options helps represent option and configuration data in a clean, high-function way. Changes can “overlay” defaults or earlier settings.

For most functions and classes, options is flexibility overkill. Not everyone wants to be a world-class gymnast, yogi, or contortionist. For most, Python’s regular function arguments, *args, **kwargs, and inheritance patterns are elegant and sufficient. options is for the top 1% that need:

  • extremely functional classes, functions, and methods,
  • with many different features and options,
  • the settings for which might be adjusted or overriden at any time,
  • yet that need “reasonable” or “intelligent” defaults, and
  • that yearn for a simple, unobtrusive API.

In those cases, Python’s built-in, inheritance-based model stops being the simple approach. Non-trivial argument-management code and complexity begins to pervade. This is where options‘s layered, delegation-based approach begins to shine. Almost regardless of how varied the options it wrangles, or how much flexibility is required, code complexity remains very flat.


Python has very flexible arguments for functions and methods, and good connection of values from classes to subclasses to methods. It doesn’t, however, connect those very well to configuration files, module defaults, method parameters, and other uses. options, in contrast, seamlessly connects all of these varied layers and cases.

For more backstory, see this StackOverflow.com discussion of how to combat “configuration sprawl”. For examples of options in use, see say, quoter, and show.