jsonablr 2.0.1

Serialize Pydantic models and non-JSONable types into a JSONable variable.

JsonAblr

JsonAblr is a Python library that allows you to encode variables into a format that represents a JSON object. It provides a simple and intuitive way to convert variables into a JSON-like structure, making it easier to store and transmit data.

This was heavily inspired by FastAPI's jsonable_encoder.

• JsonAblr now requires Pydantic 2.0 or above. If you are using Pydantic 1.x, please use JsonAblr 1.1.1.

Features

• Encode variables into a JSON-like format
• Supports various data types including strings, numbers, lists, dictionaries, dates, Pydantic models, dataclasses and more
• Easy to use
• Customizable for your own types

Installation

You can install Jsonablr using pip:

pip install jsonablr

Usage

encode helper function:

from jsonablr import encode
from datetime import datetime, timezone

my_date = datetime.now(tz=timezone.utc)

print(encode(my_date))

Output:

2023-06-26T12:30:00.000Z

JsonAblr class:

from typing import Set
from datetime import datetime
from jsonablr import JsonAblr
from pydantic import BaseModel

class DynamoDbItem(BaseModel):
    partition_key: str
    sort_key: str
    games: Set[str]
    when: datetime

encoder = JsonAblr(
    encoders={
        datetime: lambda x: x.isoformat(sep=' ', timespec='seconds')
    },
    preserve_set=True
)

item = DynamoDbItem(
    partition_key='foo',
    sort_key='bar',
    games=['minecraft', 'terraria'],
    when=datetime.now()
)

print(encoder.encode(item))

This example demonstrates using a Pydantic model to be used as AWS DynamoDB item and not to be used as JSON. The games attribute is returned as a set (when using the preserve_set=True option) to support the DynamoDB SET type. Serialising a set with json.dumps will cause an error.

Output:

{
    'partition_key': 'foo',
    'sort_key': 'bar',
    'games': {'minecraft', 'terraria'},
    'when': '2023-06-26 12:30:00'
}