Add --use-default-factory-for-optional-nested-models option

docs/cli-reference/index.md

@@ -11,7 +11,7 @@ This documentation is auto-generated from test cases.
 | 📁 [Base Options](base-options.md) | 5 | Input/output configuration |
 | 🔧 [Typing Customization](typing-customization.md) | 17 | Type annotation and import behavior |
 | 🏷️ [Field Customization](field-customization.md) | 20 | Field naming and docstring behavior |
-| 🏗️ [Model Customization](model-customization.md) | 26 | Model generation behavior |
+| 🏗️ [Model Customization](model-customization.md) | 27 | Model generation behavior |
 | 🎨 [Template Customization](template-customization.md) | 16 | Output formatting and custom rendering |
 | 📘 [OpenAPI-only Options](openapi-only-options.md) | 6 | OpenAPI-specific features |
 | ⚙️ [General Options](general-options.md) | 14 | Utilities and meta options |
@@ -160,6 +160,7 @@ This documentation is auto-generated from test cases.
 - [`--use-attribute-docstrings`](field-customization.md#use-attribute-docstrings)
 - [`--use-decimal-for-multiple-of`](typing-customization.md#use-decimal-for-multiple-of)
 - [`--use-default`](model-customization.md#use-default)
+- [`--use-default-factory-for-optional-nested-models`](model-customization.md#use-default-factory-for-optional-nested-models)
 - [`--use-default-kwarg`](model-customization.md#use-default-kwarg)
 - [`--use-double-quotes`](template-customization.md#use-double-quotes)
 - [`--use-enum-values-in-discriminator`](field-customization.md#use-enum-values-in-discriminator)

docs/cli-reference/model-customization.md

@@ -25,6 +25,7 @@
 | [`--target-python-version`](#target-python-version) | Target Python version for generated code syntax and imports.... |
 | [`--union-mode`](#union-mode) | Union mode for combining anyOf/oneOf schemas (smart or left_... |
 | [`--use-default`](#use-default) | Use default values from schema in generated models. |
+| [`--use-default-factory-for-optional-nested-models`](#use-default-factory-for-optional-nested-models) | Generate default_factory for optional nested model fields. |
 | [`--use-default-kwarg`](#use-default-kwarg) | Use default= keyword argument instead of positional argument... |
 | [`--use-frozen-field`](#use-frozen-field) | Generate frozen (immutable) field definitions for readOnly p... |
 | [`--use-one-literal-as-default`](#use-one-literal-as-default) | Use single literal value as default when enum has only one o... |
@@ -4834,6 +4835,148 @@ The `--use-default` flag configures the code generation behavior.
 
