This chapter covers the object-oriented aspects of the Groovy programming language.

1. Types

1.1. Primitive types

Groovy supports the same primitive types as defined by the Java Language Specification:

  • integral types: byte (8 bit), short (16 bit), int (32 bit) and long (64 bit)

  • floating-point types: float (32 bit) and double (64 bit)

  • the boolean type (one of true or false)

  • the char type (16 bit, usable as a numeric type, representing a UTF-16 code)

Also like Java, Groovy uses the respective wrapper classes when objects corresponding to any of the primitive types are required:

Table 1. primitive wrappers
Primitive type Wrapper class

boolean

Boolean

char

Character

short

Short

int

Integer

long

Long

float

Float

double

Double

Automatic boxing and unboxing occur when, for instance, calling a method requiring the wrapper class and passing it a primitive variable as the parameter, or vice-versa. This is similar to Java but Groovy takes the idea further.

In most scenarios, you can treat a primitive just like it was the full object wrapper equivalent. For instance, you can call .toString() or .equals(other) on a primitive. Groovy autowraps and unwraps between references and primitives as needed.

Here’s an example using int which is declared as a static field in a class (discussed shortly):

class Foo {
    static int i
}

assert Foo.class.getDeclaredField('i').type == int.class           (1)
assert Foo.i.class != int.class && Foo.i.class == Integer.class    (2)
1 Primitive type is respected in the bytecode
2 Looking at the field at runtime shows it has been autowrapped

Now you may be concerned that this means every time you use a mathematical operator on a reference to a primitive that you’ll incur the cost of unboxing and reboxing the primitive. But this is not the case, as Groovy will compile your operators into their method equivalents and uses those instead. Additionally, Groovy will automatically unbox to a primitive when calling a Java method that takes a primitive parameter and automatically box primitive method return values from Java. However, be aware there are some differences from Java’s method resolution.

1.2. Reference Types

Apart from primitives, everything else is an object and has an associated class defining its type. We’ll discuss classes, and class-related or class-like things like interfaces, traits and records shortly.

We might declare two variables, of type String and List, as follows:

String movie = 'The Matrix'
List actors = ['Keanu Reeves', 'Hugo Weaving']

1.3. Generics

Groovy carries across the same concepts with regard to generics as Java. When defining classes and methods, it is possible to use a type parameter and create a generic class, interface, method or constructor.

Usage of generic classes and methods, regardless of whether they are defined in Java or Groovy, may involve supplying a type argument.

We might declare a variable, of type "list of string", as follows:

List<String> roles = ['Trinity', 'Morpheus']

Java employs type erasure for backwards compatibility with earlier versions of Java. Dynamic Groovy can be thought of as more aggressively applying type erasure. In general, less generics type information will be checked at compile time. Groovy’s static nature employs similar checks to Java with regard to generics information.

2. Classes

Groovy classes are very similar to Java classes, and are compatible with Java ones at JVM level. They may have methods, fields and properties (think JavaBeans properties but with less boilerplate). Classes and class members can have the same modifiers (public, protected, private, static, etc.) as in Java with some minor differences at the source level which are explained shortly.

The key differences between Groovy classes and their Java counterparts are:

  • Classes or methods with no visibility modifier are automatically public (a special annotation can be used to achieve package private visibility).

  • Fields with no visibility modifier are turned into properties automatically, which results in less verbose code, since explicit getter and setter methods aren’t needed. More on this aspect will be covered in the fields and properties section.

  • Classes do not need to have the same base name as their source file definitions but it is highly recommended in most scenarios (see also the next point about scripts).

  • One source file may contain one or more classes (but if a file contains any code not in a class, it is considered a script). Scripts are just classes with some special conventions and will have the same name as their source file (so don’t include a class definition within a script having the same name as the script source file).

The following code presents an example class.

class Person {                       (1)

    String name                      (2)
    Integer age

    def increaseAge(Integer years) { (3)
        this.age += years
    }
}
1 class beginning, with the name Person
2 string field and property named name
3 method definition

2.1. Normal class

Normal classes refer to classes which are top level and concrete. This means they can be instantiated without restrictions from any other classes or scripts. This way, they can only be public (even though the public keyword may be suppressed). Classes are instantiated by calling their constructors, using the new keyword, as in the following snippet.

def p = new Person()

2.2. Inner class

Inner classes are defined within another classes. The enclosing class can use the inner class as usual. On the other side, an inner class can access members of its enclosing class, even if they are private. Classes other than the enclosing class are not allowed to access inner classes. Here is an example:

class Outer {
    private String privateStr

    def callInnerMethod() {
        new Inner().methodA()       (1)
    }

    class Inner {                   (2)
        def methodA() {
            println "${privateStr}." (3)
        }
    }
}
1 the inner class is instantiated and its method gets called
2 inner class definition, inside its enclosing class
3 even being private, a field of the enclosing class is accessed by the inner class

There are some reasons for using inner classes:

  • They increase encapsulation by hiding the inner class from other classes, which do not need to know about it. This also leads to cleaner packages and workspaces.

  • They provide a good organization, by grouping classes that are used by only one class.

  • They lead to more maintainable codes, since inner classes are near the classes that use them.

It is common for an inner class to be an implementation of some interface whose method(s) are needed by the outer class. The code below illustrates this typical usage pattern, here being used with threads.

class Outer2 {
    private String privateStr = 'some string'

    def startThread() {
       new Thread(new Inner2()).start()
    }

    class Inner2 implements Runnable {
        void run() {
            println "${privateStr}."
        }
    }
}

Note that the class Inner2 is defined only to provide an implementation of the method run to class Outer2. Anonymous inner classes help to eliminate verbosity in this case. That topic is covered shortly.

Groovy 3+ also supports Java syntax for non-static inner class instantiation, for example:

class Computer {
    class Cpu {
        int coreNumber

        Cpu(int coreNumber) {
            this.coreNumber = coreNumber
        }
    }
}

assert 4 == new Computer().new Cpu(4).coreNumber

2.2.1. Anonymous inner class

The earlier example of an inner class (Inner2) can be simplified with an anonymous inner class. The same functionality can be achieved with the following code:

class Outer3 {
    private String privateStr = 'some string'

    def startThread() {
        new Thread(new Runnable() {      (1)
            void run() {
                println "${privateStr}."
            }
        }).start()                       (2)
    }
}
1 comparing with the last example of previous section, the new Inner2() was replaced by new Runnable() along with all its implementation
2 the method start is invoked normally

Thus, there was no need to define a new class to be used just once.

2.2.2. Abstract class

Abstract classes represent generic concepts, thus, they cannot be instantiated, being created to be subclassed. Their members include fields/properties and abstract or concrete methods. Abstract methods do not have implementation, and must be implemented by concrete subclasses.

abstract class Abstract {         (1)
    String name

    abstract def abstractMethod() (2)

    def concreteMethod() {
        println 'concrete'
    }
}
1 abstract classes must be declared with abstract keyword
2 abstract methods must also be declared with abstract keyword

Abstract classes are commonly compared to interfaces. There are at least two important differences of choosing one or another. First, while abstract classes may contain fields/properties and concrete methods, interfaces may contain only abstract methods (method signatures). Moreover, one class can implement several interfaces, whereas it can extend just one class, abstract or not.

2.3. Inheritance

Inheritance in Groovy resembles inheritance in Java. It provides a mechanism for a child class (or subclass) to reuse code or properties from a parent (or super class). Classes related through inheritance form an inheritance hierarchy. Common behavior and members are pushed up the hierarchy to reduce duplication. Specializations occur in child classes.

Different forms of inheritance are supported:

  • implementation inheritance where code (methods, fields or properties) from a superclass or from one or more traits is reused by a child class

  • contract inheritance where a class promises to provide particular abstract methods defined in a superclass, or defined in one or more traits or interfaces.

2.4. Superclasses

Parent classes share visible fields, properties or methods with child classes. A child class may have at most one parent class. The extends keyword is used immediately prior to giving the superclass type.

2.5. Interfaces

An interface defines a contract that a class needs to conform to. Typically, an interface defines zero or more abstract method definitions, but does not define the method’s implementation.

Here is a Greeter interface defining one greet method:

interface Greeter {                                         (1)
    void greet(String name)                                 (2)
}
1 an interface needs to be declared using the interface keyword
2 the abstract method signature for the greet method

Such method signatures are public by default. It is an error to use protected or package-private methods in interfaces:

interface Greeter {
    protected void greet(String name)                       (1)
}
1 Using protected is a compile-time error

A class implements an interface if it defines the interface in its implements list or if any of its superclasses does:

class SystemGreeter implements Greeter {                    (1)
    void greet(String name) {                               (2)
        println "Hello $name"
    }
}

def greeter = new SystemGreeter()
assert greeter instanceof Greeter                           (3)
1 The SystemGreeter declares the Greeter interface using the implements keyword
2 Then implements the required greet method
3 Any instance of SystemGreeter is also an instance of the Greeter interface

An interface can extend another interface:

interface ExtendedGreeter extends Greeter {                 (1)
    void sayBye(String name)
}
1 the ExtendedGreeter interface extends the Greeter interface using the extends keyword

It is worth noting that for a class to be an instance of an interface, it has to be explicit. For example, the following class defines the greet method as it is declared in the Greeter interface, but does not declare Greeter in its interfaces:

class DefaultGreeter {
    void greet(String name) { println "Hello" }
}

greeter = new DefaultGreeter()
assert !(greeter instanceof Greeter)

In other words, Groovy does not define structural typing. It is however possible to make an instance of an object implement an interface at runtime, using the as coercion operator:

greeter = new DefaultGreeter()                              (1)
coerced = greeter as Greeter                                (2)
assert coerced instanceof Greeter                           (3)
1 create an instance of DefaultGreeter that does not implement the interface
2 coerce the instance into a Greeter at runtime
3 the coerced instance implements the Greeter interface

You can see that there are two distinct objects: one is the source object, a DefaultGreeter instance, which does not implement the interface. The other is an instance of Greeter that delegates to the coerced object.

Groovy traits are close to interfaces, but support other important features described elsewhere in this manual. If you need more power than offered by interfaces, consider using traits.

While interfaces typically contain abstract method definitions, non-abstract methods are also possible. Variants allowed are default, static, or private:

  • Default methods provide a mechanism to evolve API functionality. You can add new functionality to existing interfaces but maintain binary compatibility with code written for older versions of those interfaces. Shared functionality for default methods may be placed in private methods.

  • Static methods provide a mechanism to associate methods directly an interface class without it impacting the OO contracts or being overridden. You could use them for writing factory or utility methods without creating an additional utility class.

3. Class members

3.1. Constructors

Constructors are special methods used to initialize an object with a specific state. As with normal methods, it is possible for a class to declare more than one constructor, so long as each constructor has a unique type signature. If an object doesn’t require any parameters during construction, it may use a no-arg constructor. If no constructors are supplied, an empty no-arg constructor will be provided by the Groovy compiler.

