Enforcing Single Responsibility Principle in Python

Enforcing Single Responsibility Principle in Python

<strong>Single Responsibility Principle (or SRP) is one of the most important concepts in software development. The main idea of this concept is: all pieces of software must have only a single responsibility.</strong>

Single Responsibility Principle (or SRP) is one of the most important concepts in software development. The main idea of this concept is: all pieces of software must have only a single responsibility.

Why SRP is important? It is the main idea that stands behind software development. Decompose complex tasks to the set of simple building blocks to compose complex software from them back again. Just like we can compose lego or builtin functions:

print(int(input('Input number: ')))

This article will guide you through a complex process of writing simple code. I personally consider this article rather complicated and hard to percept if you do not have a solid background in python, so it is split into several parts:

  1. Definition of simple building block
  2. Problems with the functional composition in python
  3. Introduction to callable objects to solve functional composition problems
  4. Dependency injection reduces the boilerplate code of callable objects

It is okay to stop after each piece, revise it and to get to the latter one, because it will ensure you get the point, or at least first proof-readings showed so.

Please do not hesitate to suggest edits or to ask questions. This topic is not covered so well and I will be glad to clarify anything.

Defining building blocks

Let’s start with defining what are these “pieces of software” and “simplest building blocks” I am talking about?

The simplest building blocks are usually language’s expressions and statements. We can literally compose everything from it. But, we cannot fully rely on them since they are too simple. And we do not want to have repeated code all over the place. So we invent functions to abstract these simplest language constructs into something more meaningful that we can actually work with.

We expect these simplest building blocks (read “functions”) to be composable. And to be easily composable they must respect Single Responsibility Principle. Otherwise, we would have troubles. Because you can not compose things that do several things when you only need a part of them.

Functions can be complex too

Now, let’s make sure we can really rely on functions as simple building blocks.

We probably already know that functions can grow complex too and we have all seen absolutely unreadable functions like this one:

def create_objects(name, data, send=False, code=None):
data = [r for r in data if r[0] and r[1]]
keys = ['{}:{}'.format(*r) for r in data]

existing_objects = dict(Object.objects.filter(
    name=name, key__in=keys).values_list('key', 'uid'))

with transaction.commit_on_success():
    for (pid, w, uid), key in izip(data, keys):
        if key not in existing_objects:
            try:
                if pid.startswith('_'):
                    result = Result.objects.get(pid=pid)
                else:
                    result = Result.objects.filter(
                        Q(barcode=pid) | Q(oid=pid)).latest('created')
            except Result.DoesNotExist:
                logger.info("Can't find result [%s] for w [%s]", pid, w)
                continue

            try:
                t = Object.objects.get(name=name, w=w, result=result)
            except:
                if result.container.is_co:
                    code = result.container.co.num
                else:
                    code = name_code
                t = Object.objects.create(
                    name=name, w=w, key=key,
                    result=result, uid=uid, name_code=code)

                reannounce(t)

                if result.expires_date or (
                      result.registry.is_sending
                      and result.status in [Result.C, Result.W]):
                    Client().Update(result)

            if not result.is_blocked and not result.in_container:
                if send:
                    if result.status == Result.STATUS1:
                        Result.objects.filter(
                            id=result.id
                        ).update(
                             status=Result.STATUS2,
                             on_way_back_date=datetime.now())
                    else:
                        started(result)

        elif uid != existing_objects[key] and uid:
            t = Object.objects.get(name=name, key=key)
            t.uid = uid
            t.name_code = name_code
            t.save()
            reannounce(t)

It really works and powers someone’s production system. However, we can still say that this function definitely has more than one responsibility and should be refactored. But how do we make this decision?

There are different formal methods to track functions like this, including:

After we apply these methods it would be clear to us that this function is too complex. And we won’t be able to compose it easily. It is possible (and recommended) to go further and to automate this process. That’s how code-quality tools work with wemake-python-styleguide as a notable example.

Just use it. It will detect all the hidden complexity and will not allow your code to rot.

Here’s the less obvious example of a function that does several things and breaks SRP (and, sadly, things like that can not be automated at all, code reviews are the only way to find this kind of issues):

def calculate_price(products: List[Product]) -> Decimal:
"""Returns the final price of all selected products (in rubles)."""
price = 0
for product in products:
price += product.price

logger.log('Final price is: {0}', price)
return price

Look at this logger variable. How did it make its way into the function’s body? It is not an argument. It is just a hard-coded behavior. What if I do not want to log this specific price for some reason? Should I disable it with an argument flag?

In case I will try to do that, I will end up with something like this:

def calculate_price(products: List[Product], log: bool = True) -> ...
...

Congratulations, we now have a well-known anti-pattern in our code. Do not use boolean flags. They are bad.

Moreover, how do I test this function? Without this logger.log call it would be a perfectly testable pure function. Something goes in and I can predict what will go out. And now it is impure. To test that logger.log actually works I would have to mock it somehow and assert that log was created.

You may argue that logger in python has a global configuration just for this case. But, it is a dirty solution to the same problem.

Such a mess just because of a single line! The problem with this function is that it is hard to notice this double responsibility. If we rename this function from calculate_price to proper calculate_and_log_price it would become obvious that this function does not respect SRP. And the rule is that simple: if “correct and full” function name contains andor, or then – it is a good candidate for refactoring.

Ok, this is all scary and stuff, but what to do with this case in general? How can we change the behavior of this function so it will finally respect SRP?

I would say that the only way to achieve SRP is composition: compose different functions together so each of them would do just one thing, but their composition would do all the things we want.

Let’s see different patterns that we can use to compose functions in python.

Decorators

We can use the decorator pattern to compose functions together.

@log('Final price is: {0}')
def calculate_price(...) -> ...:
...

What consequences this pattern has?

  1. It not just composes, but glues functions together. This way you won’t have an ability to actually run just calculate_price without log
  2. It is static. You can not change things from the calling point. Or you have to pass arguments to the decorator function before actual function parameters
  3. It creates visual noise. When the number of decorators will grow – it would pollute our functions with a huge amount of extra lines

All in all, decorators make perfect sense in specific situations while are not suited for others. Good examples are: @login_required@contextmanager, and friends.

Functional composition

It is quite similar to the decorator pattern, the only exception is that it is applied in the runtime, not “import” time.

from logger import log

def controller(products: List[Product]):
final_price = log(calculate_price, message='Price is: {0}')(products)
...

  1. With this approach, we can easily call functions the way we actually want to call them: with our without logpart
  2. On the other hand, it creates more boilerplate and visual noise
  3. It is hard to refactor due to the higher amount of the boilerplate and since you delegate composition to the caller instead of the declaration

But, it also works for some cases. For example, I use @safe function all the time:

from returns.functions import safe

user_input = input('Input number: ')

The next line won't raise any exceptions:

safe_number = safe(int)(user_input)

You can read more about why exceptions might be harmful to your business logic in a separate article. We also provide a utility type-safe compose function in returns library that you might use for composing things at runtime.

Passing arguments

We can always just pass arguments. As easy as that!

def calculate_price(
products: List[Product],
callback=Callable[[Decimal], Decimal],
) -> Decimal:
"""Returns the final price of all selected products (in rubles)."""
price = 0
for product in products:
price += product.price

return callback(price)

And then we can invoke it:

from functools import partial

from logger import log

price_log = partial(log, 'Price is: {0}')
calculate_price(products_list, callback=price_log)

And it works great. Now our function does not know a thing about logging. It only calculates the price and returns the callback of it. We can now supply any callback, not just log. It might be any function that receives one Decimaland returns one back:

def make_discount(price: Decimal) -> Decimal:
return price * 0.95

calculate_price(products_list, callback=make_discount)

See? No problem, just compose functions the way you like. The hidden disadvantage of this method is in the nature of function arguments. We must explicitly pass them. And if the call-stack is huge, we need to pass a lot of parameters to different functions. And potentially cover different cases: we need callback A in case of a and callback B in case of b.

Of course, we can try to patch them somehow, create more functions that return more functions or pollute our code with @inject decorators everywhere, but I think that is ugly.

Unsolved problems:

  1. Mixed logic arguments and dependency arguments, because we pass them together at the same time and it hard to tell what is what
  2. Explicit arguments that can be hard or impossible to maintain if your call-stack is huge

To fix these problems, let me introduce you to the concept of callable objects.

Separating logic and dependencies

Before we start discussing callable objects, we need to discuss objects and OOP in general keeping SRP in mind. I see a major problem in OOP just inside its main idea: “Let’s combine data and behavior together”. For me, it is a clear violation of SRP, because objects by design do two things at once: they contain their state and have perform some attached behavior. Of course, we will eliminate this flaw with callable objects.

Callable objects look like regular objects with two public methods: init and call. And they follow specific rules that make them unique:

  1. Handle only dependencies in the constructor
  2. Handle only logic arguments in the call method
  3. No mutable state
  4. No other public methods or any public attributes
  5. No parent classes or subclasses

The straight-forward way to implement a callable object is something like this:

class CalculatePrice(object):
def init(self, callback: Callable[[Decimal], Decimal]) -> None:
self._callback = callback

def __call__(self, products: List[Product]) -&gt; Decimal:
    price = 0
    for product in products:
        price += product.price
    return self._callback(price)

The main difference between callable objects and functions is that callable objects have an explicit step for passing dependencies, while functions mix regular logic arguments with dependencies together (you can already notice that callable objects are just a special case of a partial function application):

# Regular functions mix regular arguments with dependencies:
calculate_price(products_list, callback=price_log)

Callable objects first handle dependencies, then regular arguments:

CalculatePrice(price_log)(products_list)

But, given example do not follow all rules we impose on callable objects. In particular, they are mutable and can have subclasses. Let’s fix that too:

from typing_extensions import final

from attr import dataclass

@final
@dataclass(frozen=True, slots=True)
class CalculatePrice(object):
_callback: Callable[[Decimal], Decimal]

def __call__(self, products: List[Product]) -&gt; Decimal:
    ...

Now with the addition of @final decorator that restricts this class to be subclassed and @dataclass decorator with frozen and slots properties our class respects all the rules we impose in the beginning.

  1. Handle only dependencies in the constructor. True, we have only declarative dependencies, the constructor is created for us by attrs
  2. Handle only logic arguments in the call method. True, by definition
  3. No mutable state. True, since we use frozen and slots
  4. No other public methods or any public attributes. Mostly true, we cannot have public attributes by declaring slots property and declarative protected instance attributes, but we still can have public methods. Consider using a linter for this
  5. No parent classes or subclasses. True, we explicitly inherit from object and marking this class final, so any subclasses will be restricted

It now may look like an object, but it is surely not a real object. It can not have any state, public methods, or attributes. But, it is great for Single Responsibility Principle. First of all, it does not have data and behavior. Just pure behavior. Secondly, it is hard to mess things up this way. You will always have a single method to call in all the objects that you have. And this is what SRP is all about. Just make sure that this method is not too complex and does one thing. Remember, no one stops you from creating protected methods to decompose call behavior.

However, we have not fixed the second problem of passing dependencies as arguments to functions (or callable objects): noisy explicitness.

Dependency injection

DI pattern is widely known and used outside of the python world. But, for some reason is not very popular inside it. I think that this is a bug that should be fixed.

Let’s see a new example. Imagine that we have postcards sending app. Users create postcards to send them to other users on specific dates: holidays, birthdays, etc. We are also interested in how many of them were sent for analytic purposes. Let’s see how this use-case will look like:

from project.postcards.repository import PostcardsForToday
from project.postcards.services import (
SendPostcardsByEmail,
CountPostcardsInAnalytics,
)

@final
@dataclass(frozen=True, slots=True)
class SendTodaysPostcardsUsecase(object):
_repository: PostcardsForToday
_email: SendPostcardsByEmail
_analytics: CountPostcardInAnalytics

def __call__(self, today: datetime) -&gt; None:
    postcards = self._repository(today)
    self._email(postcards)
    self._analytics(postcards)

Next, we have to invoke this callable class:

# Injecting dependencies:
send_postcards = SendTodaysPostcardsUsecase(
PostcardsForToday(db=Postgres('postgres://...')),
SendPostcardsByEmail(email=SendGrid('username', 'pass')),
CountPostcardInAnalytics(source=GoogleAnalytics('google', 'admin')),
)

Actually invoking postcards send:

send_postcards(datetime.now())

The problem is clearly seen in this example. We have a lot of dependencies-related boilerplate. Every time we create an instance of SendTodaysPostcardsUsecase – we have to create all its dependencies. Going all the way deep.