 ---
 
+## `--use-default-factory-for-optional-nested-models` {#use-default-factory-for-optional-nested-models}
+
+Generate default_factory for optional nested model fields.
+
+The `--use-default-factory-for-optional-nested-models` flag generates default_factory
+for optional nested model fields instead of None default:
+- Dataclasses: `field: Model | None = field(default_factory=Model)`
+- Pydantic: `field: Model | None = Field(default_factory=Model)`
+- msgspec: `field: Model | UnsetType = field(default_factory=Model)`
+
+!!! tip "Usage"
+
+    ```bash
+    datamodel-codegen --input schema.json --use-default-factory-for-optional-nested-models # (1)!
+    ```
+
+    1. :material-arrow-left: `--use-default-factory-for-optional-nested-models` - the option documented here
+
+??? example "Examples"
+
+    **Input Schema:**
+
+    ```json
+    {
+      "$schema": "http://json-schema.org/draft-07/schema#",
+      "type": "object",
+      "properties": {
+        "name": {"type": "string"},
+        "address": {"$ref": "#/$defs/Address"},
+        "contact": {"$ref": "#/$defs/Contact"}
+      },
+      "required": ["name"],
+      "$defs": {
+        "Address": {
+          "type": "object",
+          "properties": {
+            "street": {"type": "string"},
+            "city": {"type": "string"}
+          }
+        },
+        "Contact": {
+          "type": "object",
+          "properties": {
+            "email": {"type": "string"},
+            "phone": {"type": "string"}
+          }
+        }
+      }
+    }
+    ```
+
+    **Output:**
+
+    === "Pydantic v2"
+
+        ```python
+        # generated by datamodel-codegen:
+        #   filename:  default_factory_nested_model.json
+        #   timestamp: 2019-07-26T00:00:00+00:00
+
+        from __future__ import annotations
+
+        from pydantic import BaseModel, Field
+
+
+        class Address(BaseModel):
+            street: str | None = None
+            city: str | None = None
+
+
+        class Contact(BaseModel):
+            email: str | None = None
+            phone: str | None = None
+
+
+        class Model(BaseModel):
+            name: str
+            address: Address | None = Field(default_factory=Address)
+            contact: Contact | None = Field(default_factory=Contact)
+        ```
+
+    === "dataclass"
+
+        ```python
+        # generated by datamodel-codegen:
+        #   filename:  default_factory_nested_model.json
+        #   timestamp: 2019-07-26T00:00:00+00:00
+
+        from __future__ import annotations
+
+        from dataclasses import dataclass, field
+
+
+        @dataclass
+        class Address:
+            street: str | None = None
+            city: str | None = None
+
+
+        @dataclass
+        class Contact:
+            email: str | None = None
+            phone: str | None = None
+
+
+        @dataclass
+        class Model:
+            name: str
+            address: Address | None = field(default_factory=Address)
+            contact: Contact | None = field(default_factory=Contact)
+        ```
+
+    === "msgspec"
+
+        ```python
+        # generated by datamodel-codegen:
+        #   filename:  default_factory_nested_model.json
+        #   timestamp: 2019-07-26T00:00:00+00:00
+
+        from __future__ import annotations
+
+        from msgspec import UNSET, Struct, UnsetType, field
+
+
+        class Address(Struct):
+            street: str | UnsetType = UNSET
+            city: str | UnsetType = UNSET
+
+
+        class Contact(Struct):
+            email: str | UnsetType = UNSET
+            phone: str | UnsetType = UNSET
+
+
+        class Model(Struct):
+            name: str
+            address: Address | UnsetType = field(default_factory=Address)
+            contact: Contact | UnsetType = field(default_factory=Contact)
+        ```
+
+---
+
 ## `--use-default-kwarg` {#use-default-kwarg}
 
 Use default= keyword argument instead of positional argument for fields with defaults.

docs/cli-reference/quick-reference.md

@@ -94,6 +94,7 @@ datamodel-codegen [OPTIONS]
 | [`--target-python-version`](model-customization.md#target-python-version) | Target Python version for generated code syntax and imports. |
 | [`--union-mode`](model-customization.md#union-mode) | Union mode for combining anyOf/oneOf schemas (smart or left_to_right). |
 | [`--use-default`](model-customization.md#use-default) | Use default values from schema in generated models. |
+| [`--use-default-factory-for-optional-nested-models`](model-customization.md#use-default-factory-for-optional-nested-models) | Generate default_factory for optional nested model fields. |
 | [`--use-default-kwarg`](model-customization.md#use-default-kwarg) | Use default= keyword argument instead of positional argument for fields with def... |
 | [`--use-frozen-field`](model-customization.md#use-frozen-field) | Generate frozen (immutable) field definitions for readOnly properties. |
 | [`--use-one-literal-as-default`](model-customization.md#use-one-literal-as-default) | Use single literal value as default when enum has only one option. |
@@ -252,6 +253,7 @@ All options sorted alphabetically:
 - [`--use-attribute-docstrings`](field-customization.md#use-attribute-docstrings) - Generate field descriptions as attribute docstrings instead ...
 - [`--use-decimal-for-multiple-of`](typing-customization.md#use-decimal-for-multiple-of) - Generate Decimal types for fields with multipleOf constraint...
 - [`--use-default`](model-customization.md#use-default) - Use default values from schema in generated models.
+- [`--use-default-factory-for-optional-nested-models`](model-customization.md#use-default-factory-for-optional-nested-models) - Generate default_factory for optional nested model fields.
 - [`--use-default-kwarg`](model-customization.md#use-default-kwarg) - Use default= keyword argument instead of positional argument...
 - [`--use-double-quotes`](template-customization.md#use-double-quotes) - Use double quotes for string literals in generated code.
 - [`--use-enum-values-in-discriminator`](field-customization.md#use-enum-values-in-discriminator) - Use enum values in discriminator mappings for union types.

