Use SerializeAsAny for pydantic v2 models with subclasses (#2612)

tobias-bahls · koxudaxi · web-flow · commit 61e2ca7f84e6 · 2025-12-08T23:47:34.000+09:00
* Use SerializeAsAny for pydantic v2 models; ensuring that fields of child classes will be serialized as well. Fixes #1807 * Introduce parameter and support collections using SerializeAsAny * Add coverage and one extra test case --------- Co-authored-by: Koudai Aono <koxudaxi@gmail.com>
diff --git a/README.md b/README.md
@@ -393,6 +393,9 @@ Typing customization:
                         corresponding con* constrained types.
   --use-one-literal-as-default
                         Use one literal as default value for one literal field
+  --use-serialize-as-any
+                        Use pydantic.SerializeAsAny for fields with types that have subtypes
+                        (Pydantic v2 only)
   --use-specialized-enum, --no-use-specialized-enum
                         Don''t use specialized Enum class (StrEnum, IntEnum) even if the
                         target Python version supports it
diff --git a/docs/index.md b/docs/index.md
@@ -385,6 +385,9 @@ Typing customization:
                         corresponding con* constrained types.
   --use-one-literal-as-default
                         Use one literal as default value for one literal field
+  --use-serialize-as-any
+                        Use pydantic.SerializeAsAny for fields with types that have subtypes
+                        (Pydantic v2 only)
   --use-specialized-enum, --no-use-specialized-enum
                         Don''t use specialized Enum class (StrEnum, IntEnum) even if the
                         target Python version supports it
diff --git a/src/datamodel_code_generator/__init__.py b/src/datamodel_code_generator/__init__.py
@@ -418,6 +418,7 @@ def generate(  # noqa: PLR0912, PLR0913, PLR0914, PLR0915
     http_headers: Sequence[tuple[str, str]] | None = None,
     http_ignore_tls: bool = False,
     use_annotated: bool = False,
+    use_serialize_as_any: bool = False,
     use_non_positive_negative_number_constrained_types: bool = False,
     original_field_name_delimiter: str | None = None,
     use_double_quotes: bool = False,
@@ -657,6 +658,7 @@ def get_header_and_first_line(csv_file: IO[str]) -> dict[str, Any]:
         http_headers=http_headers,
         http_ignore_tls=http_ignore_tls,
         use_annotated=use_annotated,
+        use_serialize_as_any=use_serialize_as_any,
         use_non_positive_negative_number_constrained_types=use_non_positive_negative_number_constrained_types,
         original_field_name_delimiter=original_field_name_delimiter,
         use_double_quotes=use_double_quotes,
diff --git a/src/datamodel_code_generator/__main__.py b/src/datamodel_code_generator/__main__.py
@@ -424,6 +424,7 @@ def validate_all_exports_collision_strategy(cls, values: dict[str, Any]) -> dict
     http_headers: Optional[Sequence[tuple[str, str]]] = None  # noqa: UP045
     http_ignore_tls: bool = False
     use_annotated: bool = False
+    use_serialize_as_any: bool = False
     use_non_positive_negative_number_constrained_types: bool = False
     original_field_name_delimiter: Optional[str] = None  # noqa: UP045
     use_double_quotes: bool = False
@@ -853,6 +854,7 @@ def main(args: Sequence[str] | None = None) -> Exit:  # noqa: PLR0911, PLR0912,
             http_headers=config.http_headers,
             http_ignore_tls=config.http_ignore_tls,
             use_annotated=config.use_annotated,
+            use_serialize_as_any=config.use_serialize_as_any,
             use_non_positive_negative_number_constrained_types=config.use_non_positive_negative_number_constrained_types,
             original_field_name_delimiter=config.original_field_name_delimiter,
             use_double_quotes=config.use_double_quotes,
diff --git a/src/datamodel_code_generator/arguments.py b/src/datamodel_code_generator/arguments.py
@@ -354,6 +354,12 @@ def start_section(self, heading: str | None) -> None:
     action="store_true",
     default=None,
 )
+typing_options.add_argument(
+    "--use-serialize-as-any",
+    help="Use pydantic.SerializeAsAny for fields with types that have subtypes (Pydantic v2 only)",
+    action="store_true",
+    default=None,
+)
 typing_options.add_argument(
     "--use-generic-container-types",
     help="Use generic container types for type hinting (typing.Sequence, typing.Mapping). "
diff --git a/src/datamodel_code_generator/model/base.py b/src/datamodel_code_generator/model/base.py
@@ -137,6 +137,7 @@ class Config:
     parent: Optional[DataModel] = None  # noqa: UP045
     extras: dict[str, Any] = Field(default_factory=dict)
     use_annotated: bool = False
+    use_serialize_as_any: bool = False
     has_default: bool = False
     use_field_description: bool = False
     use_inline_field_description: bool = False
diff --git a/src/datamodel_code_generator/model/dataclass.py b/src/datamodel_code_generator/model/dataclass.py
@@ -183,6 +183,7 @@ def __init__(  # noqa: PLR0913, PLR0917
         use_pendulum: bool = False,  # noqa: FBT001, FBT002
         target_datetime_class: DatetimeClassType = DatetimeClassType.Datetime,
         treat_dot_as_module: bool = False,  # noqa: FBT001, FBT002
+        use_serialize_as_any: bool = False,  # noqa: FBT001, FBT002
     ) -> None:
         """Initialize type manager with datetime type mapping."""
         super().__init__(
@@ -195,6 +196,7 @@ def __init__(  # noqa: PLR0913, PLR0917
             use_pendulum,
             target_datetime_class,
             treat_dot_as_module,
+            use_serialize_as_any,
         )
 