Groovy supports two invocation styles:

  • positional parameters are used in a similar to how you would use Java constructors

  • named parameters allow you to specify parameter names when invoking the constructor.

3.1.1. Positional parameters

To create an object by using positional parameters, the respective class needs to declare one or more constructors. In the case of multiple constructors, each must have a unique type signature. The constructors can also be added to the class using the groovy.transform.TupleConstructor annotation.

Typically, once at least one constructor is declared, the class can only be instantiated by having one of its constructors called. It is worth noting that, in this case, you can’t normally create the class with named parameters. Groovy does support named parameters so long as the class contains a no-arg constructor or provides a constructor which takes a Map argument as the first (and potentially only) argument - see the next section for details.

There are three forms of using a declared constructor. The first one is the normal Java way, with the new keyword. The others rely on coercion of lists into the desired types. In this case, it is possible to coerce with the as keyword and by statically typing the variable.

class PersonConstructor {
    String name
    Integer age

    PersonConstructor(name, age) {          (1)
        this.name = name
        this.age = age
    }
}

def person1 = new PersonConstructor('Marie', 1)  (2)
def person2 = ['Marie', 2] as PersonConstructor  (3)
PersonConstructor person3 = ['Marie', 3]         (4)
1 Constructor declaration
2 Constructor invocation, classic Java way
3 Constructor usage, using coercion with as keyword
4 Constructor usage, using coercion in assignment

3.1.2. Named parameters

If no (or a no-arg) constructor is declared, it is possible to create objects by passing parameters in the form of a map (property/value pairs). This can be in handy in cases where one wants to allow several combinations of parameters. Otherwise, by using traditional positional parameters it would be necessary to declare all possible constructors. Having a constructor where the first (and perhaps only) argument is a Map argument is also supported - such a constructor may also be added using the groovy.transform.MapConstructor annotation.

class PersonWOConstructor {                                  (1)
    String name
    Integer age
}

def person4 = new PersonWOConstructor()                      (2)
def person5 = new PersonWOConstructor(name: 'Marie')         (3)
def person6 = new PersonWOConstructor(age: 1)                (4)
def person7 = new PersonWOConstructor(name: 'Marie', age: 2) (5)
1 No constructor declared
2 No parameters given in the instantiation
3 name parameter given in the instantiation
4 age parameter given in the instantiation
5 name and age parameters given in the instantiation

It is important to highlight, however, that this approach gives more power to the constructor caller, while imposing an increased responsibility on the caller to get the names and value types correct. Thus, if greater control is desired, declaring constructors using positional parameters might be preferred.

Notes:

  • While the example above supplied no constructor, you can also supply a no-arg constructor or a constructor where the first argument is a Map, most typically it’s the only argument.

  • When no (or a no-arg) constructor is declared, Groovy replaces the named constructor call by a call to the no-arg constructor followed by calls to the setter for each supplied named property.

  • When the first argument is a Map, Groovy combines all named parameters into a Map (regardless of ordering) and supplies the map as the first parameter. This can be a good approach if your properties are declared as final (since they will be set in the constructor rather than after the fact with setters).

  • You can support both named and positional construction by supply both positional constructors as well as a no-arg or Map constructor.

  • You can support hybrid construction by having a constructor where the first argument is a Map but there are also additional positional parameters. Use this style with caution.

3.2. Methods

Groovy methods are quite similar to other languages. Some peculiarities will be shown in the next subsections.

3.2.1. Method definition

A method is defined with a return type or with the def keyword, to make the return type untyped. A method can also receive any number of arguments, which may not have their types explicitly declared. Java modifiers can be used normally, and if no visibility modifier is provided, the method is public.

Methods in Groovy always return some value. If no return statement is provided, the value evaluated in the last line executed will be returned. For instance, note that none of the following methods uses the return keyword.

def someMethod() { 'method called' }                           (1)
String anotherMethod() { 'another method called' }             (2)
def thirdMethod(param1) { "$param1 passed" }                   (3)
static String fourthMethod(String param1) { "$param1 passed" } (4)
1 Method with no return type declared and no parameter
2 Method with explicit return type and no parameter
3 Method with a parameter with no type defined
4 Static method with a String parameter

3.2.2. Named parameters

Like constructors, normal methods can also be called with named parameters. To support this notation, a convention is used where the first argument to the method is a Map. In the method body, the parameter values can be accessed as in normal maps (map.key). If the method has just a single Map argument, all supplied parameters must be named.

def foo(Map args) { "${args.name}: ${args.age}" }
foo(name: 'Marie', age: 1)
Mixing named and positional parameters

Named parameters can be mixed with positional parameters. The same convention applies, in this case, in addition to the Map argument as the first argument, the method in question will have additional positional arguments as needed. Supplied positional parameters when calling the method must be in order. The named parameters can be in any position. They are grouped into the map and supplied as the first parameter automatically.

def foo(Map args, Integer number) { "${args.name}: ${args.age}, and the number is ${number}" }
foo(name: 'Marie', age: 1, 23)  (1)
foo(23, name: 'Marie', age: 1)  (2)
1 Method call with additional number argument of Integer type
2 Method call with changed order of arguments

If we don’t have the Map as the first argument, then a Map must be supplied for that argument instead of named parameters. Failure to do so will lead to groovy.lang.MissingMethodException:

def foo(Integer number, Map args) { "${args.name}: ${args.age}, and the number is ${number}" }
foo(name: 'Marie', age: 1, 23)  (1)
1 Method call throws groovy.lang.MissingMethodException: No signature of method: foo() is applicable for argument types: (LinkedHashMap, Integer) values: [[name:Marie, age:1], 23], because the named argument Map parameter is not defined as the first argument

Above exception can be avoided if we replace named arguments with an explicit Map argument:

def foo(Integer number, Map args) { "${args.name}: ${args.age}, and the number is ${number}" }
foo(23, [name: 'Marie', age: 1])  (1)
1 Explicit Map argument in place of named arguments makes invocation valid
Although Groovy allows you to mix named and positional parameters, it can lead to unnecessary confusion. Mix named and positional arguments with caution.

3.2.3. Default arguments

Default arguments make parameters optional. If the argument is not supplied, the method assumes a default value.

def foo(String par1, Integer par2 = 1) { [name: par1, age: par2] }
assert foo('Marie').age == 1

Parameters are dropped from the right, however mandatory parameters are never dropped.

def baz(a = 'a', int b, c = 'c', boolean d, e = 'e') { "$a $b $c $d $e" }

assert baz(42, true) == 'a 42 c true e'
assert baz('A', 42, true) == 'A 42 c true e'
assert baz('A', 42, 'C', true) == 'A 42 C true e'
assert baz('A', 42, 'C', true, 'E') == 'A 42 C true E'

The same rule applies to constructors as well as methods. If using @TupleConstructor, additional configuration options apply.

3.2.4. Varargs

Groovy supports methods with a variable number of arguments. They are defined like this: def foo(p1, …​, pn, T…​ args). Here foo supports n arguments by default, but also an unspecified number of further arguments exceeding n.

def foo(Object... args) { args.length }
assert foo() == 0
assert foo(1) == 1
assert foo(1, 2) == 2

This example defines a method foo, that can take any number of arguments, including no arguments at all. args.length will return the number of arguments given. Groovy allows T[] as an alternative notation to T…​. That means any method with an array as last parameter is seen by Groovy as a method that can take a variable number of arguments.

def foo(Object[] args) { args.length }
assert foo() == 0
assert foo(1) == 1
assert foo(1, 2) == 2

If a method with varargs is called with null as the vararg parameter, then the argument will be null and not an array of length one with null as the only element.

def foo(Object... args) { args }
assert foo(null) == null

If a varargs method is called with an array as an argument, then the argument will be that array instead of an array of length one containing the given array as the only element.

def foo(Object... args) { args }
Integer[] ints = [1, 2]
assert foo(ints) == [1, 2]

Another important point are varargs in combination with method overloading. In case of method overloading Groovy will select the most specific method. For example if a method foo takes a varargs argument of type T and another method foo also takes one argument of type T, the second method is preferred.

def foo(Object... args) { 1 }
def foo(Object x) { 2 }
assert foo() == 1
assert foo(1) == 2
assert foo(1, 2) == 1

3.2.5. Method selection algorithm

Dynamic Groovy supports multiple dispatch (aka multimethods). When calling a method, the actual method invoked is determined dynamically based on the run-time type of methods arguments. First the method name and number of arguments will be considered (including allowance for varargs), and then the type of each argument. Consider the following method definitions:

def method(Object o1, Object o2) { 'o/o' }
def method(Integer i, String  s) { 'i/s' }
def method(String  s, Integer i) { 's/i' }

Perhaps as expected, calling method with String and Integer parameters, invokes our third method definition.

assert method('foo', 42) == 's/i'

Of more interest here is when the types are not known at compile time. Perhaps the arguments are declared to be of type Object (a list of such objects in our case). Java would determine that the method(Object, Object) variant would be selected in all cases (unless casts were used) but as can be seen in the following example, Groovy uses the runtime type and will invoke each of our methods once (and normally, no casting is needed):

List<List<Object>> pairs = [['foo', 1], [2, 'bar'], [3, 4]]
assert pairs.collect { a, b -> method(a, b) } == ['s/i', 'i/s', 'o/o']

For each of the first two of our three method invocations an exact match of argument types was found. For the third invocation, an exact match of method(Integer, Integer) wasn’t found but method(Object, Object) is still valid and will be selected.

Method selection then is about finding the closest fit from valid method candidates which have compatible parameter types. So, method(Object, Object) is also valid for the first two invocations but is not as close a match as the variants where types exactly match. To determine the closest fit, the runtime has a notion of the distance an actual argument type is away from the declared parameter type and tries to minimise the total distance across all parameters.

The following table illustrates some factors which affect the distance calculation.

Aspect Example

Directly implemented interfaces match more closely than ones from further up the inheritance hierarchy.

