Python Type Hints are awesome


I really love Julia Evans’s blog, she writes with such excitement and enthusiasm about every new bit of technology she learns. For me, this is what makes our job so great - we can learn something new and cool almost every day.

And recently I’ve felt very excited with some of new abilities Python now have. I’m talking about including of PEP 484, which describes Type Hints, in Python 3.5 (old news, I know).

I’m a big fan of static analysis (I love when computer helps me to spot mistakes and bugs, and have even written a Linter for HAProxy configs), so when I’ve found that Python 3.5 has gained standard ability to specify types for function parameters, return values and local variables, I was eager to test it.

Just to show what I’m talking about, here are some basic example of Type Hints:

# without Type Hints
def greeting(name):
    return 'Hello, {}'.format(name)
# with Type Hints
def greeting(name: str) -> str:
    return 'Hello, {}'.format(name)

and with generics:

from typing import List

def greeting(names: List[str]) -> str:
    return 'Hello, {}'.format(', '.join(names))

Fortunately, PyCharm IDE already has full support of PEP 484, so I’ve immediately started adding Type Hints to my pet project “lyricstagger”.

After some experiments and adding types to actual code I can safely conclude that Type Hints are awesome: they let us express all the information (in a standard way, yay!) about function parameters and variable types that we already hold in our mind, so that IDE or linters can help us by checking that we are passing correct values to our functions. Using Type Hints feels much more natural to me than using one of several ways of specifying type information in docstrings.

I was writing a lot of code in Go programming language last year, and got used to all the benefits of static typing. Now I can be more confident in my Python code, because static analysis with Type Hints can catch many common mistakes even before I run any tests.

And you don’t need to stick with Python 3.5 to use type hinting, the whole typing module was published in pip and can be easily installed and used with Python 3.2+.

If you are as excited about Type Hints in Python as I am, you can read more about them in JetBrains PyCharm blog.

Tags: Python Programming

Categories: IT