An Introduction To Closures and Decorators in Python
Table of Contents
The article explains the concepts of Python closures and decorators for the reader. Earthly offers a reliable method to automate and containerize Python build processes. Learn more about Earthly.
Python is a high-level general-purpose language that supports classes as part of its built-in object-oriented programming (OOP) paradigm. Occasionally, when working with variables in Python, you may want to hide a variable without writing an unnecessary class to keep the code more maintainable. Moreover, you may want to add minor functionality to a function without creating another redundant function. In such scenarios, closures, and decorators can be used as solutions: closures for variable hiding and decorators for function modification. This guide will demonstrate the concept of closures and decorators, explain how and where to use them, and explain when it’s most appropriate to use them instead of traditional classes and unnecessary functions.
At the outset, it is imperative to mention that closures and decorators might look similar owing to the fact that both involve functions, but they serve different conceptual purposes.
Closures: A closure is a function object that remembers values in the enclosing scope, even if they are not present in memory. A closure is created when a nested function references variables from its enclosing scope. The closure “closes over” the variables it references, thus preserving their values even if the enclosing scope is no longer active. Closures are often used to implement data encapsulation and maintain the state between function calls.
Decorators: A decorator is a way to modify the behavior of a function or a class without directly changing its source code. It allows adding functionality to an existing function or class by wrapping it with another function or class. Decorators are commonly used for tasks such as logging, timing, input validation, authentication, and more. They provide a method to separate concerns and keep the code modular and reusable.
Before diving deeper into the concept of closures and decorators, let’s start with simple pedagogical examples of both using code.
A Simple Closure Example
Put simply, a closure is a function that can remember and access the values of variables from its surrounding environment, even after the execution of the outer function has finished.
In Python, closures can be created by defining a function inside another function and returning the inner function. Here is an example:
def outer_function(x):
def inner_function(y):
return x + y
return inner_function
= outer_function(5)
closure print(closure(3)) # Output: 8
In the example provided above, the ‘outer_function’ takes an argument ‘x’ and defines an inner function called ‘inner_function’ that takes another argument ‘y’ and returns the sum of ‘x’ and ‘y.’
The ‘outer_function’ then returns the inner function ‘inner_function.’ Since ‘inner_function’ is a nested function, it has access to the variables in the enclosing scope of ‘outer_function’, which in this case is the variable ‘x.’
When ‘outer_function’ is called with the argument ‘5’, it returns the function ‘inner_function’, which remembers the value of ‘x’ as ‘5’. This function object is assigned to the variable closure
.
Finally, closure
is called with argument ‘3’, which adds ‘3’ to the value of ‘x’ (which is ‘5’), resulting in output ‘8.’
Therefore, the output of the code is ‘8’, which demonstrates that the inner function ‘inner_function’ has retained the value of ‘x’ from the outer scope, even after the execution of the ‘outer_function’ has finished.
Why and How to Use Closures
Closures are used in Python for various reasons, such as:
Implementing data hiding: In Python, closures can be used to hide data within a function by defining a variable inside the outer function and then using it in the inner function. This helps in encapsulating the data and prevents it from being modified from outside the function.
Implementing decorators: Closures are also used to implement decorators in Python. Decorators are functions that modify the behavior of other functions.
Now you will learn how data hiding can be simplified with the use of closures compared to the class method. Therefore a comparative approach will be followed to explain it better. To start with, here is an example of how to use a class to hide data in Python
class SecureData:
def __init__(self, data):
self.data = data
self.password = 'secret'
def get_data(self, passwd):
if passwd == self.password:
return self.data
else:
return None
= SecureData('my sensitive data')
secure_data
print(secure_data.get_data('secret')) # Output: 'my sensitive data'
print(secure_data.get_data('wrong password')) # Output: None
The SecureData
class has an __init__
method that initializes two instance variables: data
and password
.
The get_data
method is defined within the SecureData
class, which takes a password as a parameter. It compares the provided password with the password
instance variable of the class. If they match, it returns the data
instance variable. Otherwise, it returns None
.
To retrieve the secure data, an instance of the SecureData
class is created (secure_data
). You can then call the get_data
method on the secure_data
instance, passing the password as an argument.
The output of this code is my sensitive data
when the correct password is provided and None
when an incorrect password is provided.
By using a class, the data and associated behavior are encapsulated within the class methods. The instance variables (data
and password
) are accessible and modifiable only through the defined methods of the class, providing data hiding and access control.
Now let us move to the closures. Here is an example of how to use closures to hide data in Python:
def create_secure_data(data):
= 'secret'
password
def get_data(passwd):
if passwd == password:
return data
else:
return None
return get_data
= create_secure_data('my sensitive data')
secure_data
# Now, the 'secure_data' variable contains a reference to
# the inner function 'get_data' which can access the 'password'
# and 'data' variables of the outer function 'create_secure_data'.
# To retrieve the secure data, you need to call the 'secure_data'
# function with the correct password.
print(secure_data('secret')) # Output: 'my sensitive data'
print(secure_data('wrong password'))
In this example, the create_secure_data
function takes in some data and returns an inner function get_data
. The get_data
function takes in a password as an argument and checks if it matches the password variable defined in the outer function. If the password is correct, the get_data
function returns the enclosed data variable, otherwise, it returns None
. By using closures in this way, you have hidden the data variable and only allow access to it if the correct password is provided. This provides a simple way of implementing data hiding in Python.
