Principios de programación orientada a objetos en Java: conceptos de programación orientada a objetos para principiantes

La programación orientada a objetos ofrece una forma sostenible de escribir código espagueti. Le permite acumular programas como una serie de parches.

- Paul Graham

Fundamentos de la programación orientada a objetos

La programación orientada a objetos es un paradigma de programación donde todo se representa como un objeto.

Los objetos se transmiten mensajes entre sí. Cada objeto decide qué hacer con un mensaje recibido. OOP se centra en los estados y comportamientos de cada objeto.

¿Qué son los objetos?

Un objeto es una entidad que tiene estados y comportamientos.

Por ejemplo, perro, gato y vehículo. Para ilustrarlo, un perro tiene estados como edad, color, nombre y comportamientos como comer, dormir y correr.

El estado nos dice cómo se ve el objeto o qué propiedades tiene.

El comportamiento nos dice lo que hace el objeto.

De hecho, podemos representar un perro del mundo real en un programa como un objeto de software definiendo sus estados y comportamientos.

Los objetos de software son la representación real de objetos del mundo real. La memoria se asigna en RAM siempre que se crea un objeto lógico.

Un objeto también se refiere a una instancia de una clase. Instanciar una clase significa lo mismo que crear un objeto.

Lo importante a recordar al crear un objeto es: el tipo de referencia debe ser del mismo tipo o un super tipo del tipo de objeto. Veremos qué es un tipo de referencia más adelante en este artículo.

¿Qué son las clases?

Una clase es una plantilla o plano a partir del cual se crean los objetos.

Imagina una clase como un cortador de galletas y objetos como galletas.

Las clases definen estados como variables de instancia y comportamientos como métodos de instancia.

Las variables de instancia también se conocen como variables miembro.

Las clases no consumen espacio.

Para darle una idea sobre las clases y los objetos, creemos una clase Cat que represente estados y comportamientos del Cat del mundo real.

public class Cat { /* Instance variables: states of Cat */ String name; int age; String color; String breed; /* Instance methods: behaviors of Cat */ void sleep(){ System.out.println("Sleeping"); } void play(){ System.out.println("Playing"); } void feed(){ System.out.println("Eating"); } }

Ahora hemos definido con éxito una plantilla para Cat. Digamos que tenemos dos gatos llamados Thor y Rambo.

¿Cómo podemos definirlos en nuestro programa?

Primero, necesitamos crear dos objetos de la clase Cat.

public class Main { public static void main(String[] args) { Cat thor = new Cat(); Cat rambo = new Cat(); } }

A continuación, definiremos sus estados y comportamientos.

public class Main { public static void main(String[] args) { /* Creating objects */ Cat thor = new Cat(); Cat rambo = new Cat(); /* Defining Thor cat */ thor.name = "Thor"; thor.age = 3; thor.breed = "Russian Blue"; thor.color = "Brown"; thor.sleep(); /* Defining Rambo cat */ rambo.name = "Rambo"; rambo.age = 4; rambo.breed = "Maine Coon"; rambo.color = "Brown"; rambo.play(); } }

Al igual que los ejemplos de código anteriores, podemos definir nuestra clase, instanciarla (crear objetos) y especificar los estados y comportamientos de esos objetos.

Ahora, hemos cubierto los conceptos básicos de la programación orientada a objetos. Pasemos a los principios de la programación orientada a objetos.

Principios de la programación orientada a objetos

Estos son los cuatro principios fundamentales del paradigma de programación orientada a objetos. Comprenderlos es esencial para convertirse en un programador exitoso.

  1. Encapsulamiento
  2. Herencia
  3. Abstracción
  4. Polimorfismo

Ahora veamos cada uno con más detalle.

Encapsulamiento

La encapsulación es un proceso de empaquetar código y datos en una sola unidad.

Es como una cápsula que contiene una combinación de varios medicamentos y es una técnica que ayuda a mantener protegidas las variables de instancia.

Esto se puede lograr mediante el uso de privatemodificadores de acceso a los que nada fuera de la clase pueda acceder. Para acceder a estados privados de forma segura, tenemos que proporcionar métodos de obtención y establecimiento públicos. (En Java, estos métodos deben seguir los estándares de denominación de JavaBeans).

Digamos que hay una tienda de discos que vende álbumes de música de diferentes artistas y un almacenista que los gestiona.

Si observa la figura 4, la StockKeeperclase puede acceder a los Albumestados de la clase directamente como Albumestán configurados los estados de la clase public.

