Design Pattern

Behavioral Patterns

Author

Dr. Ashish Sai

Behavioral Design Patterns

Behavioral patterns in software design focus on effective communication and the assignment of responsibilities among objects.

Behavioral Patterns {style=“font-size: 40%;”}

Pattern Description Covered
Observer Defines a one-to-many dependency between objects so that when one object changes state, all its dependents are notified and updated automatically. ✅
Strategy Defines a family of algorithms, encapsulates each one, and makes them interchangeable. Strategy lets the algorithm vary independently from clients that use it. ✅
Command Encapsulates a request as an object, thereby allowing for parameterization of clients with queues, requests, and operations. ❌
State Allows an object to alter its behavior when its internal state changes. The object will appear to change its class. ❌
Chain of Responsibility Passes the request along the chain of handlers. Upon receiving a request, each handler decides either to process the request or to pass it to the next handler in the chain. ❌
Interpreter Provides a way to evaluate language grammar or expressions. The Interpreter pattern defines a grammar for the language, as well as an interpreter that uses the grammar to interpret sentences in the language. ❌
Memento Captures and externalizes an object’s internal state so the object can be restored to this state later. ❌
Visitor Represents an operation to be performed on the elements of an object structure. Visitor lets you define a new operation without changing the classes of the elements on which it operates. ❌
Template Method Defines the skeleton of an algorithm in the method, deferring some steps to subclasses. Template Method lets subclasses redefine certain steps of an algorithm without changing its structure. ❌

Strategy Pattern

Strategy Pattern

Composition over inheritance!

Strategy Pattern is defined as:

  • Defining a family of algorithms
  • Encapsulating each algorithm
  • Making them interchangeable

Inheritance vs. Composition

  • Inheritance is not always intended for code reuse.
  • Composition offers greater flexibility in many scenarios.
  • Strategy Pattern focuses on using composition over inheritance.

Problem Statement: Duck Example

  • Consider a system with different types of ducks.
  • Each duck type has its own display method.
  • Common methods like quack are shared.
public class Duck {
    public void quack() {
        // Common quack behavior
    }
    public abstract void display();
}

Problem Statement: Duck Example

  • We have different types of ducks: wild duck, city duck, rubber duck, lets add them one by one

Problem Statement: Duck Example

  • I want a rubber duck 🐥

Can RubberDucks fly?

Problem Statement: Duck Example

  • We found two new types of ducks: MountainDuck and CloudDuck, they fly at supersonic speed 1 in a ZigZag pattern.

Problem Statement: Duck Example

  • Now the Ducks want food, WildDuck and CityDuck eat rice but MountainDuck and CloudDuck eat only quinoa whereas RubberDuck eats ?

Introducing the Strategy Pattern

  • The Strategy Pattern allows the duck’s behaviors to vary independently.
  • Encapsulates quacking and flying behaviors.

Strategy Design Pattern: Intent of Strategy Pattern

The intent of the Strategy pattern is to define a set of interchangeable algorithms or strategies that can be selected at runtime according to the needs of the context or client.

Strategy Design Pattern: Problem and Solution

  • Problem: Need for a flexible way to incorporate different behaviors or algorithms within a class and the ability to change them at runtime.

  • Solution: The Strategy pattern suggests separating the behavior into different strategy classes and using a reference to these strategies in the context class.

Problem with Inheritance: Adding Fly Method

  • Adding fly method to Duck class leads to issues.
  • Not all ducks should fly (e.g., rubber ducks).
public class Duck {
    public void fly() {
        // Flying behavior
    }
}

Strategy Pattern Solution: Encapsulating Behaviors

  • Separate fly and quack behaviors into different strategies.
  • Each duck type can have its own flying and quacking behavior.

Implementing Duck Subclasses

  • Different types of ducks inherit from Duck class.

  • Each subclass implements its own display method.

public class MountainDuck extends Duck {
    public MountainDuck() {
        quackBehavior = new Quack();
        flyBehavior = new FlyWithWings();
    }
    public void display() {
        // MallardDuck specific display
    }
}

Advantages of Strategy Pattern

  • Promotes flexible code structure.

  • Allows behaviors to change dynamically.

  • Reduces dependency on inheritance.

Decoupling Behaviors

  • Behaviors are not hard-coded in the Duck class.

  • They can vary independently from the duck type.

public class Duck {
    FlyBehavior flyBehavior;
    QuackBehavior quackBehavior;

    public void performFly() {
        flyBehavior.fly();
    }

    public void performQuack() {
        quackBehavior.quack();
    }
}

Defining Behavior Interfaces

  • Define interfaces for each behavior.

Concrete Implementations

  • Implement different flying and quacking behaviors.

Strategy Pattern in Duck Subclasses

  • Subclasses of Duck can choose different behaviors.
public class RubberDuck extends Duck {
    public RubberDuck() {
        flyBehavior = new FlyNoWay();
        quackBehavior = new Squeak();
    }
    public void display() {
        // RubberDuck specific display
    }
}

Strategy Pattern: Flexibility

  • Easy to add new behaviors without modifying existing classes.
public class JetFlyingBehavior implements FlyBehavior {
    public void fly() {
        // Jet-powered flying
    }
}

Problem: Code Duplication in Inheritance

  • Inheritance can lead to duplicated code across subclasses.

