# Problema con Class, extend y herencia

**URL:** <https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204>\
**Category:** Libraries\
**Created:** [May 13, 2019, 6:35pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204 "2019-05-13T18:35:37Z")\
**Posts on this page:** 13\
**Page:** 1

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 13, 2019, 6:35pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/1 "2019-05-13T18:35:37Z")

</div>

Hola a todos.

Este es mi primer post. Soy bastante nuevo en esto.  
Estoy haciendo una clase, supongamos “animales”. Dentro de la misma tengo un procedimiento “void hijo”, que crea una copia de si misma con algunas variaciones y la agrega una arraylist de la cual forman parte todos los “animales”. Por otro lado tengo una clase “perro” que extiende la clase “animales”. Cuando llamo al procedimiento “hijo” de la clase “perro” me crea otro “animal” y no un perro. Por ahora lo soluciono copiando todo el procedimiento “hijo” en la subclase y cambiando de “Animal nuevo = New Animal;” por “Perro nuevo = New Perro”. Estoy seguro que tiene que haber una mejor manera. Si alguien me puede ayudar se lo agradecería.

Saludos!

---

<div class="post-metadata">

**Author:** ![GoToLoop](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/gotoloop/32/86_2.png) [@GoToLoop](https://discourse.processing.org/u/GoToLoop)\
**Post date:** [May 13, 2019, 9:17pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/2 "2019-05-13T21:17:03Z")

</div>

Disclaimer: I know English & Portuguese only. 😑

And although I can understand Spanish more or less, still I can’t properly reply using it, sorry. 😞

Anyways, we can always count on Google Translate, right? 😌

> **[Google Translate](https://translate.google.com/#view=home&op=translate&sl=en&tl=es)**
>
> Google's service, offered free of charge, instantly translates words, phrases, and web pages between English and over 100 other languages.

> [@aledante](#):
>
> Cuando llamo al procedimiento **hijo()** de la clase Perro me crea otro Animal y no un Perro.

Indeed, if your method Animal::**hijo()** is something like this:

```auto
class Animal {
  Animal hijo(final String name) {
    return new Animal(name);
  }
}

```

And you’ve got a `class` that `extends` it:

```auto
class Perro extends Animal {
  Perro(final String name) {
    super(name);
  }
}

```

Java won’t “automagically” upgrade the datatype Animal in `return new Animal(name);` to the datatype of that subclass like `return new Perro(name);` in any way! 😢

As a workaround, we could use Java reflection techniques in order to always instantiate the current subclass’ Constructor: 🏗  
[Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/reflect/Constructor.html](http://Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/reflect/Constructor.html)

However, reflection is complicated! And thus it should be used only if there’s no other alternative: ⚠

> [@\[SOLVED\] Instantiating a particular class constructor using reflection](https://discourse.processing.org/t/instantiating-a-particular-class-constructor-using-reflection/7968/9):
>
> The (cast\_type) operator gotta be of a datatype compatible w/ the class you’re attempting to instantiate w/ Constructor::newInstance(). face_with_monocle BtW, here’s version 3.1. It’s now using Class.forName() in place of Class::getDeclaredClasses(): innocent /\*\* \* Reflexive Inner Class Instantiation (v3.1) \* GoToLoop (2019/Jan/30) \* \* https://Discourse.Processing.org/t/ \* instantiating-a-particular-class-constructor-using-reflection/7968/9 \* \* https://Forum.Processing.org/two/disc…

And a simpler alternative is to implement the Cloneable `interface` and its **clone()** method: 💡  
[Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Cloneable.html](http://Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Cloneable.html)  
[Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Object.html#clone()](http://Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Object.html#clone())

Here’s an example of a `class` named Pair which `implements` Cloneable: 🤓

> [@Functions argument](https://discourse.processing.org/t/functions-argument/11010/5):
>
> Or even better, use it as an actual full-fledged class, w/ constructors, fields & methods: triumph // https://Discourse.Processing.org/t/functions-argument/11010/5 // GoToLoop (v1.1) (2019/May/09) final Pair pair = new Pair(5, -10); void setup() { println(pair); // [5, -10] println(pair.calc('y')); // [-50, 540] println(pair.set(5, -10)); // [5, -10] println(pair.calc('u')); // [3, -2] println(pair.clone()); // [3, -2] exit(); } static final class…

