Oxley: Pydantic classes from JSON schema

Oxley generates Pydantic classes at runtime from user-provided JSON schema documents. Heavily indebted to packages like Python-JSONschema-Objects, Oxley enables data validation pipelines to function dynamically, and with the help of Pydantic, interface directly with popular web frameworks such as FastAPI and Starlite.

Quick start

Install from PIP:

python3 -m pip install oxley

Given a simple JSONschema document:

{
  "$id": "https://github.com/jsstevenson/oxley",
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "type": "object",
  "$defs": {
    "User": {
      "type": "object",
      "properties": {
        "username": {"type": "string"},
        "user_id": {"type": "number"}
      },
      "required": ["username", "user_id"]
    },
    "Post": {
      "type": "object",
      "properties": {
        "author": {"$ref": "#/$defs/User"},
        "content": {"type": "string"},
        "allow_responses": {"type": "boolean"}
      },
      "required": ["author", "content"]
    }
  }
}

Provide a schema and construct classes:

from oxley import ClassBuilder
schema_path = "path/to/my_jsonschema_document.json"
cb = ClassBuilder(schema_path)
User, Post = cb.build_classes()

The resulting objects are functioning Pydantic classes, providing features like runtime data validation and matching schema output.

dril = User(username="dril", user_id=99)
post = Post(author=dril, content="should i learn Letters first?  or choose the path of Numbers? a queston every baby must ask it self")
another_post = Post(author=dril)  # raises pydantic.ValidationError

Development

Clone and install dev and test dependencies:

git clone https://github.com/jsstevenson/oxley
cd oxley
# make virtual environment of your choosing
python3 -m pip install ".[dev,test]"

Install pre-commit hooks:

pre-commit install

Run tests with tox:

tox