Abstract class in Python

In Python, an abstract class is a class that is meant to be subclassed, but not instantiated. It is used as a template or blueprint for other classes, which inherit its properties and methods.

An abstract class is defined using the abc module, which stands for “Abstract Base Classes”. This module provides the infrastructure for defining abstract classes and abstract methods in Python. An abstract class can be defined by inheriting the ABC class or by decorating a class with the @abc.ABC decorator.

Here’s an example of how to define an abstract class in Python:

import abc

class Animal(metaclass=abc.ABCMeta):
    @abc.abstractmethod
    def speak(self):
        pass

In this example, we define an abstract class Animal using the abc module. The @abc.abstractmethod decorator marks the speak method as abstract, which means that any subclass of Animal must implement this method. The pass statement inside the speak method is a placeholder, indicating that the method does nothing but must be overridden in subclasses.

Here’s an example of how to define a subclass of the abstract Animal class:

class Dog(Animal):
    def speak(self):
        return "Woof!"

In this example, we define a Dog class that inherits from the abstract Animal class. We implement the speak method in Dog by returning the string “Woof!”.

Now we can create an instance of the Dog class and call its speak method:

dog = Dog()
print(dog.speak())  # "Woof!"

In this example, we create a new instance of the Dog class and call its speak method, which returns the string “Woof!”. Since Dog is a subclass of Animal, it is required to implement the speak method defined in the abstract Animal class.

The main use case for abstract classes in Python is to define a set of methods that subclasses should implement. This allows for a more structured and consistent approach to creating classes that share a common set of properties and behaviors. Abstract classes can help to prevent errors and ensure that all subclasses implement the required methods.