Fix unit tests

Author: NathalieCharbel

src/neo4j_graphrag/experimental/components/schema.py

@@ -119,10 +119,16 @@ class NodeType(BaseModel):
     @classmethod
     def validate_input_if_string(cls, data: EntityInputType) -> EntityInputType:
         if isinstance(data, str):
+            logger.info(
+                f"Converting string '{data}' to NodeType with default 'name' property "
+                f"and additional_properties=True to allow flexible property extraction."
+            )
             return {
                 "label": data,
                 # added to satisfy the model validation (min_length=1 for properties of node types)
                 "properties": [{"name": "name", "type": "STRING"}],
+                # allow LLM to extract additional properties beyond the default "name"
+                "additional_properties": True,  # type: ignore[dict-item]
             }
         return data
 
@@ -741,7 +747,18 @@ def _filter_items_without_labels(
         for item in items:
             if isinstance(item, str):
                 if item and " " not in item and not item.startswith("{"):
-                    filtered_items.append({"label": item})
+                    # Add default property for node types to satisfy min_length=1 constraint
+                    # This matches the behavior of NodeType.validate_input_if_string
+                    if item_type == "node type":
+                        filtered_items.append(
+                            {
+                                "label": item,
+                                "properties": [{"name": "name", "type": "STRING"}],
+                                "additional_properties": True,
+                            }
+                        )
+                    else:
+                        filtered_items.append({"label": item})
                 elif item:
                     logging.info(
                         f"Filtering out {item_type} with invalid label: {item}"

tests/unit/experimental/components/test_schema.py

@@ -125,15 +125,15 @@ def test_relationship_type_additional_properties_default() -> None:
     assert relationship_type.additional_properties is False
 
     # manually changing the default value
-    # impossible cases: no properties and no additional
-    with pytest.raises(ValidationError):
-        RelationshipType.model_validate(
-            {"label": "REL", "additional_properties": False}
-        )
-    with pytest.raises(ValidationError):
-        RelationshipType.model_validate(
-            {"label": "REL", "properties": [], "additional_properties": False}
-        )
+    # auto-correction: no properties and additional_properties=False -> auto-corrected to True
+    relationship_type = RelationshipType.model_validate(
+        {"label": "REL", "additional_properties": False}
+    )
+    assert relationship_type.additional_properties is True  # auto-corrected
+    relationship_type = RelationshipType.model_validate(
+        {"label": "REL", "properties": [], "additional_properties": False}
+    )
+    assert relationship_type.additional_properties is True  # auto-corrected
 
     # working case: properties and additional allowed
     relationship_type = RelationshipType.model_validate(
@@ -365,8 +365,15 @@ def valid_node_types() -> tuple[NodeType, ...]:
         NodeType(
             label="ORGANIZATION",
             description="A structured group of people with a common purpose.",
+            properties=[PropertyType(name="name", type="STRING")],
+            additional_properties=True,
+        ),
+        NodeType(
+            label="AGE",
+            description="Age of a person in years.",
+            properties=[PropertyType(name="value", type="INTEGER")],
+            additional_properties=True,
         ),
-        NodeType(label="AGE", description="Age of a person in years."),
     )
 
 

tests/unit/experimental/pipeline/config/template_pipeline/test_simple_kg_builder.py

@@ -427,7 +427,10 @@ def test_simple_kg_pipeline_config_process_schema_with_precedence_schema_dict()
     patterns = schema_dict["patterns"]
     assert len(node_types) == 2
     assert node_types[0]["label"] == "Person"
-    assert len(node_types[0]["properties"]) == 0
+    # String input gets default "name" property and additional_properties=True
+    assert len(node_types[0]["properties"]) == 1
+    assert node_types[0]["properties"][0]["name"] == "name"
+    assert node_types[0]["additional_properties"] is True
     assert node_types[1]["label"] == "Organization"
     assert len(node_types[1]["properties"]) == 1
     assert relationship_types is not None
@@ -492,7 +495,10 @@ def test_simple_kg_pipeline_config_process_schema_with_precedence_schema_object(
     patterns = schema_dict["patterns"]
     assert len(node_types) == 2
     assert node_types[0]["label"] == "Person"
-    assert len(node_types[0]["properties"]) == 0
+    # String input gets default "name" property and additional_properties=True
+    assert len(node_types[0]["properties"]) == 1
+    assert node_types[0]["properties"][0]["name"] == "name"
+    assert node_types[0]["additional_properties"] is True
     assert node_types[1]["label"] == "Organization"
     assert len(node_types[1]["properties"]) == 1
     assert relationship_types is not None