src/datamodel_code_generator/__init__.py

@@ -470,6 +470,7 @@ def generate(  # noqa: PLR0912, PLR0913, PLR0914, PLR0915
     frozen_dataclasses: bool = False,
     no_alias: bool = False,
     use_frozen_field: bool = False,
+    use_default_factory_for_optional_nested_models: bool = False,
     formatters: list[Formatter] = DEFAULT_FORMATTERS,
     settings_path: Path | None = None,
     parent_scoped_naming: bool = False,
@@ -717,6 +718,7 @@ def get_header_and_first_line(csv_file: IO[str]) -> dict[str, Any]:
         frozen_dataclasses=frozen_dataclasses,
         no_alias=no_alias,
         use_frozen_field=use_frozen_field,
+        use_default_factory_for_optional_nested_models=use_default_factory_for_optional_nested_models,
         formatters=formatters,
         encoding=encoding,
         parent_scoped_naming=parent_scoped_naming,

src/datamodel_code_generator/__main__.py

@@ -459,6 +459,7 @@ def validate_all_exports_collision_strategy(cls, values: dict[str, Any]) -> dict
     dataclass_arguments: Optional[DataclassArguments] = None  # noqa: UP045
     no_alias: bool = False
     use_frozen_field: bool = False
+    use_default_factory_for_optional_nested_models: bool = False
     formatters: list[Formatter] = DEFAULT_FORMATTERS
     parent_scoped_naming: bool = False
     disable_future_imports: bool = False
@@ -761,6 +762,7 @@ def run_generate_from_config(  # noqa: PLR0913, PLR0917
         frozen_dataclasses=config.frozen_dataclasses,
         no_alias=config.no_alias,
         use_frozen_field=config.use_frozen_field,
+        use_default_factory_for_optional_nested_models=config.use_default_factory_for_optional_nested_models,
         formatters=config.formatters,
         settings_path=settings_path,
         parent_scoped_naming=config.parent_scoped_naming,

src/datamodel_code_generator/arguments.py

@@ -594,6 +594,13 @@ def start_section(self, heading: str | None) -> None:
     action="store_true",
     default=None,
 )
+field_options.add_argument(
+    "--use-default-factory-for-optional-nested-models",
+    help="Use default_factory for optional nested model fields instead of None default. "
+    "E.g., `field: Model | None = Field(default_factory=Model)` instead of `field: Model | None = None`",
+    action="store_true",
+    default=None,
+)
 
 # ======================================================================================
 # Options for templating output

src/datamodel_code_generator/cli_options.py

@@ -92,6 +92,9 @@ class CLIOptionMeta:
     "--strip-default-none": CLIOptionMeta(name="--strip-default-none", category=OptionCategory.MODEL),
     "--dataclass-arguments": CLIOptionMeta(name="--dataclass-arguments", category=OptionCategory.MODEL),
     "--use-frozen-field": CLIOptionMeta(name="--use-frozen-field", category=OptionCategory.MODEL),
+    "--use-default-factory-for-optional-nested-models": CLIOptionMeta(
+        name="--use-default-factory-for-optional-nested-models", category=OptionCategory.MODEL
+    ),
     "--union-mode": CLIOptionMeta(name="--union-mode", category=OptionCategory.MODEL),
     "--parent-scoped-naming": CLIOptionMeta(name="--parent-scoped-naming", category=OptionCategory.MODEL),
     "--use-one-literal-as-default": CLIOptionMeta(name="--use-one-literal-as-default", category=OptionCategory.MODEL),

src/datamodel_code_generator/model/base.py

@@ -166,6 +166,7 @@ class Config:
     read_only: bool = False
     write_only: bool = False
     use_frozen_field: bool = False
+    use_default_factory_for_optional_nested_models: bool = False
 
     if not TYPE_CHECKING:
         if not PYDANTIC_V2:

src/datamodel_code_generator/model/dataclass.py

@@ -143,6 +143,19 @@ def field(self) -> str | None:
             return None
         return result
 
+    def _get_default_factory_for_nested_model(self) -> str | None:
+        """Get default_factory for nested dataclass model fields.
+
+        Returns the class name if the field type references a DataClass,
+        otherwise returns None.
+        """
+        for data_type in self.data_type.data_types or (self.data_type,):
+            if data_type.is_dict:
+                continue
+            if data_type.reference and isinstance(data_type.reference.source, DataClass):
+                return data_type.alias or data_type.reference.source.class_name
+        return None
+
     def __str__(self) -> str:
         """Generate field() call or default value representation."""
         data: dict[str, Any] = {k: v for k, v in self.extras.items() if k in self._FIELD_KEYS}
@@ -161,6 +174,16 @@ def __str__(self) -> str:
                 }
             }
 