Solving Code Duplication

  • Strategy Pattern avoids duplication by sharing behavior implementations.

Strategy Pattern in Context

  • Allows ducks to have various combinations of behaviors.
  • Easy to maintain and extend.

Strategy Design Pattern: Structure of Strategy Pattern {style=“font-size: 70%;”}

  • Context: Maintains a reference to a Strategy object and delegates it the algorithm execution.

  • Strategy: Common interface for all strategies defining the algorithm execution method.

  • ConcreteStrategy: Implements the algorithm using the Strategy interface.

Strategy Design Pattern: Implementation in Java: Context and Strategy

// Strategy Interface
interface Strategy {
    void executeAlgorithm();
}

// Context Class
class Context {
    private Strategy strategy;

    Context(Strategy strategy) {
        this.strategy = strategy;
    }

    void setStrategy(Strategy strategy) {
        this.strategy = strategy;
    }

    void executeStrategy() {
        strategy.executeAlgorithm();
    }
}

Strategy Design Pattern: Concrete Strategies in Java

// Concrete Strategy A
class ConcreteStrategyA implements Strategy {
    public void executeAlgorithm() {
        // Implement algorithm A
    }
}

// Concrete Strategy B
class ConcreteStrategyB implements Strategy {
    public void executeAlgorithm() {
        // Implement algorithm B
    }
}

Strategy Design Pattern: Applicability

Use the Strategy pattern when:

  • You have different variations of an algorithm and want to switch between them at runtime.

  • You want to avoid exposing complex, algorithm-specific data structures.

  • You want to replace inheritance with composition for behavioral variations.

Strategy Design Pattern: Pros and Cons {style=“font-size: 80%;”}

Pros:

  • Enables the Open/Closed Principle by allowing the introduction of new strategies without changing the context.

  • Simplifies unit testing by isolating algorithms.

Cons:

  • Increases the number of objects in the application.

  • Clients must be aware of the differences between strategies to select the right one.

Observer Pattern

Weather Station and Display

Understanding the Problem

  • Scenario: When an object changes its state, other objects need to be notified.
  • Challenge: Continuously checking (polling) the state of an object is inefficient.

Basics of Observer Pattern

  • Definition: A design pattern where an object, known as the subject, notifies a list of observers about its state changes.
  • Key Concept: Push vs. Pull notification (move from pull to push).

Observer Pattern: Intent

Observer pattern allows for the establishment of a subscription mechanism to notify multiple objects about any events that happen to the object they’re observing.

Observer Pattern: Problem and Solution

  • Problem: Managing knowledge about changes in a system’s state can be complex when multiple entities need updates.

  • Solution: Observer pattern offers a subscription model where subjects notify observers about changes, promoting decoupling and efficient data distribution.

UML Diagram: Basic Structure

Real-World Example: Weather Station

  • Observable: Weather Station measuring and updating weather data.
  • Observers: Displays (e.g., phone display, window display) showing updated weather.

UML Diagram: Weather Station Example

Java Implementation: Interfaces

public interface Observer {
  void update();
}

public interface Observable {
  void addObserver(Observer o);
  void removeObserver(Observer o);
  void notifyObservers();
}

Java Implementation: WeatherStation

public class WeatherStation implements Observable {
  private List<Observer> observers;
  private int temperature;

  // Methods implementation...
}

Java Implementation: PhoneDisplay

public class PhoneDisplay implements Observer {
  private WeatherStation weatherStation;

  public void update() {
    // Implementation...
  }
}

Advantages of Observer Pattern

  • Reduces Coupling: Observers are loosely coupled with the subject.
  • Real-time Update: Efficient update mechanism for state changes.

Observer Pattern: Push vs. Pull

  • Push Model: Subject sends detailed data to observers.

  • Pull Model: Observers request data from the subject.

UML Diagram: Push Model

UML Diagram: Pull Model

Java Implementation: Push Model

public interface Observer {
  void update(Object data);
}

public class ConcreteObserver implements Observer {
  public void update(Object data) {
    // Use data directly
  }
}

Java Implementation: Pull Model

public interface Observer {
  void update();
}

public class ConcreteObserver implements Observer {
  private ConcreteSubject subject;

  public void update() {
    Object data = subject.getState();
    // Use data
  }
}

Registering Observers

  • Observers must register themselves to the subject.
  • Allows dynamic addition and removal of observers.

Java Code: Observer Registration

public class Main {
  public static void main(String[] args) {
    WeatherStation station = new WeatherStation();
    PhoneDisplay display = new PhoneDisplay(station);
    station.addObserver(display);
  }
}

Benefits of Observer Pattern

  • Scalability: Easily add new observers without modifying the subject.
  • Flexibility: Supports both push and pull data models.

Observer Pattern: Limitations

  • Potential for Memory Leaks: Observers need to be explicitly removed.
  • Unexpected Updates: Observers might receive updates at unpredictable times.

Summary and Conclusion

  • Observer Pattern is crucial for state change notification in software design.
  • Offers a robust, scalable, and flexible solution for maintaining consistency across different parts of a system.
  • Suitable for various applications like UI, weather monitoring, and more.

See you in Lab! 👋🏼 {.LastSlide background-color=“#f3d8cb”}

Footnotes

  1. Really they don’t, don’t google it.↩︎