jsondaora 0.10.1

Interoperates @dataclass with json objects

jsondaora

Interoperates dataclasses and TypedDict annotations with json objects for python

---

Documentation: https://dutradda.github.io/jsondaora

Source Code: https://github.com/dutradda/jsondaora

---

Key Features

• Full compatibility with dataclasses module and TypedDict annotation
• Deserialize values from dict
• Deserialize values from bytes*
• Deserialization/serialization of chosen fields
• Deserialize the following standard types: Dict, List, Tuple, Set, Union and Any
• Dict serialization
• Direct json serialization with orjson (don't convert to dict recursively before serialization)
• Optional validation according with the json-schema specification*

* feature in development.

Requirements

• Python 3.8+
• orjson for json serialization

Installation

$ pip install jsondaora

Basic example

from typing import List, TypedDict

from jsondaora import (
    as_typed_dict,
    asdataclass,
    dataclass_asjson,
    jsondaora,
    typed_dict_asjson,
)


@jsondaora
class Person:
    name: str
    age: int

    class Music:
        name: str

    musics: List[Music]


jsondict = dict(name=b'John', age='40', musics=[dict(name='Imagine')])
person = asdataclass(jsondict, Person)

print('dataclass:')
print(person)
print(dataclass_asjson(person))
print()


# TypedDict


@jsondaora
class MusicT(TypedDict):
    name: str


@jsondaora
class PersonT(TypedDict):
    name: str
    age: int

    musics: List[MusicT]


jsondict = dict(name=b'John', age='40', musics=[dict(name='Imagine')])
person = as_typed_dict(jsondict, PersonT)

print('TypedDict:')
print(person)
print(typed_dict_asjson(person, Person))

dataclass:
Person(name='John', age=40, musics=[Person.Music(name='Imagine')])
b'{"name":"John","age":40,"musics":[{"name":"Imagine"}]}'

TypedDict:
{'name': 'John', 'age': 40, 'musics': [{'name': 'Imagine'}]}
b'{"name":"John","age":40,"musics":[{"name":"Imagine"}]}'

Example for choose fields to deserialize

from typing import List, TypedDict

from jsondaora import (
    as_typed_dict,
    asdataclass,
    dataclass_asjson,
    jsondaora,
    typed_dict_asjson,
)


@jsondaora(deserialize_fields=('name'))
class Person:
    name: str
    age: int

    class Music:
        name: str

    musics: List[Music]


jsondict = dict(name=b'John', age='40', musics=[dict(name='Imagine')])
person = asdataclass(jsondict, Person)

print('dataclass:')
print(person)
print(dataclass_asjson(person))
print()


# TypedDict


@jsondaora
class MusicT(TypedDict):
    name: str


@jsondaora(deserialize_fields=('name',))
class PersonT(TypedDict):
    name: str
    age: int

    musics: List[MusicT]


jsondict = dict(name=b'John', age='40', musics=[dict(name='Imagine')])
person = as_typed_dict(jsondict, PersonT)

print('TypedDict:')
print(person)
print(typed_dict_asjson(person, PersonT))

dataclass:
Person(name='John', age='40', musics=[{'name': 'Imagine'}])
b'{"name":"John","age":"40","musics":[{"name":"Imagine"}]}'

TypedDict:
{'name': 'John', 'musics': [{'name': 'Imagine'}], 'age': '40'}
b'{"name":"John","musics":[{"name":"Imagine"}],"age":"40"}'

Example for choose fields to serialize

from typing import List, TypedDict

from jsondaora import (
    as_typed_dict,
    asdataclass,
    dataclass_asjson,
    jsondaora,
    typed_dict_asjson,
)


@jsondaora(serialize_fields=('name', 'age'))
class Person:
    name: str
    age: int

    class Music:
        name: str

    musics: List[Music]


jsondict = dict(name='John', age=40, musics=[dict(name='Imagine')])
person = asdataclass(jsondict, Person)

print('dataclass:')
print(person)
print(dataclass_asjson(person))
print()


# TypedDict


@jsondaora
class Music(TypedDict):
    name: str


@jsondaora(serialize_fields=('age',))
class PersonT(TypedDict):
    name: str
    age: int
    musics: List[Music]


jsondict = dict(name=b'John', age='40', musics=[dict(name='Imagine')])
person = as_typed_dict(jsondict, PersonT)

print('TypedDict:')
print(person)
print(typed_dict_asjson(person, PersonT))

dataclass:
Person(name='John', age=40, musics=[Person.Music(name='Imagine')])
b'{"age":40,"name":"John"}'

TypedDict:
{'name': 'John', 'age': 40, 'musics': [{'name': 'Imagine'}]}
b'{"age":40}'

Wins Pydantic Benchmark

jsondaora is up to 3.2 times faster than pydantic on it's benchmark