Given these interface and method definitions:

interface I1 {}
interface I2 extends I1 {}
interface I3 {}
class Clazz implements I3, I2 {}

def method(I1 i1) { 'I1' }
def method(I3 i3) { 'I3' }

The directly implemented interface will match:

assert method(new Clazz()) == 'I3'

An Object array is preferred over an Object.

 
def method(Object[] arg) { 'array' }
def method(Object arg) { 'object' }

assert method([] as Object[]) == 'array'

Non-vararg variants are favored over vararg variants.

 
def method(String s, Object... vargs) { 'vararg' }
def method(String s) { 'non-vararg' }

assert method('foo') == 'non-vararg'

If two vararg variants are applicable, the one which uses the minimum number of vararg arguments is preferred.

 
def method(String s, Object... vargs) { 'two vargs' }
def method(String s, Integer i, Object... vargs) { 'one varg' }

assert method('foo', 35, new Date()) == 'one varg'

Interfaces are preferred over super classes.

 
interface I {}
class Base {}
class Child extends Base implements I {}

def method(Base b) { 'superclass' }
def method(I i) { 'interface' }

assert method(new Child()) == 'interface'

For a primitive argument type, a declared parameter type which is the same or slightly larger is preferred.

 
def method(Long l) { 'Long' }
def method(Short s) { 'Short' }
def method(BigInteger bi) { 'BigInteger' }

assert method(35) == 'Long'

In the case where two variants have exactly the same distance, this is deemed ambiguous and will cause a runtime exception:

def method(Date d, Object o) { 'd/o' }
def method(Object o, String s) { 'o/s' }

def ex = shouldFail {
    println method(new Date(), 'baz')
}
assert ex.message.contains('Ambiguous method overloading')

Casting can be used to select the desired method:

assert method(new Date(), (Object)'baz') == 'd/o'
assert method((Object)new Date(), 'baz') == 'o/s'

3.2.6. Exception declaration

Groovy automatically allows you to treat checked exceptions like unchecked exceptions. This means that you don’t need to declare any checked exceptions that a method may throw as shown in the following example which can throw a FileNotFoundException if the file isn’t found:

def badRead() {
    new File('doesNotExist.txt').text
}

shouldFail(FileNotFoundException) {
    badRead()
}

Nor will you be required to surround the call to the badRead method in the previous example within a try/catch block - though you are free to do so if you wish.

If you wish to declare any exceptions that your code might throw (checked or otherwise) you are free to do so. Adding exceptions won’t change how the code is used from any other Groovy code but can be seen as documentation for the human reader of your code. The exceptions will become part of the method declaration in the bytecode, so if your code might be called from Java, it might be useful to include them. Using an explicit checked exception declaration is illustrated in the following example:

def badRead() throws FileNotFoundException {
    new File('doesNotExist.txt').text
}

shouldFail(FileNotFoundException) {
    badRead()
}

3.3. Fields and Properties

3.3.1. Fields

A field is a member of a class, interface or trait which stores data. A field defined in a Groovy source file has:

  • a mandatory access modifier (public, protected, or private)

  • one or more optional modifiers (static, final, synchronized)

  • an optional type

  • a mandatory name

class Data {
    private int id                                  (1)
    protected String description                    (2)
    public static final boolean DEBUG = false       (3)
}
1 a private field named id, of type int
2 a protected field named description, of type String
3 a public static final field named DEBUG of type boolean

A field may be initialized directly at declaration:

class Data {
    private String id = IDGenerator.next() (1)
    // ...
}
1 the private field id is initialized with IDGenerator.next()

It is possible to omit the type declaration of a field. This is however considered a bad practice and in general it is a good idea to use strong typing for fields:

class BadPractice {
    private mapping                         (1)
}
class GoodPractice {
    private Map<String,String> mapping      (2)
}
1 the field mapping doesn’t declare a type
2 the field mapping has a strong type

The difference between the two is important if you want to use optional type checking later. It is also important as a way to document the class design. However, in some cases like scripting or if you want to rely on duck typing it may be useful to omit the type.

3.3.2. Properties

A property is an externally visible feature of a class. Rather than just using a public field to represent such features (which provides a more limited abstraction and would restrict refactoring possibilities), the typical approach in Java is to follow the conventions outlined in the JavaBeans Specification, i.e. represent the property using a combination of a private backing field and getters/setters. Groovy follows these same conventions but provides a simpler way to define the property. You can define a property with:

  • an absent access modifier (no public, protected or private)

  • one or more optional modifiers (static, final, synchronized)

  • an optional type

  • a mandatory name

Groovy will then generate the getters/setters appropriately. For example:

class Person {
    String name                             (1)
    int age                                 (2)
}
1 creates a backing private String name field, a getName and a setName method
2 creates a backing private int age field, a getAge and a setAge method

If a property is declared final, no setter is generated:

class Person {
    final String name                   (1)
    final int age                       (2)
    Person(String name, int age) {
        this.name = name                (3)
        this.age = age                  (4)
    }
}
1 defines a read-only property of type String
2 defines a read-only property of type int
3 assigns the name parameter to the name field
4 assigns the age parameter to the age field

Properties are accessed by name and will call the getter or setter transparently, unless the code is in the class which defines the property:

class Person {
    String name
    void name(String name) {
        this.name = "Wonder $name"      (1)
    }
    String title() {
        this.name                       (2)
    }
}
def p = new Person()
p.name = 'Diana'                        (3)
assert p.name == 'Diana'                (4)
p.name('Woman')                         (5)
assert p.title() == 'Wonder Woman'      (6)
1 this.name will directly access the field because the property is accessed from within the class that defines it
2 similarly a read access is done directly on the name field
3 write access to the property is done outside of the Person class so it will implicitly call setName
4 read access to the property is done outside of the Person class so it will implicitly call getName
5 this will call the name method on Person which performs a direct access to the field
6 this will call the title method on Person which performs a direct read access to the field

It is worth noting that this behavior of accessing the backing field directly is done in order to prevent a stack overflow when using the property access syntax within a class that defines the property.

It is possible to list the properties of a class thanks to the meta properties field of an instance:

class Person {
    String name
    int age
}
def p = new Person()
assert p.properties.keySet().containsAll(['name','age'])

By convention, Groovy will recognize properties even if there is no backing field provided there are getters or setters that follow the Java Beans specification. For example:

class PseudoProperties {
    // a pseudo property "name"
    void setName(String name) {}
    String getName() {}

    // a pseudo read-only property "age"
    int getAge() { 42 }

    // a pseudo write-only property "groovy"
    void setGroovy(boolean groovy) {  }
}
def p = new PseudoProperties()
p.name = 'Foo'                      (1)
assert p.age == 42                  (2)
p.groovy = true                     (3)
1 writing p.name is allowed because there is a pseudo-property name
2 reading p.age is allowed because there is a pseudo-readonly property age
3 writing p.groovy is allowed because there is a pseudo-write-only property groovy

This syntactic sugar is at the core of many DSLs written in Groovy.

Property naming conventions

It is generally recommended that the first two letters of a property name are lowercase and for multi-word properties that camel case is used. In those cases, generated getters and setters will have a name formed by capitalizing the property name and adding a get or set prefix (or optionally "is" for a boolean getter). So, getLength would be a getter for a length property and setFirstName a setter for a firstName property. isEmpty might be the getter method name for a property named empty.

Property names starting with a capital letter would have getters/setters with just the prefix added. So, the property Foo is allowed even though it isn’t following the recommended naming conventions. For this property, the accessor methods would be setFoo and getFoo. A consequence of this is that you aren’t allowed to have both a foo and a Foo property, since they would have the same named accessor methods.

The JavaBeans specification makes a special case for properties which typically might be acronyms. If the first two letters of a property name are uppercase, no capitalization is performed (or more importantly, no decapitalization is done if generating the property name from the accessor method name). So, getURL would be the getter for a URL property.

Because of the special "acronym handling" property naming logic in the JavaBeans specification, the conversion to and from a property name are non-symmetrical. This leads to some strange edge cases. Groovy adopts a naming convention that avoids one ambiguity that might seem a little strange but was popular at the time of Groovy’s design and has remained (so far) for historical reasons. Groovy looks at the second letter of a property name. If that is a capital, the property is deemed to be one of the acronym style properties and no capitalization is done, otherwise normal capitalization is done. Although we never recommend it, it does allow you to have what might seem like "duplicate named" properties, e.g. you can have aProp and AProp, or pNAME and PNAME. The getters would be getaProp and getAProp, and getpNAME and getPNAME respectively.
Modifiers on a property

We have already seen that properties are defined by omitting the visibility modifier. In general, any other modifiers, e.g. transient would be copied across to the field. Two special cases are worth noting:

  • final, which we saw earlier is for read-only properties, is copied onto the backing field but also causes no setter to be defined

  • static is copied onto the backing field but also causes the accessor methods to be static

If you wish to have a modifier like final also carried over to the accessor methods, you can write your properties long hand or consider using a split property definition.

Annotations on a property

Annotations, including those associated with AST transforms, are copied on to the backing field for the property. This allows AST transforms which are applicable to fields to be applied to properties, e.g.:

class Animal {
    int lowerCount = 0
    @Lazy String name = { lower().toUpperCase() }()
    String lower() { lowerCount++; 'sloth' }
}

def a = new Animal()
assert a.lowerCount == 0  (1)
assert a.name == 'SLOTH'  (2)
assert a.lowerCount == 1  (3)
1 Confirms no eager initialization
2 Normal property access
3 Confirms initialization upon property access
Split property definition with an explicit backing field

Groovy’s property syntax is a convenient shorthand when your class design follows certain conventions which align with common JavaBean practice. If your class doesn’t exactly fit these conventions, you can certainly write the getter, setter and backing field long hand like you would in Java. However, Groovy does provide a split definition capability which still provides a shortened syntax while allowing slight adjustments to the conventions. For a split definition, you write a field and a property with the same name and type. Only one of the field or property may have an initial value.

For split properties, annotations on the field remain on the backing field for the property. Annotations on the property part of the definition are copied onto the getter and setter methods.

This mechanism allows a number of common variations that property users may wish to use if the standard property definition doesn’t exactly fit their needs. For example, if the backing field should be protected rather than private:

class HasPropertyWithProtectedField {
    protected String name  (1)
    String name            (2)
}
1 Protected backing field for name property instead of normal private one
2 Declare name property

Or, the same example but with a package-private backing field:

class HasPropertyWithPackagePrivateField {
    String name                (1)
    @PackageScope String name  (2)
}
1 Declare name property
2 Package-private backing field for name property instead of normal private one

As a final example, we may wish to apply method-related AST transforms, or in general, any annotation to the setters/getters, e.g. to have the accessors be synchronized:

class HasPropertyWithSynchronizedAccessorMethods {
    private String name        (1)
    @Synchronized String name  (2)
}
1 Backing field for name property
2 Declare name property with annotation for setter/getter
Explicit accessor methods

The automatic generation of accessor methods doesn’t occur if there is an explicit definition of the getter or setter in the class. This allows you to modify the normal behavior of such a getter or setter if needed. Inherited accessor methods aren’t normally considered but if an inherited accessor method is marked final, that will also cause no generation of an additional accessor method to honor the final requirement of no subclassing of such methods.

4. Annotations

4.1. Annotation definition

An annotation is a kind of special interface dedicated at annotating elements of the code. An annotation is a type which superinterface is the java.lang.annotation.Annotation interface. Annotations are declared in a very similar way to interfaces, using the @interface keyword:

@interface SomeAnnotation {}

An annotation may define members in the form of methods without bodies and an optional default value. The possible member types are limited to:

For example:

@interface SomeAnnotation {
    String value()                          (1)
}
@interface SomeAnnotation {
    String value() default 'something'      (2)
}
@interface SomeAnnotation {
    int step()                              (3)
}
@interface SomeAnnotation {
    Class appliesTo()                       (4)
}
@interface SomeAnnotation {}
@interface SomeAnnotations {
    SomeAnnotation[] value()                (5)
}
enum DayOfWeek { mon, tue, wed, thu, fri, sat, sun }
@interface Scheduled {
    DayOfWeek dayOfWeek()                   (6)
}
1 an annotation defining a value member of type String
2 an annotation defining a value member of type String with a default value of something
3 an annotation defining a step member of type the primitive type int
4 an annotation defining a appliesTo member of type Class
5 an annotation defining a value member which type is an array of another annotation type
6 an annotation defining a dayOfWeek member which type is the enumeration type DayOfWeek

Unlike in the Java language, in Groovy, an annotation can be used to alter the semantics of the language. It is especially true of AST transformations which will generate code based on annotations.

