I am looking for a way to pass a method by reference. I understand that Java does not pass methods as parameters, however, I would like to get an alternative.
I've been told interfaces are the alternative to passing methods as parameters but I don't understand how an interface can act as a method by reference. If I understand correctly an interface is simply an abstract set of methods that are not defined. I don't want to send an interface that needs to be defined every time because several different methods could call the same method with the same parameters.
What I would like to accomplish is something similar to this:
public void setAllComponents(Component[] myComponentArray, Method myMethod) {
for (Component leaf : myComponentArray) {
if (leaf instanceof Container) { //recursive call if Container
Container node = (Container) leaf;
setAllComponents(node.getComponents(), myMethod);
} //end if node
myMethod(leaf);
} //end looping through components
}
invoked such as:
setAllComponents(this.getComponents(), changeColor());
setAllComponents(this.getComponents(), changeSize());
Edit: as of Java 8, lambda expressions are a nice solution as other answers have pointed out. The answer below was written for Java 7 and earlier...
Take a look at the command pattern.
// NOTE: code not tested, but I believe this is valid java...
public class CommandExample
{
public interface Command
{
public void execute(Object data);
}
public class PrintCommand implements Command
{
public void execute(Object data)
{
System.out.println(data.toString());
}
}
public static void callCommand(Command command, Object data)
{
command.execute(data);
}
public static void main(String... args)
{
callCommand(new PrintCommand(), "hello world");
}
}
Edit: as Pete Kirkham points out, there's another way of doing this using a Visitor. The visitor approach is a little more involved - your nodes all need to be visitor-aware with an acceptVisitor() method - but if you need to traverse a more complex object graph then it's worth examining.
In Java 8, you can now pass a method more easily using Lambda Expressions and Method References. First, some background: a functional interface is an interface that has one and only one abstract method, although it can contain any number of default methods (new in Java 8) and static methods. A lambda expression can quickly implement the abstract method, without all the unnecessary syntax needed if you don't use a lambda expression.
Without lambda expressions:
obj.aMethod(new AFunctionalInterface() {
#Override
public boolean anotherMethod(int i)
{
return i == 982
}
});
With lambda expressions:
obj.aMethod(i -> i == 982);
Here is an excerpt from the Java tutorial on Lambda Expressions:
Syntax of Lambda Expressions
A lambda expression consists of the following:
A comma-separated list of formal parameters enclosed in parentheses. The CheckPerson.test method contains one parameter, p,
which represents an instance of the Person class.Note: You
can omit the data type of the parameters in a lambda expression. In
addition, you can omit the parentheses if there is only one parameter.
For example, the following lambda expression is also valid:
p -> p.getGender() == Person.Sex.MALE
&& p.getAge() >= 18
&& p.getAge() <= 25
The arrow token, ->
A body, which consists of a single expression or a statement block. This example uses the following expression:
p.getGender() == Person.Sex.MALE
&& p.getAge() >= 18
&& p.getAge() <= 25
If you specify a single expression, then the Java runtime evaluates the expression and then returns its value. Alternatively,
you can use a return statement:
p -> {
return p.getGender() == Person.Sex.MALE
&& p.getAge() >= 18
&& p.getAge() <= 25;
}
A return statement is not an expression; in a lambda expression, you must enclose statements in braces ({}). However, you do not have
to enclose a void method invocation in braces. For example, the
following is a valid lambda expression:
email -> System.out.println(email)
Note that a lambda expression looks a lot like a method declaration;
you can consider lambda expressions as anonymous methods—methods
without a name.
Here is how you can "pass a method" using a lambda expression:
interface I {
public void myMethod(Component component);
}
class A {
public void changeColor(Component component) {
// code here
}
public void changeSize(Component component) {
// code here
}
}
class B {
public void setAllComponents(Component[] myComponentArray, I myMethodsInterface) {
for(Component leaf : myComponentArray) {
if(leaf instanceof Container) { // recursive call if Container
Container node = (Container)leaf;
setAllComponents(node.getComponents(), myMethodInterface);
} // end if node
myMethodsInterface.myMethod(leaf);
} // end looping through components
}
}
class C {
A a = new A();
B b = new B();
public C() {
b.setAllComponents(this.getComponents(), component -> a.changeColor(component));
b.setAllComponents(this.getComponents(), component -> a.changeSize(component));
}
}
Class C can be shortened even a bit further by the use of method references like so:
class C {
A a = new A();
B b = new B();
public C() {
b.setAllComponents(this.getComponents(), a::changeColor);
b.setAllComponents(this.getComponents(), a::changeSize);
}
}
Since Java 8 there is a Function<T, R> interface (docs), which has method
R apply(T t);
You can use it to pass functions as parameters to other functions. T is the input type of the function, R is the return type.