And inspired by my previous sketch above, here’s “Cloneable Inheritance”: 🤠

```auto
/**
 * Cloneable Inheritance (v1.0.1)
 * GoToLoop (2019/May/13)
 * Discourse.Processing.org/t/problema-con-class-extend-y-herencia/11204/2
 */

import java.util.List;
final List<Animal> animales = new ArrayList<Animal>();

void setup() {
  animales.add(new Animal("Madre"));
  println(animales);

  animales.add(animales.get(animales.size() - 1).hijo("Hijo"));
  println(animales);

  animales.add(new Perro("Rex"));
  println(animales);

  animales.add(animales.get(animales.size() - 1).hijo("Perla"));
  println(animales);

  exit();
}

class Animal implements Cloneable {
  String name;

  Animal() {
  }

  Animal(final String name) {
    set(name);
  }

  Animal set(final String nick) {
    name = nick;
    return this;
  }

  Animal hijo(final String name) {
    //return new Animal(name);
    return clone().set(name);
  }

  @Override Animal clone() {
    try {
      return (Animal) super.clone();
    }
    catch (final CloneNotSupportedException e) {
      throw new RuntimeException(e);
    }
  }

  @Override String toString() {
    return name + " el " + getClass().getSimpleName();
  }
}

class Perro extends Animal {
  Perro() {
  }

  Perro(final String name) {
    super(name);
  }
}

```

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 13, 2019, 11:36pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/3 "2019-05-13T23:36:51Z")

</div>

I really appreciate your response. I thanks you for taking the job of translating my question. I understand english! I´m not so good talking it, or writing it. I´m trying to understand your response, and din´t do it yet! I don´t know how to publish code in the right format, but more or less this is what i’m doing. I would spect to have:

Ball  
Subclass  
Ball  
Subclass

but I get:

Ball  
Subclass  
Ball  
Ball

```auto
    /**
     * 
     * 
     * aledante@hotmail.com
     */
    import org.gicentre.utils.geom.*; 
    import java.util.*;
    HashGrid<Ball> balls;
    Set<Ball> paranacer=new HashSet<Ball>();
    
    void setup() {
      size(400, 400);
    
      balls = new HashGrid(width, height, 10);
    
      balls.add(new Ball());
      // println(balls);
      balls.add(new Subclass());
      // println(balls);
    }
    
    void draw() {
      
      // I add all to a set.
      for (Iterator<Ball> iterator = balls.iterator(); iterator.hasNext(); ) {
        Ball ball = iterator.next();
        paranacer.add(ball);
      }
      // I make to execute hijo on all the ser.
      for (Iterator<Ball> iterator = paranacer.iterator(); iterator.hasNext(); ) {
        Ball ball = iterator.next();
        ball.hijo();
      }
      for (Iterator<Ball> iterator = balls.iterator(); iterator.hasNext(); ) {
        Ball ball = iterator.next();
        println(ball.nombre);
      }
    
      exit();
    }
    
    class Ball implements Locatable {
      PVector location; 
      String nombre;
    
      Ball() {
        location = new PVector(40, 20);
        nombre ="Ball";
      }
      void hijo() {
        Ball hijo = new Ball();
        balls.add(hijo);
      }
    
      public PVector getLocation()
      {
        return location;
      }
    }
    
    class Subclass extends Ball {
      Subclass() {
        super();
        nombre="Subclass";
      }
    }

```

---

<div class="post-metadata">

**Author:** ![GoToLoop](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/gotoloop/32/86_2.png) [@GoToLoop](https://discourse.processing.org/u/GoToLoop)\
**Post date:** [May 14, 2019, 5:29am UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/4 "2019-05-14T05:29:44Z")

</div>

> [@aledante](#):
>
> I thank you for taking the job of translating my question.

I didn’t need to translate anything, nor use Google Translate. 😉

> [@aledante](#):
>
> I would expect to have: Ball, Subclass, Ball, Subclass

Like I had already explained, you’re gonna need to implement **clone()** in your own `class`, rather than relying on the [`new` operator](https://processing.org/reference/new.html):

So instead of:

```auto
Ball hijo() {
  return new Ball();
}

```

Go w/ something like this:

```auto
Ball hijo() {
  return clone();
}

@Override Ball clone() {
  try {
    final Ball hijo = (Ball) super.clone();
    hijo.location = location.get();
    return hijo;
  }
  catch (final CloneNotSupportedException e) {
    throw new RuntimeException(e);
  }
}

```

Notice I had to do: `hijo.location = location.get();`. Here’s the reason why: ⚠

- Your class Ball got 2 fields: _location_ & _nombre_.
- The 1st refers to a PVector, while the 2nd refers to a String.
- Datatype String is immutable. Its content can’t be changed (at least not w/o some hardcore hacking 😈).
- However, datatype PVector is mutable. Its fields _x_, _y_ & _z_ can be freely reassigned.
- If we don’t also clone each mutable field from a class when we call **clone()**, modifying those mutable fields will reflect on both the original object and all of its clones! 😨
- That’s why I invoke the method PVector::**get()** over the cloned field _location_, so the original and the cloned field _location_ won’t share the same PVector object: 🤓
- [ProcessingJS.org/reference/PVector\_get\_/](http://ProcessingJS.org/reference/PVector_get_/)

