Module Hygiene

Simple tools to help with Python namespace hygiene!

Description

This is an opinionated project! I prefer Python modules to only include what's necessary for users. If I'm in an IDE and type from module import <TAB>, I want only API-level elements to be available! Many Python projects use a leading underscore to specify private-like objects and modules. Still, it's sometimes nice to keep things simple without multiple _module definitions in a package.

This package is my solution! It currently provides a single function, cleanup, which returns code to delete all locals which are not in a provided "export list".

Usage

See the example usage below. Note that we need to import T to add a proper typed signature for the function square. Unfortunately, this means that T is also available to anyone who imports this module.

By default, cleanup assumes the name of the module's export list is '__export__'. If you want to choose a different export list name, pass that name as a str to cleanup!

"""module.py

A python module which exports a single function, `square`.
"""

from hygiene import cleanup
from typing  import T 

__export__ = [
    "square",
]

def square(x: T) -> T:
    """Returns the square of x!"""
    return x ** 2

if __name__ != "__main__":
    cleanup()

This cleanup approach will likely change how you write Python modules. If you need a Python package throughout your module, like numpy, you likely import numpy as np at the top of your module, and use np in various functions throughout the module. If you cleanup at the end of the module, your code will break on execution because np will no longer exist!

If you choose to use cleanup, then you will need to import modules at the function level. Personally, I like this better anyways! Every dependency is right next to where it's used. One major downside of this approach is you need to parse your source code for import statements to track all dependencies.

"""another.py

Another module which exports a single function, `abssqrt`!
"""

from hygiene import cleanup
from typing  import T

# Don't do this!
# from numpy import abs, sqrt

__export_list__ = [
    "abssqrt",
]

def abssqrt(x: T) -> T:
    """Returns the square root of the absolute value of `x`!"""
    from numpy import abs, sqrt
    return sqrt(abs(x))

if __name__ != "__main__":
    cleanup(export = __export_list__)