4.1.1. Annotation placement

An annotation can be applied on various elements of the code:

@SomeAnnotation                 (1)
void someMethod() {
    // ...
}

@SomeAnnotation                 (2)
class SomeClass {}

@SomeAnnotation String var      (3)
1 @SomeAnnotation applies to the someMethod method
2 @SomeAnnotation applies to the SomeClass class
3 @SomeAnnotation applies to the var variable

In order to limit the scope where an annotation can be applied, it is necessary to declare it on the annotation definition, using the java.lang.annotation.Target annotation. For example, here is how you would declare that an annotation can be applied to a class or a method:

import java.lang.annotation.ElementType
import java.lang.annotation.Target

@Target([ElementType.METHOD, ElementType.TYPE])     (1)
@interface SomeAnnotation {}                        (2)
1 the @Target annotation is meant to annotate an annotation with a scope.
2 @SomeAnnotation will therefore only be allowed on TYPE or METHOD

The list of possible targets is available in the java.lang.annotation.ElementType.

Groovy does not support the java.lang.annotation.ElementType#TYPE_PARAMETER and java.lang.annotation.ElementType#TYPE_PARAMETER element types which were introduced in Java 8.

4.1.2. Annotation member values

When an annotation is used, it is required to set at least all members that do not have a default value. For example:

@interface Page {
    int statusCode()
}

@Page(statusCode=404)
void notFound() {
    // ...
}

However it is possible to omit value= in the declaration of the value of an annotation if the member value is the only one being set:

@interface Page {
    String value()
    int statusCode() default 200
}

@Page(value='/home')                    (1)
void home() {
    // ...
}

@Page('/users')                         (2)
void userList() {
    // ...
}

@Page(value='error',statusCode=404)     (3)
void notFound() {
    // ...
}
1 we can omit the statusCode because it has a default value, but value needs to be set
2 since value is the only mandatory member without a default, we can omit value=
3 if both value and statusCode need to be set, it is required to use value= for the default value member

4.1.3. Retention policy

The visibility of an annotation depends on its retention policy. The retention policy of an annotation is set using the java.lang.annotation.Retention annotation:

import java.lang.annotation.Retention
import java.lang.annotation.RetentionPolicy

@Retention(RetentionPolicy.SOURCE)                   (1)
@interface SomeAnnotation {}                         (2)
1 the @Retention annotation annotates the @SomeAnnotation annotation
2 so @SomeAnnotation will have a SOURCE retention

The list of possible retention targets and description is available in the java.lang.annotation.RetentionPolicy enumeration. The choice usually depends on whether you want an annotation to be visible at compile time or runtime.

4.1.4. Closure annotation parameters

An interesting feature of annotations in Groovy is that you can use a closure as an annotation value. Therefore annotations may be used with a wide variety of expressions and still have IDE support. For example, imagine a framework where you want to execute some methods based on environmental constraints like the JDK version or the OS. One could write the following code:

class Tasks {
    Set result = []
    void alwaysExecuted() {
        result << 1
    }
    @OnlyIf({ jdk>=6 })
    void supportedOnlyInJDK6() {
        result << 'JDK 6'
    }
    @OnlyIf({ jdk>=7 && windows })
    void requiresJDK7AndWindows() {
        result << 'JDK 7 Windows'
    }
}

For the @OnlyIf annotation to accept a Closure as an argument, you only have to declare the value as a Class:

@Retention(RetentionPolicy.RUNTIME)
@interface OnlyIf {
    Class value()                    (1)
}

To complete the example, let’s write a sample runner that would use that information:

class Runner {
    static <T> T run(Class<T> taskClass) {
        def tasks = taskClass.newInstance()                                         (1)
        def params = [jdk: 6, windows: false]                                       (2)
        tasks.class.declaredMethods.each { m ->                                     (3)
            if (Modifier.isPublic(m.modifiers) && m.parameterTypes.length == 0) {   (4)
                def onlyIf = m.getAnnotation(OnlyIf)                                (5)
                if (onlyIf) {
                    Closure cl = onlyIf.value().newInstance(tasks,tasks)            (6)
                    cl.delegate = params                                            (7)
                    if (cl()) {                                                     (8)
                        m.invoke(tasks)                                             (9)
                    }
                } else {
                    m.invoke(tasks)                                                 (10)
                }
            }
        }
        tasks                                                                       (11)
    }
}
1 create a new instance of the class passed as an argument (the task class)
2 emulate an environment which is JDK 6 and not Windows
3 iterate on all declared methods of the task class
4 if the method is public and takes no arguments
5 try to find the @OnlyIf annotation
6 if it is found get the value and create a new Closure out of it
7 set the delegate of the closure to our environment variable
8 call the closure, which is the annotation closure. It will return a boolean
9 if it is true, call the method
10 if the method is not annotated with @OnlyIf, execute the method anyway
11 after that, return the task object

Then the runner can be used this way:

def tasks = Runner.run(Tasks)
assert tasks.result == [1, 'JDK 6'] as Set

4.2. Meta-annotations

4.2.1. Declaring meta-annotations

Meta-annotations, also known as annotation aliases are annotations that are replaced at compile time by other annotations (one meta-annotation is an alias for one or more annotations). Meta-annotations can be used to reduce the size of code involving multiple annotations.

Let’s start with a simple example. Imagine you have the @Service and @Transactional annotations and that you want to annotate a class with both:

@Service
@Transactional
class MyTransactionalService {}

Given the multiplication of annotations that you could add to the same class, a meta-annotation could help by reducing the two annotations with a single one having the very same semantics. For example, we might want to write this instead:

@TransactionalService                           (1)
class MyTransactionalService {}
1 @TransactionalService is a meta-annotation

A meta-annotation is declared as a regular annotation but annotated with @AnnotationCollector and the list of annotations it is collecting. In our case, the @TransactionalService annotation can be written:

import groovy.transform.AnnotationCollector

@Service                                        (1)
@Transactional                                  (2)
@AnnotationCollector                            (3)
@interface TransactionalService {
}
1 annotate the meta-annotation with @Service
2 annotate the meta-annotation with @Transactional
3 annotate the meta-annotation with @AnnotationCollector

4.2.2. Behavior of meta-annotations

Groovy supports both precompiled and source form meta-annotations. This means that your meta-annotation may be precompiled, or you can have it in the same source tree as the one you are currently compiling.

INFO: Meta-annotations are a Groovy-only feature. There is no chance for you to annotate a Java class with a meta-annotation and hope it will do the same as in Groovy. Likewise, you cannot write a meta-annotation in Java: both the meta-annotation definition and usage have to be Groovy code. But you can happily collect Java annotations and Groovy annotations within your meta-annotation.

When the Groovy compiler encounters a class annotated with a meta-annotation, it replaces it with the collected annotations. So, in our previous example, it will replace @TransactionalService with @Transactional and @Service:

def annotations = MyTransactionalService.annotations*.annotationType()
assert (Service in annotations)
assert (Transactional in annotations)

The conversion from a meta-annotation to the collected annotations is performed during the semantic analysis compilation phase. 

In addition to replacing the alias with the collected annotations, a meta-annotation is capable of processing them, including arguments.

4.2.3. Meta-annotation parameters

Meta-annotations can collect annotations which have parameters. To illustrate this, we will imagine two annotations, each of them accepting one argument:

@Timeout(after=3600)
@Dangerous(type='explosive')

And suppose that you want to create a meta-annotation named @Explosive:

@Timeout(after=3600)
@Dangerous(type='explosive')
@AnnotationCollector
public @interface Explosive {}

By default, when the annotations are replaced, they will get the annotation parameter values as they were defined in the alias. More interesting, the meta-annotation supports overriding specific values:

@Explosive(after=0)                 (1)
class Bomb {}
1 the after value provided as a parameter to @Explosive overrides the one defined in the @Timeout annotation

If two annotations define the same parameter name, the default processor will copy the annotation value to all annotations that accept this parameter:

@Retention(RetentionPolicy.RUNTIME)
public @interface Foo {
   String value()                                   (1)
}
@Retention(RetentionPolicy.RUNTIME)
public @interface Bar {
    String value()                                  (2)
}

@Foo
@Bar
@AnnotationCollector
public @interface FooBar {}                         (3)

@Foo('a')
@Bar('b')
class Bob {}                                        (4)

assert Bob.getAnnotation(Foo).value() == 'a'        (5)
println Bob.getAnnotation(Bar).value() == 'b'       (6)

@FooBar('a')
class Joe {}                                        (7)
assert Joe.getAnnotation(Foo).value() == 'a'        (8)
println Joe.getAnnotation(Bar).value() == 'a'       (9)
1 the @Foo annotation defines the value member of type String
2 the @Bar annotation also defines the value member of type String
3 the @FooBar meta-annotation aggregates @Foo and @Bar
4 class Bob is annotated with @Foo and @Bar
5 the value of the @Foo annotation on Bob is a
6 while the value of the @Bar annotation on Bob is b
7 class Joe is annotated with @FooBar
8 then the value of the @Foo annotation on Joe is a
9 and the value of the @Bar annotation on Joe is also a

In the second case, the meta-annotation value was copied in both @Foo and @Bar annotations.

It is a compile time error if the collected annotations define the same members with incompatible types. For example if on the previous example @Foo defined a value of type String but @Bar defined a value of type int.

It is however possible to customize the behavior of meta-annotations and describe how collected annotations are expanded. We’ll look at how to do that shortly but first there is an advanced processing option to cover.

4.2.4. Handling duplicate annotations in meta-annotations

The @AnnotationCollector annotation supports a mode parameter which can be used to alter how the default processor handles annotation replacement in the presence of duplicate annotations.

INFO: Custom processors (discussed next) may or may not support this parameter.

As an example, suppose you create a meta-annotation containing the @ToString annotation and then place your meta-annotation on a class that already has an explicit @ToString annotation. Should this be an error? Should both annotations be applied? Does one take priority over the other? There is no correct answer. In some scenarios it might be quite appropriate for any of these answers to be correct. So, rather than trying to preempt one correct way to handle the duplicate annotation issue, Groovy lets you write your own custom meta-annotation processors (covered next) and lets you write whatever checking logic you like within AST transforms - which are a frequent target for aggregating. Having said that, by simply setting the mode, a number of commonly expected scenarios are handled automatically for you within any extra coding. The behavior of the mode parameter is determined by the AnnotationCollectorMode enum value chosen and is summarized in the following table.

