API Star

🛠 The Web API toolkit. 🛠

Build Status codecov Package version


Community: https://discuss.apistar.org/ 🤔 💭 🤓 💬 😎

Repository: https://github.com/encode/apistar 💻


API Star is a toolkit for working with OpenAPI or Swagger schemas. It allows you to:

You can use it to build static documentation, integrate it within a Web framework, or use it as the client library for interacting with other APIs.

Quickstart

Install API Star:

$ pip3 install apistar

Let's take a look at some of the functionality the toolkit provides...

We'll start by creating an OpenAPI schema, schema.yaml:

openapi: 3.0.0
info:
  title: Widget API
  version: '1.0'
  description: An example API for widgets
servers:
  - url: https://www.example.org/
paths:
  /widgets:
    get:
      summary: List all the widgets.
      operationId: listWidgets
      parameters:
      - in: query
        name: search
        description: Filter widgets by this search term.
        schema:
          type: string

Let's also create a configuration file apistar.yml:

schema:
  path: schema.yaml
  format: openapi

We're now ready to start using the apistar command line tool.

We can validate our OpenAPI schema:

$ apistar validate
✓ Valid OpenAPI schema.

Or build developer documentation for our API:

$ apistar docs --serve
✓ Documentation available at "http://127.0.0.1:8000/" (Ctrl+C to quit)

We can also make API requests to the server referenced in the schema:

$ apistar request listWidgets search=cogwheel

Where did the server go?

With version 0.6 onwards the API Star project is being focused as a framework-agnositic suite of API tooling. The plan is to build out this functionality in a way that makes it appropriate for use either as a stand-alone tool, or together with a large range of frameworks.

The 0.5 branch remains available on GitHub, and can be installed from PyPI with pip install apistar==0.5.41. Any further development of the API Star server would likely need to be against a fork of that, under a new maintainer.

If you're looking for a high-performance Python-based async framework, then I would instead recommend Starlette.