What Are Decorators
A decorator is a higher-order function that takes another function as an argument, adds some functionality to it, and returns a new function without modifying the original function’s source code. Decorators allow you to modify the behavior of functions or classes by wrapping them inside another function. Here is an example:
def decorator_function(func):
def wrapper_function():
print("Before function is called.")
func()print("After function is called.")
return wrapper_function
@decorator_function
def hello():
print("Hello, world!")
hello()
Run the above code and you will receive an output similar to this:
Before function is called.
Hello, world!
After function is called.
In the example provided above, you define a decorator function ‘decorator_function’ that takes a function ‘func’ as an argument and returns a new function ‘wrapper_function.’ The ‘wrapper_function’ adds some functionality to the original function ‘func.’ The ‘@decorator_function’ syntax is used to decorate the ‘hello’ function with the ‘decorator_function’ decorator. When ‘hello’ is called, it is actually calling the ‘wrapper_function’ returned by ‘decorator_function.’ This allows modifying the behavior of hello
without changing its source code directly.
When ‘hello’ is called, it prints “Before function is called.” using the print function, then it calls the original func (which in this case is print(“Hello, world!”)), and finally it prints “After function is called.”
Exploring Decorators in Detail
In Python, decorators enable meta-programming, which refers to a programming technique in which the code can manipulate other code at either compile-time or run-time. So as an example of metaprogramming technique, decorators provide a means to modify the behavior of functions or classes without modifying their source code.
Here’s an example of how to use decorators for meta-programming in Python:
def debug(func):
def wrapper(*args, **kwargs):
print(f"Calling {func.__name__} with args: {args}, \
kwargs: {kwargs}")
= func(*args, **kwargs)
result print(f"{func.__name__} returned: {result}")
return result
return wrapper
@debug
def add(a, b):
return a + b
= add(3, 5)
result print(result)
In this example, you define a decorator called debug
that adds debugging information to the decorated function. The debug
decorator wraps the original function (add
) with a closure function called wrapper
. The wrapper
function prints the function name, arguments, and keyword arguments before calling the original function. It also prints the returned result after executing the original function.
By using the @debug
syntax, you can apply the debug
decorator to your add
function. This enhances your add
function with debugging information without modifying its original implementation.
When you call add(3, 5)
, the debug
decorator intercepts the function call, prints the debugging information, executes the original add
function, captures the result, prints the result, and returns it.
So, the output of the example would be:
Calling add with args: (3, 5), kwargs: {}
add returned: 8
8
Now, to test the add
function further, let us feed the keyword arguments to the add
function. For example, run the following modified code:
= add(a="pin", b="point")
result2 print(result2)
And you will get the following output, printing the keys and :
Calling add with args: (), kwargs: {'a': 'pin', 'b': 'point'}
add returned: pinpoint
pinpoint
You see once you created your debug decorator it can be applied to any function which takes arguments or keyword arguments.
In summary, decorators are a powerful feature in Python that can be used for meta-programming, enabling you to modify the behavior of functions and classes without altering their source code.
Having understood the meta-programming example, you should turn to another use, i.e., runtime modification of a function.
Runtime Modification of a Class Using Decorators in Python
In Python, decorators can be used for runtime modification of a class as well. Decorators can add or modify behavior of a class without modifying its source code. Here you will understand how to implement the singleton pattern using decorators. In software design, the singleton pattern is a pattern where only a single instance of a class is instantiated and used throughout the designed system. Here’s an example code:
def make_singleton(cls):
= {}
instances
def wrapper(*args, **kwargs):
if cls not in instances:
= cls(*args, **kwargs)
instances[cls] return instances[cls]
return wrapper
@make_singleton
class DatabaseConnection:
def __init__(self, url):
self.url = url
print("Initializing database connection.")
def execute_query(self, query):
print(f"Executing query: {query} on database: {self.url}")
# Create two instances of DatabaseConnection
= DatabaseConnection("https://example.com/db")
connection1 = DatabaseConnection("https://example.com/db")
connection2
print(connection1 is connection2) # Output: True
In this example, you define a decorator called make_singleton
that transforms a class into a singleton. The make_singleton
decorator wraps the class with a closure function called wrapper
. The wrapper
function maintains a dictionary instances
to keep track of instances of the class. When creating a new instance, the decorator checks if an instance of the class already exists. If it does, the existing instance is returned. Otherwise, a new instance is created and stored in the instances
dictionary.
By using the @make_singleton
syntax, you can apply the make_singleton
decorator to your desired class. This modifies the behavior of the class by ensuring that only a single instance of the class is created and shared.
