Lambda Expressions

Introduction#

Lambda expressions provide a clear and concise way of implementing a single-method interface using an expression. They allow you to reduce the amount of code you have to create and maintain. While similar to anonymous classes, they have no type information by themselves. Type inference needs to happen.

Method references implement functional interfaces using existing methods rather than expressions. They belong to the lambda family as well.

Syntax#

() -> { return expression; } // Zero-arity with function body to return a value.
() -> expression // Shorthand for the above declaration; there is no semicolon for expressions.
() -> { function-body } // Side-effect in the lambda expression to perform operations.
parameterName -> expression // One-arity lambda expression. In lambda expressions with only one argument, the parenthesis can be removed.
(Type parameterName, Type secondParameterName, …) -> expression // lambda evaluating an expression with parameters listed to the left
(parameterName, secondParameterName, …) -> expression // Shorthand that removes the parameter types for the parameter names. Can only be used in contexts that can be inferred by the compiler where the given parameter list size matches one (and only one) of the size of the functional interfaces expected.

Using Lambda Expressions to Sort a Collection

Sorting lists

Prior to Java 8, it was necessary to implement the java.util.Comparator interface with an anonymous (or named) class when sorting a list¹:

List<Person> people = ...
Collections.sort(
    people,
    new Comparator<Person>() {
        public int compare(Person p1, Person p2){
            return p1.getFirstName().compareTo(p2.getFirstName());
        }
    }
);

Starting with Java 8, the anonymous class can be replaced with a lambda expression. Note that the types for the parameters p1 and p2 can be left out, as the compiler will infer them automatically:

Collections.sort(
    people, 
    (p1, p2) -> p1.getFirstName().compareTo(p2.getFirstName())
);

The example can be simplified by using Comparator.comparing and method references expressed using the :: (double colon) symbol.

Collections.sort(
    people,
    Comparator.comparing(Person::getFirstName)
);

A static import allows us to express this more concisely, but it is debatable whether this improves overall readability:

import static java.util.Collections.sort;
import static java.util.Comparator.comparing;
//...
sort(people, comparing(Person::getFirstName));

Comparators built this way can also be chained together. For example, after comparing people by their first name, if there are people with the same first name, the thenComparing method with also compare by last name:

sort(people, comparing(Person::getFirstName).thenComparing(Person::getLastName));

^{1 - Note that Collections.sort(…) only works on collections that are subtypes of List. The Set and Collection APIs do not imply any ordering of the elements.}

Sorting maps

You can sort the entries of a HashMap by value in a similar fashion. (Note that a LinkedHashMap must be used as the target. The keys in an ordinary HashMap are unordered.)

Map<String, Integer> map = new HashMap();  // ... or any other Map class
// populate the map
map = map.entrySet()
    .stream()
    .sorted(Map.Entry.<String, Integer>comparingByValue())
    .collect(Collectors.toMap(k -> k.getKey(), v -> v.getValue(),
                              (k, v) -> k, LinkedHashMap::new));

Introduction to Java lambdas

Functional Interfaces

Lambdas can only operate on a functional interface, which is an interface with just one abstract method. Functional interfaces can have any number of default or static methods. (For this reason, they are sometimes referred to as Single Abstract Method Interfaces, or SAM Interfaces).

interface Foo1 {
    void bar();
}

interface Foo2 {
    int bar(boolean baz);
}

interface Foo3 {
    String bar(Object baz, int mink);
}

interface Foo4 {
    default String bar() { // default so not counted
        return "baz";
    }
    void quux();
}

When declaring a functional interface the @FunctionalInterface annotation can be added. This has no special effect, but a compiler error will be generated if this annotation is applied to an interface which is not functional, thus acting as a reminder that the interface should not be changed.