Mode

Description

DUPLICATE

Annotations from the annotation collection will always be inserted. After all transforms have been run, it will be an error if multiple annotations (excluding those with SOURCE retention) exist.

PREFER_COLLECTOR

Annotations from the collector will be added and any existing annotations with the same name will be removed.

PREFER_COLLECTOR_MERGED

Annotations from the collector will be added and any existing annotations with the same name will be removed but any new parameters found within existing annotations will be merged into the added annotation.

PREFER_EXPLICIT

Annotations from the collector will be ignored if any existing annotations with the same name are found.

PREFER_EXPLICIT_MERGED

Annotations from the collector will be ignored if any existing annotations with the same name are found but any new parameters on the collector annotation will be added to existing annotations.

4.2.5. Custom meta-annotation processors

A custom annotation processor will let you choose how to expand a meta-annotation into collected annotations. The behaviour of the meta-annotation is, in this case, totally up to you. To do this, you must:

To illustrate this, we are going to explore how the meta-annotation @CompileDynamic is implemented.

@CompileDynamic is a meta-annotation that expands itself to @CompileStatic(TypeCheckingMode.SKIP). The problem is that the default meta annotation processor doesn’t support enums and the annotation value TypeCheckingMode.SKIP is one.

The naive implementation here would not work:

@CompileStatic(TypeCheckingMode.SKIP)
@AnnotationCollector
public @interface CompileDynamic {}

Instead, we will define it like this:

@AnnotationCollector(processor = "org.codehaus.groovy.transform.CompileDynamicProcessor")
public @interface CompileDynamic {
}

The first thing you may notice is that our interface is no longer annotated with @CompileStatic. The reason for this is that we rely on the processor parameter instead, that references a class which will generate the annotation.

Here is how the custom processor is implemented:

CompileDynamicProcessor.groovy
@CompileStatic                                                                  (1)
class CompileDynamicProcessor extends AnnotationCollectorTransform {            (2)
    private static final ClassNode CS_NODE = ClassHelper.make(CompileStatic)    (3)
    private static final ClassNode TC_NODE = ClassHelper.make(TypeCheckingMode) (4)

    List<AnnotationNode> visit(AnnotationNode collector,                        (5)
                               AnnotationNode aliasAnnotationUsage,             (6)
                               AnnotatedNode aliasAnnotated,                    (7)
                               SourceUnit source) {                             (8)
        def node = new AnnotationNode(CS_NODE)                                  (9)
        def enumRef = new PropertyExpression(
            new ClassExpression(TC_NODE), "SKIP")                               (10)
        node.addMember("value", enumRef)                                        (11)
        Collections.singletonList(node)                                         (12)
    }
}
1 our custom processor is written in Groovy, and for better compilation performance, we use static compilation
2 the custom processor has to extend org.codehaus.groovy.transform.AnnotationCollectorTransform
3 create a class node representing the @CompileStatic annotation type
4 create a class node representing the TypeCheckingMode enum type
5 collector is the @AnnotationCollector node found in the meta-annotation. Usually unused.
6 aliasAnnotationUsage is the meta-annotation being expanded, here it is @CompileDynamic
7 aliasAnnotated is the node being annotated with the meta-annotation
8 sourceUnit is the SourceUnit being compiled
9 we create a new annotation node for @CompileStatic
10 we create an expression equivalent to TypeCheckingMode.SKIP
11 we add that expression to the annotation node, which is now @CompileStatic(TypeCheckingMode.SKIP)
12 return the generated annotation

In the example, the visit method is the only method which has to be overridden. It is meant to return a list of annotation nodes that will be added to the node annotated with the meta-annotation. In this example, we return a single one corresponding to @CompileStatic(TypeCheckingMode.SKIP).

5. Traits

Traits are a structural construct of the language which allows:

  • composition of behaviors

  • runtime implementation of interfaces

  • behavior overriding

  • compatibility with static type checking/compilation

They can be seen as interfaces carrying both default implementations and state. A trait is defined using the trait keyword:

trait FlyingAbility {                           (1)
        String fly() { "I'm flying!" }          (2)
}
1 declaration of a trait
2 declaration of a method inside a trait

Then it can be used like a normal interface using the implements keyword:

class Bird implements FlyingAbility {}          (1)
def b = new Bird()                              (2)
assert b.fly() == "I'm flying!"                 (3)
1 Adds the trait FlyingAbility to the Bird class capabilities
2 instantiate a new Bird
3 the Bird class automatically gets the behavior of the FlyingAbility trait

Traits allow a wide range of capabilities, from simple composition to testing, which are described thoroughly in this section.

5.1. Methods

5.1.1. Public methods

Declaring a method in a trait can be done like any regular method in a class:

trait FlyingAbility {                           (1)
        String fly() { "I'm flying!" }          (2)
}
1 declaration of a trait
2 declaration of a method inside a trait

5.1.2. Abstract methods

In addition, traits may declare abstract methods too, which therefore need to be implemented in the class implementing the trait:

trait Greetable {
    abstract String name()                              (1)
    String greeting() { "Hello, ${name()}!" }           (2)
}
1 implementing class will have to declare the name method
2 can be mixed with a concrete method

Then the trait can be used like this:

class Person implements Greetable {                     (1)
    String name() { 'Bob' }                             (2)
}

def p = new Person()
assert p.greeting() == 'Hello, Bob!'                    (3)
1 implement the trait Greetable
2 since name was abstract, it is required to implement it
3 then greeting can be called

5.1.3. Private methods

Traits may also define private methods. Those methods will not appear in the trait contract interface:

trait Greeter {
    private String greetingMessage() {                      (1)
        'Hello from a private method!'
    }
    String greet() {
        def m = greetingMessage()                           (2)
        println m
        m
    }
}
class GreetingMachine implements Greeter {}                 (3)
def g = new GreetingMachine()
assert g.greet() == "Hello from a private method!"          (4)
try {
    assert g.greetingMessage()                              (5)
} catch (MissingMethodException e) {
    println "greetingMessage is private in trait"
}
1 define a private method greetingMessage in the trait
2 the public greet message calls greetingMessage by default
3 create a class implementing the trait
4 greet can be called
5 but not greetingMessage
Traits only support public and private methods. Neither protected nor package private scopes are supported.

5.1.4. Final methods

If we have a class implementing a trait, conceptually implementations from the trait methods are "inherited" into the class. But, in reality, there is no base class containing such implementations. Rather, they are woven directly into the class. A final modifier on a method just indicates what the modifier will be for the woven method. While it would likely be considered bad style to inherit and override or multiply inherit methods with the same signature but a mix of final and non-final variants, Groovy doesn’t prohibit this scenario. Normal method selection applies and the modifier used will be determined from the resulting method. You might consider creating a base class which implements the desired trait(s) if you want trait implementation methods that can’t be overridden.

5.2. The meaning of this

this represents the implementing instance. Think of a trait as a superclass. This means that when you write:

trait Introspector {
    def whoAmI() { this }
}
class Foo implements Introspector {}
def foo = new Foo()

then calling:

foo.whoAmI()

will return the same instance:

assert foo.whoAmI().is(foo)

5.3. Interfaces

Traits may implement interfaces, in which case the interfaces are declared using the implements keyword:

interface Named {                                       (1)
    String name()
}
trait Greetable implements Named {                      (2)
    String greeting() { "Hello, ${name()}!" }
}
class Person implements Greetable {                     (3)
    String name() { 'Bob' }                             (4)
}

def p = new Person()
assert p.greeting() == 'Hello, Bob!'                    (5)
assert p instanceof Named                               (6)
assert p instanceof Greetable                           (7)
1 declaration of a normal interface
2 add Named to the list of implemented interfaces
3 declare a class that implements the Greetable trait
4 implement the missing name method
5 the greeting implementation comes from the trait
6 make sure Person implements the Named interface
7 make sure Person implements the Greetable trait

5.4. Properties

A trait may define properties, like in the following example:

trait Named {
    String name                             (1)
}
class Person implements Named {}            (2)
def p = new Person(name: 'Bob')             (3)
assert p.name == 'Bob'                      (4)
assert p.getName() == 'Bob'                 (5)
1 declare a property name inside a trait
2 declare a class which implements the trait
3 the property is automatically made visible
4 it can be accessed using the regular property accessor
5 or using the regular getter syntax

5.5. Fields

5.5.1. Private fields

Since traits allow the use of private methods, it can also be interesting to use private fields to store state. Traits will let you do that:

trait Counter {
    private int count = 0                   (1)
    int count() { count += 1; count }       (2)
}
class Foo implements Counter {}             (3)
def f = new Foo()
assert f.count() == 1                       (4)
assert f.count() == 2
1 declare a private field count inside a trait
2 declare a public method count that increments the counter and returns it
3 declare a class that implements the Counter trait
4 the count method can use the private field to keep state
This is a major difference with Java 8 virtual extension methods. While virtual extension methods do not carry state, traits can. Moreover, traits in Groovy are supported starting with Java 6, because their implementation does not rely on virtual extension methods. This means that even if a trait can be seen from a Java class as a regular interface, that interface will not have default methods, only abstract ones.

5.5.2. Public fields

Public fields work the same way as private fields, but in order to avoid the diamond problem, field names are remapped in the implementing class:

trait Named {
    public String name                      (1)
}
class Person implements Named {}            (2)
def p = new Person()                        (3)
p.Named__name = 'Bob'                       (4)
1 declare a public field inside the trait
2 declare a class implementing the trait
3 create an instance of that class
4 the public field is available, but renamed

The name of the field depends on the fully qualified name of the trait. All dots (.) in package are replaced with an underscore (_), and the final name includes a double underscore. So if the type of the field is String, the name of the package is my.package, the name of the trait is Foo and the name of the field is bar, in the implementing class, the public field will appear as:

String my_package_Foo__bar
While traits support public fields, it is not recommended to use them and considered as a bad practice.

5.6. Composition of behaviors

Traits can be used to implement multiple inheritance in a controlled way. For example, we can have the following traits:

trait FlyingAbility {                           (1)
        String fly() { "I'm flying!" }          (2)
}
trait SpeakingAbility {
    String speak() { "I'm speaking!" }
}

And a class implementing both traits:

class Duck implements FlyingAbility, SpeakingAbility {} (1)

def d = new Duck()                                      (2)
assert d.fly() == "I'm flying!"                         (3)
assert d.speak() == "I'm speaking!"                     (4)
1 the Duck class implements both FlyingAbility and SpeakingAbility
2 creates a new instance of Duck
3 we can call the method fly from FlyingAbility
4 but also the method speak from SpeakingAbility

Traits encourage the reuse of capabilities among objects, and the creation of new classes by the composition of existing behavior.