         datetime_map = (
diff --git a/src/datamodel_code_generator/model/msgspec.py b/src/datamodel_code_generator/model/msgspec.py
@@ -407,6 +407,7 @@ def __init__(  # noqa: PLR0913, PLR0917
         use_pendulum: bool = False,  # noqa: FBT001, FBT002
         target_datetime_class: DatetimeClassType | None = None,
         treat_dot_as_module: bool = False,  # noqa: FBT001, FBT002
+        use_serialize_as_any: bool = False,  # noqa: FBT001, FBT002
     ) -> None:
         """Initialize type manager with optional datetime type mapping."""
         super().__init__(
@@ -419,6 +420,7 @@ def __init__(  # noqa: PLR0913, PLR0917
             use_pendulum,
             target_datetime_class,
             treat_dot_as_module,
+            use_serialize_as_any,
         )
 
         datetime_map = (
diff --git a/src/datamodel_code_generator/model/pydantic/types.py b/src/datamodel_code_generator/model/pydantic/types.py
@@ -175,6 +175,7 @@ def __init__(  # noqa: PLR0913, PLR0917
         use_pendulum: bool = False,  # noqa: FBT001, FBT002
         target_datetime_class: DatetimeClassType | None = None,
         treat_dot_as_module: bool = False,  # noqa: FBT001, FBT002
+        use_serialize_as_any: bool = False,  # noqa: FBT001, FBT002
     ) -> None:
         """Initialize the DataTypeManager with Pydantic v1 type mappings."""
         super().__init__(
@@ -187,6 +188,7 @@ def __init__(  # noqa: PLR0913, PLR0917
             use_pendulum,
             target_datetime_class,
             treat_dot_as_module,
+            use_serialize_as_any,
         )
 