@FunctionalInterface
interface Foo5 {
    void bar();
}

@FunctionalInterface
interface BlankFoo1 extends Foo3 { // inherits abstract method from Foo3
}

@FunctionalInterface
interface Foo6 {
    void bar();
    boolean equals(Object obj); // overrides one of Object's method so not counted
}

Conversely, this is not a functional interface, as it has more than one abstract method:

interface BadFoo {
    void bar();
    void quux(); // <-- Second method prevents lambda: which one should 
                 // be considered as lambda?
}

This is also not a functional interface, as it does not have any methods:

interface BlankFoo2 { }

Take note of the following. Suppose you have

interface Parent { public int parentMethod(); }

and

interface Child extends Parent { public int ChildMethod(); }

Then Child cannot be a functional interface since it has two specified methods.

Java 8 also provides a number of generic templated functional interfaces in the package java.util.function. For example, the built-in interface Predicate<T> wraps a single method which inputs a value of type T and returns a boolean.

Lambda Expressions

The basic structure of a Lambda expression is:

fi will then hold a singleton instance of a class, similar to an anonymous class, which implements FunctionalInterface and where the one method’s definition is { System.out.println("Hello"); }. In other words, the above is mostly equivalent to:

FunctionalInterface fi = new FunctionalInterface() {
    @Override
    public void theOneMethod() {
        System.out.println("Hello");
    }
};

The lambda is only “mostly equivalent” to the anonymous class because in a lambda, the meaning of expressions like this, super or toString() reference the class within which the assignment takes place, not the newly created object.

You cannot specify the name of the method when using a lambda—but you shouldn’t need to, because a functional interface must have only one abstract method, so Java overrides that one.

In cases where the type of the lambda is not certain, (e.g. overloaded methods) you can add a cast to the lambda to tell the compiler what its type should be, like so:

Object fooHolder = (Foo1) () -> System.out.println("Hello");
System.out.println(fooHolder instanceof Foo1); // returns true

If the functional interface’s single method takes parameters, the local formal names of these should appear between the brackets of the lambda. There is no need to declare the type of the parameter or return as these are taken from the interface (although it is not an error to declare the parameter types if you want to). Thus, these two examples are equivalent:

Foo2 longFoo = new Foo2() {
    @Override
    public int bar(boolean baz) {
        return baz ? 1 : 0;
    }
};
Foo2 shortFoo = (x) -> { return x ? 1 : 0; };

The parentheses around the argument can be omitted if the function only has one argument:

Foo2 np = x -> { return x ? 1 : 0; }; // okay
Foo3 np2 = x, y -> x.toString() + y // not okay

Implicit Returns

If the code placed inside a lambda is a Java expression rather than a statement, it is treated as a method which returns the value of the expression. Thus, the following two are equivalent:

IntUnaryOperator addOneShort = (x) -> (x + 1);
IntUnaryOperator addOneLong = (x) -> { return (x + 1); };

Accessing Local Variables (value closures)

Since lambdas are syntactic shorthand for anonymous classes, they follow the same rules for accessing local variables in the enclosing scope; the variables must be treated as final and not modified inside the lambda.

IntUnaryOperator makeAdder(int amount) {
    return (x) -> (x + amount); // Legal even though amount will go out of scope
                                // because amount is not modified
}

IntUnaryOperator makeAccumulator(int value) {
    return (x) -> { value += x; return value; }; // Will not compile
}

If it is necessary to wrap a changing variable in this way, a regular object that keeps a copy of the variable should be used. Read more in https://stackoverflow.com/documentation/java/91/lambda-expressions/14441/java-closures-with-lambda-expressions

Accepting Lambdas

Because a lambda is an implementation of an interface, nothing special needs to be done to make a method accept a lambda: any function which takes a functional interface can also accept a lambda.

public void passMeALambda(Foo1 f) {
    f.bar();
}
passMeALambda(() -> System.out.println("Lambda called"));

The Type of a Lambda Expression

A lambda expression, by itself, does not have a specific type. While it is true that the types and number of parameters, along with the type of a return value can convey some type information, such information will only constrain what types it can be assigned to. The lambda receives a type when it is assigned to a functional interface type in one of the following ways:

Direct assignment to a functional type, e.g. myPredicate = s -> s.isEmpty()
Passing it as a parameter that has a functional type, e.g. stream.filter(s -> s.isEmpty())
Returning it from a function that returns a functional type, e.g. return s -> s.isEmpty()
Casting it to a functional type, e.g. (Predicate<String>) s -> s.isEmpty()