¿Qué pasa si el encargado de las existencias crea un álbum y establece los estados en valores negativos? Esto puede ser hecho intencionalmente o no por un almacenista.

Para ilustrar, veamos un programa Java de muestra que explica el diagrama y la declaración anteriores.

Clase de álbum:

public class Album { public String name; public String artist; public double price; public int numberOfCopies; public void sellCopies(){ if(numberOfCopies > 0){ numberOfCopies--; System.out.println("One album has sold!"); } else{ System.out.println("No more albums available!"); } } public void orderCopies(int num){ numberOfCopies += num; } }

Clase StockKeeper:

public class StockKeeper { public String name; public StockKeeper(String name){ this.name = name; } public void manageAlbum(Album album, String name, String artist, double price, int numberOfCopies){ /* Defining states and behaviors for album */ album.name = name; album.artist = artist; album.price = price; album.numberOfCopies = numberOfCopies; /* Printing album details */ System.out.println("Album managed by :"+ this.name); System.out.println("Album details::::::::::"); System.out.println("Album name : " + album.name); System.out.println("Album artist : " + album.artist); System.out.println("Album price : " + album.price); System.out.println("Album number of copies : " + album.numberOfCopies); } }

Clase principal:

public class Main { public static void main(String[] args) { StockKeeper johnDoe = new StockKeeper("John Doe"); /* Stock keeper creates album and assigns negative values for price and number of copies available */ johnDoe.manageAlbum(new Album(), "Slippery When Wet", "Bon Jovi", -1000.00, -50); } }

Salida:

Album managed by :John Doe Album details:::::::::: Album name : Slippery When Wet Album artist : Bon Jovi Album price : -1000.0 Album number of copies : -50

El precio del álbum y el número de copias no pueden ser valores negativos. ¿Cómo podemos evitar esta situación? Aquí es donde usamos la encapsulación.

En este escenario, podemos impedir que el almacenista asigne valores negativos. Si intentan asignar valores negativos para el precio del álbum y el número de copias, los asignaremos como 0.0 y 0.

Clase de álbum:

public class Album { private String name; private String artist; private double price; private int numberOfCopies; public void sellCopies(){ if(numberOfCopies > 0){ numberOfCopies--; System.out.println("One album has sold!"); } else{ System.out.println("No more albums available!"); } } public void orderCopies(int num){ numberOfCopies += num; } public String getName() { return name; } public void setName(String name) { this.name = name; } public String getArtist() { return artist; } public void setArtist(String artist) { this.artist = artist; } public double getPrice() { return price; } public void setPrice(double price) { if(price > 0) { this.price = price; } else { this.price = 0.0; } } public int getNumberOfCopies() { return numberOfCopies; } public void setNumberOfCopies(int numberOfCopies) { if(numberOfCopies > 0) { this.numberOfCopies = numberOfCopies; } else { this.numberOfCopies = 0; } } }

Clase StockKeeper:

public class StockKeeper { private String name; StockKeeper(String name){ setName(name); } public void manageAlbum(Album album, String name, String artist, double price, int numberOfCopies){ /* Defining states and behaviors for album */ album.setName(name); album.setArtist(artist); album.setPrice(price); album.setNumberOfCopies(numberOfCopies); /* Printing album details */ System.out.println("Album managed by :"+ getName()); System.out.println("Album details::::::::::"); System.out.println("Album name : " + album.getName()); System.out.println("Album artist : " + album.getArtist()); System.out.println("Album price : " + album.getPrice()); System.out.println("Album number of copies : " + album.getNumberOfCopies()); } public String getName() { return name; } public void setName(String name) { this.name = name; } }

Clase principal:

public class Main { public static void main(String[] args) { StockKeeper johnDoe = new StockKeeper("John Doe"); /* Stock keeper creates album and assigns negative values for price and number of copies available */ johnDoe.manageAlbum(new Album(), "Slippery When Wet", "Bon Jovi", -1000.00, -50); } }

Salida:

Album managed by :John Doe Album details:::::::::: Album name : Slippery When Wet Album artist : Bon Jovi Album price : 0.0 Album number of copies : 0

With encapsulation, we’ve blocked our stock keeper from assigning negative values, meaning we have control over the data.

Advantages of encapsulation in Java

  1. We can make a class read-only or write-only: for a read-only class, we should provide only a getter method. For a write-only class, we should provide only a setter method.
  2. Control over the data: we can control the data by providing logic to setter methods, just like we restricted the stock keeper from assigning negative values in the above example.
  3. Data hiding: other classes can’t access private members of a class directly.