5.7. Overriding default methods

Traits provide default implementations for methods, but it is possible to override them in the implementing class. For example, we can slightly change the example above, by having a duck which quacks:

class Duck implements FlyingAbility, SpeakingAbility {
    String quack() { "Quack!" }                         (1)
    String speak() { quack() }                          (2)
}

def d = new Duck()
assert d.fly() == "I'm flying!"                         (3)
assert d.quack() == "Quack!"                            (4)
assert d.speak() == "Quack!"                            (5)
1 define a method specific to Duck, named quack
2 override the default implementation of speak so that we use quack instead
3 the duck is still flying, from the default implementation
4 quack comes from the Duck class
5 speak no longer uses the default implementation from SpeakingAbility

5.8. Extending traits

5.8.1. Simple inheritance

Traits may extend another trait, in which case you must use the extends keyword:

trait Named {
    String name                                     (1)
}
trait Polite extends Named {                        (2)
    String introduce() { "Hello, I am $name" }      (3)
}
class Person implements Polite {}
def p = new Person(name: 'Alice')                   (4)
assert p.introduce() == 'Hello, I am Alice'         (5)
1 the Named trait defines a single name property
2 the Polite trait extends the Named trait
3 Polite adds a new method which has access to the name property of the super-trait
4 the name property is visible from the Person class implementing Polite
5 as is the introduce method

5.8.2. Multiple inheritance

Alternatively, a trait may extend multiple traits. In that case, all super traits must be declared in the implements clause:

trait WithId {                                      (1)
    Long id
}
trait WithName {                                    (2)
    String name
}
trait Identified implements WithId, WithName {}     (3)
1 WithId trait defines the id property
2 WithName trait defines the name property
3 Identified is a trait which inherits both WithId and WithName

5.9. Duck typing and traits

5.9.1. Dynamic code

Traits can call any dynamic code, like a normal Groovy class. This means that you can, in the body of a method, call methods which are supposed to exist in an implementing class, without having to explicitly declare them in an interface. This means that traits are fully compatible with duck typing:

trait SpeakingDuck {
    String speak() { quack() }                      (1)
}
class Duck implements SpeakingDuck {
    String methodMissing(String name, args) {
        "${name.capitalize()}!"                     (2)
    }
}
def d = new Duck()
assert d.speak() == 'Quack!'                        (3)
1 the SpeakingDuck expects the quack method to be defined
2 the Duck class does implement the method using methodMissing
3 calling the speak method triggers a call to quack which is handled by methodMissing

5.9.2. Dynamic methods in a trait

It is also possible for a trait to implement MOP methods like methodMissing or propertyMissing, in which case implementing classes will inherit the behavior from the trait, like in this example:

trait DynamicObject {                               (1)
    private Map props = [:]
    def methodMissing(String name, args) {
        name.toUpperCase()
    }
    def propertyMissing(String name) {
        props.get(name)
    }
    void setProperty(String name, Object value) {
        props.put(name, value)
    }
}

class Dynamic implements DynamicObject {
    String existingProperty = 'ok'                  (2)
    String existingMethod() { 'ok' }                (3)
}
def d = new Dynamic()
assert d.existingProperty == 'ok'                   (4)
assert d.foo == null                                (5)
d.foo = 'bar'                                       (6)
assert d.foo == 'bar'                               (7)
assert d.existingMethod() == 'ok'                   (8)
assert d.someMethod() == 'SOMEMETHOD'               (9)
1 create a trait implementing several MOP methods
2 the Dynamic class defines a property
3 the Dynamic class defines a method
4 calling an existing property will call the method from Dynamic
5 calling a non-existing property will call the method from the trait
6 will call setProperty defined on the trait
7 will call getProperty defined on the trait
8 calling an existing method on Dynamic
9 but calling a non-existing method thanks to the trait methodMissing

5.10. Multiple inheritance conflicts

5.10.1. Default conflict resolution

It is possible for a class to implement multiple traits. If some trait defines a method with the same signature as a method in another trait, we have a conflict:

trait A {
    String exec() { 'A' }               (1)
}
trait B {
    String exec() { 'B' }               (2)
}
class C implements A,B {}               (3)
1 trait A defines a method named exec returning a String
2 trait B defines the very same method
3 class C implements both traits

In this case, the default behavior is that the method from the last declared trait in the implements clause wins. Here, B is declared after A so the method from B will be picked up:

def c = new C()
assert c.exec() == 'B'

5.10.2. User conflict resolution

In case this behavior is not the one you want, you can explicitly choose which method to call using the Trait.super.foo syntax. In the example above, we can ensure the method from trait A is invoked by writing this:

class C implements A,B {
    String exec() { A.super.exec() }    (1)
}
def c = new C()
assert c.exec() == 'A'                  (2)
1 explicit call of exec from the trait A
2 calls the version from A instead of using the default resolution, which would be the one from B

5.11. Runtime implementation of traits

5.11.1. Implementing a trait at runtime

Groovy also supports implementing traits dynamically at runtime. It allows you to "decorate" an existing object using a trait. As an example, let’s start with this trait and the following class:

trait Extra {
    String extra() { "I'm an extra method" }            (1)
}
class Something {                                       (2)
    String doSomething() { 'Something' }                (3)
}
1 the Extra trait defines an extra method
2 the Something class does not implement the Extra trait
3 Something only defines a method doSomething

Then if we do:

def s = new Something()
s.extra()

the call to extra would fail because Something is not implementing Extra. It is possible to do it at runtime with the following syntax:

def s = new Something() as Extra                        (1)
s.extra()                                               (2)
s.doSomething()                                         (3)
1 use of the as keyword to coerce an object to a trait at runtime
2 then extra can be called on the object
3 and doSomething is still callable
When coercing an object to a trait, the result of the operation is not the same instance. It is guaranteed that the coerced object will implement both the trait and the interfaces that the original object implements, but the result will not be an instance of the original class.

5.11.2. Implementing multiple traits at once

Should you need to implement several traits at once, you can use the withTraits method instead of the as keyword:

trait A { void methodFromA() {} }
trait B { void methodFromB() {} }

class C {}

def c = new C()
c.methodFromA()                     (1)
c.methodFromB()                     (2)
def d = c.withTraits A, B           (3)
d.methodFromA()                     (4)
d.methodFromB()                     (5)
1 call to methodFromA will fail because C doesn’t implement A
2 call to methodFromB will fail because C doesn’t implement B
3 withTraits will wrap c into something which implements A and B
4 methodFromA will now pass because d implements A
5 methodFromB will now pass because d also implements B
When coercing an object to multiple traits, the result of the operation is not the same instance. It is guaranteed that the coerced object will implement both the traits and the interfaces that the original object implements, but the result will not be an instance of the original class.

5.12. Chaining behavior

Groovy supports the concept of stackable traits. The idea is to delegate from one trait to the other if the current trait is not capable of handling a message. To illustrate this, let’s imagine a message handler interface like this:

interface MessageHandler {
    void on(String message, Map payload)
}

Then you can compose a message handler by applying small behaviors. For example, let’s define a default handler in the form of a trait:

trait DefaultHandler implements MessageHandler {
    void on(String message, Map payload) {
        println "Received $message with payload $payload"
    }
}

Then any class can inherit the behavior of the default handler by implementing the trait:

class SimpleHandler implements DefaultHandler {}

Now what if you want to log all messages, in addition to the default handler? One option is to write this:

class SimpleHandlerWithLogging implements DefaultHandler {
    void on(String message, Map payload) {                                  (1)
        println "Seeing $message with payload $payload"                     (2)
        DefaultHandler.super.on(message, payload)                           (3)
    }
}
1 explicitly implement the on method
2 perform logging
3 continue by delegating to the DefaultHandler trait

This works but this approach has drawbacks:

  1. the logging logic is bound to a "concrete" handler

  2. we have an explicit reference to DefaultHandler in the on method, meaning that if we happen to change the trait that our class implements, code will be broken

As an alternative, we can write another trait which responsibility is limited to logging:

trait LoggingHandler implements MessageHandler {                            (1)
    void on(String message, Map payload) {
        println "Seeing $message with payload $payload"                     (2)
        super.on(message, payload)                                          (3)
    }
}
1 the logging handler is itself a handler
2 prints the message it receives
3 then super makes it delegate the call to the next trait in the chain

Then our class can be rewritten as this:

class HandlerWithLogger implements DefaultHandler, LoggingHandler {}
def loggingHandler = new HandlerWithLogger()
loggingHandler.on('test logging', [:])

which will print:

Seeing test logging with payload [:]
Received test logging with payload [:]

As the priority rules imply that LoggerHandler wins because it is declared last, then a call to on will use the implementation from LoggingHandler. But the latter has a call to super, which means the next trait in the chain. Here, the next trait is DefaultHandler so both will be called:

The interest of this approach becomes more evident if we add a third handler, which is responsible for handling messages that start with say:

trait SayHandler implements MessageHandler {
    void on(String message, Map payload) {
        if (message.startsWith("say")) {                                    (1)
            println "I say ${message - 'say'}!"
        } else {
            super.on(message, payload)                                      (2)
        }
    }
}
1 a handler specific precondition
2 if the precondition is not met, pass the message to the next handler in the chain

Then our final handler looks like this:

class Handler implements DefaultHandler, SayHandler, LoggingHandler {}
def h = new Handler()
h.on('foo', [:])
h.on('sayHello', [:])

Which means:

  • messages will first go through the logging handler

  • the logging handler calls super which will delegate to the next handler, which is the SayHandler

  • if the message starts with say, then the handler consumes the message

  • if not, the say handler delegates to the next handler in the chain

This approach is very powerful because it allows you to write handlers that do not know each other and yet let you combine them in the order you want. For example, if we execute the code, it will print:

Seeing foo with payload [:]
Received foo with payload [:]
Seeing sayHello with payload [:]
I say Hello!

but if we move the logging handler to be the second one in the chain, the output is different:

class AlternateHandler implements DefaultHandler, LoggingHandler, SayHandler {}
h = new AlternateHandler()
h.on('foo', [:])
h.on('sayHello', [:])

prints:

Seeing foo with payload [:]
Received foo with payload [:]
I say Hello!

The reason is that now, since the SayHandler consumes the message without calling super, the logging handler is not called anymore.

5.12.1. Semantics of super inside a trait

If a class implements multiple traits and a call to an unqualified super is found, then:

  1. if the class implements another trait, the call delegates to the next trait in the chain

  2. if there isn’t any trait left in the chain, super refers to the super class of the implementing class (this)

For example, it is possible to decorate final classes thanks to this behavior:

trait Filtering {                                       (1)
    StringBuilder append(String str) {                  (2)
        def subst = str.replace('o','')                 (3)
        super.append(subst)                             (4)
    }
    String toString() { super.toString() }              (5)
}
def sb = new StringBuilder().withTraits Filtering       (6)
sb.append('Groovy')
assert sb.toString() == 'Grvy'                          (7)
1 define a trait named Filtering, supposed to be applied on a StringBuilder at runtime
2 redefine the append method
3 remove all 'o’s from the string
4 then delegate to super
5 in case toString is called, delegate to super.toString
6 runtime implementation of the Filtering trait on a StringBuilder instance
7 the string which has been appended no longer contains the letter o

In this example, when super.append is encountered, there is no other trait implemented by the target object, so the method which is called is the original append method, that is to say the one from StringBuilder. The same trick is used for toString, so that the string representation of the proxy object which is generated delegates to the toString of the StringBuilder instance.