Until any such assignment to a functional type is made, the lambda does not have a definite type. To illustrate, consider the lambda expression o -> o.isEmpty(). The same lambda expression can be assigned to many different functional types:

Predicate<String> javaStringPred = o -> o.isEmpty();
Function<String, Boolean> javaFunc = o -> o.isEmpty();
Predicate<List> javaListPred = o -> o.isEmpty();
Consumer<String> javaStringConsumer = o -> o.isEmpty(); // return value is ignored!
com.google.common.base.Predicate<String> guavaPredicate = o -> o.isEmpty();

Now that they are assigned, the examples shown are of completely different types even though the lambda expressions looked the same, and they cannot be assigned to each other.

Method References

Method references allow predefined static or instance methods that adhere to a compatible functional interface to be passed as arguments instead of an anonymous lambda expression.

Assume that we have a model:

class Person {
    private final String name;
    private final String surname;

    public Person(String name, String surname){
        this.name = name;
        this.surname = surname;
    }

    public String getName(){ return name; }
    public String getSurname(){ return surname; }
}

List<Person> people = getSomePeople();

Instance method reference (to an arbitrary instance)

people.stream().map(Person::getName)

The equivalent lambda:

people.stream().map(person -> person.getName())

In this example, a method reference to the instance method getName() of type Person, is being passed. Since it’s known to be of the collection type, the method on the instance (known later) will be invoked.

Instance method reference (to a specific instance)

people.forEach(System.out::println);

Since System.out is an instance of PrintStream, a method reference to this specific instance is being passed as an argument.

The equivalent lambda:

people.forEach(person -> System.out.println(person));

Static method reference

Also for transforming streams we can apply references to static methods:

List<Integer> numbers = Arrays.asList(1, 2, 3, 4, 5, 6);
numbers.stream().map(String::valueOf)

This example passes a reference to the static valueOf() method on the String type. Therefore, the instance object in the collection is passed as an argument to valueOf().

The equivalent lambda:

 numbers.stream().map(num -> String.valueOf(num))

Reference to a constructor

List<String> strings = Arrays.asList("1", "2", "3");
strings.stream().map(Integer::new)

Read Collect Elements of a Stream into a Collection to see how to collect elements to collection.

The single String argument constructor of the Integer type is being used here, to construct an integer given the string provided as the argument. In this case, as long as the string represents a number, the stream will be mapped to Integers. The equivalent lambda:

strings.stream().map(s -> new Integer(s));

Cheat-Sheet

| Method Reference Format | Code | Equivalent Lambda | | --- | --- | --- | --- | | Static method | `TypeName::method` | `(args) -> TypeName.method(args)` | | Non-static method (on instance^*) | `instance::method` | `(args) -> instance.method(args)` | | Non-static method (no instance) | `TypeName::method` | `(instance, args) -> instance.method(args)` | | Constructor^** | `TypeName::new` | `(args) -> new TypeName(args)` | | Array constructor | `TypeName[]::new` | `(int size) -> new TypeName[size]` |

^* instance can be any expression that evaluates to a reference to an instance, e.g. getInstance()::method, this::method

^** If TypeName is a non-static inner class, constructor reference is only valid within the scope of an outer class instance

Implementing multiple interfaces

Sometimes you may want to have a lambda expression implementing more than one interface. This is mostly useful with marker interfaces (such as java.io.Serializable) since they don’t add abstract methods.

For example, you want to create a TreeSet with a custom Comparator and then serialize it and send it over the network. The trivial approach:

TreeSet<Long> ts = new TreeSet<>((x, y) -> Long.compare(y, x));

doesn’t work since the lambda for the comparator does not implement Serializable. You can fix this by using intersection types and explicitly specifying that this lambda needs to be serializable:

TreeSet<Long> ts = new TreeSet<>(
    (Comparator<Long> & Serializable) (x, y) -> Long.compare(y, x));

If you’re frequently using intersection types (for example, if you’re using a framework such as Apache Spark where almost everything has to be serializable), you can create empty interfaces and use them in your code instead:

public interface SerializableComparator extends Comparator<Long>, Serializable {}

public class CustomTreeSet {
  public CustomTreeSet(SerializableComparator comparator) {}
}

This way you’re guaranteed that the passed comparator will be serializable.

