Python @property Decorator

Introduction

A very crucial component of Object Oriented Programming is the concept of Getters, Setters, and Deleters. Python eases the use of these functions with the help of Property Decorators, and in this article, we’ll explore just that!

We’ll cover Getters, Setters, and Deleters, why we need them, and how we can use them to ensure ease of use for you and the developers who use your class in their projects.

Class Without Getters and Setters

The use of python property decorators can be best explained by first defining the problem.

Consider the following code –

We first define a class Person which has attributes firstName, middleName, lastName and fullname. An instance of Person is instantiated, and the first name of the person is changed to something else. Users of this class would expect the attribute fullname to change dynamically, but in this scenario, it remains the same, which makes the class practically unusable.

[Example 1]

This code gives the output:

As you can see, the attribute fullname did not get updated when we changed person1’s first name. An easy way of fixing this is by removing the line:

And by adding the function fullname() in the Person class

Now, any time we want to use the person's full name, we’ll have to call the function fullname(), which will return the value as a string.

The problem with this fix is that we’ll have to call the newly created fullname() function as a function instead of calling it as an attribute.

So, print statements will go from this:

To this:

If our codebase is small, making this change won’t be a problem, but many issues arise as our application scales to over a few thousand lines of code. People using your class will also have to change all instances of

to

Which is not practical. If you try to call the attribute directly, that is, without the parentheses, instead of calling the function, it will return the function itself, with its address –

To fix this in the ‘object oriented’ way, languages like Java use Getter functions. The good news is that we can implement Getter functions in Python with property decorators!

Now that you’ve been convinced that this topic is useful, we’ll go back to the basics, and build the concepts from the ground up.

What is property() in Python?

The property() method returns a property attribute from the defined Getter, Setter, and Deleter.

Before we get to Getters, Setters and Deleters, it would be useful to recall the scope of variables in object-oriented programming -

Variable Scope

When class attributes and functions are defined in the class, they can be defined as public, private, or protected variables/functions. In languages like C++ and Java, we cannot access private attributes outside the class they are defined in. Protected values of class X can be accessed by classes that inherit from class X.

In python, however, the concept of public, private, and protected attributes does not exist, so as a naming convention, variables that start with an underscore character are considered non-public variables. They can technically still be accessed outside the class, so keep that in mind!

Example of a non-private variable being accessed outside its class (this is to be avoided!)

[Example 2]

Output:

Getters, Setters, and Deleters

In object-oriented programming, when a class gets created by convention, the attributes are made private. It is a good practice to make them private, so they don’t get changed accidentally when called by objects.

These attributes can then be retrieved, manipulated or deleted by getters, setters, and deleters.

Declaring a property()

The property() function -

The property() method takes 4 arguments:

• fget – function that acts as a Getter
• fset – function that acts as a Setter
• fdel – function that acts as a Deleter
• doc – String will be parsed into a docstring that explains what the returned property attribute does (like a docstring below a function that explains the function).

All four are set to None by default.

The property() function returns a property object. On printing it, we get -

This property object gets invoked whenever the instantiated object fetches, changes, or deletes the attribute(s) outside the class. For example, instead of calling the Setter function setName(), we can directly do obj.name = "new name".

The following example will make this notion clearer. It implements the same logic from example 3 but with the property() method instead.

[Example 4]

Output: - Exactly the same as [Example 3]

Each attribute can have its property attribute. As an exercise, try to add a _fullname attribute and its respective getter, setter, and deleter with the property() function.

If the doc argument is not mentioned, the docstring under the Getter function is used for the entire property attribute.

Using the @property Decorator

A python decorator is a function that accepts another function as an argument.

A decorator's point is to enhance the function passed to it. It returns the modified function, which can then be called normally.

Decorators in python are very powerful and help make classes easier to understand and use. They help add new features to a python object without having to modify the structure of that object.

The @property decorator does the same thing as the property() method but makes the code cleaner and easier to read. The following code implements @property, but does the same as examples 3 and 4.

[Example 5]

Output: - Exactly the same as [Example 3]

As you can see, there’s no need to define a property attribute object using property(). The @property tells us exactly what each function does and whether it is a getter, setter, or deleter.

The @property decorator makes sense out of context provided and then applies the appropriate function. For example, the first time it is used, it is used in a print function, so it calls the object getter. The second time name is used when changing the person’s name from Mike to Jason. Although we have used the same call, person.name, it invokes the setter and passes ‘Jason’ as an argument to the setter. That’s pretty neat.

Now, armed with the @property decorator, we’ll wrap this article by solving the problem stated at the very beginning of the article

[Example 1] – the one involving fullname

The following steps will fix the problem using Python’s Property Decorator.

• Remove the fullname attribute from __init__()
• Create the fullname() function the usual way
• Add the line ‘@property’ above the function definition
• Call fullname as if you were calling the object’s attribute

[Example 6]

We’ve also defined the setter for fullname for you, so as an exercise try changing the first, middle and last names using the fullname setter.

Code Output:

And voila! That fixes our issue. Now all class users can continue using fullname the way they were using while getting the freedom to dynamically alter the full name by changing the person’s first, middle, or last name.