5.13. Advanced features

5.13.1. SAM type coercion

If a trait defines a single abstract method, it is candidate for SAM (Single Abstract Method) type coercion. For example, imagine the following trait:

trait Greeter {
    String greet() { "Hello $name" }        (1)
    abstract String getName()               (2)
}
1 the greet method is not abstract and calls the abstract method getName
2 getName is an abstract method

Since getName is the single abstract method in the Greeter trait, you can write:

Greeter greeter = { 'Alice' }               (1)
1 the closure "becomes" the implementation of the getName single abstract method

or even:

void greet(Greeter g) { println g.greet() } (1)
greet { 'Alice' }                           (2)
1 the greet method accepts the SAM type Greeter as parameter
2 we can call it directly with a closure

5.13.2. Differences with Java 8 default methods

In Java 8, interfaces can have default implementations of methods. If a class implements an interface and does not provide an implementation for a default method, then the implementation from the interface is chosen. Traits behave the same but with a major difference: the implementation from the trait is always used if the class declares the trait in its interface list and that it doesn’t provide an implementation even if a super class does.

This feature can be used to compose behaviors in a very precise way, in case you want to override the behavior of an already implemented method.

To illustrate the concept, let’s start with this simple example:

import groovy.test.GroovyTestCase
import groovy.transform.CompileStatic
import org.codehaus.groovy.control.CompilerConfiguration
import org.codehaus.groovy.control.customizers.ASTTransformationCustomizer
import org.codehaus.groovy.control.customizers.ImportCustomizer

class SomeTest extends GroovyTestCase {
    def config
    def shell

    void setup() {
        config = new CompilerConfiguration()
        shell = new GroovyShell(config)
    }
    void testSomething() {
        assert shell.evaluate('1+1') == 2
    }
    void otherTest() { /* ... */ }
}

In this example, we create a simple test case which uses two properties (config and shell) and uses those in multiple test methods. Now imagine that you want to test the same, but with another distinct compiler configuration. One option is to create a subclass of SomeTest:

class AnotherTest extends SomeTest {
    void setup() {
        config = new CompilerConfiguration()
        config.addCompilationCustomizers( ... )
        shell = new GroovyShell(config)
    }
}

It works, but what if you have actually multiple test classes, and that you want to test the new configuration for all those test classes? Then you would have to create a distinct subclass for each test class:

class YetAnotherTest extends SomeTest {
    void setup() {
        config = new CompilerConfiguration()
        config.addCompilationCustomizers( ... )
        shell = new GroovyShell(config)
    }
}

Then what you see is that the setup method of both tests is the same. The idea, then, is to create a trait:

trait MyTestSupport {
    void setup() {
        config = new CompilerConfiguration()
        config.addCompilationCustomizers( new ASTTransformationCustomizer(CompileStatic) )
        shell = new GroovyShell(config)
    }
}

Then use it in the subclasses:

class AnotherTest extends SomeTest implements MyTestSupport {}
class YetAnotherTest extends SomeTest2 implements MyTestSupport {}
...

It would allow us to dramatically reduce the boilerplate code, and reduces the risk of forgetting to change the setup code in case we decide to change it. Even if setup is already implemented in the super class, since the test class declares the trait in its interface list, the behavior will be borrowed from the trait implementation!

This feature is in particular useful when you don’t have access to the super class source code. It can be used to mock methods or force a particular implementation of a method in a subclass. It lets you refactor your code to keep the overridden logic in a single trait and inherit a new behavior just by implementing it. The alternative, of course, is to override the method in every place you would have used the new code.

It’s worth noting that if you use runtime traits, the methods from the trait are always preferred to those of the proxied object: 
class Person {
    String name                                         (1)
}
trait Bob {
    String getName() { 'Bob' }                          (2)
}

def p = new Person(name: 'Alice')
assert p.name == 'Alice'                                (3)
def p2 = p as Bob                                       (4)
assert p2.name == 'Bob'                                 (5)
1 the Person class defines a name property which results in a getName method
2 Bob is a trait which defines getName as returning Bob
3 the default object will return Alice
4 p2 coerces p into Bob at runtime
5 getName returns Bob because getName is taken from the trait
Again, don’t forget that dynamic trait coercion returns a distinct object which only implements the original interfaces, as well as the traits.

5.14. Differences with mixins

There are several conceptual differences with mixins, as they are available in Groovy. Note that we are talking about runtime mixins, not the @Mixin annotation which is deprecated in favour of traits.

First of all, methods defined in a trait are visible in bytecode:

  • internally, the trait is represented as an interface (without default or static methods) and several helper classes

  • this means that an object implementing a trait effectively implements an interface

  • those methods are visible from Java

  • they are compatible with type checking and static compilation

Methods added through a mixin are, on the contrary, only visible at runtime:

class A { String methodFromA() { 'A' } }        (1)
class B { String methodFromB() { 'B' } }        (2)
A.metaClass.mixin B                             (3)
def o = new A()
assert o.methodFromA() == 'A'                   (4)
assert o.methodFromB() == 'B'                   (5)
assert o instanceof A                           (6)
assert !(o instanceof B)                        (7)
1 class A defines methodFromA
2 class B defines methodFromB
3 mixin B into A
4 we can call methodFromA
5 we can also call methodFromB
6 the object is an instance of A
7 but it’s not an instanceof B

The last point is actually a very important and illustrates a place where mixins have an advantage over traits: the instances are not modified, so if you mixin some class into another, there isn’t a third class generated, and methods which respond to A will continue responding to A even if mixed in.

5.15. Static methods, properties and fields

The following instructions are subject to caution. Static member support is work in progress and still experimental. The information below is valid for 5.0.0-alpha-9 only.

It is possible to define static methods in a trait, but it comes with numerous limitations:

  • Traits with static methods cannot be compiled statically or type checked. All static methods, properties and field are accessed dynamically (it’s a limitation from the JVM).

  • Static methods do not appear within the generated interfaces for each trait.

  • The trait is interpreted as a template for the implementing class, which means that each implementing class will get its own static methods, properties and fields. So a static member declared on a trait doesn’t belong to the Trait, but to its implementing class.

  • You should typically not mix static and instance methods of the same signature. The normal rules for applying traits apply (including multiple inheritance conflict resolution). If the method chosen is static but some implemented trait has an instance variant, a compilation error will occur. If the method chosen is the instance variant, the static variant will be ignored (the behavior is similar to static methods in Java interfaces for this case).

Let’s start with a simple example:

trait TestHelper {
    public static boolean CALLED = false        (1)
    static void init() {                        (2)
        CALLED = true                           (3)
    }
}
class Foo implements TestHelper {}
Foo.init()                                      (4)
assert Foo.TestHelper__CALLED                   (5)
1 the static field is declared in the trait
2 a static method is also declared in the trait
3 the static field is updated within the trait
4 a static method init is made available to the implementing class
5 the static field is remapped to avoid the diamond issue

As usual, it is not recommended to use public fields. Anyway, should you want this, you must understand that the following code would fail:

Foo.CALLED = true

because there is no static field CALLED defined on the trait itself. Likewise, if you have two distinct implementing classes, each one gets a distinct static field:

class Bar implements TestHelper {}              (1)
class Baz implements TestHelper {}              (2)
Bar.init()                                      (3)
assert Bar.TestHelper__CALLED                   (4)
assert !Baz.TestHelper__CALLED                  (5)
1 class Bar implements the trait
2 class Baz also implements the trait
3 init is only called on Bar
4 the static field CALLED on Bar is updated
5 but the static field CALLED on Baz is not, because it is distinct

5.16. Inheritance of state gotchas

We have seen that traits are stateful. It is possible for a trait to define fields or properties, but when a class implements a trait, it gets those fields/properties on a per-trait basis. So consider the following example:

trait IntCouple {
    int x = 1
    int y = 2
    int sum() { x+y }
}

The trait defines two properties, x and y, as well as a sum method. Now let’s create a class which implements the trait:

class BaseElem implements IntCouple {
    int f() { sum() }
}
def base = new BaseElem()
assert base.f() == 3

The result of calling f is 3, because f delegates to sum in the trait, which has state. But what if we write this instead?

class Elem implements IntCouple {
    int x = 3                                       (1)
    int y = 4                                       (2)
    int f() { sum() }                               (3)
}
def elem = new Elem()
1 Override property x
2 Override property y
3 Call sum from trait

If you call elem.f(), what is the expected output? Actually it is:

assert elem.f() == 3

The reason is that the sum method accesses the fields of the trait. So it is using the x and y values defined in the trait. If you want to use the values from the implementing class, then you need to dereference fields by using getters and setters, like in this last example:

trait IntCouple {
    int x = 1
    int y = 2
    int sum() { getX()+getY() }
}

class Elem implements IntCouple {
    int x = 3
    int y = 4
    int f() { sum() }
}
def elem = new Elem()
assert elem.f() == 7

5.17. Self types

5.17.1. Type constraints on traits

Sometimes you will want to write a trait that can only be applied to some type. For example, you may want to apply a trait on a class that extends another class which is beyond your control, and still be able to call those methods. To illustrate this, let’s start with this example:

class CommunicationService {
    static void sendMessage(String from, String to, String message) {       (1)
        println "$from sent [$message] to $to"
    }
}

class Device { String id }                                                  (2)

trait Communicating {
    void sendMessage(Device to, String message) {
        CommunicationService.sendMessage(id, to.id, message)                (3)
    }
}

class MyDevice extends Device implements Communicating {}                   (4)

def bob = new MyDevice(id:'Bob')
def alice = new MyDevice(id:'Alice')
bob.sendMessage(alice,'secret')                                             (5)
1 A Service class, beyond your control (in a library, …​) defines a sendMessage method
2 A Device class, beyond your control (in a library, …​)
3 Defines a communicating trait for devices that can call the service
4 Defines MyDevice as a communicating device
5 The method from the trait is called, and id is resolved

It is clear, here, that the Communicating trait can only apply to Device. However, there’s no explicit contract to indicate that, because traits cannot extend classes. However, the code compiles and runs perfectly fine, because id in the trait method will be resolved dynamically. The problem is that there is nothing that prevents the trait from being applied to any class which is not a Device. Any class which has an id would work, while any class that does not have an id property would cause a runtime error.

The problem is even more complex if you want to enable type checking or apply @CompileStatic on the trait: because the trait knows nothing about itself being a Device, the type checker will complain saying that it does not find the id property.