Lambdas and Execute-around Pattern

There are several good examples of using lambdas as a FunctionalInterface in simple scenarios. A fairly common use case that can be improved by lambdas is what is called the Execute-Around pattern. In this pattern, you have a set of standard setup/teardown code that is needed for multiple scenarios surrounding use case specific code. A few common example of this are file io, database io, try/catch blocks.

interface DataProcessor {
    void process( Connection connection ) throws SQLException;;
}

public void doProcessing( DataProcessor processor ) throws SQLException{
    try (Connection connection = DBUtil.getDatabaseConnection();) {
        processor.process(connection);
        connection.commit();
    } 
}

Then to call this method with a lambda it might look like:

public static void updateMyDAO(MyVO vo) throws DatabaseException {
    doProcessing((Connection conn) -> MyDAO.update(conn, ObjectMapper.map(vo)));
}

This is not limited to I/O operations. It can apply to any scenario where similar setup/tear down tasks are applicable with minor variations. The main benefit of this Pattern is code re-use and enforcing DRY (Don’t Repeat Yourself).

Using lambda expression with your own functional interface

Lambdas are meant to provide inline implementation code for single method interfaces and the ability to pass them around as we have been doing with normal variables. We call them Functional Interface.

For example, writing a Runnable in anonymous class and starting a Thread looks like:

//Old way
new Thread(
        new Runnable(){
            public void run(){
                System.out.println("run logic...");
            }
        }
).start();

//lambdas, from Java 8
new Thread(
        ()-> System.out.println("run logic...")
).start();

Now, in line with above, lets say you have some custom interface:

interface TwoArgInterface {
    int operate(int a, int b);
}

How do you use lambda to give implementation of this interface in your code? Same as Runnable example shown above. See the driver program below:

public class CustomLambda {
    public static void main(String[] args) {

        TwoArgInterface plusOperation = (a, b) -> a + b;
        TwoArgInterface divideOperation = (a,b)->{
            if (b==0) throw new IllegalArgumentException("Divisor can not be 0");
            return a/b;
        };

        System.out.println("Plus operation of 3 and 5 is: " + plusOperation.operate(3, 5));
        System.out.println("Divide operation 50 by 25 is: " + divideOperation.operate(50, 25));

    }
}

`return` only returns from the lambda, not the outer method

The return method only returns from the lambda, not the outer method.

Beware that this is different from Scala and Kotlin!

void threeTimes(IntConsumer r) {
  for (int i = 0; i < 3; i++) {
    r.accept(i);
  }
}

void demo() {
  threeTimes(i -> {
    System.out.println(i);
    return; // Return from lambda to threeTimes only!
  });
}

This can lead to unexpected behavior when attempting to write own language constructs, as in builtin constructs such as for loops return behaves differently:

void demo2() {
  for (int i = 0; i < 3; i++) {
    System.out.println(i);
    return; // Return from 'demo2' entirely
  }
}

In Scala and Kotlin, demo and demo2 would both only print 0. But this is not more consistent. The Java approach is consistent with refactoring and the use of classes - the return in the code at the top, and the code below behaves the same:

void demo3() {
  threeTimes(new MyIntConsumer());
}

class MyIntConsumer implements IntConsumer {
  public void accept(int i) {
    System.out.println(i);
    return;
  }
}

Therefore, the Java return is more consistent with class methods and refactoring, but less with the for and while builtins, these remain special.

