Pydantic computed field example values

[rsaleev]

Initial Checks

• I have searched Google & GitHub for similar requests and couldn't find anything
• I have read and followed the docs and still think this feature is missing

Description

Maybe my approach is not right but I use computed_field wrapper for injecting ORM values in root model and it will be for sure a cool feature to have examples like for Field, less code and integration with FastAPI/OpenAPI

Affected Components

• Compatibility between releases
• Data validation/parsing
• Data serialization - .model_dump() and .model_dump_json()
• JSON Schema
• Dataclasses
• Model Config
• Field Types - adding or changing a particular data type
• Function validation decorator
• Generic Models
• Other Model behaviour - model_construct(), pickling, private attributes, ORM mode
• Plugins and integration with other tools - mypy, FastAPI, python-devtools, Hypothesis, VS Code, PyCharm, etc.

[sydney-runkle]

Hi @rsaleev,

Thanks for your feedback. Could you provide an example MRE code snippet so that we can better understand what you're asking for? Thanks 👍

[rsaleev]

Hi @rsaleev,

Thanks for your feedback. Could you provide an example MRE code snippet so that we can better understand what you're asking for? Thanks 👍

computed_field(alias="marketPlaceAccountName", return_type=str)
 def marketplace_acccount_name(self):
        return self.marketplace.name

No examples available for OpenAPI, only attr will be shown with field type

 order_number: Optional[str] = Field(
        serialization_alias="orderNumber", examples=[None, "1234567"]
    )

Examples will be shown in rendered OpenAPI docs

[sydney-runkle]

@rsaleev,

Thanks for the response. This seems like a reasonable feature request - I think we could also include json_schema_extra and a few other attributes from Field for computed_field.

[TiansuYu]

Started looking into this issue without hard commitment (Also leave the comment as a marker, so I can find the page again).

[alexmojaki]

PR: #8013

[adriangb]

I see we merged #8013 but I do want to point out that Annotated and Field in the return type works:

from typing import Annotated

from pydantic import BaseModel, Field, computed_field


class Model(BaseModel):
    a: int
    b: int

    @computed_field
    @property
    def c(self) -> Annotated[int, Field(description='The sum of a and b', examples=[3, 5])]:
        return self.a + self.b

print(Model.model_json_schema(mode='serialization')['properties']['c'])
# {'description': 'The sum of a and b', 'examples': [3, 5], 'readOnly': True, 'title': 'C', 'type': 'integer'}

This avoids having to duplicate the entirety of Field into computed_field. Nonetheless I do think having the parameters in computed_field is more ergonomic for end users.