+        if (
+            self.use_default_factory_for_optional_nested_models
+            and not self.required
+            and (self.default is None or self.default is UNDEFINED)
+            and "default_factory" not in data
+        ):
+            nested_model_name = self._get_default_factory_for_nested_model()
+            if nested_model_name:
+                data["default_factory"] = nested_model_name
+
         if not data:
             return ""
 

src/datamodel_code_generator/model/msgspec.py

@@ -247,7 +247,7 @@ def field(self) -> str | None:
             return None
         return result
 
-    def __str__(self) -> str:
+    def __str__(self) -> str:  # noqa: PLR0912
         """Generate field() call or default value representation."""
         data: dict[str, Any] = {k: v for k, v in self.extras.items() if k in self._FIELD_KEYS}
         if self.alias:
@@ -284,6 +284,17 @@ def __str__(self) -> str:
             else:
                 data["default_factory"] = type(default_value).__name__
 
+        if (
+            self.use_default_factory_for_optional_nested_models
+            and not self.required
+            and (self.default is None or self.default is UNDEFINED)
+            and "default_factory" not in data
+        ):
+            nested_model_name = self._get_default_factory_for_optional_nested_model()
+            if nested_model_name:
+                data["default_factory"] = nested_model_name
+                data.pop("default", None)
+
         if not data:
             return ""
 
@@ -412,6 +423,19 @@ def _get_default_as_struct_model(self) -> str | None:
                 )
         return None
 
+    def _get_default_factory_for_optional_nested_model(self) -> str | None:
+        """Get default_factory for optional nested Struct model fields.
+
+        Returns the class name if the field type references a Struct,
+        otherwise returns None.
+        """
+        for data_type in self.data_type.data_types or (self.data_type,):
+            if data_type.is_dict:
+                continue
+            if data_type.reference and isinstance(data_type.reference.source, Struct):
+                return data_type.alias or data_type.reference.source.class_name
+        return None
+
 
 class DataTypeManager(_DataTypeManager):
     """Type manager for msgspec Struct models."""

src/datamodel_code_generator/model/pydantic/base_model.py

@@ -152,6 +152,19 @@ def _get_default_as_pydantic_model(self) -> str | None:
                 )
         return None
 
+    def _get_default_factory_for_optional_nested_model(self) -> str | None:
+        """Get default_factory for optional nested Pydantic model fields.
+
+        Returns the class name if the field type references a BaseModel,
+        otherwise returns None.
+        """
+        for data_type in self.data_type.data_types or (self.data_type,):
+            if data_type.is_dict:
+                continue
+            if data_type.reference and isinstance(data_type.reference.source, BaseModelBase):
+                return data_type.alias or data_type.reference.source.class_name
+        return None
+
     def _process_data_in_str(self, data: dict[str, Any]) -> None:
         if self.const:
             data["const"] = True
@@ -204,6 +217,14 @@ def __str__(self) -> str:  # noqa: PLR0912
         else:
             default_factory = data.pop("default_factory", None)
 
+        if (
+            default_factory is None
+            and self.use_default_factory_for_optional_nested_models
+            and not self.required
+            and (self.default is None or self.default is UNDEFINED)
+        ):
+            default_factory = self._get_default_factory_for_optional_nested_model()
+
         self.__dict__["_computed_default_factory"] = default_factory
 
         field_arguments = sorted(f"{k}={v!r}" for k, v in data.items() if v is not None)

src/datamodel_code_generator/parser/base.py

@@ -734,6 +734,7 @@ def __init__(  # noqa: PLR0913, PLR0915
         frozen_dataclasses: bool = False,
         no_alias: bool = False,
         use_frozen_field: bool = False,
+        use_default_factory_for_optional_nested_models: bool = False,
         formatters: list[Formatter] = DEFAULT_FORMATTERS,
         parent_scoped_naming: bool = False,
         dataclass_arguments: DataclassArguments | None = None,
@@ -875,6 +876,7 @@ def __init__(  # noqa: PLR0913, PLR0915
         self.type_mappings: dict[tuple[str, str], str] = Parser._parse_type_mappings(type_mappings)
         self.read_only_write_only_model_type: ReadOnlyWriteOnlyModelType | None = read_only_write_only_model_type
         self.use_frozen_field: bool = use_frozen_field
+        self.use_default_factory_for_optional_nested_models: bool = use_default_factory_for_optional_nested_models
 
     @property
     def field_name_model_type(self) -> ModelType: