Skip to content

Functional Serializers

This module contains related classes and functions for serialization.

PlainSerializer dataclass

Plain serializers use a function to modify the output of serialization.

This is particularly helpful when you want to customize the serialization for annotated types. Consider an input of list, which will be serialized into a space-delimited string.

from typing import List

from typing_extensions import Annotated

from pydantic import BaseModel, PlainSerializer

CustomStr = Annotated[
    List, PlainSerializer(lambda x: ' '.join(x), return_type=str)
]

class StudentModel(BaseModel):
    courses: CustomStr

student = StudentModel(courses=['Math', 'Chemistry', 'English'])
print(student.model_dump())
#> {'courses': 'Math Chemistry English'}

Attributes:

Name Type Description
func SerializerFunction

The serializer function.

return_type Any

The return type for the function. If omitted it will be inferred from the type annotation.

when_used Literal['always', 'unless-none', 'json', 'json-unless-none']

Determines when this serializer should be used. Accepts a string with values 'always', 'unless-none', 'json', and 'json-unless-none'. Defaults to 'always'.

WrapSerializer dataclass

Wrap serializers receive the raw inputs along with a handler function that applies the standard serialization logic, and can modify the resulting value before returning it as the final output of serialization.

For example, here's a scenario in which a wrap serializer transforms timezones to UTC and utilizes the existing datetime serialization logic.

from datetime import datetime, timezone
from typing import Any, Dict

from typing_extensions import Annotated

from pydantic import BaseModel, WrapSerializer

class EventDatetime(BaseModel):
    start: datetime
    end: datetime

def convert_to_utc(value: Any, handler, info) -> Dict[str, datetime]:
    # Note that `helper` can actually help serialize the `value` for further custom serialization in case it's a subclass.
    partial_result = handler(value, info)
    if info.mode == 'json':
        return {
            k: datetime.fromisoformat(v).astimezone(timezone.utc)
            for k, v in partial_result.items()
        }
    return {k: v.astimezone(timezone.utc) for k, v in partial_result.items()}

UTCEventDatetime = Annotated[EventDatetime, WrapSerializer(convert_to_utc)]

class EventModel(BaseModel):
    event_datetime: UTCEventDatetime

dt = EventDatetime(
    start='2024-01-01T07:00:00-08:00', end='2024-01-03T20:00:00+06:00'
)
event = EventModel(event_datetime=dt)
print(event.model_dump())
'''
{
    'event_datetime': {
        'start': datetime.datetime(
            2024, 1, 1, 15, 0, tzinfo=datetime.timezone.utc
        ),
        'end': datetime.datetime(
            2024, 1, 3, 14, 0, tzinfo=datetime.timezone.utc
        ),
    }
}
'''

print(event.model_dump_json())
'''
{"event_datetime":{"start":"2024-01-01T15:00:00Z","end":"2024-01-03T14:00:00Z"}}
'''

Attributes:

Name Type Description
func WrapSerializerFunction

The serializer function to be wrapped.

return_type Any

The return type for the function. If omitted it will be inferred from the type annotation.

when_used Literal['always', 'unless-none', 'json', 'json-unless-none']

Determines when this serializer should be used. Accepts a string with values 'always', 'unless-none', 'json', and 'json-unless-none'. Defaults to 'always'.

field_serializer

field_serializer(
    *fields: str,
    mode: Literal["plain", "wrap"] = "plain",
    return_type: Any = PydanticUndefined,
    when_used: Literal[
        "always", "unless-none", "json", "json-unless-none"
    ] = "always",
    check_fields: bool | None = None
) -> Callable[[Any], Any]

Decorator that enables custom field serialization.

In the below example, a field of type set is used to mitigate duplication. A field_serializer is used to serialize the data as a sorted list.

from typing import Set

from pydantic import BaseModel, field_serializer

class StudentModel(BaseModel):
    name: str = 'Jane'
    courses: Set[str]

    @field_serializer('courses', when_used='json')
    def serialize_courses_in_order(courses: Set[str]):
        return sorted(courses)

student = StudentModel(courses={'Math', 'Chemistry', 'English'})
print(student.model_dump_json())
#> {"name":"Jane","courses":["Chemistry","English","Math"]}

See Custom serializers for more information.

Four signatures are supported:

  • (self, value: Any, info: FieldSerializationInfo)
  • (self, value: Any, nxt: SerializerFunctionWrapHandler, info: FieldSerializationInfo)
  • (value: Any, info: SerializationInfo)
  • (value: Any, nxt: SerializerFunctionWrapHandler, info: SerializationInfo)

Parameters:

Name Type Description Default
fields str

Which field(s) the method should be called on.

()
mode Literal['plain', 'wrap']

The serialization mode.

  • plain means the function will be called instead of the default serialization logic,
  • wrap means the function will be called with an argument to optionally call the default serialization logic.
'plain'
return_type Any

Optional return type for the function, if omitted it will be inferred from the type annotation.

PydanticUndefined
when_used Literal['always', 'unless-none', 'json', 'json-unless-none']

Determines the serializer will be used for serialization.

'always'
check_fields bool | None

Whether to check that the fields actually exist on the model.

None

Returns:

Type Description
Callable[[Any], Any]

The decorator function.

Source code in pydantic/functional_serializers.py
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
def field_serializer(
    *fields: str,
    mode: Literal['plain', 'wrap'] = 'plain',
    return_type: Any = PydanticUndefined,
    when_used: Literal['always', 'unless-none', 'json', 'json-unless-none'] = 'always',
    check_fields: bool | None = None,
) -> Callable[[Any], Any]:
    """Decorator that enables custom field serialization.