Inheritance

Let’s say that the record shop we discussed above also sells Blu-ray movies.

As you can see in the above diagram, there are many common states and behaviors (common code) between Album and Movie.

When implementing this class diagram into code, are you going to write (or copy & paste) the entire code for Movie? If you do, you are repeating yourself. How can you avoid code duplication?

This is where we use inheritance.

Inheritance is a mechanism in which one object acquires all the states and behaviors of a parent object.

Inheritance uses a parent-child relationship (IS-A relationship).

So what exactly is inherited?

Visibility/access modifiers impact what gets inherited from one class to another.

In Java, as a rule of thumb we make instance variables private and instance methods public .

In this case, we can safely say that the following are inherited:

  1. public instance methods.
  2. private instance variables (private instance variables can be accessed only through public getter and setter methods).

Types of Inheritance in Java

There are five types of inheritance in Java. They are single, multilevel, hierarchical, multiple, and hybrid.

Class allows single, multilevel and hierarchical inheritances. Interface allows multiple and hybrid inheritances.

A class can extend only one class however it can implement any number of interfaces. An interface can extend more than one interfaces.

Relationships

I. IS-A relationship

An IS-A relationship refers to inheritance or implementation.

a. Generalization

Generalization uses an IS-A relationship from a specialization class to generalization class.

II. HAS-A relationship

An instance of one class HAS-A reference to an instance of another class.

a. Aggregation

In this relationship, the existence of class A and B are not dependent on each other.

For this aggregation part, we going to see an example of the Student class and the ContactInfo class.

class ContactInfo { private String homeAddress; private String emailAddress; private int telephoneNumber; //12025550156 } public class Student { private String name; private int age; private int grade; private ContactInfo contactInfo;//Student HAS-A ContactInfo public void study() { System.out.println("Study"); } }

Student HAS-A ContactInfo. ContactInfo can be used in other places – for example, a company's Employee class can also use this ContactInfo class. So Student can exist without ContactInfo and ContactInfo can exist without Student . This type of relationship is known as aggregation.

b. Composition

In this relationship, class B can not exist without class A – but class A can exist without class B.

To give you an idea about composition, let's see an example of the Student class and the StudentId class.

class StudentId { private String idNumber;//A-123456789 private String bloodGroup; private String accountNumber; } public class Student { private String name; private int age; private int grade; private StudentId studentId;//Student HAS-A StudentId public void study() { System.out.println("Study"); } }

Student HAS-A StudentId. Student can exist without StudentId but StudentId can not exist without Student. This type of relationship is known as composition.

Now, let’s back to our previous record shop example that we discussed above.

We can implement this diagram in Java to avoid code duplication.

Advantages of inheritance

  1. Code reuse: the child class inherits all instance members of the parent class.
  2. You have more flexibility to change code: changing code in place is enough.
  3. You can use polymorphism: method overriding requires IS-A relationship.

Abstraction

Abstraction is a process of hiding the implementation details and showing only functionality to the user.

A common example of abstraction is that pressing the accelerator will increase the speed of a car. But the driver doesn’t know how pressing the accelerator increases the speed – they don't have to know that.

Technically abstract means something incomplete or to be completed later.

In Java, we can achieve abstraction in two ways: abstract class (0 to 100%) and interface (100%).

The keyword abstract can be applied to classes and methods. abstract and final or static can never be together.

I. Abstract class

An abstract class is one that contains the keyword abstract.

Abstract classes can’t be instantiated (can’t create objects of abstract classes). They can have constructors, static methods, and final methods.

II. Abstract methods

An abstract method is one that contains the keyword abstract.

An abstract method doesn’t have implementation (no method body and ends up with a semi colon). It shouldn’t be marked as private.

III. Abstract class and Abstract methods

  • If at least one abstract method exists inside a class then the whole class should be abstract.
  • We can have an abstract class with no abstract methods.
  • We can have any number of abstract as well as non-abstract methods inside an abstract class at the same time.
  • The first concrete sub class of an abstract class must provide implementation to all abstract methods.
  • If this doesn't happen, then the sub class also should be marked as abstract.

In a real world scenario, the implementation will be provided by someone who is unknown to end users. Users don’t know the implementation class and the actual implementation.

Let’s consider an example of abstract concept usage.

abstract class Shape { public abstract void draw(); } class Circle extends Shape{ public void draw() { System.out.println("Circle!"); } } public class Test { public static void main(String[] args) { Shape circle = new Circle(); circle.draw(); } }

When do we want to mark a class as abstract?