         self.type_map: dict[Types, DataType] = self.type_map_factory(
diff --git a/src/datamodel_code_generator/model/pydantic_v2/imports.py b/src/datamodel_code_generator/model/pydantic_v2/imports.py
@@ -12,3 +12,4 @@
 IMPORT_NAIVE_DATETIME = Import.from_full_path("pydantic.NaiveDatetime")
 IMPORT_BASE64STR = Import.from_full_path("pydantic.Base64Str")
 # IMPORT_BASE64STR: Used for OpenAPI strings with format "byte" (base64 encoded characters).
+IMPORT_SERIALIZE_AS_ANY = Import.from_full_path("pydantic.SerializeAsAny")
diff --git a/src/datamodel_code_generator/model/pydantic_v2/types.py b/src/datamodel_code_generator/model/pydantic_v2/types.py
@@ -14,18 +14,96 @@
     IMPORT_AWARE_DATETIME,
     IMPORT_BASE64STR,
     IMPORT_NAIVE_DATETIME,
+    IMPORT_SERIALIZE_AS_ANY,
+)
+from datamodel_code_generator.types import (
+    DataType,
+    PythonVersion,
+    PythonVersionMin,
+    StrictTypes,
+    Types,
 )
-from datamodel_code_generator.types import DataType, StrictTypes, Types
 
 if TYPE_CHECKING:
-    from collections.abc import Sequence
+    from collections.abc import Iterator, Sequence
+
+    from datamodel_code_generator.imports import Import
+
+
+class PydanticV2DataType(DataType):
+    """Pydantic v2-specific DataType with SerializeAsAny support."""
+
+    def _should_wrap_with_serialize_as_any(self) -> bool:
+        if not self.use_serialize_as_any:
+            return False
+
+        assert self.reference is not None
+
+        from datamodel_code_generator.model.base import DataModel  # noqa: PLC0415
+
+        return any(isinstance(child, DataModel) and child.fields for child in self.reference.children)
+
+    def _get_wrapped_reference_type_hint(self, type_: str) -> str:
+        if self._should_wrap_with_serialize_as_any():
+            return f"SerializeAsAny[{type_}]"
+
+        return type_
+
+    @property
+    def imports(self) -> Iterator[Import]:
+        """Yield imports including SerializeAsAny when needed."""
+        yield from super().imports
+
+        if "SerializeAsAny" in self.type_hint:
+            yield IMPORT_SERIALIZE_AS_ANY
 
 
 class DataTypeManager(_DataTypeManager):
     """Type manager for Pydantic v2 with pattern key support."""
 
     PATTERN_KEY: ClassVar[str] = "pattern"
 
+    def __init__(  # noqa: PLR0913, PLR0917
+        self,
+        python_version: PythonVersion = PythonVersionMin,
+        use_standard_collections: bool = False,  # noqa: FBT001, FBT002
+        use_generic_container_types: bool = False,  # noqa: FBT001, FBT002
+        strict_types: Sequence[StrictTypes] | None = None,
+        use_non_positive_negative_number_constrained_types: bool = False,  # noqa: FBT001, FBT002
+        use_union_operator: bool = False,  # noqa: FBT001, FBT002
+        use_pendulum: bool = False,  # noqa: FBT001, FBT002
+        target_datetime_class: DatetimeClassType | None = None,
+        treat_dot_as_module: bool = False,  # noqa: FBT001, FBT002
+        use_serialize_as_any: bool = False,  # noqa: FBT001, FBT002
+    ) -> None:
+        """Initialize with pydantic v2-specific DataType."""
+        super().__init__(
+            python_version=python_version,
+            use_standard_collections=use_standard_collections,
+            use_generic_container_types=use_generic_container_types,
+            strict_types=strict_types,
+            use_non_positive_negative_number_constrained_types=use_non_positive_negative_number_constrained_types,
+            use_union_operator=use_union_operator,
+            use_pendulum=use_pendulum,
+            target_datetime_class=target_datetime_class,
+            treat_dot_as_module=treat_dot_as_module,
+            use_serialize_as_any=use_serialize_as_any,
+        )
+
+        # Override the data_type with our pydantic v2 version
+        from pydantic import create_model  # noqa: PLC0415
+
+        self.data_type: type[DataType] = create_model(
+            "PydanticV2ContextDataType",
+            python_version=(PythonVersion, python_version),
+            use_standard_collections=(bool, use_standard_collections),
+            use_generic_container=(bool, use_generic_container_types),
+            use_union_operator=(bool, use_union_operator),
+            treat_dot_as_module=(bool, treat_dot_as_module),
+            use_serialize_as_any=(bool, use_serialize_as_any),
+            __base__=PydanticV2DataType,
+        )
+
     def type_map_factory(
         self,
         data_type: type[DataType],
diff --git a/src/datamodel_code_generator/model/types.py b/src/datamodel_code_generator/model/types.py
@@ -77,6 +77,7 @@ def __init__(  # noqa: PLR0913, PLR0917
         use_pendulum: bool = False,  # noqa: FBT001, FBT002
         target_datetime_class: DatetimeClassType | None = None,
         treat_dot_as_module: bool = False,  # noqa: FBT001, FBT002
+        use_serialize_as_any: bool = False,  # noqa: FBT001, FBT002
     ) -> None:
         """Initialize type manager with basic type mapping."""
         super().__init__(
@@ -89,6 +90,7 @@ def __init__(  # noqa: PLR0913, PLR0917
             use_pendulum,
             target_datetime_class,
             treat_dot_as_module,
+            use_serialize_as_any,
         )
 
