Methods
Navigate Language Fundamentals topic: ) |
Methods are how we communicate with objects. When we invoke or call a method we are asking the object to carry out a task. We can say methods implement the behaviour of objects. For each method we need to give a name, we need to define its input parameters and we need to define its return type. We also need to set its visibility (private, package, or public). If the method throws an Exception, that needs to be declared as well. It is called a method definition. The syntax of method definition is:
class
MyClass { ...public
ReturnType methodName(ParamOneType param1, ParamTwoType param2)throws
ExceptionName { ReturnType retType; ...return
retType; } ... }
We can declare that the method does not return anything using the void
Java keyword. For example:
![]() |
Code section 3.50: Method without returned data.
|
When the method returns nothing, the return
keyword at the end of the method is optional. When the execution flow reaches the return
keyword, the method execution is stopped and the execution flow returns to the caller method. The return
keyword can be used anywhere in the method as long as there is a way to execute the instructions below:
![]() |
Code section 3.51: return keyword location.
|
In the code section 3.51, the return
keyword at line 5 is well placed because the instructions below can be reached when a
is negative or equal to 0. However, the return
keyword at line 8 is badly placed because the instructions below can't be reached.
Question 3.9: Consider the following code:
![]() |
Question 3.9: Compiler error.
|
The code above will return a compiler error. Why?
![]() |
Answer 3.9: Compiler error.
|
The method is supposed to return a int
but when a
is negative or equal to 0, it returns nothing.
Parameter passing[edit]
We can pass any primitive data types or objects to a method but the two are not processed the same way.
Primitive type parameter[edit]
The primitive types are passed in by value. It means that as soon as the primitive type is passed in, there is no more link between the value inside the method and the source variable:
![]() |
Code section 3.52: A method modifying a variable.
|
|
|
As you can see in code section 3.53, the modifyValue()
method has not modified the value of i
.
Object parameter[edit]
The object references are passed by value. It means that:
- There is no more link between the reference inside the method and the source reference,
- The source object itself and the object itself inside the method are still the same.
You must understand the difference between the reference of an object and the object itself. A object reference is the link between a variable name and an instance of object:
Object obj ⇔ new Object() |
An object reference is a pointer, an address to the object instance.
The object itself is the value of its attributes inside the object instance:
obj.firstName | ⇒ | "James" |
obj.lastName | ⇒ | "Gosling" |
obj.birthDay | ⇒ | "May 19" |
Take a look at the example above:
![]() |
Code section 3.54: A method modifying an object.
|
|
|
The name has changed because the method has changed the object itself and not the reference. Now take a look at the other example:
![]() |
Code section 3.56: A method modifying an object reference.
|
|
|
The name has not changed because the method has changed the reference and not the object itself. The behavior is the same as if the method was in-lined and the parameters were assigned to new variable names:
|
|
Functions[edit]
In Java, functions (methods really) are just like in C++ except that they must be declared inside a class and objects are passed by value of reference pointing to that object. You cannot create pointers to a function but Java has events which really are function pointers under the hood for when you need that type of functionality.
Return parameter[edit]
So as we can see, a method may or may not return a value. If the method does not return a value we use the void
Java keyword.
Same as the parameter passing, the method can return a primitive type or an object reference. So a method can return only one value. What if you want to return more than one value from a method. You can always pass in an object reference to the method, and let the method modify the object properties. The modified values can be considered as an output value from the method. However you can also create an Object array inside the method, assign the return values and return the array to the caller. You could have a problem however, if you want to mix primitive data types and object references as the output values from the method.
There is a better approach. Defines special return object with the needed return values. Create that object inside the method, assign the values and return the reference to this object. This special object is "bound" to this method and used only for returning values, so do not use a public class. The best way is to use a nested class, see example below:
![]() |
Code listing 3.10: Multiple returned variables.
|
In the above example the getPersonInfoById
method returns an object reference that contains both values of the name and the age. See below how you may use that object:
![]() |
Code section 3.59: Retrieving the values.
|
Question 3.10: Consider the following code:
![]() |
Question 3.10: Compiler error.
|
The code above will return a compiler error. Why?
![]() |
Answer 3.10: Compiler error.
|
The method is supposed to return a int
but at line 4, it returns c
, which is a String.
Special method, the constructor[edit]
The constructor is a special method called automatically when an object is created with the new
keyword. Constructor does not have a return value and its name is the same as the class name. Each class must have a constructor. If we do not define one, the compiler will create a default so called empty constructor automatically.
![]() |
Code listing 3.11: Automatically created constructor.
|
Static methods[edit]
A static method is a method that can be called without an object instance. It can be called on the class directly. For example, the valueOf(String)
method of the Integer
class is a static method:
![]() |
Code section 3.60: Static method.
|
As a consequence, it cannot use the non-static methods of the class but it can use the static ones. The same way, it cannot use the non-static attributes of the class but it can use the static ones:
![]() |
Code section 3.61: Static attribute.
|
You can notice that when you use System.out.println()
, out
is a static attribute of the System
class. A static attribute is related to a class, not to any object instance, so there is only one value for all the object instances. This attribute is unique in the whole Java Virtual Machine. All the object instances use the same attribute:
|
|
Question 3.11: Visit the Oracle JavaDoc of the class java.lang.Integer
.
How many static fields this class has?
4.
int MAX_VALUE
,int MIN_VALUE
,int SIZE
andClass<Integer> TYPE
.
- To learn how to overload and override a method, see Overloading Methods and Constructors.