In your example you need to pass a function that takes Component type as an input and returns nothing - Void. In this case Function<T, R> is not the best choice, since there is no autoboxing of Void type. The interface you are looking for is called Consumer<T> (docs) with method
void accept(T t);
It would look like this:
public void setAllComponents(Component[] myComponentArray, Consumer<Component> myMethod) {
for (Component leaf : myComponentArray) {
if (leaf instanceof Container) {
Container node = (Container) leaf;
setAllComponents(node.getComponents(), myMethod);
}
myMethod.accept(leaf);
}
}
And you would call it using method references:
setAllComponents(this.getComponents(), this::changeColor);
setAllComponents(this.getComponents(), this::changeSize);
Assuming that you have defined changeColor() and changeSize() methods in the same class.
If your method happens to accept more than one parameter, you can use BiFunction<T, U, R> - T and U being types of input parameters and R being return type. There is also BiConsumer<T, U> (two arguments, no return type). Unfortunately for 3 and more input parameters, you have to create an interface by yourself. For example:
public interface Function4<A, B, C, D, R> {
R apply(A a, B b, C c, D d);
}
Use the java.lang.reflect.Method object and call invoke
First define an Interface with the method you want to pass as a parameter
public interface Callable {
public void call(int param);
}
Implement a class with the method
class Test implements Callable {
public void call(int param) {
System.out.println( param );
}
}
// Invoke like that
Callable cmd = new Test();
This allows you to pass cmd as parameter and invoke the method call defined in the interface
public invoke( Callable callable ) {
callable.call( 5 );
}
While this is not yet valid for Java 7 and below, I believe that we should look to the future and at least recognize the changes to come in new versions such as Java 8.
Namely, this new version brings lambdas and method references to Java (along with new APIs, which are another valid solution to this problem. While they still require an interface no new objects are created, and extra classfiles need not pollute output directories due to different handling by the JVM.
Both flavors(lambda and method reference) require an interface available with a single method whose signature is used:
public interface NewVersionTest{
String returnAString(Object oIn, String str);
}
Names of methods will not matter from here on. Where a lambda is accepted, a method reference is as well. For example, to use our signature here:
public static void printOutput(NewVersionTest t, Object o, String s){
System.out.println(t.returnAString(o, s));
}
This is just a simple interface invocation, up until the lambda1 gets passed:
public static void main(String[] args){
printOutput( (Object oIn, String sIn) -> {
System.out.println("Lambda reached!");
return "lambda return";
}
);
}
This will output:
Lambda reached!
lambda return
Method references are similar. Given:
public class HelperClass{
public static String testOtherSig(Object o, String s){
return "real static method";
}
}
and main:
public static void main(String[] args){
printOutput(HelperClass::testOtherSig);
}
the output would be real static method. Method references can be static, instance, non-static with arbitrary instances, and even constructors. For the constructor something akin to ClassName::new would be used.
1 This is not considered a lambda by some, as it has side effects. It does illustrate, however, the use of one in a more straightforward-to-visualize fashion.
Last time I checked, Java is not capable of natively doing what you want; you have to use 'work-arounds' to get around such limitations. As far as I see it, interfaces ARE an alternative, but not a good alternative. Perhaps whoever told you that was meaning something like this:
public interface ComponentMethod {
public abstract void PerfromMethod(Container c);
}
public class ChangeColor implements ComponentMethod {
#Override
public void PerfromMethod(Container c) {
// do color change stuff
}
}
public class ChangeSize implements ComponentMethod {
#Override
public void PerfromMethod(Container c) {
// do color change stuff
}
}
public void setAllComponents(Component[] myComponentArray, ComponentMethod myMethod) {
for (Component leaf : myComponentArray) {
if (leaf instanceof Container) { //recursive call if Container
Container node = (Container) leaf;
setAllComponents(node.getComponents(), myMethod);
} //end if node
myMethod.PerfromMethod(leaf);
} //end looping through components
}
Which you'd then invoke with:
setAllComponents(this.getComponents(), new ChangeColor());
setAllComponents(this.getComponents(), new ChangeSize());
If you don't need these methods to return something, you could make them return Runnable objects.
