Serving REST APIs¶

Note

This topic describes how to create a simple REST API server with built-in Swagger support. For REST

Shortcut for creating API`s¶

micro.rest.api.make_api(app, callbacks=(), name=None, description=None, version=None)[source]¶

Create API server instance

Parameters:	app (`micro.web.Micro`) – micro.core.app.Micro instance callbacks (iterable) – iterable of callbacks which should be run before API initializes name (str) – Name of this API (required) description (str) – Description of this API (required) version (str) – Version of this API (required)
Returns:	API instance, a `micro.rest.api.API` subclass

Pre-created system routes¶

Any API server created with make_api() will have next routes serving the content:

/swagger.json - serving the Swagger API description JSON object (see Swagger specification)
/health-check - serving the two bytes ‘UP’, which is suitable for checking that API server is up and running

API servers which are running in debug mode (see micro.rest.api_debug_mode) will have the next routes pre-defined:

/_debug - which serves the redirect into Swagger UI interface
/_debug/ - under this prefix we serve Swagger UI interface static files. We have a pretty convenient swagger task in

our Rakefile for cloning Swagger UI distribution into the right place for serving its static.

Running API server in development for debugging purposes¶

The following code runs an API server without any declared endpoints in debug mode.

from micro.core import Micro

from micro.rest import make_api

app = Micro()

api_instance = make_api(app, name='Micro.Rest example server',
    description='An example API server without any endpoints', version='1')
api_instance.debug(host='0.0.0.0', port=8080, name='TestAPI')

This runs API server listening on 0.0.0.0:8080 with log prefix ‘TestAPI’

Running API server in production¶

The class micro.rest.api.API (and its subclasses which are created on fly with make_api() method) are full compatible with aiohttp gunicorn worker