Below’s my new attempt sketch “Cloneable Locatable”. Any doubts about it just ask: 😇

```auto
/**
 * Cloneable Locatable (v1.0.1)
 * GoToLoop (2019/May/14)
 * Discourse.Processing.org/t/problema-con-class-extend-y-herencia/11204/4
 */

import org.gicentre.utils.geom.HashGrid;
import org.gicentre.utils.geom.Locatable;

import java.util.Collection;
//import java.util.Set;

static final int SIZE = 10;
final Collection<Bola> temp = new ArrayList<Bola>();

Collection<Bola> pelotas;
//Set<Bola> pelotas;
//HashGrid<Bola> pelotas;

void setup() {
  size(400, 400);
  noLoop();

  pelotas = new HashGrid<Bola>(width, height, SIZE);

  pelotas.add(new Bola());
  pelotas.add(new Bolita());
}

void draw() {
  background((color) random(#000000));
  getSurface().setTitle("Frame: " + frameCount);
  println("\nPelotas: " + pelotas.size());
  for (final Bola b : pelotas) println(b);
}

void mousePressed() {
  temp.clear();
  for (final Bola b : pelotas) temp.add(b.hijo());
  pelotas.addAll(temp);
  redraw = true;
}

class Bola implements Locatable, Cloneable {
  PVector location = new PVector(40, 20);
  String nombre = getClass().getSimpleName();

  Bola hijo() {
    return clone();
  }

  @Override Bola clone() {
    try {
      final Bola pelota = (Bola) super.clone();
      pelota.location = location.get();
      return pelota;
    }
    catch (final CloneNotSupportedException e) {
      throw new RuntimeException(e);
    }
  }

  @Override PVector getLocation() {
    return location;
  }

  @Override String toString() {
    return nombre + ": " + location;
  }
}

class Bolita extends Bola {
}

```

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 27, 2019, 3:26am UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/5 "2019-05-27T03:26:09Z")

</div>

Thanks again. Great answer, extremely complete, and clear.  
Some questions:

1. I made the example code very simple to explain my problem. My class have several variables. Some int, another PVector, some float, and another class. Only PVector and class are mutables, isn´t it?

So

```auto
  @Override Bola clone() {
    try {
      final Bola pelota = (Bola) super.clone();
      pelota.location = location.get();
pelota.vel = location.get(); // This is the other PVector
pelota.comp=comp.copia(); // This is a method that I have to make to copy my other class
      return pelota;
    }

```

Is it correct?

1. Is it necessary to use “@Override”. In Processing reference says that is not… But I see it a lot. 🧐

TXH!😁

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 27, 2019, 4:15am UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/6 "2019-05-27T04:15:37Z")

</div>

Ehhh… And why you do this:

```auto
  @Override PVector getLocation() {

    return location;
  }

```

---

<div class="post-metadata">

**Author:** ![GoToLoop](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/gotoloop/32/86_2.png) [@GoToLoop](https://discourse.processing.org/u/GoToLoop)\
**Post date:** [May 27, 2019, 7:03am UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/7 "2019-05-27T07:03:52Z")

</div>

> [@aledante](#):
>
> Is it necessary to use `@Override`.

`@Override`’s just a labeling annotation. You can remove them all: 👋  
[Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Override.html](http://Docs.Oracle.com/en/java/javase/11/docs/api/java.base/java/lang/Override.html)

> [@aledante](#):
>
> `pelota.vel = location.get(); // This is the other PVector`

In general the cloning assignment should use the same field: ©  
`pelota.vel = vel.get();`

However, if your intention is indeed to give some other initial value to a newborn cloned field, that’s OK AFAIK. 👶

Only you know better whether the cloned field Bola::_vel_ should be born using its old value or a diff. 1. 🧙

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 27, 2019, 4:57pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/8 "2019-05-27T16:57:45Z")

</div>

You are my hero!  
Thanks😀

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 31, 2019, 12:39pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/9 "2019-05-31T12:39:03Z")

</div>

Last little question. Is color mutable or inmutable? THX

---

<div class="post-metadata">

**Author:** ![GoToLoop](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/gotoloop/32/86_2.png) [@GoToLoop](https://discourse.processing.org/u/GoToLoop)\
**Post date:** [May 31, 2019, 1:33pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/10 "2019-05-31T13:33:29Z")

</div>

> [@aledante](#):
>
> Is _color_ mutable or immutable?

