API

Swagger.py will dynamically build an object model from a Swagger-enabled RESTful API.

Here is a simple example using the Asterisk REST Interface

#!/usr/bin/env python

import json

from swaggerpython.client import SwaggerClient
from swaggerpython.http_client import SynchronousHttpClient

http_client = SynchronousHttpClient()
http_client.set_basic_auth('localhost', 'hey', 'peekaboo')

ari = SwaggerClient(
    "http://localhost:8088/ari/api-docs/resources.json",
    http_client=http_client)

ws = ari.events.eventWebsocket(app='hello')

for msg_str in iter(lambda: ws.recv(), None):
    msg_json = json.loads(msg_str)
    if msg_json['type'] == 'StasisStart':
        channelId = msg_json['channel']['id']
        ari.channels.answer(channelId=channelId)
        ari.channels.play(channelId=channelId,
                          media='sound:hello-world')
        ari.channels.continueInDialplan(channelId=channelId)

swagger-codegen

There are the beginnings of a Mustache-based code generator, but it’s not functional… yet.

Data model

The data model presented by the swagger_model module is nearly identical to the original Swagger API resource listing and API declaration. This means that if you add extra custom metadata to your docs (such as a _author or _copyright field), they will carry forward into the object model. I recommend prefixing custom fields with an underscore, to avoid collisions with future versions of Swagger.

There are a few meaningful differences.

• Resource listing

• The file and base_dir fields have been added, referencing the original .json file.

• The objects in a resource_listing’s api array contains a field api_declaration, which is the processed result from the referenced API doc.

• API declaration

• A file field has been added, referencing the original .json file.