Mastering Python: 7 Methods for Writing Clear, Organized, and Environment friendly Code

Picture by Creator

Have you ever ever in contrast your Python code to that of skilled builders and felt a stark distinction? Regardless of studying Python from on-line assets, there’s usually a niche between newbie and expert-level code. That is as a result of skilled builders adhere to greatest practices established by the group. These practices are sometimes missed in on-line tutorials however are essential for large-scale purposes. On this article, I can be sharing 7 ideas that I take advantage of in my manufacturing code for clearer and extra organized code.

1. Sort Hinting and Annotations

 
Python is a dynamically typed programming language, the place the variable varieties are inferred at runtime. Whereas it permits for flexibility, it considerably reduces code readability and understanding in a collaborative setting.

Python gives assist for kind hinting in perform declarations that function an annotation of the perform argument varieties and the return varieties. Despite the fact that Python does not implement these varieties throughout runtime, it is nonetheless useful as a result of it makes your code simpler to know for different individuals (and your self!).

Beginning with a fundamental instance, right here is an easy perform declaration with kind hinting:
 

def sum(a: int, b: int) -> int:
	return a + b

Right here, regardless that the perform is pretty self-explanatory, we see that the perform parameters and return values are denoted as int kind. The perform physique may very well be a single line, as right here, or a number of hundred traces. But, we will perceive the pre-conditions and return varieties simply by wanting on the perform declaration.

It is vital to know that these annotations are only for readability and steerage; they do not implement the categories throughout execution. So, even in the event you go in values of various varieties, like strings as a substitute of integers, the perform will nonetheless run. However be cautious: in the event you do not present the anticipated varieties, it’d result in surprising habits or errors throughout runtime. For example, within the supplied instance, the perform sum() expects two integers as arguments. However in the event you attempt to add a string and an integer, Python will throw a runtime error. Why? As a result of it does not know add a string and an integer collectively! It is like making an attempt so as to add apples and oranges – it simply does not make sense. Nevertheless, if each arguments are strings, it’ll concatenate them with none situation.

Here is the clarified model with take a look at instances:
 

print(sum(2,5)) # 7
# print(sum('hi there', 2)) # TypeError: can solely concatenate str (not "int") to str
# print(sum(3,'world')) # TypeError: unsupported operand kind(s) for +: 'int' and 'str'
print(sum('hi there', 'world')) # helloworld 

Typing Library for Superior Sort Hinting

 
For superior annotations, Python contains the typing normal library. Allow us to see its use in a extra attention-grabbing strategy.
 

from typing import Union, Tuple, Record
import numpy as np

def sum(variable: Union[np.ndarray, List]) -> float:
	whole = 0
	# perform physique to calculate the sum of values in iterable
	return whole

 
Right here, we alter the identical summation perform that now accepts a numpy array or checklist iterable. It computes and returns their sum as a floating-point worth. We make the most of the Union annotation from the typing library to specify the potential varieties that the variable parameter can settle for.

Allow us to additional change the perform declaration to indicate that the checklist members must also be of kind float.
 

def sum(variable: Union[np.ndarray, List[float]]) -> float:
	whole = 0
	# perform physique to calculate the sum of values in iterable
	return whole

These are just a few newbie examples to assist perceive kind hinting in Python. As tasks develop, and codebases change into extra modular, kind annotations considerably improve readability and maintainability. The typing library gives a wealthy set of options together with Optionally available, varied iterables, Generics, and assist for custom-defined varieties, empowering builders to precise complicated information buildings and relationships with precision and readability.

2. Writing Defensive Capabilities and Enter Validation

 
Despite the fact that type-hinting appears useful, it’s nonetheless error-prone because the annotations will not be enforced. These are simply additional documentation for the builders however the perform will nonetheless be executed if totally different argument varieties are used. Due to this fact, there’s a must implement the pre-conditions for a perform and code in a defensive method. Therefore, we manually verify these varieties and lift acceptable errors if the circumstances are violated.

