How to parse and read "_id" field from and to a pydantic model?

Question

I am trying to parse MongoDB data to a pydantic schema but fail to read its _id field which seem to just disappear from the schema.
The issue is definitely related to the underscore in front of the object attribute. I can't change _id field name since that would imply not parsing the field at all.
Please find below the code I use (using int instead of ObjectId for the sake of simplification)

from pydantic import BaseModel

class User_1(BaseModel):
    _id: int

data_1 = {"_id": 1}

parsed_1 = User_1(**data_1)
print(parsed_1.schema())

class User_2(BaseModel):
    id: int

data_2 = {"id": 1}

parsed_2 = User_2(**data_2)
print(parsed_2.schema())

User_1 is parsed successfully since its _id field is required but can't be read afterwards. User_2 works in the above example by fails if attached to Mongo which doesn't provide any id field but _id.

Output of the code above reads as follows:

User_1  {'title': 'User_1', 'type': 'object', 'properties': {}}
User_2  {'title': 'User_2', 'type': 'object', 'properties': {'id': {'title': 'Id', 'type': 'integer'}}, 'required': ['id']}

Samuel Colvin · Accepted Answer · 2020-01-02 12:11:17Z

36

you need to use an alias for that field name

from pydantic import BaseModel, Field

class User_1(BaseModel):
    id: int = Field(..., alias='_id')

See the docs here.

answered Jan 2, 2020 at 12:11

Samuel Colvin

13.5k7 gold badges64 silver badges78 bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

miksus Over a year ago

Thanks! For those who wish to convert an object back to dict containing the _id, just use User_1(_id=1234).dict(by_alias=True) so you end up with a dict having the _id key.

Thiago Lages de Alencar Over a year ago

using Field didn't work, i used Config.fields instead

GoodGuyGregory 2 days ago

could you explain this more?

Thiago Lages de Alencar yesterday

In old versions of pydantic you could configure more details by creating a class named "Config" inside your model (docs.pydantic.dev/1.10/usage/model_config). Until version 2.10 they would still mention the old way (docs.pydantic.dev/2.10/concepts/config). But you shouldn't use it nowadays

Zeailer · Accepted Answer · 2023-12-28 15:24:40Z

6

I know that you had asked this ages ago, but for anyone else still struggling with it here's a suggested solution by MongoDb

When applying their solution we get the following code.

from typing import Optional, Annotated
from pydantic import BaseModel, Field, BeforeValidator

PyObjectId = Annotated[str, BeforeValidator(str)]

class User_1(BaseModel):
    id: Optional[PyObjectId] = Field(alias="_id", default=None)

All the validation and model conversions work just fine, without any class Config, or other workarounds. Hope this helps somebody.

answered Dec 28, 2023 at 15:24

Zeailer

761 silver badge3 bronze badges

Comments

GoodGuyGregory · Accepted Answer · 2025-11-20 19:39:42Z

Okay so I recently needed to create a data object reference with Pydantic. I was able to get

Object Definition

class User(BaseModel):
    '''
        User Model for Deserialization With MongoDB
    '''
    id: str
    username: str
    api_key: str
    expiration: datetime
    last_used: datetime 
    created_at: datetime
    updated_at: datetime

Attempted Unpacking Object

I gave up on trying to unpack the object with the following, clearly there is something stupid I am missing here.

user = User(id=str(found_user["_id"]), **found_user)

Error Output

_id
  Input should be a valid string [type=string_type, input_value=ObjectId('6942069420'), input_type=ObjectId]
    For further information visit https://errors.pydantic.dev/2.12/v/string_type

Explicit Object Fields

Adding the fields explicitly. This feels stupid but works.

 user = User(id=str(found_user['_id']),
                            username=found_user['username'],
                            api_key=found_user['api_key'],
                            expiration=found_user['expiration'],
                            last_used=found_user['last_used'],
                            created_at=found_user['created_at'],
                            updated_at=found_user['updated_at'])

Eric Aya · Accepted Answer · 2024-03-16 15:06:17Z

-1

The other two answers don't work for me, I don't know why. Maybe the versions are deprecated, here is the updated answer:

from pydantic import Field
from bson import ObjectId 
from typing import Optional

id: Optional[ObjectId] = Field(alias='_id')
title: str
cover: str

class Config:
    json_encoders = {
        ObjectId: str
    }
    arbitrary_types_allowed = True

edited Mar 16, 2024 at 15:06

Eric Aya

70.2k36 gold badges190 silver badges266 bronze badges

answered Mar 16, 2024 at 7:52

hem charan

1252 silver badges9 bronze badges

Collectives™ on Stack Overflow

How to parse and read "_id" field from and to a pydantic model?

4 Answers 4

4 Comments

Comments

Object Definition

Attempted Unpacking Object

Explicit Object Fields

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

4 Comments

Comments

Object Definition

Attempted Unpacking Object

Explicit Object Fields

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related