private Runnable methodName (final int arg) {
return (new Runnable() {
public void run() {
// do stuff with arg
}
});
}
Then use it like:
private void otherMethodName (Runnable arg){
arg.run();
}
Java-8 onwards
Java 8 onwards, you can provide the implementation of the abstract method of a functional interface (an interface that has only one abstract method) using a lambda expression and pass the same to a method as a parameter.
#FunctionalInterface
interface ArithmeticFunction {
public int calcualate(int a, int b);
}
public class Main {
public static void main(String args[]) {
ArithmeticFunction addition = (a, b) -> a + b;
ArithmeticFunction subtraction = (a, b) -> a - b;
int a = 20, b = 5;
System.out.println(perform(addition, a, b));
// or
System.out.println(perform((x, y) -> x + y, a, b));
System.out.println(perform(subtraction, a, b));
// or
System.out.println(perform((x, y) -> x - y, a, b));
}
static int perform(ArithmeticFunction function, int a, int b) {
return function.calcualate(a, b);
}
}
Output:
25
25
15
15
ONLINE DEMO
Learn more about it from Method References.
I didn't find any example explicit enough for me on how to use java.util.function.Function for simple method as parameter function. Here is a simple example:
import java.util.function.Function;
public class Foo {
private Foo(String parameter) {
System.out.println("I'm a Foo " + parameter);
}
public static Foo method(final String parameter) {
return new Foo(parameter);
}
private static Function parametrisedMethod(Function<String, Foo> function) {
return function;
}
public static void main(String[] args) {
parametrisedMethod(Foo::method).apply("from a method");
}
}
Basically you have a Foo object with a default constructor. A method that will be called as a parameter from the parametrisedMethod which is of type Function<String, Foo>.
Function<String, Foo> means that the function takes a String as parameter and return a Foo.
The Foo::Method correspond to a lambda like x -> Foo.method(x);
parametrisedMethod(Foo::method) could be seen as x -> parametrisedMethod(Foo.method(x))
The .apply("from a method") is basically to do parametrisedMethod(Foo.method("from a method"))
Which will then return in the output:
>> I'm a Foo from a method
The example should be running as is, you can then try more complicated stuff from the above answers with different classes and interfaces.
Java do have a mechanism to pass name and call it. It is part of the reflection mechanism.
Your function should take additional parameter of class Method.
public void YouMethod(..... Method methodToCall, Object objWithAllMethodsToBeCalled)
{
...
Object retobj = methodToCall.invoke(objWithAllMethodsToBeCalled, arglist);
...
}
I did not found any solution here that show how to pass method with parameters bound to it as a parameter of a method. Bellow is example of how you can pass a method with parameter values already bound to it.
Step 1: Create two interfaces one with return type, another without. Java has similar interfaces but they are of little practical use because they do not support Exception throwing.
public interface Do {
void run() throws Exception;
}
public interface Return {
R run() throws Exception;
}
Example of how we use both interfaces to wrap method call in transaction. Note that we pass method with actual parameters.
