Last Updated: August 12, 2019
·
9.706K
· lucho

Python dict to object

Python objects are actually dicts, so it is fairly simple to represent a dict structure as an object. I find this useful when I have a configuration-like structures with multiple levels of nesting. It is far more easier to use such structure as config.x.y.z than config['x']['y']['z'].

The code would look like this:

class DictToObject(object):

    def __init__(self, dictionary):
        def _traverse(key, element):
            if isinstance(element, dict):
                return key, DictToObject(element)
            else:
                return key, element

        objd = dict(_traverse(k, v) for k, v in dictionary.iteritems())
        self.__dict__.update(objd)

self.__dict__.update will update the object's internal dict representation with the supplied new structure and the recursive call of DictToObject in _traverse will take care of any nested dicts.

Example usage:

# let's pretend that you have loaded a json, yaml or other file
config_dict = {
    'group1': {
        'server1': {
            'apps': ('nginx', 'mysql'),
            'cpus': 4
        },
        'maintenance': True
    },
    'firewall_version': '1.2.3',
    'python2.7': True
}

config = DictToObject(config_dict)
assert config.group1.server1.apps == ('nginx', 'mysql')
assert config.group1.maintenance
assert config.firewall_version == '1.2.3'
assert config.__dict__['python2.7'] # or use
assert config.__getattribute__('python2.7')

Complete version of the code could be found in this gist: Python dict to object