "annotation_type_declaration",
+            "record_declaration",
         ],
         module_node_types=["program"],
         package_indicators=[],  # Java uses package declarations
         call_node_types=["method_invocation"],
         import_node_types=COMMON_DECLARATION_IMPORT,
         import_from_node_types=COMMON_DECLARATION_IMPORT,  # Java uses same node for imports
+        # Pre-formatted Tree-sitter queries for comprehensive Java parsing
+        function_query="""
+        (method_declaration
+            name: (identifier) @name) @function
+        (constructor_declaration
+            name: (identifier) @name) @function
+        """,
+        class_query="""
+        (class_declaration
+            name: (identifier) @name) @class
+        (interface_declaration
+            name: (identifier) @name) @class
+        (enum_declaration
+            name: (identifier) @name) @class
+        (annotation_type_declaration
+            name: (identifier) @name) @class
+        (record_declaration
+            name: (identifier) @name) @class
+        """,
+        call_query="""
+        (method_invocation
+            name: (identifier) @name) @call
+        (method_invocation
+            object: (_)
+            name: (identifier) @name) @call
+        (object_creation_expression
+            type: (type_identifier) @name) @call
+        """,
     ),
     "cpp": create_lang_config(
         file_extensions=[

codebase_rag/parsers/call_processor.py

@@ -285,6 +285,22 @@ def _get_call_target_name(self, call_node: Node) -> str | None:
                 return convert_operator_symbol_to_name(operator_text)
 
         # For 'method_invocation' in Java
+        if call_node.type == "method_invocation":
+            # Get the object (receiver) part
+            object_node = call_node.child_by_field_name("object")
+            name_node = call_node.child_by_field_name("name")
+
+            if name_node and name_node.text:
+                method_name = str(name_node.text.decode("utf8"))
+
+                if object_node and object_node.text:
+                    object_text = str(object_node.text.decode("utf8"))
+                    return f"{object_text}.{method_name}"
+                else:
+                    # No object, likely this.method() or static method
+                    return method_name
+
+        # General case for other languages
         if name_node := call_node.child_by_field_name("name"):
             text = name_node.text
             if text is not None:
@@ -327,6 +343,10 @@ def _ingest_function_calls(
         captures = cursor.captures(caller_node)
         call_nodes = captures.get("call", [])
 
+        logger.debug(
+            f"Found {len(call_nodes)} call nodes in {language} for {caller_qn}"
+        )
+
         for call_node in call_nodes:
             if not isinstance(call_node, Node):
                 continue
@@ -345,9 +365,15 @@ def _ingest_function_calls(
             if not call_name:
                 continue
 
-            callee_info = self._resolve_function_call(
-                call_name, module_qn, local_var_types, class_context
-            )
+            # Use Java-specific resolution for Java method calls
+            if language == "java" and call_node.type == "method_invocation":
+                callee_info = self._resolve_java_method_call(
+                    call_node, module_qn, local_var_types, language
+                )
+            else:
+                callee_info = self._resolve_function_call(
+                    call_name, module_qn, local_var_types, class_context
+                )
             if not callee_info:
                 # Check if it's a built-in JavaScript method
                 builtin_info = self._resolve_builtin_call(call_name)
@@ -1027,3 +1053,26 @@ def _is_method(self, func_node: Node, lang_config: LanguageConfig) -> bool:
                 return True
             current = current.parent
         return False
+
+    def _resolve_java_method_call(
+        self,
+        call_node: Node,
+        module_qn: str,
+        local_var_types: dict[str, str],
+        language: str,
+    ) -> tuple[str, str] | None:
+        """Resolve Java method calls using the JavaTypeInferenceEngine."""
+        # Get the Java type inference engine from the main type inference engine
+        java_engine = self.type_inference.java_type_inference
+
+        # Use the Java engine to resolve the method call
+        result = java_engine.resolve_java_method_call(
+            call_node, local_var_types, module_qn
+        )
+
+        if result:
+            logger.debug(
+                f"Java method call resolved: {call_node.text.decode('utf8') if call_node.text else 'unknown'} -> {result[1]}"
+            )
+
+        return result

codebase_rag/parsers/definition_processor.py

@@ -43,13 +43,15 @@ def __init__(
         function_registry: Any,
         simple_name_lookup: dict[str, set[str]],
         import_processor: ImportProcessor,
+        module_qn_to_file_path: dict[str, Path],
     ):
         self.ingestor = ingestor
         self.repo_path = repo_path
         self.project_name = project_name
         self.function_registry = function_registry
         self.simple_name_lookup = simple_name_lookup
         self.import_processor = import_processor
+        self.module_qn_to_file_path = module_qn_to_file_path
         self.class_inheritance: dict[str, list[str]] = {}
 
     def _get_node_type_for_inheritance(self, qualified_name: str) -> str:
@@ -121,6 +123,9 @@ def process_file(
                     [self.project_name] + list(relative_path.parent.parts)
                 )
 
+            # Populate the module QN to file path mapping for efficient lookups
+            self.module_qn_to_file_path[module_qn] = file_path
+
             self.ingestor.ensure_node_batch(
                 "Module",
                 {
@@ -1151,6 +1156,16 @@ def _ingest_classes_and_methods(
                     node_type, class_qn, parent_class_qn
                 )
 
+            # Handle Java interface implementations
+            if class_node.type == "class_declaration":
+                implemented_interfaces = self._extract_implemented_interfaces(
+                    class_node, module_qn
+                )
+                for interface_qn in implemented_interfaces:
+                    self._create_implements_relationship(
+                        node_type, class_qn, interface_qn
+                    )
+
             body_node = class_node.child_by_field_name("body")
             if not body_node:
                 continue
@@ -1327,6 +1342,35 @@ def _extract_parent_classes(self, class_node: Node, module_qn: str) -> list[str]
                     )
             return parent_classes
 