//example - when passed method does not return any value
public void tx(final Do func) throws Exception {
connectionScope.beginTransaction();
try {
func.run();
connectionScope.commit();
} catch (Exception e) {
connectionScope.rollback();
throw e;
} finally {
connectionScope.close();
}
}
//Invoke code above by
tx(() -> api.delete(6));
Another example shows how to pass a method that actually returns something
public R tx(final Return func) throws Exception {
R r=null;
connectionScope.beginTransaction();
try {
r=func.run();
connectionScope.commit();
} catch (Exception e) {
connectionScope.rollback();
throw e;
} finally {
connectionScope.close();
}
return r;
}
//Invoke code above by
Object x= tx(() -> api.get(id));
Example of solution with reflection, passed method must be public
import java.lang.reflect.Method;
import java.lang.reflect.InvocationTargetException;
public class Program {
int i;
public static void main(String[] args) {
Program obj = new Program(); //some object
try {
Method method = obj.getClass().getMethod("target");
repeatMethod( 5, obj, method );
}
catch ( NoSuchMethodException | IllegalAccessException | InvocationTargetException e) {
System.out.println( e );
}
}
static void repeatMethod (int times, Object object, Method method)
throws IllegalAccessException, InvocationTargetException {
for (int i=0; i<times; i++)
method.invoke(object);
}
public void target() { //public is necessary
System.out.println("target(): "+ ++i);
}
}
Use the Observer pattern (sometimes also called Listener pattern):
interface ComponentDelegate {
void doSomething(Component component);
}
public void setAllComponents(Component[] myComponentArray, ComponentDelegate delegate) {
// ...
delegate.doSomething(leaf);
}
setAllComponents(this.getComponents(), new ComponentDelegate() {
void doSomething(Component component) {
changeColor(component); // or do directly what you want
}
});
new ComponentDelegate()... declares an anonymous type implementing the interface.
Here is a basic example:
public class TestMethodPassing
{
private static void println()
{
System.out.println("Do println");
}
private static void print()
{
System.out.print("Do print");
}
private static void performTask(BasicFunctionalInterface functionalInterface)
{
functionalInterface.performTask();
}
#FunctionalInterface
interface BasicFunctionalInterface
{
void performTask();
}
public static void main(String[] arguments)
{
performTask(TestMethodPassing::println);
performTask(TestMethodPassing::print);
}
}
Output:
Do println
Do print
I'm not a java expert but I solve your problem like this:
#FunctionalInterface
public interface AutoCompleteCallable<T> {
String call(T model) throws Exception;
}
I define the parameter in my special Interface
public <T> void initialize(List<T> entries, AutoCompleteCallable getSearchText) {.......
//call here
String value = getSearchText.call(item);
...
}
Finally, I implement getSearchText method while calling initialize method.
initialize(getMessageContactModelList(), new AutoCompleteCallable() {
#Override
public String call(Object model) throws Exception {
return "custom string" + ((xxxModel)model.getTitle());
}
})
I appreciate the answers above but I was able to achieve the same behavior using the method below; an idea borrowed from Javascript callbacks. I'm open to correction though so far so good (in production).
The idea is to use the return type of the function in the signature, meaning that the yield has to be static.
Below is a function that runs a process with a timeout.
public static void timeoutFunction(String fnReturnVal) {
Object p = null; // whatever object you need here
String threadSleeptime = null;
Config config;
try {
config = ConfigReader.getConfigProperties();
threadSleeptime = config.getThreadSleepTime();
} catch (Exception e) {
log.error(e);
log.error("");
log.error("Defaulting thread sleep time to 105000 miliseconds.");
log.error("");
threadSleeptime = "100000";
}
ExecutorService executor = Executors.newCachedThreadPool();
Callable<Object> task = new Callable<Object>() {
public Object call() {
// Do job here using --- fnReturnVal --- and return appropriate value
return null;
}
};
Future<Object> future = executor.submit(task);
try {
p = future.get(Integer.parseInt(threadSleeptime), TimeUnit.MILLISECONDS);
} catch (Exception e) {
log.error(e + ". The function timed out after [" + threadSleeptime
+ "] miliseconds before a response was received.");
} finally {
// if task has started then don't stop it
future.cancel(false);
}
}
private static String returnString() {
return "hello";
}
public static void main(String[] args) {
timeoutFunction(returnString());
}
I am a beginner and i try to teach myself clean coding. I want to pass a function as a parameter that I can reuse a method without to repeat code. As an example I have this:
public class Dog {
private String name;
private int id;
private List<String> characteristic;
public List<String> getCharacteristic() {
return characteristic;
}
public void setCharacteristic(List<String> characteristic) {
this.characteristic = characteristic;
}
}
public class Check{
private List<Dog> dogs = new ArrayList();
public void iterate() {
while (dogs.size() > 0) {
for (Dog dog : dogs) {
List<String> restChara = new ArrayList<>();
restChara= checkChara(dog, restChara);
if (restChara.size()>0) {
dog.setCharacteristic(restChara);
} else {
dogs.remove(dog);
}
}
}
}
private List<String> checkChara(Dog dog, List<String> restChara) {
for (String chara : dog.getCharacteristic()) {
boolean charaChecked = doSomething(chara);
if (!charaChecked) {
restChara.add(chara);
} else {
dog.getCharacteristic().remove(chara);
}
}
return restChara;
}
private boolean doSomething(String chara){
//do sth.
return true;
}
private boolean doSomething2(String chara){
//do sth.
return true;
}
}
How would you define the method checkChara in order to use different functions within it?
