Serving¶
This section shows you how to run your APIs in RESTArt.
As a concrete example, consider the following Greeting API:
# helloworld.py
from restart.api import RESTArt
from restart.resource import Resource
api = RESTArt()
@api.route(methods=['GET'])
class Greeting(Resource):
name = 'greeting'
def read(self, request):
return {'hello': 'world'}
Development¶
The Pythonic Way¶
As a Pythonista, chances are you like to run the API just as a normal Python script. That’s good!
The Pythonic way you want is supported by RESTArt, at the cost of a little wrapper code with the help of the Service class:
# runserver.py
from restart.serving import Service
from helloworld import api
service = Service(api)
if __name__ == '__main__':
service.run()
Now, you can run the API like this:
$ python runserver.py
The Command Line Utility¶
To make the serving step as simple as possible, RESTArt also provides a command line utility called restart. You may have seen it in Quickstart. Yes! It’s born for serving, and you will not be disappointed to use it:
$ restart helloworld:api
That’s all. Isn’t it amazing?
restart has only one argument:
| Argument | Example | Description |
|---|---|---|
| entrypoint | helloworld:api | A string in the form module_path:api
where api is the central RESTArt API
object and module_path is the path to
the module where api is defined. |
For the options supported by reatart, see the help messages:
$ restart --help
Deployment¶
RESTArt’s primary deployment platform is WSGI, the Python standard for web servers and applications.
To make RESTArt APIs easy to deploy, it’s recommended to create a file named wsgi.py as follows:
# wsgi.py
from restart.serving import Service
from helloworld import api
application = Service(api)
Then use awesome WSGI servers to communicate with the application callable.