What Interface Segregation Principle (ISP) is?

The Interface Segregation Principle is one of the SOLID principles in software engineering. The ISP states that clients should not be forced to depend on interfaces they do not use. To better understand the above statement let's take a look at the following example.

Imagine you have a big box with many different toys inside. Each toy has a specific purpose and functionality. Now let's say you are a child who wants to play with a particular toy from that box.

The Interface Segregation Principle (ISP) is like saying that you, as the child, shouldn't be forced to take the entire big box of toys when you only want to play with one specific toy. Instead, you should be given the option to pick and choose the toys you want to play with.

In the context of software development, think of the big box of toys as a software interface that provides multiple methods or functionalities. The toys inside represent the individual methods or functionalities provided by that interface.

The ISP tells us that if a client (a piece of code or a class) only needs to use a specific set of methods from an interface, it shouldn't be forced to depend on the entire interface. It should be able to choose and depend only on the methods it needs.

Code Example

Now, let's take the previous example a step further to illustrate the differences and demonstrate how applying the Interface Segregation Principle (ISP) can help us. I will provide code examples, starting with a poorly implemented version without using ISP and then followed by an improved version utilizing ISP. This will highlight the benefits of using ISP in our code design.

Poorly Implemented Version (without ISP):

// ToyBox with a big interface containing all toy functionalities
public interface IToyBox
{
    void PlayWithCarToy();
    void PlayWithDollToy();
    void PlayWithTrainToy();
    // ... and more toy functionalities
}

public class ToyBox : IToyBox
{
    public void PlayWithCarToy()
    {
        Console.WriteLine("Playing with a car toy...");
    }

    public void PlayWithDollToy()
    {
        Console.WriteLine("Playing with a doll toy...");
    }

    public void PlayWithTrainToy()
    {
        Console.WriteLine("Playing with a train toy...");
    }

    // ... implementations for other toy functionalities
}

public class Child
{
    private readonly IToyBox toyBox;

    public Child(IToyBox toyBox)
    {
        this.toyBox = toyBox;
    }

    public void Play()
    {
        // The child class is forced to depend on the entire ToyBox interface,
        // even though they only want to play with a specific toy.
        toyBox.PlayWithCarToy();
    }
}

In the above poorly implemented version, the ToyBox class has a big interface containing all toy functionalities. The Child class depends on this entire interface. even though we only want to play with a car toy. This violates the ISP since the client is forced to depend on interfaces they do not use.

Restructured Version (using ISP):

// Smaller, more focused interfaces for individual toy functionalities
public interface ICarToy
{
    void PlayWithCarToy();
}

public interface IDollToy
{
    void PlayWithDollToy();
}

public interface ITrainToy
{
    void PlayWithTrainToy();
}

public class CarToy : ICarToy
{
    public void PlayWithCarToy()
    {
        Console.WriteLine("Playing with a car toy...");
    }
}

public class DollToy : IDollToy
{
    public void PlayWithDollToy()
    {
        Console.WriteLine("Playing with a doll toy...");
    }
}

public class TrainToy : ITrainToy
{
    public void PlayWithTrainToy()
    {
        Console.WriteLine("Playing with a train toy...");
    }
}

public class Child
{
    private readonly ICarToy carToy;

    public Child(ICarToy carToy)
    {
        this.carToy = carToy;
    }

    public void Play()
    {
        // The child class can now depend only on the specific
        // toy interface they want to play with,
        // adhering to the Interface Segregation Principle (ISP).
        carToy.PlayWithCarToy();
    }
}

In this restructured version, we create smaller more focused interfaces representing individual toy functionalities. The Child class can now depend only on the specific toy interface they want to play with, aligning with the ISP. This allows for better code organization, reduced dependencies and improved flexibility in extending or modifying the codebase.

Conclusion

In conclusion, adhering to the Interface Segregation Principle (ISP) brings forth a multitude of benefits. By following ISP, we can achieve a codebase that exhibits reduced coupling, improved modularity, enhanced maintainability, better testability, and flexible extensibility.

By reducing coupling, ISP minimizes dependencies between modules or classes, leading to code that is more independent and modular.

Additionally, ISP improves maintainability by limiting the impact of changes. Clients only rely on the specific methods they use, so modifications to other parts of the codebase do not affect them. This reduces the risk of unintended side effects and makes the system easier to maintain and evolve.

Furthermore, ISP enhances testability by enabling focused unit testing. Smaller more specific interfaces allow for targeted tests.

Lastly, ISP enables flexible extensibility by introducing new interfaces and implementing them as needed, we can easily add new functionality without modifying existing code.

To sum up, by embracing the Interface Segregation Principle (ISP), we foster a software design characterized by loose coupling, modularity, maintainability, testability and adaptability.