My first thought was to pass the function as a parameter (i think it would be in C# delegates)
Thank you very much!
EDIT:
I think I found another pattern strategy design pattern
https://www.freecodecamp.org/news/the-strategy-pattern-explained-using-java-bc30542204e0/
Java does not support “directly” nested methods. Many functional programming languages support method within method. But you can achieve nested method functionality in Java 7 or older version by define local classes, class within method so this does compile. And in java 8 and newer version you achieve it by lambda expression.
Method 1 (Using anonymous subclasses)
It is an inner class without a name and for which only a single object is created. An anonymous inner class can be useful when making an instance of an object with certain “extras” such as overloading methods of a class or interface, without having to actually subclass a class.
//Java program implements method inside method
public class GFG {
// create a local interface with one abstract
// method run()
interface myInterface {
void run();
}
// function have implements another function run()
static void Foo()
{
// implement run method inside Foo() function
myInterface r = new myInterface() {
public void run()
{
System.out.println("geeksforgeeks");
};
};
r.run();
}
public static void main(String[] args)
{
Foo();
}
}
Method 2 (Using local classes)
You can also implement a method inside a local class. A class created inside a method is called local inner class. If you want to invoke the methods of local inner class, you must instantiate this class inside method.
// Java program implements method inside method
public class GFG {
// function have implementation of another
// function inside local class
static void Foo()
{
// local class
class Local {
void fun()
{
System.out.println("geeksforgeeks");
}
}
new Local().fun();
}
public static void main(String[] args)
{
Foo();
}
}
Method 3 (Using a lambda expression)
Lambda expressions basically express instances of functional interfaces (An interface with single abstract method is called functional interface. An example is java.lang.Runnable). lambda expressions implement the only abstract function and therefore implement functional interfaces.
// Java program implements method inside method
public class GFG {
interface myInterface {
void run();
}
// function have implements another function
// run() using Lambda expression
static void Foo()
{
// Lambda expression
myInterface r = () ->
{
System.out.println("geeksforgeeks");
};
r.run();
}
public static void main(String[] args)
{
Foo();
}
}
I don't exaclty know what you mean, but i figure out it could be something like that:
List<Runnable> runMyStuff = new ArrayList<Runnable>();
String variable= "Hallo"; //needs to be effectively final
runMyStuff.add(() -> {
System.out.println(variable);
doSomething(variable);
});
runMyStuff.add(() ->{
System.out.println("This is a test");
});
runMyStuff.add(() ->{
System.out.println("2 + 2 = " + (2+2) );
});
runMyStuff.get(0).run();
runMyStuff.get(2).run();
runMyStuff.get(0).run();
runMyStuff.get(1).run();
runMyStuff.get(2).run();
will result in :
Hallo
2 + 2 = 4
Hallo
This is a test
2 + 2 = 4
When you put variables or Passing Parameters in those runnable methods, they need to be effectively final or you pass them in a container.
You can re-run each method.
And within those Methods you can execute other methods.
NOTE:
If you want to have return Parameters you could do the same with Callable and than ".call()" instead of run.
