A python-based pillar `sls` file that uses grains data to create a `Network` object. The network object is returned in a dictionary that is included in the render context.

net.conf

addr={{ pillar['shoe']['network'].addr }}

shoe.sls

#!py
from somewhere import Network

def run():
    # the grains variable is available inside the run() function.
    return {
        'shoe': {
            'mirror': 'http://apt:3142/archive.ubuntu.com/ubuntu',
            'network': Network(grains['nodename']),
        },
    }