bobbie

Version
Status
Documentation
Tools
Compatibility
Stats

---

What is bobbie?

Named after Roberta Draper ("Bobbie") from "The Expanse," bobbie provides a lightweight and easy-to-use way to store and access configuration settings for a Python project.

Why use bobbie?

There are numerous options for storing configuration settings in Python. So, what makes bobbie different?

• Flexible: a Settings instance is easily built from a dict, Python module, or file (ini, json, toml, and yaml formats are supported).
• Lightweight: an efficient codebase ensures a very small memory footprint.
• Intuitive: a create class method constructs Settings (from any data source).
• Convenient: unlike configparser, automatic data and type validation is performed when Settings is created.

The comparison table below shows how bobbie compares to the other major options that store configuration options for Python projects.

Getting started

Installation

To install bobbie, use pip:

pip install bobbie

Create a Settings instance

bobbie supports several ways to create a Settings instance. You can either: * If the settings are in a file, call a class method specific to the file type (from_yaml in the second example below); * Call a class method for the general type of source data (from_dict or from_file); * Call the create class method and it will automatically figure out the appropriate constructor; or * If the settings data is in a dict-like object, pass it as the first argument to Settings.

From a dict

import bobbie

configuration = {
  'general': {
    'verbose': False,
    'seed': 43,
    'parallelize': False},
  'files': {
    'source_format': 'csv',
    'file_encoding': 'windows-1252',
    'float_format': '%.4f',
    'export_results': True}}
# You can pick a specific constructor method.
settings = bobbie.Settings.from_dict(configuration)
# Or, may use the general `create` method.
settings = bobbie.Settings.create(configuration)
# Or, just send a `dict` to `Settings` itself.
settings = bobbie.Settings(configuration)

From a file

import bobbie

# You may use the general `create` method.
settings = bobbie.Settings.create('settings_file.yaml')
# Or, the `from_file` method.
settings = bobbie.Settings.from_file('settings_file.yaml')
# Or, the `from_yaml` method. They all do the same thing.
settings = bobbie.Settings.from_yaml('settings_file.yaml')

If the file is a Python module, it must contain a variable named settings in the module namespace (unless you change the global setting for the variable name).

Type inference

ini and json files often store numbers, booleans, and lists as strings, so bobbie converts their values to the matching Python types:

• '42' becomes 42 and '2.5' becomes 2.5.
• 'true', 'yes', 'false', and 'no' (in any case) become True or False.
• '[x, y, z]' and 'x, y, z' become ['x', 'y', 'z'].
• '{a: 1, b: [x, y]}' becomes {'a': 1, 'b': ['x', 'y']}.

Items in lists and dicts are converted the same way and may be nested. Put an item in quotes (such as ["a, b", '1']) to keep it as a string.

toml, yaml, and Python module values are not converted because those formats already store their own types. So, a toml value of "[a, b]" stays a string.

Contributing

Contributors are always welcome. Feel free to grab an issue to work on or make a suggested improvement. If you wish to contribute, please read the Contribution Guide and Code of Conduct.

Similar projects

There are a lot of great packages for storing project settings. The table below shows the features of the leading libraries.

Feature Comparison of Python Configuration Libraries

Library	Typing	Secrets	dict	ini	json	py	toml	yaml
bobbie	✅		✅	✅	✅	✅	✅	✅
configParser				✅
dynaconf	✅	✅		✅	✅		✅	✅
Parser-it	✅			✅	✅		✅	✅
python-decouple	✅	✅		✅
pyconfig		✅
pydantic-settings	✅	✅

As you can see, bobbie lacks a method for storing passwords, encryption keys, and other secrets. That is because it is focused on internal Python projects and the goal of keeping its resource usage as low as possible. So, if you need secrets stored in your project settings, there are several good options linked above that you should explore.

License

Use of this repository is authorized under the Apache Software License 2.0.