EDIT:
Under the assumption you mean character Check or something like that
Example for charaCheck passable Method With Interface:
public interface CharacterChecker{
//is a template returns boolean, need a String param
public boolean call(String chara);
}
A method that executes a passes method of the type of "CHaracterChecker"
public static void excecutePassedMethod(CharacterChecker checker, String chara) {
System.out.println(chara + ": " + checker.call(chara));
}
Two different implementations of a "Character Checker"
CharacterChecker goodChecker = new CharacterChecker() {
#Override
public boolean call(String chara) {
return "good".equals(chara);
}
};
CharacterChecker lazyCheker = new CharacterChecker() {
#Override
public boolean call(String chara) {
return "lazy".equals(chara);
}
};
The Methods that are Passed the "method" (More like anonymos class object with the method)
excecutePassedMethod(goodChecker, "bad");
excecutePassedMethod(goodChecker, "good");
excecutePassedMethod(goodChecker, "jolly");
excecutePassedMethod(lazyCheker, "frisky");
excecutePassedMethod(lazyCheker, "lazy");
result will be:
bad: false
good: true
jolly: false
frisky: false
lazy: true
public interface Function {
double apply(double arg);
Function derivative();
String toString();
}
public interface Integrable extends Function {
Function integrate();
}
public class Sum implements Function {
private Function func1, func2;
public Sum(Function func1, Function func2) {
this.func1 = func1;
this.func2 = func2;
}
#Override
public double apply(double arg) {
return func1.apply(arg) + func2.apply(arg);
}
#Override
public Function derivative() {
return new Sum(func1.derivative(), func2.derivative());
}
#Override
public String toString() {
return func1.toString() + " + " + func2.toString();
}
#Override
public Function integrate() {
//TODO only allow if (this instanceof Integrable)
try {
return new Sum(((Integrable) func1).integrate(), ((Integrable) func2).integrate());
} catch (ClassCastException e) {
throw new RuntimeException("could not call integrate on one of the functions of this sum, as it is not of type Integrable");
}
}
}
I'm trying to make the Sum class above, but it should only be of type Integrable if both functions are also Integrable. Otherwise, it should just be a Function.
Is there any way to do this efficiently, or is it better to make it Integrable by default and check the 2 fields in integrate()?
I would say that that the parameter of Sum must take an Integrable in that case.
You can create two classes - Sum and IntegrableSum (uhh.. need a better name)
class Sum implements Function {
public Sum(Function func1, Function func2) {
....
}
}
class IntegrableSum implements Integrable {
public IntegrableSum(Integrable integrable1, Integrable integrable2) {
....
}
}
I'd suggest you create adapter classes that implement the Integrable interface, provide default implementations to the respective abstract methods of your choice. But I don't think you can create a class in Java based on conditions as classes created in Java is direct, I mean you should know what your class is all about before creation.
If we consider the following class to count object added in an HashSet :
public class CountingHashSet<E> extends HashSet<E> {
public int addCount = 0;
#Override
public boolean add(E e) {
addCount +=1;
return super.add(e);
}
#Override
public boolean addAll(Collection<?
extends E> c) {
addCount += c.size();
return super.addAll(c);
}
}
Then, the JUnit test failed :
#Test
public void testCount() {
CountingHashSet<Integer> s = new CountingHashSet<>();
s.addAll(Arrays.asList(1, 2, 3, 4, 5));
for (int i = 6; i <= 10; ++i)
s.add(i);
assertEquals(10, s.addCount);
}
I get the following :
java.lang.AssertionError: expected:<10> but was <15>
Why I get 15 ? To my mind s.addAll(myCollection) call super.addAll(c) and if I look into the source code of hashSet, I saw that addAll(c) call add(e) to add each element. But why super.addAll(c) call the add method that I redefined ? (that's why I get 15 instead of 10)
You're treating inheritance as if it were composition. It's not. The calls don't end up being "add() on the HashSet" - they end up being "add() on the current object".
But why super.addAll(c) call the add method that I redefined ?
Because that's how virtual methods behave. addAll just calls add(), which will use the most overridden implementation in the actual type. That's how polymorphism always works. Let's write a simpler example:
class Superclass {
public void foo() {
bar();
}
public void bar() {
System.out.println("Superclass.bar()");
}
}
class Subclass extends Superclass {
#Override
public void bar() {
System.out.println("Subclass.bar()");
}
}
public class Test {
public static void main(String [] args) {
Superclass x = new Subclass();
x.foo(); // Prints Subclass.bar()
}
}
Is the result of Subclass.bar() what you'd expect from this example? If so, what do you expect the difference would be in your version? Just because you're calling super.addAll() doesn't mean that the object is suddenly in "non-overriding" mode or anything like that.
That's how polymorphism works. Your object is of type CountingHashSet, so a call to add will call CountingHashSet.add, even from the super type.