The under perform exhibits how curiosity is calculated utilizing the enter parameters.
 

def calculate_interest(principal, fee, years):
	return principal * fee * years

 
It’s a easy operation, but will this perform work for each potential answer? No, not for the sting instances the place the invalid values are handed as enter. We have to make sure that the enter values are certain inside a legitimate vary for the perform to execute appropriately. In essence, some pre-conditions should be glad for the perform implementation to be appropriate.

We do that as follows:
 

from typing import Union

def calculate_interest(
	principal: Union[int, float],
	fee: float,
	years: int
) -> Union[int, float]:
	if not isinstance(principal, (int, float)):
    	    increase TypeError("Principal should be an integer or float")
	if not isinstance(fee, float):
    	    increase TypeError("Charge should be a float")
	if not isinstance(years, int):
    	    increase TypeError("Years should be an integer")
	if principal <= 0:
    	    increase ValueError("Principal should be constructive")
	if fee <= 0:
    	    increase ValueError("Charge should be constructive")
	if years <= 0:
    	    increase ValueError("Years should be constructive")

	curiosity = principal * fee * years
	return curiosity

Be aware, that we use conditional statements for enter validation. Python additionally has assertion statements which can be generally used for this goal. Nevertheless, assertions for enter validation will not be a greatest apply as they’ll disabled simply and can result in surprising behaviour in manufacturing. Using express Python conditional expressions is preferable for enter validation and implementing pre-conditions, post-conditions, and code invariants.

3. Lazy Loading with Turbines and Yield Statements

Think about a state of affairs, the place you might be supplied with a big dataset of paperwork. You might want to course of the paperwork and carry out sure operations on every doc. Nevertheless, because of the giant dimension, you can’t load all of the paperwork in reminiscence and pre-process them concurrently.

A potential answer is to solely load a doc in reminiscence when required and course of solely a single doc at a time, additionally referred to as lazy loading. Despite the fact that we all know what paperwork we’ll want, we don’t load a useful resource till it’s required. There isn’t any must retain the majority of paperwork in reminiscence when they don’t seem to be in lively use in our code. That is precisely how turbines and yield statements strategy the issue.

Turbines enable lazy-loading that improves the reminiscence effectivity of Python code execution. Values are generated on the fly as wanted, lowering reminiscence footprint and rising execution pace.
 

import os

def load_documents(listing):
	for document_path in os.listdir(listing):
    	    with open(document_path) as _file:
        	        yield _file

def preprocess_document(doc):
	filtered_document = None
	# preprocessing code for the doc saved in filtered_document
	return filtered_document

listing = "docs/"
for doc in load_documents(listing):
	preprocess_document(doc)

 
Within the above perform, the load_documents perform makes use of the yield key phrase. The tactic returns an object of kind <class generator>. Once we iterate over this object, it continues execution from the place the final yield assertion is. Due to this fact, a single doc is loaded and processed, enhancing Python code effectivity.

4. Stopping Reminiscence Leaks utilizing Context Managers

For any language, environment friendly use of assets is of major significance. We solely load one thing in reminiscence when required as defined above by using turbines. Nevertheless, it’s equally vital to shut a useful resource when it’s not wanted by our program. We have to stop reminiscence leaks and carry out correct useful resource teardown to save lots of reminiscence.

Context managers simplify the frequent use case of useful resource setup and teardown. It is very important launch assets when they don’t seem to be required anymore, even in case of exceptions and failures. Context managers cut back the chance of reminiscence leaks utilizing automated cleanup whereas retaining the code concise and readable.

