django-dto-field 0.1.0b1

Storing DTO data in easy-to-use and production-ready Django Model Field with fast [de]serialization.

django-dto-field

Storing DTO data in easy-to-use and production-ready Django Model Field with fast [de]serialization.

A custom Django Model Field that can serialize and deserialize for different types of DTO (Data-Transfer-Object). DTOField detect DTO type and serialize the most efficiently. Also it has own binary container to maintain data integrity and economical storage.

✨ Features

• All DB support (if DB support BinaryFiled)
• Zip field for size saving (optional)
• Supports pure dict as DTO
• Supports dataclass as DTO
• Supports pydantic as DTO
• Supports marshmallow as DTO
• Supports attrs as DTO
• Supports adapdtix as DTO

⬇️ Install

$ pip install django-dto-field

🚀 Quick start

Choose your DTO: dict, dataclass

dict

We support pure dict for using it like DTO object. The main different between dict and other DTO types is that dict has no validation and DTOField uses it by default.

Also we have some benchmarks with default JSONField. And they shows that DTOField is 2.4x faster on write operation (thanks to amazing msgspec lib).

1. Set DTOField to your model:

>>> from django.db.models import CharField
>>> from django_dto_field import DTOField

>>> class Country(Model):
...    name = CharField()
...    city = DTOField()

2. Use it:

>>> city = {"name": "Capitol", "districts": [f"d{i}" for i in range(1, 14)]}
>>> created_country = Country.objects.create(name="Panem", city=city)

>>> from_db_country = Country.objects.get(pk=created_country.pk)
>>> assert isinstance(from_db_country.city, dict)
>>> assert from_db_country.city == city

dataclass

We support dataclass object as DTO. You can use it simply without thinking of serialization, deserialization and validation. DTOField will do it for you.

1. Set dataclass schema:

>>> from dataclasses import dataclass

>>> @dataclass
... class City:
...     name: str
...     districts: list[str]

2. Set DTOField to your model and add schema.

>>> from django.db.models import CharField
>>> from django_dto_field import DTOField

>>> class Country(Model):
...    name = CharField()
...    city = DTOField(schema=City)

3. Use it:

>>> city = City(name="Capitol", districts=[f"d{i}" for i in range(1, 14)])
>>> created_country = Country.objects.create(name="Panem", city=city)

>>> from_db_country = Country.objects.get(pk=created_country.pk)
>>> assert isinstance(from_db_country.city, City)
>>> assert from_db_country.city == city

Validation error will raise if wrong schema will appear (what if schema change?):

>>> @dataclass
... class WrongCity:
...     addresses: list[str]
...

>>> wrong_city = WrongCity(addresses=[])
>>> from_db_country.city = wrong_city
Traceback (most recent call last):
    ...
ValidationError: given value 'WrongCity(addresses=[])' did not match schema 'City'

What if schema will change?

The goal of DTOField is to give you an easy tool for working with different DTO's objects but not how to manipulate them. This is why we will never give migration features.

But it's not a problem! Often you can make migration with 3 steps:

1. Define the new DTOField with your new schema.
2. Write new custom Django migration to get, change and save updated data to new field.
3. Delete old field.

🤗 Author

Made with love by @skv0zsneg