multiarg-dispatch 0.1.0

A replacement function for single dispatch that allows dispatching based on type hints for all arguments.

multiarg-dispatch

multiarg-dispatch is a Python package that extends the functionality of functools.singledispatch to allow dispatching based on type hints for all arguments, not just the first one. It provides a simple decorator to define generic functions with multiple implementations depending on argument types.

Features

• Dispatch functions based on the types of all arguments, including keyword arguments.
• Supports union types in type hints.
• Raises a warning if arguments have default values (since defaults are not considered during dispatch).
• Type checking enforced at registration: all parameters must have type hints.
• Fully compatible with Python >3.11.
• Note that the registry uses strong references, so for garbage collection do not forget to delete the function that uses multidispatch.
• Note that local classes cannot be used as type hints, since these are accessible globally to retrieve as type hint.

Installation

Install from PyPI:

pip install multiarg-dispatch

Or with uv:

uv add multiarg-dispatch

Usage

from multiarg_dispatch import multidispatch, DispatchWarning

@multidispatch
def test_func(a, b=None):
    return "default"

@test_func.register
def _(a: int, b: str = "default") -> str:
    return f"int:{a},str:{b}"

@test_func.register
def _(a: str, b: list = []) -> str:
    if b is None:
        b = []
    return f"str:{a},list:{b}"

@test_func.register
def _(a: float, b: str | list = "default") -> str:
    return f"float:{a},union:{b}"

print(test_func(10))            # int:10,str:default
print(test_func("hello", []))   # str:hello,list:[]
print(test_func(3.14, "extra")) # float:3.14,union:extra

API

multidispatch(func)

Decorator to make a function multi-dispatch capable.

• register(func): Register a new implementation based on type hints.
• dispatch(cls): Retrieve the implementation for given types.
• registry: Read-only view of all registered implementations.

DispatchWarning

Warning raised when dispatching might be affected by defaults.

Contributing

Contributions are welcome! Please follow these guidelines:

1. Fork the repository.
2. Create a feature branch (git checkout -b feature/my-feature).
3. Make your changes.
4. Ensure all tests pass (pytest recommended).
5. Open a Pull Request with a clear description of your changes.

Please follow PEP 8 style guidelines and include type hints where appropriate.

License

This project is licensed under the MIT License. See the LICENSE file for details.