And all this boilerplate seems redundant. We have already specified all types of expected dependencies in our class. And transitive dependencies in our class’s dependencies, and so on. Why do we have to duplicate this code once again?

Actually, we don’t have to. We can use some kind of DI framework. I can personally recommend dependencies or punq. Their main difference is in how they resolve dependencies: dependencies uses names and punq uses types. We would go with punq for this example.

Do not forget to install it:

pip install punq

Now our code can be simplified so we won’t have to mess with dependencies. We create a single place where all the dependencies are registered:

# project/implemented.py

import punq

container = punq.Container()

Low level dependencies:

container.register(Postgres)
container.register(SendGrid)
container.register(GoogleAnalytics)

Intermediate dependencies:

container.register(PostcardsForToday)
container.register(SendPostcardsByEmail)
container.register(CountPostcardInAnalytics)

End dependencies:

container.register(SendTodaysPostcardsUsecase)

And then use it everywhere:

from project.implemented import container

send_postcards = container.resolve(SendTodaysPostcardsUsecase)
send_postcards(datetime.now())

There’s literally no repeated boilerplate, readability, and type-safety out-of-the-box. We now do not have to manually wire any dependencies together. They will be wired by annotations by punq. Just type your declarative fields in callable objects the way you need, register dependencies in the container, and you are ready to go.

Of course, there are some advanced typing patterns for better Inversion of Control, but it is better covered in punq's docs.

When not to use callable objects

It is quite obvious that all programming concepts have their limitations.

Callable objects should not be used in the infrastructure layer of your application. Since there are too many existing APIs that do not support this kind of classes and API. Use it inside your business logic to make it more readable and maintainable.

Consider adding returns library to the mix, so you can get rid of exceptions as well.

Conclusion

We came a long way. From absolutely messy functions that do scary things to simple callable objects with dependency injection that respect Single Responsibility Principle. We have discovered different tools, practices, and patterns along the way.

But did our efforts make a big change? The most important question to ask yourself: is my code better after all this refactoring?

My answer is: yes. This made a significant change for me. I can compose simple building blocks into complex use-cases with ease. It is typed, testable, and readable.

Originally published by Nikita Sobolev at https://sobolevn.me/2019/03/enforcing-srp

Learn More

☞ Complete Python Bootcamp: Go from zero to hero in Python 3

☞ Complete Python Masterclass

☞ Learn Python by Building a Blockchain & Cryptocurrency

☞ Python and Django Full Stack Web Developer Bootcamp

☞ The Python Bible™ | Everything You Need to Program in Python

☞ Learning Python for Data Analysis and Visualization

☞ Python for Financial Analysis and Algorithmic Trading

☞ The Modern Python 3 Bootcamp

Python GUI Programming Projects using Tkinter and Python 3

Python GUI Programming Projects using Tkinter and Python 3

Python GUI Programming Projects using Tkinter and Python 3

Description
Learn Hands-On Python Programming By Creating Projects, GUIs and Graphics

Python is a dynamic modern object -oriented programming language
It is easy to learn and can be used to do a lot of things both big and small
Python is what is referred to as a high level language
Python is used in the industry for things like embedded software, web development, desktop applications, and even mobile apps!
SQL-Lite allows your applications to become even more powerful by storing, retrieving, and filtering through large data sets easily
If you want to learn to code, Python GUIs are the best way to start!

I designed this programming course to be easily understood by absolute beginners and young people. We start with basic Python programming concepts. Reinforce the same by developing Project and GUIs.

Why Python?