Assets can have a number of variants akin to database connections, locks, threads, community connections, reminiscence entry, and file handles. Let’s concentrate on the only case: file handles. The problem right here is making certain that every file opened is closed precisely as soon as. Failure to shut a file can result in reminiscence leaks, whereas making an attempt to shut a file deal with twice leads to runtime errors. To handle this, file handles needs to be wrapped inside a try-except-finally block. This ensures that the file is closed correctly, no matter whether or not an error happens throughout execution. Here is how the implementation would possibly look:
 

file_path = "instance.txt"
file = None

attempt:
	file = open(file_path, 'r')

	contents = file.learn()
	print("File contents:", contents)

lastly:
	if file just isn't None:
    	file.shut()

 
Nevertheless, Python gives a extra elegant answer utilizing context managers, which deal with useful resource administration robotically. Here is how we will simplify the above code utilizing the file context supervisor:
 

file_path = "instance.txt"
with open(file_path, 'r') as file:
	contents = file.learn()
	print("File contents:", contents)

On this model, we needn’t explicitly shut the file. The context supervisor takes care of it, stopping potential reminiscence leaks.

​​Whereas Python gives built-in context managers for file dealing with, we will additionally create our personal for {custom} lessons and features. For sophistication-based implementation, we outline __enter__ and __exit__ dunder strategies. Here is a fundamental instance:
 

class CustomContextManger:
	def __enter__(self):
    	    # Code to create occasion of useful resource
    	    return self

	def __exit__(self, exc_type, exc_value, traceback):
    	    # Teardown code to shut useful resource
     	    return None

 
Now, we will use this practice context supervisor inside ‘with’ blocks:

with CustomContextManger() as _cm:
	print("Customized Context Supervisor Useful resource might be accessed right here")

 
This strategy maintains the clear and concise syntax of context managers whereas permitting us to deal with assets as wanted.

5. Separation of Concern with Decorators

 
We frequently see a number of features with the identical logic applied explicitly. It is a prevalent code scent, and extreme code duplication makes the code troublesome to keep up and unscalable. Decorators are used to encapsulate related performance in a single place. When the same performance is for use by a number of different features, we will cut back code duplication by implementing frequent performance inside a decorator. It follows Side-Oriented Programming (AOP) and the Single Accountability precept.

Decorators are closely used within the Python net frameworks akin to Django, Flask and FastAPI. Let me clarify the effectiveness of decorators through the use of it as a middleware in Python for logging. In a manufacturing setting, we have to know the way lengthy it takes to service a request. It’s a frequent use case and can be shared throughout all endpoints. So, allow us to implement a easy decorator-based middleware that can log the time taken to service a request.

The dummy perform under is used to service a person request.
 

def service_request():
	# Perform physique representing complicated computation
	return True

Now, we have to log the time it takes for this perform to execute. A method is so as to add logging inside this perform as follows:
 

import time

def service_request():
	start_time = time.time()
	# Perform physique representing complicated computation
	print(f"Time Taken: {time.time() - start_time}s")
	return True

 
Whereas this strategy works, it results in code duplication. If we add extra routes, we would need to repeat the logging code in every perform. This will increase code duplication as this shared logging performance must be added to every implementation. We take away this with using decorators.

The logging middleware can be applied as under:
 

def request_logger(func):
	def wrapper(*args, **kwargs):
    	    start_time = time.time()
    	    res = func()
    	    print(f"Time Taken: {time.time() - start_time}s")
    	    return res
	return wrapper

 
On this implementation, the outer perform is the decorator, which accepts a perform as enter. The interior perform implements the logging performance, and the enter perform known as inside the wrapper.

Now, we merely enhance the unique service_request perform with our request_logger decorator:
 

@request_logger
def service_request():
	# Perform physique representing complicated computation
	return True

 
Utilizing the @ image passes the service_request perform to the request_logger decorator. It logs the time taken and calls the unique perform with out modifying its code. This separation of issues permits us to simply add logging to different service strategies in the same method like this:
 

@request_logger
def service_request():
	# Perform physique representing complicated computation
	return True

@request_logger
def service_another_request():
	# Perform physique
	return True

6. Match Case Statements

Match statements have been launched in Python3.10 so it’s a pretty new addition to the Python syntax. It permits for easier and extra readable sample matching, stopping extreme boilerplate and branching within the typical if-elif-else statements.