         self.type_map: dict[Types, DataType] = type_map_factory(self.data_type)
diff --git a/src/datamodel_code_generator/parser/base.py b/src/datamodel_code_generator/parser/base.py
@@ -530,6 +530,7 @@ def __init__(  # noqa: PLR0913, PLR0915
         http_headers: Sequence[tuple[str, str]] | None = None,
         http_ignore_tls: bool = False,
         use_annotated: bool = False,
+        use_serialize_as_any: bool = False,
         use_non_positive_negative_number_constrained_types: bool = False,
         original_field_name_delimiter: str | None = None,
         use_double_quotes: bool = False,
@@ -575,6 +576,7 @@ def __init__(  # noqa: PLR0913, PLR0915
             use_pendulum=use_pendulum,
             target_datetime_class=target_datetime_class,
             treat_dot_as_module=treat_dot_as_module,
+            use_serialize_as_any=use_serialize_as_any,
         )
         self.data_model_type: type[DataModel] = data_model_type
         self.data_model_root_type: type[DataModel] = data_model_root_type
@@ -673,6 +675,7 @@ def __init__(  # noqa: PLR0913, PLR0915
         if self.use_annotated and not self.field_constraints:  # pragma: no cover
             msg = "`use_annotated=True` has to be used with `field_constraints=True`"
             raise Exception(msg)  # noqa: TRY002
+        self.use_serialize_as_any: bool = use_serialize_as_any
         self.use_non_positive_negative_number_constrained_types = use_non_positive_negative_number_constrained_types
         self.use_double_quotes = use_double_quotes
         self.allow_responses_without_content = allow_responses_without_content
diff --git a/src/datamodel_code_generator/parser/graphql.py b/src/datamodel_code_generator/parser/graphql.py
@@ -185,6 +185,7 @@ def __init__(  # noqa: PLR0913
         dataclass_arguments: DataclassArguments | None = None,
         type_mappings: list[str] | None = None,
         read_only_write_only_model_type: ReadOnlyWriteOnlyModelType | None = None,
+        use_serialize_as_any: bool = False,
     ) -> None:
         """Initialize the GraphQL parser with configuration options."""
         super().__init__(
@@ -274,6 +275,7 @@ def __init__(  # noqa: PLR0913
             dataclass_arguments=dataclass_arguments,
             type_mappings=type_mappings,
             read_only_write_only_model_type=read_only_write_only_model_type,
+            use_serialize_as_any=use_serialize_as_any,
         )
 