The Python coding language integrates well with other platforms – and runs on virtually all modern devices. If you’re new to coding, you can easily learn the basics in this fast and powerful coding environment. If you have experience with other computer languages, you’ll find Python simple and straightforward. This OSI-approved open-source language allows free use and distribution – even commercial distribution.

When and how do I start a career as a Python programmer?

In an independent third party survey, it has been revealed that the Python programming language is currently the most popular language for data scientists worldwide. This claim is substantiated by the Institute of Electrical and Electronic Engineers, which tracks programming languages by popularity. According to them, Python is the second most popular programming language this year for development on the web after Java.

Python Job Profiles
Software Engineer
Research Analyst
Data Analyst
Data Scientist
Software Developer
Python Salary

The median total pay for Python jobs in California, United States is $74,410, for a professional with one year of experience
Below are graphs depicting average Python salary by city
The first chart depicts average salary for a Python professional with one year of experience and the second chart depicts the average salaries by years of experience
Who Uses Python?

This course gives you a solid set of skills in one of today’s top programming languages. Today’s biggest companies (and smartest startups) use Python, including Google, Facebook, Instagram, Amazon, IBM, and NASA. Python is increasingly being used for scientific computations and data analysis
Take this course today and learn the skills you need to rub shoulders with today’s tech industry giants. Have fun, create and control intriguing and interactive Python GUIs, and enjoy a bright future! Best of Luck
Who is the target audience?

Anyone who wants to learn to code
For Complete Programming Beginners
For People New to Python
This course was designed for students with little to no programming experience
People interested in building Projects
Anyone looking to start with Python GUI development
Basic knowledge
Access to a computer
Download Python (FREE)
Should have an interest in programming
Interest in learning Python programming
Install Python 3.6 on your computer
What will you learn
Build Python Graphical User Interfaces(GUI) with Tkinter
Be able to use the in-built Python modules for their own projects
Use programming fundamentals to build a calculator
Use advanced Python concepts to code
Build Your GUI in Python programming
Use programming fundamentals to build a Project
Signup Login & Registration Programs
Quizzes
Assignments
Job Interview Preparation Questions
& Much More

Guide to Python Programming Language

Guide to Python Programming Language

Guide to Python Programming Language

Description
The course will lead you from beginning level to advance in Python Programming Language. You do not need any prior knowledge on Python or any programming language or even programming to join the course and become an expert on the topic.

The course is begin continuously developing by adding lectures regularly.

Please see the Promo and free sample video to get to know more.

Hope you will enjoy it.

Basic knowledge
An Enthusiast Mind
A Computer
Basic Knowledge To Use Computer
Internet Connection
What will you learn
Will Be Expert On Python Programming Language
Build Application On Python Programming Language

Python Programming Tutorials For Beginners

Python Programming Tutorials For Beginners

Python Programming Tutorials For Beginners

Description
Hello and welcome to brand new series of wiredwiki. In this series i will teach you guys all you need to know about python. This series is designed for beginners but that doesn't means that i will not talk about the advanced stuff as well.

As you may all know by now that my approach of teaching is very simple and straightforward.In this series i will be talking about the all the things you need to know to jump start you python programming skills. This series is designed for noobs who are totally new to programming, so if you don't know any thing about

programming than this is the way to go guys Here is the links to all the videos that i will upload in this whole series.

In this video i will talk about all the basic introduction you need to know about python, which python version to choose, how to install python, how to get around with the interface, how to code your first program. Than we will talk about operators, expressions, numbers, strings, boo leans, lists, dictionaries, tuples and than inputs in python. With

Lots of exercises and more fun stuff, let's get started.

Download free Exercise files.

Dropbox: https://bit.ly/2AW7FYF

Who is the target audience?

First time Python programmers
Students and Teachers
IT pros who want to learn to code
Aspiring data scientists who want to add Python to their tool arsenal
Basic knowledge
Students should be comfortable working in the PC or Mac operating system
What will you learn
know basic programming concept and skill
build 6 text-based application using python
be able to learn other programming languages
be able to build sophisticated system using python in the future

To know more: