Simple configuration system based on recursively merging dicts.
Zenconf is an unopinionated config management system. You put dicts in, it recursively merges them, then returns a single dict with values from later dicts overwriting values from earlier dicts.
A default implementation shows how to pull values from a dictionary of defaults, a config file, environment variables and command line parameters. Simply create your own custom config class and add values from whatever data sources you like in your chosen order of precedence. Then, request the merged dictionary and use it throughout the rest of your code.
- Simple. Just add dicts (or OrderedDicts) from wherever you like, and get the merged dictionary back.
- No constraints on using a particular config file system, arg parser, etc.
- Key names can be normalised by applying a function to keys recursively to make for easier comparison between keys from different data sources (e.g. from environment variables or yaml files, where one uses underscores to separate words, the other hyphens). By default keys will be converted to lowercase, have hyphens converted to underscores and then have leading underscores removed.
- Support for filtering by, and stripping an app-specific prefix from keys (configurable per data source). This means that if your app is called MYAPP, only environment variables with a prefix of MYAPP_ could be added, e.g. MYAPP_LOG_LEVEL=debug could override a commandline argument –log-level.
Clone the repo then run ./setup.py install.
Or, install from pypi with:
pip install zenconf
Run tests with:
Usage is simple. Just instantiate an instance of zenconf.MergedConfig:
- The appname parameter can be used to namespace keys (such as environment variables).
- The dict_boundary parameter specifies a string that indicates that we should look up the next string fragment in a subdictionary. E.g. using a default of ‘__’, the string LOG__LEVEL refers to config[‘log’][‘level’].
Next, add dicts via the add() method.
To get the merged dict, just call get_merged_config().
See comments in the code for more information about parameters. Also see the example in the examples directory of one way to use the class.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.