Pydantic - Upgrading object to another model

Question

I have a NewUser model that is something that the end user inputs, I want to update the object to a UserInDB so that I can pass it to my db engine (DynamoDB, which expects a dict)

At the moment I'm calling .dict twice, which doesn't feel like the correct way to do it

from pydantic import BaseModel, Field
from datetime import datetime
from typing import Optional
from uuid import uuid4
    
class NewUser(BaseModel):
    name: str
    email: str
    company_name: Optional[str]
    
class UserInDB(NewUser):
    hash: str = Field(default_factory=uuid4)
    range = 'DATA'
    created_at: datetime = Field(default_factory=datetime.now)
    
#...
#Emulating what an end user would send
user = NewUser(name='Example', company_name='example', email='[email protected]')
    
#Is calling dict twice way to do it?
user_in_db = UserInDB(**user.dict()).dict()
db.create_user(user_in_db)

Answer 1

You could try to define an __init__ and the code would look nicer (to me at least).

from pydantic import BaseModel, Field
from datetime import datetime
from typing import Optional
from uuid import uuid4

class NewUser(BaseModel):
    name: str
    email: str
    company_name: Optional[str]

class UserInDB(NewUser):
    hash: str = Field(default_factory= lambda: uuid4())
    range = 'DATA'
    created_at: datetime = Field(default_factory= lambda: datetime.now())

    def __init__(self, user: NewUser):
        super().__init__(**user.dict())

#...
# Emulating what an end user would send
user = NewUser(name='Example', company_name='example', email='[email protected]')

# This looks probably better
user_in_db = UserInDB(user)

Answer 2

I'm not a fan of modifying Pydantic's __init__, so I'd suggest either of these two alternative methods.

• to_db within NewUser means you can create the dict simply by calling user.to_db()
• from_user within UserInDB means you can create the dict by calling UserInDB.from_user(user)

Personally I'd go with the first method, but both have their advantages.

Also, note how you can use Model.parse_obj(Other) instead of Model(**Other.dict()).

class NewUser(BaseModel):
    name: str
    email: str
    company_name: Optional[str] = None

    def to_db(self) -> dict:
        return UserInDB.parse_obj(self).dict(by_alias=True)


class UserInDB(NewUser):
    hash: str = Field(default_factory=uuid4)
    db_range = Field('DATA', alias="range") # Avoid mirroring built-in `range`
    created_at: datetime = Field(default_factory=datetime.now)

    @classmethod
    def from_user(cls, user: NewUser) -> dict:
        return cls.parse_obj(user).dict(by_alias=True)


user = NewUser(
    name='Example',
    company_name='example',
    email='[email protected]',
)
 
user_in_db = user.to_db()
# or
user_in_db = UserInDB.from_user(user)

db.create_user(user_in_db)