For pattern-matching, match case statements are the extra pure manner of writing it as they don’t essentially must return boolean values as in conditional statements. The next instance from the Python documentation exhibits how match case statements supply flexibility over conditional statements.
 

def make_point_3d(pt):
	match pt:
    	    case (x, y):
        		return Point3d(x, y, 0)
    	    case (x, y, z):
        		return Point3d(x, y, z)
    	    case Point2d(x, y):
        		return Point3d(x, y, 0)
    	    case Point3d(_, _, _):
        		return pt
    	    case _:
        		increase TypeError("not a degree we assist")

 
As per the documentation, with out sample matching, this perform’s implementation would require a number of isinstance() checks, one or two len() calls, and a extra convoluted management circulation. Underneath the hood, the match instance and the normal Python model translate into related code. Nevertheless, with familiarity with sample matching, the match case strategy is prone to be most popular because it gives a clearer and extra pure syntax.

Total, match case statements supply an improved different for sample matching, which is able to doubtless change into extra prevalent in newer codebases.

7. Exterior Configuration Information

In manufacturing, the vast majority of our code depends on exterior configuration parameters like API keys, passwords, and varied settings. Hardcoding these values straight into the code is taken into account poor apply for scalability and safety causes. As an alternative, it is essential to maintain configurations separate from the code itself. We generally obtain this utilizing configuration information akin to JSON or YAML to retailer these parameters, making certain they’re simply accessible to the code with out being straight embedded inside it.

An on a regular basis use case is database connections which have a number of connection parameters. We are able to preserve these parameters in a separate YAML file.
 

# config.yaml
database:
  host: localhost
  port: 5432
  username: myuser
  password: mypassword
  dbname: mydatabase

To deal with this configuration, we outline a category referred to as DatabaseConfig:
 

class DatabaseConfig:
	def __init__(self, host, port, username, password, dbname):
    	    self.host = host
    	    self.port = port
    	    self.username = username
    	    self.password = password
    	    self.dbname = dbname

	@classmethod
	def from_dict(cls, config_dict):
    	    return cls(**config_dict)

Right here, the from_dict class methodology serves as a builder methodology for the DatabaseConfig class, permitting us to create a database configuration occasion from a dictionary.

In our major code, we will make use of parameter hydration and the builder methodology to create a database configuration. By studying the exterior YAML file, we extract the database dictionary and use it to instantiate the config class:
 

import yaml

def load_config(filename):
	with open(filename, "r") as file:
    	return yaml.safe_load(file)

config = load_config("config.yaml")
db_config = DatabaseConfig.from_dict(config["database"])

 
This strategy eliminates the necessity for hardcoding database configuration parameters straight into the code. It additionally gives an enchancment over utilizing argument parsers, as we not must go a number of parameters each time we run our code. Furthermore, by accessing the config file path by an argument parser, we will make sure that the code stays versatile and does not depend on hardcoded paths. This methodology facilitates simpler administration of configuration parameters, which might be modified at any time with out requiring adjustments to the codebase.

Ending Notes

 
On this article, we mentioned a number of the greatest practices used within the business for production-ready code. These are frequent business practices that alleviate a number of issues one can face in real-life conditions.

Nonetheless, it’s value noting that regardless of all such greatest practices, documentation, docstrings, and test-driven growth are by far essentially the most important practices. It is very important take into consideration what a perform is meant to do after which doc all design selections and implementations for the longer term as individuals engaged on a codebase change over time. In case you have any insights or practices you swear by, please don’t hesitate to tell us within the remark part under.
 
 

Kanwal Mehreen Kanwal is a machine studying engineer and a technical author with a profound ardour for information science and the intersection of AI with medication. She co-authored the book “Maximizing Productiveness with ChatGPT”. As a Google Technology Scholar 2022 for APAC, she champions variety and educational excellence. She’s additionally acknowledged as a Teradata Range in Tech Scholar, Mitacs Globalink Analysis Scholar, and Harvard WeCode Scholar. Kanwal is an ardent advocate for change, having based FEMCodes to empower ladies in STEM fields.