Fix bug in handling of graphql empty list defaults (#2948)

* adding tests to highlight the graphql bug

* fixing test to actually compare files (previously was not)

* fix previous test that did not specify assert_func

* Fix empty list default for GraphQL list fields

* Fix GraphQL empty list default handling

* Skip empty list default tests on black 22

---------

Co-authored-by: Koudai Aono <koxudaxi@gmail.com>

Author: rpmcginty

src/datamodel_code_generator/model/pydantic/base_model.py

@@ -126,9 +126,22 @@ def _get_strict_field_constraint_value(self, constraint: str, value: Any) -> Any
             return value
         return int(value)
 
-    def _get_default_as_pydantic_model(self) -> str | None:
+    def _get_default_as_pydantic_model(self) -> str | None:  # noqa: PLR0911, PLR0912
         if isinstance(self.default, WrappedDefault):
             return f"lambda :{self.default!r}"
+        if self.data_type.is_list and len(self.data_type.data_types) == 1:
+            data_type_child = self.data_type.data_types[0]
+            if (
+                data_type_child.reference
+                and isinstance(data_type_child.reference.source, BaseModelBase)
+                and isinstance(self.default, list)
+            ):
+                if not self.default:
+                    return STANDARD_LIST
+                return (  # pragma: no cover
+                    f"lambda :[{data_type_child.alias or data_type_child.reference.source.class_name}."
+                    f"{self._PARSE_METHOD}(v) for v in {self.default!r}]"
+                )
         for data_type in self.data_type.data_types or (self.data_type,):
             # TODO: Check nested data_types
             if data_type.is_dict:
@@ -220,7 +233,7 @@ def __str__(self) -> str:  # noqa: PLR0912
             elif isinstance(discriminator, dict):  # pragma: no cover
                 data["discriminator"] = discriminator["propertyName"]
 
-        if self.required:
+        if self.required and not self.has_default:
             default_factory = None
         elif self.default is not UNDEFINED and self.default is not None and "default_factory" not in data:
             default_factory = self._get_default_as_pydantic_model()
@@ -249,7 +262,7 @@ def __str__(self) -> str:  # noqa: PLR0912
 
         if self.use_annotated:
             field_arguments = self._process_annotated_field_arguments(field_arguments)
-        elif self.required:
+        elif self.required and not default_factory:
             field_arguments = ["...", *field_arguments]
         elif not default_factory:
             default_repr = repr_set_sorted(self.default) if isinstance(self.default, set) else repr(self.default)

tests/data/expected/main/graphql/empty_list_default.py

@@ -0,0 +1,34 @@
+# generated by datamodel-codegen:
+#   filename:  empty_list_default.graphql
+#   timestamp: 2019-07-26T00:00:00+00:00
+
+from __future__ import annotations
+
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+type Boolean = bool
+"""
+The `Boolean` scalar type represents `true` or `false`.
+"""
+
+
+type String = str
+"""
+The `String` scalar type represents textual data, represented as UTF-8 character sequences. The String type is most often used by GraphQL to represent free-form human-readable text.
+"""
+
+
+class Container(BaseModel):
+    name: String
+    typename__: Literal['Container'] | None = Field('Container', alias='__typename')
+
+
+class PodSpec(BaseModel):
+    container_list: list[Container] = Field(default_factory=list)
+    container_list_or_none: list[Container | None] = Field(default_factory=list)
+    container_or_none_list_or_none: list[Container | None] | None = Field(
+        default_factory=list
+    )
+    typename__: Literal['PodSpec'] | None = Field('PodSpec', alias='__typename')

tests/data/expected/main/graphql/pydantic_v2_empty_list_default.py

@@ -0,0 +1,34 @@
+# generated by datamodel-codegen:
+#   filename:  empty_list_default.graphql
+#   timestamp: 2019-07-26T00:00:00+00:00
+
+from __future__ import annotations
+
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+type Boolean = bool
+"""
+The `Boolean` scalar type represents `true` or `false`.
+"""
+
+
+type String = str
+"""
+The `String` scalar type represents textual data, represented as UTF-8 character sequences. The String type is most often used by GraphQL to represent free-form human-readable text.
+"""
+
+
+class Container(BaseModel):
+    name: String
+    typename__: Literal['Container'] | None = Field('Container', alias='__typename')
+
+
+class PodSpec(BaseModel):
+    container_list: list[Container] = Field(default_factory=list)
+    container_list_or_none: list[Container | None] = Field(default_factory=list)
+    container_or_none_list_or_none: list[Container | None] | None = Field(
+        default_factory=list
+    )
+    typename__: Literal['PodSpec'] | None = Field('PodSpec', alias='__typename')

tests/data/expected/main/openapi/empty_list_default.py

@@ -4,14 +4,12 @@
 
 from __future__ import annotations
 
-from typing import List, Optional
-
 from pydantic import BaseModel, Field
 
 
 class Container(BaseModel):
-    name: Optional[str] = None
+    name: str | None = None
 
 
 class PodSpec(BaseModel):
-    containers: Optional[List[Container]] = Field(default_factory=list)
+    containers: list[Container] | None = Field(default_factory=list)

tests/data/expected/main/openapi/pydantic_v2_empty_list_default.py

@@ -4,14 +4,12 @@
 
 from __future__ import annotations
 
-from typing import List, Optional
-
 from pydantic import BaseModel, Field
 
 
 class Container(BaseModel):
-    name: Optional[str] = None
+    name: str | None = None
 
 
 class PodSpec(BaseModel):
-    containers: Optional[List[Container]] = Field(default_factory=list)
+    containers: list[Container] | None = Field(default_factory=list)

tests/data/graphql/empty_list_default.graphql

@@ -0,0 +1,10 @@
+
+input Container {
+  name: String!
+}
+
+input PodSpec {
+  container_list: [Container!]! = []
+  container_list_or_none: [Container]! = []
+  container_or_none_list_or_none: [Container] = []
+}

tests/main/graphql/test_main_graphql.py

@@ -7,7 +7,13 @@
 import black
 import pytest
 
-from tests.main.conftest import DEFAULT_VALUES_DATA_PATH, GRAPHQL_DATA_PATH, LEGACY_BLACK_SKIP, run_main_and_assert
+from tests.main.conftest import (
+    DEFAULT_VALUES_DATA_PATH,
+    EXPECTED_GRAPHQL_PATH,
+    GRAPHQL_DATA_PATH,
+    LEGACY_BLACK_SKIP,
+    run_main_and_assert,
+)
 from tests.main.graphql.conftest import assert_file_content
 
 if TYPE_CHECKING:
@@ -104,6 +110,40 @@ def test_main_use_default_kwarg(output_file: Path) -> None:
     )
 
 
+@pytest.mark.parametrize(
+    ("output_model", "expected_output"),
+    [
+        (
+            "pydantic.BaseModel",
+            "empty_list_default.py",
+        ),
+        (
+            "pydantic_v2.BaseModel",
+            "pydantic_v2_empty_list_default.py",
+        ),
+    ],
+)
+@pytest.mark.skipif(
+    black.__version__.split(".")[0] in {"19", "22"},
+    reason="Installed black doesn't support Python 3.12 target version",
+)
+def test_main_graphql_empty_list_default(output_model: str, expected_output: str, output_file: Path) -> None:
+    """Test GraphQL generation with empty list default values."""
+    run_main_and_assert(
+        input_path=GRAPHQL_DATA_PATH / "empty_list_default.graphql",
+        output_path=output_file,
+        assert_func=assert_file_content,
+        expected_file=EXPECTED_GRAPHQL_PATH / expected_output,
+        input_file_type="graphql",
+        extra_args=[
+            "--output-model-type",
+            output_model,
+            "--target-python-version",
+            "3.12",
+        ],
+    )
+
+
 @pytest.mark.skipif(
     black.__version__.split(".")[0] == "19",
     reason="Installed black doesn't support the old style",

tests/main/openapi/test_main_openapi.py

@@ -2808,6 +2808,7 @@ def test_main_openapi_empty_list_default(output_model: str, expected_output: str
         input_path=OPEN_API_DATA_PATH / "empty_list_default.yaml",
         output_path=output_file,
         expected_file=EXPECTED_OPENAPI_PATH / expected_output,
+        assert_func=assert_file_content,
         input_file_type="openapi",
         extra_args=[
             "--output-model-type",