JSON Schema

This page outlines the approach to use JSON schema for providing an user-interface for the python-based packages. It is based on the work by Prof Paul Grogan on the TAT-C and NOS-T projects.

From email correspondence with Paul

Primary supporting libraries:

pydantic is the base utility that helps interchange Python objects and JSON objects (with data typing/validation).

geojson-pydantic supplies some of the pydantic schemas conforming to the GeoJSON standard.

Other supporting libraries:

typing, typing_extensions, enum, uuid

fastapi_utils provides the APIModel schema which helps align schemas with web APIs. For example, pythonic convention uses snake case for variables (my_attribute) while web services more commonly use camel case (myAttribute) – APIModel automatically provides aliases for all snake case variables. There may be some other benefits bit it is largely for ease of use rather than new functionality.

Functionality is (in order of importance):

1. serialization/deserialization from JSON,
2. validation checks,
3. code interfaces that specify required/expected functions.

#1 helps define interoperability standards (data must be serialized to be exchanged between hosts). #2 helps refine the specific syntax allowed/disallowed. #3 helps with internal development.

pydantic is the base utility that helps interchange Python objects and JSON objects (with data typing/validation).

geojson-pydantic supplies some of the pydantic schemas conforming to the GeoJSON standard.

fastapi_utils provides the APIModel schema which helps align schemas with web APIs. For example, pythonic convention uses snake case for variables (my_attribute) while web services more commonly use camel case (myAttribute) – APIModel automatically provides aliases for all snake case variables. There may be some other benefits bit it is largely for ease of use rather than new functionality.

Implementation

In TAT-C, the src/tatc folder is for the Python library; app/ is source code for the web service application which relies on the core TAT-C library. So src/tatc/schemas are all of the base objects independent of any particular application and app/tatc_app/... contains additional schemas for the HTTP web services described here https://tatc.code-lab.org/docs

For NOS-T, nost_tools/ is the root directory for the library (functionally similar to src/tatc) and examples/ has some example applications that rely on the base library. The nost_tools/schemas.py defines base schemas for messages exchanged between applications. Any example application will need application-specific events that could also be defined with other schemas to establish a standard format.

Relevant publication:

• Paul T. Grogan, Josue I. Tapiaa, "Using JSON Schema to define a system modeling vocabulary: The Tradespace Analysis Tool for Constellations (TAT-C)", 2023 Conference on Systems Engineering Research. Grogan_CSER23_TATC.pdf