drf-toon

Native TOON (Token Oriented Object Notation) parser and renderer for Django REST Framework.

TOON is a token-efficient serialization format designed for LLM contexts. It can result in 30–60% fewer tokens than JSON.

drf-toon is built on toons, a high-performance Rust implementation meant to mimic the native python json package.

Compatibility

drf-toon supports Django 5.2 LTS on Python 3.10–3.13, matching Django's own support matrix. Every supported Python version is exercised in CI.

Support for additional Django releases can be added later by extending the CI matrix.

Installation

pip install drf-toon

Usage

Add the parser and renderer to your DRF settings:

# settings.py
REST_FRAMEWORK = {
    "DEFAULT_PARSER_CLASSES": [
        "drf_toon.parsers.TOONParser",
        "rest_framework.parsers.JSONParser",
    ],
    "DEFAULT_RENDERER_CLASSES": [
        "drf_toon.renderers.TOONRenderer",
        "rest_framework.renderers.JSONRenderer",
    ],
}

Or enable them per-view:

from rest_framework.views import APIView
from rest_framework.response import Response

from drf_toon.parsers import TOONParser
from drf_toon.renderers import TOONRenderer


class UserView(APIView):
    parser_classes = [TOONParser]
    renderer_classes = [TOONRenderer]

    def get(self, request):
        return Response({"name": "Alice", "age": 30})

Both classes use the application/toon media type. Send requests with Content-Type: application/toon to parse, and Accept: application/toon (or ?format=toon) to render.

Development

Install the dev dependencies and run the checks with uv:

uv sync --dev

uv run ruff check .          # lint
uv run ruff format .         # format
uv run ruff format --check . # format check (CI)
uv run mypy                  # type check
uv run pytest                # tests + coverage

Optionally install the git hooks:

uv run pre-commit install --install-hooks

This wires up both a pre-commit hook (autofix + format) and a pre-push hook that blocks the push if the code is not properly linted/formatted.

License

MIT