+        # Look for superclass in Java class definition (extends clause)
+        if class_node.type == "class_declaration":
+            superclass_node = class_node.child_by_field_name("superclass")
+            if superclass_node:
+                # Java superclass is a single type identifier
+                if superclass_node.type == "type_identifier":
+                    parent_text = superclass_node.text
+                    if parent_text:
+                        parent_name = parent_text.decode("utf8")
+                        # Resolve to full qualified name if possible
+                        resolved_superclass = (
+                            self._resolve_class_name(parent_name, module_qn)
+                            or f"{module_qn}.{parent_name}"
+                        )
+                        parent_classes.append(resolved_superclass)
+                else:
+                    # Look for type_identifier children in superclass node
+                    for child in superclass_node.children:
+                        if child.type == "type_identifier":
+                            parent_text = child.text
+                            if parent_text:
+                                parent_name = parent_text.decode("utf8")
+                                resolved_superclass_child = (
+                                    self._resolve_class_name(parent_name, module_qn)
+                                    or f"{module_qn}.{parent_name}"
+                                )
+                                parent_classes.append(resolved_superclass_child)
+                                break
+
         # Look for superclasses in Python class definition
         superclasses_node = class_node.child_by_field_name("superclasses")
         if superclasses_node:
@@ -1343,11 +1387,11 @@ def _extract_parent_classes(self, class_node: Node, module_qn: str) -> list[str]
                                 parent_classes.append(import_map[parent_name])
                             else:
                                 # Try to resolve within same module
-                                resolved_parent = self._resolve_class_name(
-                                    parent_name, module_qn
+                                resolved_python_parent: str | None = (
+                                    self._resolve_class_name(parent_name, module_qn)
                                 )
-                                if resolved_parent is not None:
-                                    parent_classes.append(resolved_parent)
+                                if resolved_python_parent is not None:
+                                    parent_classes.append(resolved_python_parent)
                                 else:
                                     # Fallback: assume same module
                                     parent_classes.append(f"{module_qn}.{parent_name}")
@@ -2491,3 +2535,59 @@ def _build_assignment_arrow_function_qualified_name(
             return f"{module_qn}.{'.'.join(path_parts)}.{function_name}"
         else:
             return f"{module_qn}.{function_name}"
+
+    def _extract_implemented_interfaces(
+        self, class_node: Node, module_qn: str
+    ) -> list[str]:
+        """Extract implemented interface names from a Java class definition."""
+        implemented_interfaces: list[str] = []
+
+        # Look for interfaces field in Java class declaration
+        interfaces_node = class_node.child_by_field_name("interfaces")
+        if interfaces_node:
+            # The interfaces node contains a super_interfaces structure
+            # which has a type_list with comma-separated interface types
+            self._extract_java_interface_names(
+                interfaces_node, implemented_interfaces, module_qn
+            )
+
+        return implemented_interfaces
+
+    def _extract_java_interface_names(
+        self, interfaces_node: Node, interface_list: list[str], module_qn: str
+    ) -> None:
+        """Extract interface names from Java interfaces clause using tree-sitter."""
+        for child in interfaces_node.children:
+            if child.type == "type_list":
+                # Type list contains the actual interface types
+                for type_child in child.children:
+                    if type_child.type == "type_identifier":
+                        interface_name = type_child.text
+                        if interface_name:
+                            interface_name_str = interface_name.decode("utf8")
+                            # Resolve to fully qualified name
+                            resolved_interface = (
+                                self._resolve_class_name(interface_name_str, module_qn)
+                                or f"{module_qn}.{interface_name_str}"
+                            )
+                            interface_list.append(resolved_interface)
+            elif child.type == "type_identifier":
+                # Direct type identifier (fallback case)
+                interface_name = child.text
+                if interface_name:
+                    interface_name_str = interface_name.decode("utf8")
+                    resolved_interface = (
+                        self._resolve_class_name(interface_name_str, module_qn)
+                        or f"{module_qn}.{interface_name_str}"
+                    )
+                    interface_list.append(resolved_interface)
+
+    def _create_implements_relationship(
+        self, class_type: str, class_qn: str, interface_qn: str
+    ) -> None:
+        """Create an IMPLEMENTS relationship between a class and an interface."""
+        self.ingestor.ensure_relationship_batch(
+            (class_type, "qualified_name", class_qn),
+            "IMPLEMENTS",
+            ("Interface", "qualified_name", interface_qn),
+        )

codebase_rag/parsers/factory.py

@@ -45,6 +45,9 @@ def __init__(
         self.simple_name_lookup = simple_name_lookup
         self.ast_cache = ast_cache
 
+        # Mapping from module qualified names to file paths for efficient lookups
+        self.module_qn_to_file_path: dict[str, Path] = {}
+
         # Create processors with proper dependencies
         self._import_processor: ImportProcessor | None = None
         self._structure_processor: StructureProcessor | None = None
@@ -104,6 +107,7 @@ def definition_processor(self) -> DefinitionProcessor:
                 function_registry=self.function_registry,
                 simple_name_lookup=self.simple_name_lookup,
                 import_processor=self.import_processor,
+                module_qn_to_file_path=self.module_qn_to_file_path,
             )
         return self._definition_processor
 
@@ -118,6 +122,7 @@ def type_inference(self) -> TypeInferenceEngine:
                 project_name=self.project_name,
                 ast_cache=self.ast_cache,
                 queries=self.queries,
+                module_qn_to_file_path=self.module_qn_to_file_path,
             )
         return self._type_inference