Modify data representation and field order of response model

[jd-solanki]

First Check

• I added a very descriptive title to this issue.
• I used the GitHub search to find a similar issue and didn't find it.
• I searched the FastAPI documentation, with the integrated search.
• I already searched in Google "How to X in FastAPI" and didn't find any information.
• I already read and followed all the tutorial in the docs and didn't find an answer.
• I already checked if it is not related to FastAPI but to Pydantic.
• I already checked if it is not related to FastAPI but to Swagger UI.
• I already checked if it is not related to FastAPI but to ReDoc.

Commit to Help

• I commit to help with one of those options 👆

Example Code

# TIP: This is full app. Run using `uvicorn main:app --port 5000 --reload`

from datetime import datetime
from typing import List

from fastapi import FastAPI
from pydantic import BaseModel

# Fake Data (database)

data = [
    {
        "id": 1,
        "username": "johndoe007",
        "created_at": datetime.now(),
        "updated_at": datetime.now(),
    }
]


# Pydantic
class DemoBase(BaseModel):
    username: str


class DemoOut(DemoBase):
    id: int
    created_at: datetime
    updated_at: datetime


# FastAPI
app = FastAPI()


@app.get("/", response_model=List[DemoOut])
def read_root():
    return data

Description

If you visit the app you will get below response:

[
    {
        "username": "johndoe007",
        "id": 1,
        "created_at": "2021-09-13T12:57:55.200998",
        "updated_at": "2021-09-13T12:57:55.201006"
    }
]

It is working fine but the problem is when we want below:

1. order the data sent by response_model

In the above response, id is returned at 2nd position but I want to return at the 1st position. Requirement for this can be:

• This is how I want my API to respond
• As a full stack developer I directly render the response in the table and my id column is placed in 2nd column which should be in 1st row.

2. Update the data representation

In the above response, we get the date as 2021-09-13T12:57:55.200998 but I want to return the formatted date which should be human-readable. This is just a single example but there can be others as explained in this discussion.

Wanted Solution

I guess we can create another magic class that lets us define the field order and value representation but that might introduce code duplication. Sorry but I don't have any perfect solution.

Wanted Code

No idea

Alternatives

None. I might have to raise the same issue to pydantic.

Operating System

Linux

Operating System Details

Linux Mint 20.2 Cinnamon

FastAPI Version

0.68.0

Python Version

3.9.x

Additional Context

None

[NoCrEaTiVe]

@jd-0001

class DemoOut(BaseModel):
    id: int
    username: str
    created_at: datetime
    updated_at: datetime
    class Config:
        orm_mode = True
        validate_assignment = True

    @root_validator
    def change_created_at(cls, values):
        values["created_at"] = values["created_at"].strftime("%Y-%m-%dT%H:%M:%SZ")
        values['created_at'] = datetime.datetime.strptime(
            values["created_at"], "%Y-%m-%dT%H:%M:%SZ"
        )
        return values

    @root_validator
    def change_updated_at(cls, values):
        values["updated_at"] = values["updated_at"].strftime("%Y-%m-%dT%H:%M:%SZ")
        values['updated_at'] = datetime.datetime.strptime(
            values["updated_at"], "%Y-%m-%dT%H:%M:%SZ"
        )
        return values`

[jd-solanki]

Hi @NoCrEaTiVe

I know I can use validators but this is not what they are made for. As per their name they are for validation.

[Kludex]

pydantic/pydantic#593

[adm0104]

@jd-solanki I had this same issue and I think I've got a reasonable workaround. Make an empty subclass that inherits all of its properties from your desired superclass and a "mirror" superclass that has the additional properties that you want. Based on my testing, the properties on "actual_sub_foo" in the below example will populate in reverse order both on the FastAPI Swagger page and in any responses. See below:

from pydantic import BaseModel

class base_foo(BaseModel):
    prop_2: int
    prop_3: int

class mirror_sub_foo(BaseModel):
    prop_1: int

class actual_sub_foo(base_foo, mirror_sub_foo):
    None

In this example, if you use "actual_sub_foo" as your response_model, the API response should look like:

{
    "prop_1": some_int,
    "prop_2": some_int,
    "prop_3": some_int
}