When you create multiple instances of the class, the decorator intercepts the class instantiation and returns the same instance for all of them. This is because the decorator converts the class into a singleton, allowing only one instance to be created. As a result, all references to the class will point to the same object.
This approach is useful when you need to have a single instance of a class shared across different parts of your program. By applying the make_singleton
decorator, you can ensure that there is only one instance of the class throughout your codebase.
This approach can be used to add any kind of behavior to a function or class at runtime. For example, you can use a decorator to add logging, error handling, performance monitoring, or any other functionality that you want to apply to multiple functions.
Decorators for Timing Purposes in Python
In Python, decorators can be used for timing purposes to measure the execution time of a function. Timing a function can be useful for optimizing its performance or measuring its efficiency. Here’s an example of how to use decorators for timing purposes:
import time
def timer(func):
def wrapper(*args, **kwargs):
= time.time()
start_time = func(*args, **kwargs)
result = time.time()
end_time print(f"Execution time: {end_time - start_time}")
return result
return wrapper
@timer
def my_function():
2)
time.sleep(
my_function()
Run the above code, you will receive an output similar to this:
Execution time: 2.0001144409179688
You observe that applying the decorator @timer to my_funcation()
also prints the time duration of the execution of the function.
Applying Multiple Decorators to a Single Function
You can apply multiple decorators to a single function in Python. When you apply multiple decorators to a function, the decorators are applied from the inside out. That is, the innermost decorator is applied first, followed by the next innermost, and so on until the outermost decorator is applied. Here’s an example of how to apply multiple decorators to a single function:
def decorator1(func):
def wrapper():
print("Before decorator1")
func()print("After decorator1")
return wrapper
def decorator2(func):
def wrapper():
print("Before decorator2")
func()print("After decorator2")
return wrapper
@decorator1
@decorator2
def my_function():
print("my_function")
my_function()
In the above example, two decorators, decorator1
and decorator2
, are defined, which add some print statements before and after the decorated function is called. Then, both decorators are applied to the my_function
function using the @decorator1
and @decorator2
syntax.
When you call my_function()
, the output will be:
Before decorator1
Before decorator2
my_function
After decorator2
After decorator1
As you can see, the decorators are applied from the inside out, so decorator2
is applied first, followed by decorator1
.
Performing Type Checking of Function Parameters
As the most advanced application of decorators in Python, you can also use decorators to perform type-checking of function parameters. This approach can be useful if you want to add type-checking to existing functions without modifying their source code. Here’s an example of how to use a decorator to do type checking of function parameters:
def type_check(func):
def wrapper(*args, **kwargs):
# iterate over the function arguments and their types
for arg, arg_type in zip(args, func.__annotations__.values()):
# check if the argument type is correct
if not isinstance(arg, arg_type):
raise TypeError(f"Argument {arg} has incorrect type \
{type(arg)}")
# call the original function with the given arguments
return func(*args, **kwargs)
return wrapper
This decorator function takes the original function as an argument and returns a new wrapper function that can take any number of positional and keyword arguments *args
and **kwargs
. It then iterates over the function arguments and their types, which are specified using function annotations. For each argument, it checks if the argument type is correct using the isinstance()
function. If the argument type is incorrect, it raises a TypeError
with a helpful error message. If all the arguments are of the correct type, it calls the original function with the given arguments and returns its result.
Here’s an example usage of this decorator:
@type_check
def add(x: int, y: int) -> int:
return x + y
print(add(1, 2)) # Output: 3
print(add("1", "2"))
# Raises: TypeError: Argument 1 has incorrect type <class 'str'>
In this example, the add()
function is decorated with the type_check()
decorator. The function takes two integer arguments, x
and y
, and returns their sum as an integer. The function annotations specify the argument types and the return type of the function. When you call the add()
function with two integer arguments, it returns their sum as expected. However, when you call the function with two string arguments, the decorator raises a TypeError
with a helpful error message indicating that the first argument has an incorrect type.
This approach can be used to add type-checking to any function that you want to modify. You can apply the type_check
decorator to any function that you want to check, and it will automatically perform type-checking on its arguments.
In summary, decorators can be used for performing type checking of function parameters in Python. This approach allows you to add type checking to existing functions without modifying their source code and can be helpful in catching type errors before the code is executed.
You have seen multiple uses of decorators above. So it can be summarised that decorators are a powerful feature in Python that can be used for runtime modification of a function or class, allowing you to add or modify behavior without changing their source code.
Conclusions
Closures and decorators in Python let you write flexible and potent code. Closures remember values from their outer scope even when that’s done running. Decorators let you tweak functions or classes without messing with their source code. Put closures and decorators together for even more magic, boosting your code’s functionality and reusability.
Also if you’re looking to further streamline your Python app builds, you might want to give Earthly a try! It help optimize your build process, making it even more efficient and effective.
Earthly Cloud: Consistent, Fast Builds, Any CI
Consistent, repeatable builds across all environments. Advanced caching for faster builds. Easy integration with any CI. 6,000 build minutes per month included.