Typomancy is a Python package for parsing string inputs and converting them into data types expected by a Python function or class using type annotations.
The purpose is to facilitate the interaction between user inputs supplied through a public-facing interface (e.g., HTML page) and a Python-based backend. Since end users shouldn't need to know about expected datatypes, their inputs are loosely-structured strings. The typical solution would be for functions to parse the input and perform the typecasting. This is fine, but in the case of inherited classes, it seems cleaner to perform the typecasting generically to avoid requiring all children to do similar typecasting.
The package makes use of Python's typing library in addition to Python's built-in types (int, str, bool, etc.). Typomancy
performs opinionated typecasting for non-specific type annotations, casting the argument to a built-in type that satisfies the
annotation. For example, Collection and Iterable would cause the input to be cast to tuple. Similarly, Collection[str]
causes the input to be cast to a tuple with str elements.