Because of this, the following two are equivalent in Java:

IntStream.range(1, 4)
    .map(x -> x * x)
    .forEach(System.out::println);
IntStream.range(1, 4)
    .map(x -> { return x * x; })
    .forEach(System.out::println);

Furthermore, the use of try-with-resources is safe in Java:

class Resource implements AutoCloseable {
  public void close() { System.out.println("close()"); }
}

void executeAround(Consumer<Resource> f) {
  try (Resource r = new Resource()) {
    System.out.print("before ");
    f.accept(r);
    System.out.print("after ");
  }
}

void demo4() {
  executeAround(r -> {
    System.out.print("accept() ");
    return; // Does not return from demo4, but frees the resource.
  });
}

will print before accept() after close(). In the Scala and Kotlin semantics, the try-with-resources would not be closed, but it would print before accept() only.

Java Closures with lambda expressions.

A lambda closure is created when a lambda expression references the variables of an enclosing scope (global or local). The rules for doing this are the same as those for inline methods and anonymous classes.

Local variables from an enclosing scope that are used within a lambda have to be final. With Java 8 (the earliest version that supports lambdas), they don’t need to be declared final in the outside context, but must be treated that way. For example:

int n = 0; // With Java 8 there is no need to explicit final
Runnable r = () -> { // Using lambda
    int i = n;
    // do something
};

This is legal as long as the value of the n variable is not changed. If you try to change the variable, inside or outside the lambda, you will get the following compilation error:

“local variables referenced from a lambda expression must be final or effectively final“.

For example:

int n = 0;
Runnable r = () -> { // Using lambda
    int i = n;
    // do something
};
n++; // Will generate an error.

If it is necessary to use a changing variable within a lambda, the normal approach is to declare a final copy of the variable and use the copy. For example

int n = 0;
final int k = n; // With Java 8 there is no need to explicit final
Runnable r = () -> { // Using lambda
    int i = k;
    // do something
};
n++;      // Now will not generate an error
r.run();  // Will run with i = 0 because k was 0 when the lambda was created

Naturally, the body of the lambda does not see the changes to the original variable.

Note that Java does not support true closures. A Java lambda cannot be created in a way that allows it to see changes in the environment in which it was instantiated. If you want to implement a closure that observes or makes changes to its environment, you should simulate it using a regular class. For example:

// Does not compile ...
public IntUnaryOperator createAccumulator() {
    int value = 0;
    IntUnaryOperator accumulate = (x) -> { value += x; return value; };
    return accumulate;
}

The above example will not compile for reasons discussed previously. We can work around the compilation error as follows:

// Compiles, but is incorrect ...
public class AccumulatorGenerator {
    private int value = 0;

    public IntUnaryOperator createAccumulator() {
        IntUnaryOperator accumulate = (x) -> { value += x; return value; };
        return accumulate;
    }
}

The problem is that this breaks the design contract for the IntUnaryOperator interface which states that instances should be functional and stateless. If such a closure is passed to built-in functions that accept functional objects, it is liable to cause crashes or erroneous behavior. Closures that encapsulate mutable state should be implemented as regular classes. For example.

// Correct ...
public class Accumulator {
   private int value = 0;

   public int accumulate(int x) {
      value += x;
      return value;
   }
}

Lambda - Listener Example

Anonymous class listener

Before Java 8, it’s very common that an anonymous class is used to handle click event of a JButton, as shown in the following code. This example shows how to implement an anonymous listener within the scope of btn.addActionListener.

JButton btn = new JButton("My Button");
btn.addActionListener(new ActionListener() {
    @Override
    public void actionPerformed(ActionEvent e) {
        System.out.println("Button was pressed");
    }
});

Lambda listener

Because the ActionListener interface defines only one method actionPerformed(), it is a functional interface which means there’s a place to use Lambda expressions to replace the boilerplate code. The above example can be re-written using Lambda expressions as follows:

JButton btn = new JButton("My Button");
btn.addActionListener(e -> {
    System.out.println("Button was pressed");
});

Traditional style to Lambda style

Traditional way

interface MathOperation{
    boolean unaryOperation(int num);
}

public class LambdaTry {
    public static void main(String[] args) {
        MathOperation isEven = new MathOperation() {
            @Override
            public boolean unaryOperation(int num) {
                return num%2 == 0;
            }
        };
        
        System.out.println(isEven.unaryOperation(25));
        System.out.println(isEven.unaryOperation(20));
    }
}