    In the below example, a field of type `set` is used to mitigate duplication. A `field_serializer` is used to serialize the data as a sorted list.

    ```python
    from typing import Set

    from pydantic import BaseModel, field_serializer

    class StudentModel(BaseModel):
        name: str = 'Jane'
        courses: Set[str]

        @field_serializer('courses', when_used='json')
        def serialize_courses_in_order(courses: Set[str]):
            return sorted(courses)

    student = StudentModel(courses={'Math', 'Chemistry', 'English'})
    print(student.model_dump_json())
    #> {"name":"Jane","courses":["Chemistry","English","Math"]}
    ```

    See [Custom serializers](../concepts/serialization.md#custom-serializers) for more information.

    Four signatures are supported:

    - `(self, value: Any, info: FieldSerializationInfo)`
    - `(self, value: Any, nxt: SerializerFunctionWrapHandler, info: FieldSerializationInfo)`
    - `(value: Any, info: SerializationInfo)`
    - `(value: Any, nxt: SerializerFunctionWrapHandler, info: SerializationInfo)`

    Args:
        fields: Which field(s) the method should be called on.
        mode: The serialization mode.

            - `plain` means the function will be called instead of the default serialization logic,
            - `wrap` means the function will be called with an argument to optionally call the
               default serialization logic.
        return_type: Optional return type for the function, if omitted it will be inferred from the type annotation.
        when_used: Determines the serializer will be used for serialization.
        check_fields: Whether to check that the fields actually exist on the model.

    Returns:
        The decorator function.
    """

    def dec(
        f: Callable[..., Any] | staticmethod[Any, Any] | classmethod[Any, Any, Any],
    ) -> _decorators.PydanticDescriptorProxy[Any]:
        dec_info = _decorators.FieldSerializerDecoratorInfo(
            fields=fields,
            mode=mode,
            return_type=return_type,
            when_used=when_used,
            check_fields=check_fields,
        )
        return _decorators.PydanticDescriptorProxy(f, dec_info)

    return dec

model_serializer

model_serializer(
    __f: Callable[..., Any] | None = None,
    *,
    mode: Literal["plain", "wrap"] = "plain",
    when_used: Literal[
        "always", "unless-none", "json", "json-unless-none"
    ] = "always",
    return_type: Any = PydanticUndefined
) -> Callable[[Any], Any]

Decorator that enables custom model serialization.

This is useful when a model need to be serialized in a customized manner, allowing for flexibility beyond just specific fields.

An example would be to serialize temperature to the same temperature scale, such as degrees Celsius.

from typing import Literal

from pydantic import BaseModel, model_serializer

class TemperatureModel(BaseModel):
    unit: Literal['C', 'F']
    value: int

    @model_serializer()
    def serialize_model(self):
        if self.unit == 'F':
            return {'unit': 'C', 'value': int((self.value - 32) / 1.8)}
        return {'unit': self.unit, 'value': self.value}

temperature = TemperatureModel(unit='F', value=212)
print(temperature.model_dump())
#> {'unit': 'C', 'value': 100}

See Custom serializers for more information.

Parameters:

Name Type Description Default
__f Callable[..., Any] | None

The function to be decorated.

None
mode Literal['plain', 'wrap']

The serialization mode.

  • 'plain' means the function will be called instead of the default serialization logic
  • 'wrap' means the function will be called with an argument to optionally call the default serialization logic.
'plain'
when_used Literal['always', 'unless-none', 'json', 'json-unless-none']

Determines when this serializer should be used.

'always'
return_type Any

The return type for the function. If omitted it will be inferred from the type annotation.

PydanticUndefined

Returns:

Type Description
Callable[[Any], Any]

The decorator function.

Source code in pydantic/functional_serializers.py
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
def model_serializer(
    __f: Callable[..., Any] | None = None,
    *,
    mode: Literal['plain', 'wrap'] = 'plain',
    when_used: Literal['always', 'unless-none', 'json', 'json-unless-none'] = 'always',
    return_type: Any = PydanticUndefined,
) -> Callable[[Any], Any]:
    """Decorator that enables custom model serialization.

    This is useful when a model need to be serialized in a customized manner, allowing for flexibility beyond just specific fields.

    An example would be to serialize temperature to the same temperature scale, such as degrees Celsius.

    ```python
    from typing import Literal

    from pydantic import BaseModel, model_serializer

    class TemperatureModel(BaseModel):
        unit: Literal['C', 'F']
        value: int

        @model_serializer()
        def serialize_model(self):
            if self.unit == 'F':
                return {'unit': 'C', 'value': int((self.value - 32) / 1.8)}
            return {'unit': self.unit, 'value': self.value}

    temperature = TemperatureModel(unit='F', value=212)
    print(temperature.model_dump())
    #> {'unit': 'C', 'value': 100}
    ```

    See [Custom serializers](../concepts/serialization.md#custom-serializers) for more information.

    Args:
        __f: The function to be decorated.
        mode: The serialization mode.

            - `'plain'` means the function will be called instead of the default serialization logic
            - `'wrap'` means the function will be called with an argument to optionally call the default
                serialization logic.
        when_used: Determines when this serializer should be used.
        return_type: The return type for the function. If omitted it will be inferred from the type annotation.

    Returns:
        The decorator function.
    """

    def dec(f: Callable[..., Any]) -> _decorators.PydanticDescriptorProxy[Any]:
        dec_info = _decorators.ModelSerializerDecoratorInfo(mode=mode, return_type=return_type, when_used=when_used)
        return _decorators.PydanticDescriptorProxy(f, dec_info)

    if __f is None:
        return dec
    else:
        return dec(__f)  # type: ignore