  1. To force sub classes to implement abstract methods.
  2. To stop having actual objects of that class.
  3. To keep having a class reference.
  4. To retain common class code.

Interface

An interface is a blueprint of a class.

An interface is 100% abstract. No constructors are allowed here. It represents an IS-A relationship.

NOTE: Interfaces only define required methods. We can not retain common code.

An interface can have only abstract methods, not concrete methods. By default, interface methods are public and abstract. So inside the interface, we don’t need to specify public and abstract.

So when a class implements an interface’s method without specifying the access level of that method, the compiler will throw an error stating “Cannot reduce the visibility of the inherited method from interface”. So that implemented method’s access level must be set to public.

By default, interface variables are public, static and final.

For instance:

interface Runnable { int a = 10; //similar to: public static final int a = 10; void run(); //similar to: public abstract void run(); } public class InterfaceChecker implements Runnable{ public static void main(String[] args) { Runnable.a = 5;//The final field Runnable.a cannot be assigned. } }

Let’s see an example that explains the interface concept:

interface Drawable { void draw(); } class Circle implements Drawable{ public void draw() { System.out.println("Circle!"); } } public class InterfaceChecker { public static void main(String[] args) { Drawable circle = new Circle(); circle.draw(); } }

Default and Static methods in Interfaces

Usually we implement interface methods in a separate class. Let’s say we are required to add a new method in an interface. Then we must implement that method in that separate class, too.

To overcome this issue Java 8 introduced default and static methods that implement methods inside an interface, unlike abstract methods.

  • Default method
public interface DefaultInterface { void sleep(); default void run() { System.out.println("I'm running!"); } } public class InterfaceCheckers implements DefaultInterface{ public void sleep() { System.out.println("Sleeping..."); } public static void main(String[] args) { InterfaceCheckers checker = new InterfaceCheckers(); checker.run(); checker.sleep(); } } /* Output: I'm running! Sleeping... */
  • Static method

Similar to static methods of classes, we can call them by their interface’s name.

public interface DefaultInterface { void sleep(); static void run() { System.out.println("I'm running!"); } } public class InterfaceCheckers implements DefaultInterface{ public void sleep() { System.out.println("Sleeping..."); } public static void main(String[] args) { InterfaceCheckers checker = new InterfaceCheckers(); DefaultInterface.run(); checker.sleep(); } } /* Output: I'm running! Sleeping... */
  • Marker interface

It’s an empty interface. For instance, Serializable, Cloneable, and Remote interfaces.

public interface Serializable { //No fields or methods }

Advantages of interfaces

  • They help us use multiple inheritance in Java.
  • They provide abstraction.
  • They provide loose coupling: objects are independent from one another.

When do we want to change a class to an interface?

  1. To force sub classes to implement abstract methods.
  2. To stop having actual objects of that class.
  3. To keep having a class reference.

NOTE: Remember, we can’t retain common code inside the interface.

If you want to define potentially required methods and common code, use an abstract class.

If you just want to define a required method, use an interface.

Polymorphism

Polymorphism is the ability of an object to take on many forms.

Polymorphism in OOP occurs when a super class references a sub class object.

All Java objects are considered to be polymorphic as they share more than one IS-A relationship (at least all objects will pass the IS-A test for their own type and for the class Object).

We can access an object through a reference variable. A reference variable can be of only one type. Once declared, the type of a reference variable cannot be changed.

A reference variable can be declared as a class or interface type.

A single object can be referred to by reference variables of many different types as long as they are the same type or a super type of the object.

Method overloading

If a class has multiple methods that have same name but different parameters, this is known as method overloading.

Method overloading rules:

  1. Must have a different parameter list.
  2. May have different return types.
  3. May have different access modifiers.
  4. May throw different exceptions.
class JavaProgrammer{ public void code() { System.out.println("Coding in C++"); } public void code(String language) { System.out.println("Coding in "+language); } } public class MethodOverloader { public static void main(String[] args) { JavaProgrammer gosling = new JavaProgrammer(); gosling.code(); gosling.code("Java"); } } /* Output: Coding in C++ Coding in Java */

NOTE: Static methods can also be overloaded.

class Addition { public static int add(int a,int b) { return a+b; } public static int add(int a,int b,int c) { return a+b+c; } } public class PolyTest { public static void main(String[] args) { System.out.println(Addition.add(5, 5)); System.out.println(Addition.add(2, 4, 6)); } }

NOTE: We can overload the main() method but the Java Virtual Machine (JVM) calls the main() method that receives String arrays as arguments.

public class PolyTest { public static void main() { System.out.println("main()"); } public static void main(String args) { System.out.println("String args"); } public static void main(String[] args) { System.out.println("String[] args"); } } //Output: String[] args

Rules to follow for polymorphism

Compile time rules

  1. Compiler only knows reference type.
  2. It can only look in reference type for methods.
  3. Outputs a method signature.