Lambda style

Remove class name and functional interface body.

public class LambdaTry { public static void main(String[] args) { MathOperation isEven = (int num) -> { return num%2 == 0; };
```
     System.out.println(isEven.unaryOperation(25));
     System.out.println(isEven.unaryOperation(20));
 }
```
}
Optional type declaration

MathOperation isEven = (num) -> { return num%2 == 0; };
Optional parenthesis around parameter, if it is single parameter

MathOperation isEven = num -> { return num%2 == 0; };
Optional curly braces, if there is only one line in function body
Optional return keyword, if there is only one line in function body

MathOperation isEven = num -> num%2 == 0;

Lambdas and memory utilization

Since Java lambdas are closures, they can “capture” the values of variables in the enclosing lexical scope. While not all lambdas capture anything — simple lambdas like s -> s.length() capture nothing and are called stateless — capturing lambdas require a temporary object to hold the captured variables. In this code snippet, the lambda () -> j is a capturing lambda, and may cause an object to be allocated when it is evaluated:

public static void main(String[] args) throws Exception {
    for (int i = 0; i < 1000000000; i++) {
        int j = i;
        doSomethingWithLambda(() -> j);
    }
}

Although it might not be immediately obvious since the new keyword doesn’t appear anywhere in the snippet, this code is liable to create 1,000,000,000 separate objects to represent the instances of the () -> j lambda expression. However, it should also be noted that future versions of Java¹ may be able to optimize this so that at runtime the lambda instances were reused, or were represented in some other way.

^{1 - For instance, Java 9 introduces an optional “link” phase to the Java build sequence which will provide the opportunity for doing global optimizations like this.}

Using lambda expressions & predicates to get a certain value(s) from a list

Starting with Java 8, you can use lambda expressions & predicates.

Example: Use a lambda expressions & a predicate to get a certain value from a list. In this example every person will be printed out with the fact if they are 18 and older or not.

Person Class:

public class Person {
    private String name;
    private int age;

    public Person(String name, int age) {
        this.name = name;
        this.age = age;
    }

    public int getAge() { return age; }
    public String getName() { return name; }
}

The built-in interface Predicate from the java.util.function.Predicate packages is a functional interface with a boolean test(T t) method.

Example Usage:

import java.util.ArrayList;
import java.util.List;
import java.util.function.Predicate;

public class LambdaExample {
    public static void main(String[] args) {
        List<Person> personList = new ArrayList<Person>();
        personList.add(new Person("Jeroen", 20));
        personList.add(new Person("Jack", 5));
        personList.add(new Person("Lisa", 19));

        print(personList, p -> p.getAge() >= 18);
    }

    private static void print(List<Person> personList, Predicate<Person> checker) {
        for (Person person : personList) {
            if (checker.test(person)) {
                System.out.print(person + " matches your expression.");
            } else {
                System.out.println(person  + " doesn't match your expression.");
            }
        }
    }
}

The print(personList, p -> p.getAge() >= 18); method takes a lambda expression (because the Predicate is used a parameter) where you can define the expression that is needed. The checker’s test method checks if this expression is correct or not: checker.test(person).

You can easily change this to something else, for example to print(personList, p -> p.getName().startsWith("J"));. This will check if the person’s name starts with a “J”.

Introduction#

Syntax#

Using Lambda Expressions to Sort a Collection

Sorting lists

Sorting maps

Introduction to Java lambdas

Functional Interfaces

Lambda Expressions

Implicit Returns

Accessing Local Variables (value closures)

Accepting Lambdas

The Type of a Lambda Expression

Method References

Instance method reference (to an arbitrary instance)

Instance method reference (to a specific instance)

Static method reference

Reference to a constructor

Cheat-Sheet

Implementing multiple interfaces

Lambdas and Execute-around Pattern

Using lambda expression with your own functional interface

return only returns from the lambda, not the outer method

Java Closures with lambda expressions.

Lambda - Listener Example

Traditional style to Lambda style

Lambdas and memory utilization

Using lambda expressions & predicates to get a certain value(s) from a list

`return` only returns from the lambda, not the outer method