Treat dataclasses like dictionaries

DataDict is available at PyPI. You can install it using Pip:

$ python -m pip install datadict

You can use a DataDict dataclass as a drop-in replacement for dataclasses.dataclass in the standard library:

from datadict import dataclass

@dataclass(order=True)
class Point:
    x: int
    y: int

For instances of the new data class, you can access attributes using square brackets, similar to dictionary access:

>>> point = Point(1, 2)
>>> point
Point(x=1, y=2)

>>> point.x
1
>>> point["x"]
1

>>> point["y"] = 3
>>> point
Point(x=1, y=3)

You can also convert the dataclass instance to a proper dictionary:

>>> point.asdict()
{'x': 1, 'y': 3}

You can always download the latest version of DataDict from GitHub. DataDict uses Flit as a build tool.

To install DataDict from the downloaded source, run Flit through pip:

$ python -m pip install .

If you want to change and play with the DataDict source code, you should install it in editable mode:

$ python -m pip install -w .[dev,test]