configurator

This is a Python library for building a configuration store from one or more layered configuration sources. These are most commonly files, with YAML, TOML and JSON support included and other formats easily added. The sources don't have to be files, and support is included for both environment variables and command line options.

In addition to an easy to use interface, configuration information is also made available as nested, simple python data types so that you can validate the schema of your configuration using the tool of your choice.

Quickstart

To install the library, go for:

pip install configurator[yaml,toml]

Here's how you would handle a layered set of defaults, system-wide config and then optional per-user config:

from configurator import Config

defaults = Config({
    'cache': {
        'location': '/tmp/my_app',
        'max_files': 100,
    },
    'banner': 'default banner',
    'threads': 1,
})
system = Config.from_path('/etc/my_app/config.yaml')
user = Config.from_path('~/.my_app.yaml', optional=True)
config = defaults + system + user

Now, if we wanted configuration from the environment and command line arguments to override those provided in configuration files, we could do so as follows:

import os
from argparse import ArgumentParser
from configurator import convert, target, required

config.merge(os.environ, {
    convert('MYAPP_THREADS', int): 'threads',
    required('MYAPP_CACHE_DIRECTORY'): 'cache.location',
})

parser = ArgumentParser()
parser.add_argument('--threads', type=int)
parser.add_argument('--max-files', type=int)
args = parser.parse_args()

config.merge(args, {
    'threads': 'threads',
    'max_files': 'cache.max_files',
})

To check the configuration we've accumulated is sensible we can use a data validation library such as Voluptuous:

from os.path import exists
from voluptuous import Schema, All, Required, PathExists

schema = Schema({
    'cache': {'location': All(str, PathExists()), 'max_files': int},
    'banner': Required(str),
    'threads': Required(int),
    })

schema(config.data)

So, with all of the above, we could use the following sources of configuration:

>>> import os, sys
>>> print(open('/etc/my_app/config.yaml').read())
cache:
  location: /var/my_app/
<BLANKLINE>
>>> os.environ['MYAPP_THREADS']
'2'
>>> os.environ['MYAPP_CACHE_DIRECTORY']
'/var/logs/myapp/'
>>> sys.argv
['myapp.py', '--threads', '3', '--max-files', '200']

With the above sources of configuration, we'd end up with a configuration store that we can use as follows:

>>> config.cache.location
'/var/logs/myapp/'
>>> config.cache.max_files
200
>>> config.banner
'default banner'
>>> config.threads
3