Skip to main content


Project description


SuperCell is a collection of best practices for creating with Tornado based HTTP APIs. It resembles many ideas from Coda Hale’s Dropwizard, a similar library for JVM based HTTP APIs.

TravisImage CoverAlls


Code is worth more than words, so this is a simple example demonstrating the idea:

from tornado import options
from schematics.models import Model
from schematics.types import StringType
import supercell.api as s

options.define('defaultName', default='Your Name', type=str,
               help='Default name shown when no name was given')

options.define('template', default='templates/hello-world.txt', type=str,
               help='Name for the hello world template')

# model
class Saying(Model):
    id = IntType()
    content = StringType()

class SimpleExampleHandler(s.RequestHandler):

    def get(self):
        template = self.config['HELLO_WORLD_TEMPLATE']
        name = self.get_argument('name', self.config['DEFAULT_NAME'])
        raise s.Return(Saying(id=12, content=template.render(name=name)))

class MyService(s.Service):

    def bootstrap(self):

    def run(self):

        # add our handler
        self.add_handler('/hello-world', SimpleExampleHandler,
                         {'HELLO_WORLD_TEMPLATE': self.config.template,
                          'DEFAULT_NAME': self.config.defaultName},

if __name__ == '__main__':


Configuration is managed via Tornado options. In addition to setting the options via command line, parsing config files is enabled as well. Config files are named using a combination of user- and hostname. Running as user ‘daniel’ on host ‘dev.local’ would create a configfile named ‘daniel_dev.local.cfg’.

The resolution order is determined by the Environment.configFilePaths. In the example above we have set this to ./etc/ and /etc/myservice/. So when starting the application the config files in the local directory is parsed before the system administrator’s config file. After that, command line parameters are being parsed. So all config files are parsed in that order:

./etc/config ./etc/daniel_dev.local.cfg /etc/myservice/config /etc/myservice/daniel_dev.local.cfg

None of the files have to exist in order to start the application.

The default configuration of a supercell application contains several values for communications such as ports for the main and the administrative interface and may contain settings for the underlying Tornado Application.

Managed Objects

Often you need some central instance to talk to other services to. In order to simplify this, you can add a ManagedObject to the Environment, that you then can retrieve in your handler implementation. Currently there is one built-in version for handling AsynchronousHTTPClient instances. This will use the PyCURL client, if available and fallback to the SimpleHTTPClient otherwise.

In order to add this to your environment you simply have to:

def run(self):
    env.add_managed_object('http_client', AsynchHTTPClient)

In you handler code you may then get access to the client via the environment:

client = self.environment.http_client


A simple decorator allows for fine grained control of public and browser caches. The basic idea is to:

class MyHandler(s.RequestHandler):

    def get(self):
        raise s.Ok()

This will allow all caches to store a local copy of your response and serve it for 10 minutes. By default the must_revalidate option is set, so that a cache may not serve a stale copy but revalidate it when it is expired. So in this example the header is set to:

Cache-Control: max-age=600, must-revalidate

For a detailed description of the available options see the Caching Tutorial <> and the official RFC2616, sec 14.9 <>.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for supercell, version 0.4.0.beta5
Filename, size File type Python version Upload date Hashes
Filename, size supercell-0.4.0.beta5.tar.gz (23.2 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page