optional.py

An implementation of the Optional object in Python


Keywords
optional, datatype, library, clean-code, optionals, python, python27, python3
License
MIT
Install
pip install optional.py==2.0.0

Documentation

Optional.py

Build Status test lint typecheck format editorconfig License Python Versions Contributors Open Source Love Downloads

An Implementation of the Optional Object for Python

Why

There is a difference between None as Empty and None as the result for an Error. A common bad practice is to return None to indicate the absence of something. Doing this introduces ambiguity into you code.

For example:

thing = stuff.getSomeThing().getAnotherThing()

What will happen if the result from getSomeThing returns None? We will get an AttributeError: 'NoneType' object has no attribute 'getAnotherThing'.

What can you do to prevent these kinds of exceptions? You can write defensively:

something = stuff.getSomeThing()
if something is not None:
    thing = something.getAnotherThing()

However, if we add to our chain, you can imagine how the nesting of defensive checks adds up quickly. These defensive checks obfuscate our actual business logic, decreasing readability. Furthermore, defensive checking is an error prone process, because it is easy to forget to check a required condition.

So we present you with an Optional object as an alternative.

Install

Compatible with Python 3.10 and up!

pip install optional.py

Usage

  1. You can import it using:

    from optional import Nothing, Option, Optional, Something
  2. You can set it to empty:

    instead of: 🙀

    return None

    you can do: 😸

    return Optional.empty()

    or

    return Optional.of()
  3. You can set it to have content:

    instead of: 🙀

    return "thing"

    you can do: 😸

    return Optional.of("thing")
  4. You can check if its present:

    instead of: 🙀

    if thing is not None:

    you can do: 😸

    thing = some_func_returning_an_optional()
    if thing:
  5. You can check if its empty:

    instead of: 🙀

    if thing is None:

    you can do: 😸

    thing = some_func_returning_an_optional()
    if not thing:
  6. You can match against the result and destructure the value:

    instead of: 🙀

    print(thing)

    you can do: 😼

    match some_func_returning_an_optional():
        case Something(thing):
            print(thing)
  7. You can match against an empty optional, but can't destructure the value:

    instead of: 😿

    if thing is None:
        print(None) # very odd

    you can do: 😼

    match some_func_returning_an_optional()
        case Nothing():
            print("We didn't get a thing!")
  8. You can compare two optionals: 😸

    Optional.empty() == Optional.empty() # True
    Optional.of("thing") == Optional.of("thing") # True
    Optional.of("thing") == Optional.empty() # False
    Optional.of("thing") == Optional.of("PANTS") # False

Tests

There is complete test coverage and they pass in all Python versions 3.10 and up.

Running Unit Tests

First, install poetry using the instructions located here.

Then, install the requirements using:

poetry install

You can run the tests (with coverage) using:

poetry run pytest