         self.data_model_scalar_type = data_model_scalar_type
@@ -498,6 +500,7 @@ def parse_field(
             alias=alias,
             strip_default_none=self.strip_default_none,
             use_annotated=self.use_annotated,
+            use_serialize_as_any=self.use_serialize_as_any,
             use_field_description=self.use_field_description,
             use_inline_field_description=self.use_inline_field_description,
             use_default_kwarg=self.use_default_kwarg,
diff --git a/src/datamodel_code_generator/parser/jsonschema.py b/src/datamodel_code_generator/parser/jsonschema.py
@@ -530,6 +530,7 @@ def __init__(  # noqa: PLR0913
         http_headers: Sequence[tuple[str, str]] | None = None,
         http_ignore_tls: bool = False,
         use_annotated: bool = False,
+        use_serialize_as_any: bool = False,
         use_non_positive_negative_number_constrained_types: bool = False,
         original_field_name_delimiter: str | None = None,
         use_double_quotes: bool = False,
@@ -620,6 +621,7 @@ def __init__(  # noqa: PLR0913
             http_headers=http_headers,
             http_ignore_tls=http_ignore_tls,
             use_annotated=use_annotated,
+            use_serialize_as_any=use_serialize_as_any,
             use_non_positive_negative_number_constrained_types=use_non_positive_negative_number_constrained_types,
             original_field_name_delimiter=original_field_name_delimiter,
             use_double_quotes=use_double_quotes,
@@ -888,6 +890,7 @@ def get_object_field(  # noqa: PLR0913
             strip_default_none=self.strip_default_none,
             extras=self.get_field_extras(field),
             use_annotated=self.use_annotated,
+            use_serialize_as_any=self.use_serialize_as_any,
             use_field_description=self.use_field_description,
             use_inline_field_description=self.use_inline_field_description,
             use_default_kwarg=self.use_default_kwarg,
@@ -1899,6 +1902,7 @@ def parse_array_fields(
             strip_default_none=self.strip_default_none,
             extras=self.get_field_extras(obj),
             use_annotated=self.use_annotated,
+            use_serialize_as_any=self.use_serialize_as_any,
             use_field_description=self.use_field_description,
             use_inline_field_description=self.use_inline_field_description,
             original_name=None,
diff --git a/src/datamodel_code_generator/parser/openapi.py b/src/datamodel_code_generator/parser/openapi.py
@@ -238,6 +238,7 @@ def __init__(  # noqa: PLR0913
         http_headers: Sequence[tuple[str, str]] | None = None,
         http_ignore_tls: bool = False,
         use_annotated: bool = False,
+        use_serialize_as_any: bool = False,
         use_non_positive_negative_number_constrained_types: bool = False,
         original_field_name_delimiter: str | None = None,
         use_double_quotes: bool = False,
@@ -328,6 +329,7 @@ def __init__(  # noqa: PLR0913
             http_headers=http_headers,
             http_ignore_tls=http_ignore_tls,
             use_annotated=use_annotated,
+            use_serialize_as_any=use_serialize_as_any,
             use_non_positive_negative_number_constrained_types=use_non_positive_negative_number_constrained_types,
             original_field_name_delimiter=original_field_name_delimiter,
             use_double_quotes=use_double_quotes,
@@ -614,6 +616,7 @@ def parse_all_parameters(
                         strip_default_none=self.strip_default_none,
                         extras=self.get_field_extras(object_schema) if object_schema else {},
                         use_annotated=self.use_annotated,
+                        use_serialize_as_any=self.use_serialize_as_any,
                         use_field_description=self.use_field_description,
                         use_inline_field_description=self.use_inline_field_description,
                         use_default_kwarg=self.use_default_kwarg,
diff --git a/src/datamodel_code_generator/types.py b/src/datamodel_code_generator/types.py
diff --git a/tests/data/expected/main/openapi/serialize_as_any_pydantic_v2.py b/tests/data/expected/main/openapi/serialize_as_any_pydantic_v2.py
diff --git a/tests/data/openapi/serialize_as_any.yaml b/tests/data/openapi/serialize_as_any.yaml
diff --git a/tests/main/openapi/test_main_openapi.py b/tests/main/openapi/test_main_openapi.py