- Processing’s `color` is syntactic sugar for Java’s `int` keyword.
- Java got 8 primitive datatypes represented by 8 keywords.
- All primitive types are immutable. In Java they’re not even objects.

> **[Primitive Data Types (The Java™ Tutorials \>        
            Learning...](https://docs.oracle.com/javase/tutorial/java/nutsandbolts/datatypes.html)**
>
> This beginner Java tutorial describes fundamentals of programming in the Java programming language

> **[color / Reference](https://processing.org/reference/color_datatype.html)**
>
> Datatype for storing color values. Colors may be assigned with get() and color() or they may be specified directly using hexadecimal notation such as #FFCC00 or 0xFFFFCC00.…

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 31, 2019, 3:04pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/11 "2019-05-31T15:04:47Z")

</div>

Hello again,  
Excuse me. But i can not get ir right.  
In the extended class I want to make some changes over child objects.  
Why i cannot invoke clone method on extended class? Why I cannot invoke super class?  
I show the code you send me with the modifications i would like to do, that shows me mismatch errors.  
The problem is in the last few lines.

```auto
/**
 * Cloneable Locatable (v1.0)
 * GoToLoop (2019/Jan/) 
 * Discourse.Processing.org/t/problema-con-class-extend-y-herencia/11204/4 
 */

import org.gicentre.utils.geom.HashGrid;
import org.gicentre.utils.geom.Locatable;
import java.util.List;
import java.util.Set;
static final int SIZE = 10;
final List<Bola> temp = new ArrayList<Bola>();

Set<Bola> pelotas;

// HashGrid<Bola> pelotas;
void setup() {
  size(400, 400);
  noLoop();
  pelotas = new HashGrid<Bola>(width, height, SIZE);
  pelotas.add(new Bola());
  pelotas.add(new Bolita());
}

void draw() {

  background((color) random(#000000));
  getSurface().setTitle("Frame: " + frameCount);
  println(ENTER + "Pelotas: " + pelotas.size());
  for (final Bola b : pelotas) println(b);
}

void mousePressed() {

  temp.clear();
  for (final Bola b : pelotas) temp.add(b.hijo());
  pelotas.addAll(temp);
  redraw = true;
}
class Bola implements Locatable, Cloneable {
  PVector location = new PVector(40, 20);
  String nombre = getClass().getSimpleName();

  Bola hijo() {
    Bola hijo=clone();
    return hijo;
  }

  @Override Bola clone() {

    try {

      final Bola pelota = (Bola) super.clone();
      pelota.location = location.get();
      return pelota;
    }
    catch (final CloneNotSupportedException e) {
      throw new RuntimeException(e);
    }
  }

  @Override PVector getLocation() {
    return location;
  }

  @Override String toString() {
    return nombre + ": " + location;
  }
}

class Bolita extends Bola {
  Bolita hijo() {
    Bolita hijo=super.hijo();
    hijo.nombre=hijo.nombre+"2";
    return hijo;
  }
  Bolita tween() {
    return clone();
  }
}

```

---

<div class="post-metadata">

**Author:** ![GoToLoop](https://yyz2.discourse-cdn.com/flex036/user_avatar/discourse.processing.org/gotoloop/32/86_2.png) [@GoToLoop](https://discourse.processing.org/u/GoToLoop)\
**Post date:** [May 31, 2019, 7:35pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/12 "2019-05-31T19:35:07Z")

</div>

- Obviously, method Bola::**hijo()** returns an object of datatype Bola.
- And it can’t be converted to its derived datatype Bolita.
- However, it doesn’t need to, b/c the inherited field _nombre_ exists in datatype Bola too:

```auto
class Bolita extends Bola {
  @Override Bola hijo() {
    final Bola hijo = super.hijo();
    hijo.nombre += '2';
    return hijo;
  }
}

```

- However, in case subclass Bolita would add its own fields, besides those it had inherited from Bola, you’d probably need to `@Override` Bola::**clone()** in subclass Bolita, if any of those new fields happen to be mutable.

---

<div class="post-metadata">

**Author:** ![aledante](https://avatars.discourse-cdn.com/v4/letter/a/ad7895/32.png) [@aledante](https://discourse.processing.org/u/aledante)\
**Post date:** [May 31, 2019, 11:08pm UTC](https://discourse.processing.org/t/problema-con-class-extend-y-herencia/11204/13 "2019-05-31T23:08:47Z")

</div>

You are like my Super Hero! 😁