Convention over configuration Object Schemas for python


License
MIT
Install
pip install schemey==3.3.3

Documentation

Schemey - Json Schemas for Python.

This project allows for generation of json schemas based on python classes, or python classes based on json schemas. It also allows for generation of validated dataclasses, where setters cannot violate the invariants established in a schema.

It uses the fantastic JSON Schema library for python. (Though older versions did not.)

The general idea is that the framework should not insist on any particular data structure or paradigm - it is designed to be extensible, and out of the box support is provided for iterable types, dataclasses, enums, timestamps and primitives.

Serialization is provided using marshy.

Current test coverage is at 100%

Why did you build this?

There were gaps in the functionality of existing solutions (Like pydantic) that made using them untenable for my use cases.

Installation

pip install schemey

Concepts

  • A Schema contains a link between a JSON Schema and a Python Type
  • A Validator is used to validate python objects using a schema
  • A SchemaContext is used to generate python objects for json schemas / vice versa
  • A SchemaFactory is used to plug new translation rules into a SchemaContext (more below)

Examples

Hello World

This demonstrates generating a validator for a dataclass.

Validated Dataclass

This demonstrates generating a validated dataclass

Validated Fields

This demonstrates adding custom validation rules to dataclass fields

Custom Class Validations

This demonstrates adding fully custom marshalling and validations for a class

Custom JSON Schema Validations

This demonstrates creating custom json schema validations for things not natively supported by json schema. For example, checking a date against the current time, or that a property of an object is less than another property of that object.

Beginning with a JSON Schema

This demonstrates starting with a json schema and generating python dataclasses from it.

Configuring the Context itself

Schemey uses Injecty for configuration. The default configuration is here

For example, for a project named no_more_uuids, I may add a file injecty_config_no_more_uuids/__init__.py:

from schemey.factory.schema_factory_abc import SchemaFactoryABC
from schemey.factory.uuid_factory import UuidFactory

priority = 120  # Applied after default


def configure(context):
    context.deregister_impl(SchemaFactoryABC, UuidFactory)

Installing local development dependencies

python setup.py install easy_install "schemey[dev]"

Release Procedure

status

The typical process here is:

  • Create a PR with changes. Merge these to main (The Quality workflows make sure that your PR meets the styling, linting, and code coverage standards).
  • New releases created in github are automatically uploaded to pypi