Skip to main content

Enforce annotations in your python code

Project description

Polyforce

Polyforce

🔥 Enforce static typing in your codebase at runtime 🔥

Test Suite Package version Supported Python versions


Documentation: https://polyforce.tarsild.io 📚

Source Code: https://github.com/tarsil/polyforce


Motivation

During software development we face issues where we don't know what do pass as specific parameters or even return of the functions itself.

Tools like mypy for example, allow you to run static checking in your code and therefore allowing to type your codebase properly but it does not enforce it when running.

For those coming from hevily static typed languages like Java, .net and many others, Python can be overwhelming and sometimes confusing because of its versatility.

Polyforce was created to make sure you:

  • Don't forget to type your functions and variables.
  • Validates the typing in runtime.
  • Don't forget thr return annotations.

Wouldn't be cool to have something like this:

What if my function that expects a parameter of type string, if something else is passed could simply fail, as intended?

This is where Polyforce enters.

The library

Polyforce was designed to enforce the static typing everywhere in your code base. From functions to parameters.

It was also designed to make sure the typing is enforced at runtime.

In other words, if you declare a type string and decide to pass an integer, it will blow throw and intended error.

The library offers two ways of implementing the solution.

Installation

$ pip install polyforce

How to use it

Let us see some scenarios where the conventional python is applied and then where Polyforce can make the whole difference for you.

Conventional Python

Let us start with a simple python function.

Simple function

def my_function(name: str):
    return name

In the normal python world, this wouldn't make any difference, and let us be honest, if you don't care about mypy or any related tools, this will work without any issues.

This will also allow this to run without any errors:

my_function("Polyfactory") # returns "Polyfactory"
my_function(1) # returns 1
my_function(2.0) # returns 2.0

The example above is 100% valid for that specific function and all values passed will be returned equaly valid and the reson for this is because Python does not enforce the static typing so the str declared for the parameter name is merely visual.

With objects

class MyClass:

    def my_function(self, name: str):
        return name

And then this will be also valid.

my_class = MyClass()

my_class.my_function("Polyfactory") # returns "Polyfactory"
my_class.my_function(1) # returns 1
my_class.my_function(2.0) # returns 2.0

I believe you understand the gist of what is being referred here. So, what if there was a solution where we actually enforce the typing at runtime? Throw some errors when something is missing from the typing and also when the wrong type is being sent into a function?

Enters Polyforce

Polyforce

Now, let us use the same examples used before but using Polyforce and see what happens?

Simple function

from polyforce import polycheck


@polycheck()
def my_function(name: str):
    return name

The example above it will throw a ReturnSignatureMissing or a MissingAnnotation because the missing return annotation of the function or a parameter annotation respectively.

my_function("Polyforce") # Throws an exception

The correct way would be:

from polyforce import polycheck


@polycheck()
def my_function(name: str) -> str:
    return name

So what if now you pass a value that is not of type string?

my_function(1) # Throws an exception

This will also throw a TypeError exception because you are trying to pass a type int into a declared type str.

With objects

The same level of validations are applied within class objects too.

from polyforce import PolyModel


class MyClass(PolyModel):

    def __init__(self, name, age: int):
        ...

    def my_function(self, name: str):
        return name

The example above it will throw a ReturnSignatureMissing and a MissingAnnotation because the missing return annotation for both init and the function as well as the missing types for the parameters in both.

The correct way would be:

from polyforce import PolyModel


class MyClass(PolyModel):

    def __init__(self, name: str, age: int) -> None:
        ...

    def my_function(self, name: str) -> str:
        return name

The Polyforce

As you can see, utilising the library is very simple and very easy, in fact, it was never so easy to enforce statuc typing in python.

For classes, you simply need to import the PolyModel.

from polyforce import PolyModel

And to use the decorator you simply can:

from polyforce import polycheck

PolyModel vs polycheck

When using PolyModel, there is no need to apply the polycheck decorator. The PolyModel is smart enough to apply the same level of validations as the polycheck.

When using the PolyModel you can use normal python as you would normally do and that means classmethod, staticmethod and normal functions.

This like this, the polycheck is used for all the functions that are not inside a class.

Limitations

For now, Polyforce is not looking at native magic methods (usually start and end with double underscore). In the future it is planned to understand those on a class level.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

polyforce-0.3.0.tar.gz (16.7 kB view hashes)

Uploaded Source

Built Distribution

polyforce-0.3.0-py3-none-any.whl (22.8 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page