Run time rules

  1. At runtime, JVM follows exact runtime type (object type) to find method.
  2. Must match compile time method signature to method in actual object’s class.

Method overriding

If a subclass has the same method as declared in the super class, this is known as method overriding.

Method overriding rules:

  1. Must have the same parameter list.
  2. Must have the same return type: although a covariant return allows us to change the return type of the overridden method.
  3. Must not have a more restrictive access modifier: may have a less restrictive access modifier.
  4. Must not throw new or broader checked exceptions: may throw narrower checked exceptions and may throw any unchecked exception.
  5. Only inherited methods may be overridden (must have IS-A relationship).

Example for method overriding:

public class Programmer { public void code() { System.out.println("Coding in C++"); } } public class JavaProgrammer extends Programmer{ public void code() { System.out.println("Coding in Java"); } } public class MethodOverridder { public static void main(String[] args) { Programmer ben = new JavaProgrammer(); ben.code(); } } /* Output: Coding in Java */

NOTE: Static methods can’t be overridden because methods are overridden at run time. Static methods are associated with classes while instance methods are associated with objects. So in Java, the main() method also can’t be overridden.

NOTE: Constructors can be overloaded but not overridden.

Object types and reference types

class Person{ void eat() { System.out.println("Person is eating"); } } class Student extends Person{ void study() { System.out.println("Student is studying"); } } public class InheritanceChecker { public static void main(String[] args) { Person alex = new Person();//New Person "is a" Person alex.eat(); Student jane = new Student();//New Student "is a" Student jane.eat(); jane.study(); Person mary = new Student();//New Student "is a" Person mary.eat(); //Student chris = new Person(); //New Person isn't a Student. } }

In Person mary = new Student(); , this object creation is perfectly fine.

mary is a Person type reference variable and new Student() will create a new Student object.

mary can’t access study() in compile time because the compiler only knows the reference type. Since there is no study() in the reference type class, it can’t access it. But in runtime mary is going to be the Student type (Runtime type/ object type).

Please review this post for more information on runtime types.

In this case, we can convince the compiler by saying “at runtime, mary will be Student type, so please allow me to call it”. How can we convince the compiler like this? This is where we use casting.

We can make mary a Student type in compile time and can call study() by casting it.

((Student)mary).study();

We’ll learn about casting next.

Object type casting

Java type casting is classified into two types:

  1. Widening casting (implicit): automatic type conversion.
  2. Narrowing casting (explicit): need explicit conversion.

In primitives, long is a larger type than int . Like in objects, the parent class is a larger type than the child class.

The reference variable only refers to an object. Casting a reference variable doesn’t change the object on the heap but it labels the same object in another way by means of instance members accessibility.

I. Widening casting

Superclass superRef = new Subclass();

II. Narrowing casting

Subclass ref = (Subclass) superRef;

We have to be careful when narrowing. When narrowing, we convince the compiler to compile without any error. If we convince it wrongly, we will get a run time error (usually ClassCastException).

In order to perform narrowing correctly, we use the instanceof operator. It checks for an IS-A relationship.

class A { public void display(){ System.out.println("Class A"); } } class B extends A{ public void display(){ System.out.println("Class B"); } } public class Test { public static void main(String[] args) { A objA = new B(); if(objA instanceof B){ ((B)objA).display(); } } } /** * Output: Class B */ 

As I already stated before, we must remember one important thing when creating an object using the new keyword: the reference type should be the same type or a super type of the object type.

Conclusion

Gracias a todos por leer. Espero que este artículo te haya ayudado.

Le recomiendo encarecidamente que lea más artículos relacionados sobre la programación orientada a objetos.

Consulte mi serie de artículos originales sobre Medium: principios de programación orientada a objetos en Java

No dude en hacerme saber si tiene alguna pregunta.

El sueño no es lo que ves mientras duermes, es algo que no te deja dormir.

- APJ Abdul Kalam, Alas de fuego: una autobiografía

Gracias.

¡Feliz codificación!