问题

我有一个公共类Dog。

方法1：

public Dog getDog(Dog dog){
    if(dog.size > 0){
      return dog;
    }
    return null;
}

方法2：

public Dog getNewDog(){
   return new Dog();
}

最后，将方法返回值重新赋值：

Dog mydog = new Dog();
Dog d2 = new Dog();

Dog c = mydog.getDog(d2);   // 我们是将一个引用变量赋回另一个引用吗？
Dog k = mydog.getNewDog(); // 我们是将一个新的Dog对象本身赋给Dog类型的变量。

现在，为什么我们总是将方法的返回值赋给引用变量？我在Java中到处都看到这个。


<details>
<summary>英文:</summary>

I have a public class Dog.

Method 1:

    public Dog getDog(Dog dog){
        if(dog.size&gt;0){
          return dog;
        }
        return null;
    }
        

Method 2:
    
    public Dog getNewDog(){
       return new Dog();
    }

Finally, assigning the method return value back:-

    Dog mydog = new Dog();
    Dog d2 = new Dog();
    
    Dog c = mydog.getDog(d2);   // are we assigning a reference variable back to another reference?
    Dog k = mydog.getNewDog(); // Are we assigning a new dog object itself to the Dog type variable.


now, why do we always assign back value from a method to a Reference variable? I have seen this everywhere in Java.


</details>


# 答案1
**得分**: 1

让我试着回答你提出的注释中的问题：
A. 为什么我们总是将方法的返回值分配回一个引用变量？- 不一定正确。
有些类一旦给定了一个值，就无法更改它。这些被称为“不可变”。例如，`String` 是不可变的，所以一旦给它赋值，就无法更改该字符串。让我们看下面的例子：

    String s1 = "1";
    String s2 = s1.concat("2")
    s2.concat("3");
    System.out.println(s2); // 输出 12

输出是 **12**。原因是，每当调用更改值的方法时，您必须将其分配回去，否则您将丢失新值。对于一个不可变的对象，正如您所见，有必要重新分配。但是，这些是特殊的类。在 `Dog` 的情况下，您不必每次更改其属性时都进行分配。除非您明确需要创建一个新的 `Dog` 对象。

B. 我们是将一个新的狗对象本身分配给了狗类型的变量吗？
无论何时声明一个类型为引用的变量（它扩展自 Object 类），它都将被赋予 null。如果您尝试访问其属性，将会导致 `NullPointerException`。因此，请记住您必须为其分配一个值。

注意，`Dog k = mydog.getNewDog();` 在您的代码中等同于 `Dog k = new Dog();`。

C. 我们是将一个引用变量再次分配回另一个引用吗？
让我们看看您的代码：

    public Dog getDog(Dog dog){
        if(dog.size > 0){
            return dog;
        }
        return null;
    }
            
    Dog d2 = new Dog();
    Dog c = mydog.getDog(d2);

对于这段代码：
 - 如果 `dog.size > 0` 为真，则 `d2` 和 `c` 将引用同一个对象。因此，您对其中一个进行的任何更改都将反映在两者上。
 - 如果 `dog.size > 0` 为假，则 `c` 将被赋值为 `null`。

<details>
<summary>英文:</summary>

Let me try to answer the questions you have put as comments:
A. Why do we always assign back value from a method to a Reference variable? - Not necessarily true.
There are some classes once given a value, you can&#39;t change it. This are called `Immutable`. For instance, `String` is immutable, so once you assign it a value you can&#39;t change that same string. Let&#39;s see the following 
example:	

    String s1 = &quot;1&quot;;
    String s2 = s1.concat(&quot;2&quot;)
    s2.concat(&quot;3&quot;);
    System.out.println(s2); // print out 12
	
The output is **12**. The reason is that, whenever you call a method that changes the value you have to assign it back or you will lose the new value.
For an `Immutable` object, as you can see, it is necessary that you reassign. But, this are special classes. In the case of `Dog`, you don&#39;t have to assign it every time change its properties. Unless you explicitly needed to create a new `Dog` Object.

B. Are we assigning a new dog object itself to the Dog type variable?
Whenever you declare a variable of type reference(that extends Object), it will be assigned null. And if you try to access its properties, it results in `NullPointerException`. So, keep in mind that you have to assign a value to it.

Note that `Dog k = mydog.getNewDog();` is equivalent to saying `Dog k = new Dog();` in your code.

C. Are we assigning a reference variable back to another reference?
Let&#39;s see your code: 
	

    public Dog getDog(Dog dog){
    	if(dog.size&gt;0){
    		return dog;
    	}
    	return null;
    }
    		
    Dog d2 = new Dog();
    Dog c = mydog.getDog(d2);

For this piece of code:
 - if `dog.size &gt; 0` is true, `d2` and `c` will be referencing to the same object. So, any change you make to one of them will reflect on both.
 - if `dog.size &gt; 0` is false, `c` will be assigned `null`

</details>



# 答案2
**得分**: 0

所以你正在将新的 Dog 对象分配给变量 c 和 k，这将导致覆盖先前保存的狗。

虽然不一定非要这样做，但至少在三种情况下这样做可能很有用：
 - 有时这种方式可以减少代码行数
 - 如果你之前链接的对象在其他地方有使用，可能不希望到处都进行更改。如果一个方法返回一个新的（经过转换的）对象，你可以确保旧对象仍然可以在其他地方使用
 - 这种模式可以用于构建器或链式调用。如果一个方法返回一只狗（无论是新的还是旧的），你可以直接在同一行上继续编写更多的方法调用。例如：mydog.getNewDog().bark();

<details>
<summary>英文:</summary>

So what you are doing is assignig new Dog objects to variable c and k, which results in overwriting the previously saved dogs.

It does not have to be done like this, but it can be useful in at least three cases:
 - sometimes this way results in less lines of code
 - if you used your previously linked object in other places, you might not want to change it everywhere. If a method returns a new (transformed) object, you can make sure the old object can still be used in other places
 - this pattern could be used for builders or chaining. if a method returns a dog (no matter if it is a new or an old one) you may directly continue writing more method calls on the same line. Example: mydog.getNewDog().bark();

</details>