One possibility is to explicitly add a getId method in the trait, but it would not solve all issues. What if a method requires this as a parameter, and actually requires it to be a Device?

class SecurityService {
    static void check(Device d) { if (d.id==null) throw new SecurityException() }
}

If you want to be able to call this in the trait, then you will explicitly need to cast this into a Device. This can quickly become unreadable with explicit casts to this everywhere.

5.17.2. The @SelfType annotation

In order to make this contract explicit, and to make the type checker aware of the type of itself, Groovy provides a @SelfType annotation that will:

  • let you declare the types that a class that implements this trait must inherit or implement

  • throw a compile-time error if those type constraints are not satisfied

So in our previous example, we can fix the trait using the @groovy.transform.SelfType annotation:

@SelfType(Device)
@CompileStatic
trait Communicating {
    void sendMessage(Device to, String message) {
        SecurityService.check(this)
        CommunicationService.sendMessage(id, to.id, message)
    }
}

Now if you try to implement this trait on a class that is not a device, a compile-time error will occur:

class MyDevice implements Communicating {} // forgot to extend Device

The error will be:

class 'MyDevice' implements trait 'Communicating' but does not extend self type class 'Device'

In conclusion, self types are a powerful way of declaring constraints on traits without having to declare the contract directly in the trait or having to use casts everywhere, maintaining separation of concerns as tight as it should be.

5.17.3. Differences with Sealed annotation (incubating)

Both @Sealed and @SelfType restrict classes which use a trait but in orthogonal ways. Consider the following example:

interface HasHeight { double getHeight() }
interface HasArea { double getArea() }

@SelfType([HasHeight, HasArea])                       (1)
@Sealed(permittedSubclasses=[UnitCylinder,UnitCube])  (2)
trait HasVolume {
    double getVolume() { height * area }
}

final class UnitCube implements HasVolume, HasHeight, HasArea {
    // for the purposes of this example: h=1, w=1, l=1
    double height = 1d
    double area = 1d
}

final class UnitCylinder implements HasVolume, HasHeight, HasArea {
    // for the purposes of this example: h=1, diameter=1
    // radius=diameter/2, area=PI * r^2
    double height = 1d
    double area = Math.PI * 0.5d**2
}

assert new UnitCube().volume == 1d
assert new UnitCylinder().volume == 0.7853981633974483d
1 All usages of the HasVolume trait must implement or extend both HasHeight and HasArea
2 Only UnitCube or UnitCylinder can use the trait

For the degenerate case where a single class implements a trait, e.g.:

final class Foo implements FooTrait {}

Then, either:

@SelfType(Foo)
trait FooTrait {}

or:

@Sealed(permittedSubclasses='Foo') (1)
trait FooTrait {}
1 Or just @Sealed if Foo and FooTrait are in the same source file

could express this constraint. Generally, the former of these is preferred.

5.18. Limitations

5.18.1. Compatibility with AST transformations

Traits are not officially compatible with AST transformations. Some of them, like @CompileStatic will be applied on the trait itself (not on implementing classes), while others will apply on both the implementing class and the trait. There is absolutely no guarantee that an AST transformation will run on a trait as it does on a regular class, so use it at your own risk!

5.18.2. Prefix and postfix operations

Within traits, prefix and postfix operations are not allowed if they update a field of the trait:

trait Counting {
    int x
    void inc() {
        x++                             (1)
    }
    void dec() {
        --x                             (2)
    }
}
class Counter implements Counting {}
def c = new Counter()
c.inc()
1 x is defined within the trait, postfix increment is not allowed
2 x is defined within the trait, prefix decrement is not allowed

A workaround is to use the += operator instead.

6. Record classes (incubating)

Record classes, or records for short, are a special kind of class useful for modelling plain data aggregates. They provide a compact syntax with less ceremony than normal classes. Groovy already has AST transforms such as @Immutable and @Canonical which already dramatically reduce ceremony but records have been introduced in Java and record classes in Groovy are designed to align with Java record classes.

For example, suppose we want to create a Message record representing an email message. For the purposes of this example, let’s simplify such a message to contain just a from email address, a to email address, and a message body. We can define such a record as follows:

record Message(String from, String to, String body) { }

We’d use the record class in the same way as a normal class, as shown below:

def msg = new Message('me@myhost.com', 'you@yourhost.net', 'Hello!')
assert msg.toString() == 'Message[from=me@myhost.com, to=you@yourhost.net, body=Hello!]'

The reduced ceremony saves us from defining explicit fields, getters and toString, equals and hashCode methods. In fact, it’s a shorthand for the following rough equivalent:

final class Message extends Record {
    private final String from
    private final String to
    private final String body
    private static final long serialVersionUID = 0

    /* constructor(s) */

    final String toString() { /*...*/ }

    final boolean equals(Object other) { /*...*/ }

    final int hashCode() { /*...*/ }

    String from() { from }
    // other getters ...
}

Note the special naming convention for record getters. They are the same name as the field (rather than the often common JavaBean convention of capitalized with a "get" prefix). Rather than referring to a record’s fields or properties, the term component is typically used for records. So our Message record has from, to, and body components.

Like in Java, you can override the normally implicitly supplied methods by writing your own:

record Point3D(int x, int y, int z) {
    String toString() {
        "Point3D[coords=$x,$y,$z]"
    }
}

assert new Point3D(10, 20, 30).toString() == 'Point3D[coords=10,20,30]'

You can also use generics with records in the normal way. For example, consider the following Coord record definition:

record Coord<T extends Number>(T v1, T v2){
    double distFromOrigin() { Math.sqrt(v1()**2 + v2()**2 as double) }
}

It can be used as follows:

def r1 = new Coord<Integer>(3, 4)
assert r1.distFromOrigin() == 5
def r2 = new Coord<Double>(6d, 2.5d)
assert r2.distFromOrigin() == 6.5d

6.1. Special record features

6.1.1. Compact constructor

Records have an implicit constructor. This can be overridden in the normal way by providing your own constructor - you need to make sure you set all the fields if you do this. However, for succinctness, a compact constructor syntax can be used where the parameter declaration part of a normal constructor is elided. For this special case, the normal implicit constructor is still provided but is augmented by the supplied statements in the compact constructor definition:

public record Warning(String message) {
    public Warning {
        Objects.requireNonNull(message)
        message = message.toUpperCase()
    }
}

def w = new Warning('Help')
assert w.message() == 'HELP'

6.1.2. Serializability

Groovy native records follow the special conventions for serializability which apply to Java records. Groovy record-like classes (discussed below) follow normal Java class serializability conventions.

6.2. Groovy enhancements

6.2.1. Argument defaults

Groovy supports default values for constructor arguments. This capability is also available for records as shown in the following record definition which has default values for y and color:

record ColoredPoint(int x, int y = 0, String color = 'white') {}

Arguments when left off (dropping one or more arguments from the right) are replaced with their defaults values as shown in the following example:

assert new ColoredPoint(5, 5, 'black').toString() == 'ColoredPoint[x=5, y=5, color=black]'
assert new ColoredPoint(5, 5).toString() == 'ColoredPoint[x=5, y=5, color=white]'
assert new ColoredPoint(5).toString() == 'ColoredPoint[x=5, y=0, color=white]'

This processing follows normal Groovy conventions for default arguments for constructors, essentially automatically providing the constructors with the following signatures:

ColoredPoint(int, int, String)
ColoredPoint(int, int)
ColoredPoint(int)

Named arguments may also be used (default values also apply here):

assert new ColoredPoint(x: 5).toString() == 'ColoredPoint[x=5, y=0, color=white]'
assert new ColoredPoint(x: 0, y: 5).toString() == 'ColoredPoint[x=0, y=5, color=white]'

You can disable default argument processing as shown here:

@TupleConstructor(defaultsMode=DefaultsMode.OFF)
record ColoredPoint2(int x, int y, String color) {}
assert new ColoredPoint2(4, 5, 'red').toString() == 'ColoredPoint2[x=4, y=5, color=red]'

This will produce a single constructor as per the default with Java. It will be an error if you drop off arguments in this scenario.

You can force all properties to have a default value as shown here:

@TupleConstructor(defaultsMode=DefaultsMode.ON)
record ColoredPoint3(int x, int y = 0, String color = 'white') {}
assert new ColoredPoint3(y: 5).toString() == 'ColoredPoint3[x=0, y=5, color=white]'

Any property/field without an explicit initial value will be given the default value for the argument’s type (null, or zero/false for primitives).

Diving deeper

We previously described a Message record and displayed it’s rough equivalent. Groovy in fact steps through an intermediate stage where the record keyword is replaced by the class keyword and an accompanying @RecordType annotation:

@RecordType
class Message {
    String from
    String to
    String body
}

Then @RecordType itself is processed as a meta-annotation (annotation collector) and expanded into its constituent sub-annotations such as @TupleConstructor, @POJO, @RecordBase, and others. This is in some sense an implementation detail which can often be ignored. However, if you wish to customise or configure the record implementation, you may wish to drop back to the @RecordType style or augment your record class with one of the constituent sub-annotations.

6.2.2. Declarative toString customization

As per Java, you can customize a record’s toString method by writing your own. If you prefer a more declarative style, you can alternatively use Groovy’s @ToString transform to override the default record toString. As an example, you can a three-dimensional point record as follows:

package threed

import groovy.transform.ToString

@ToString(ignoreNulls=true, cache=true, includeNames=true,
          leftDelimiter='[', rightDelimiter=']', nameValueSeparator='=')
record Point(Integer x, Integer y, Integer z=null) { }

assert new Point(10, 20).toString() == 'threed.Point[x=10, y=20]'

We customise the toString by including the package name (excluded by default for records) and by caching the toString value since it won’t change for this immutable record. We are also ignoring null values (the default value for z in our definition).

We can have a similar definition for a two-dimensional point:

package twod

import groovy.transform.ToString

@ToString(ignoreNulls=true, cache=true, includeNames=true,
          leftDelimiter='[', rightDelimiter=']', nameValueSeparator='=')
record Point(Integer x, Integer y) { }

assert new Point(10, 20).toString() == 'twod.Point[x=10, y=20]'

We can see here that without the package name it would have the same toString as our previous example.

6.2.3. Obtaining a list of the record component values

We can obtain the component values from a record as a list like so:

record Point(int x, int y, String color) { }

def p = new Point(100, 200, 'green')
def (x, y, c) = p.toList()
assert x == 100
assert y == 200
assert c == 'green'

You can use @RecordOptions(toList=false) to disable this feature.

6.2.4. Obtaining a map of the record component values

We can obtain the component values from a record as a map like so:

record Point(int x, int y, String color) { }

def p = new Point(100, 200, 'green')
assert p.toMap() == [x: 100, y: 200, color: 'green']

You can use @RecordOptions(toMap=false) to disable this feature.

6.2.5. Obtaining the number of components in a record

We can obtain the number of components in a record like so:

record Point(int x, int y, String color) { }

def p = new Point(100, 200, 'green')
assert p.size() == 3

You can use @RecordOptions(size=false) to disable this feature.

6.2.6. Obtaining the nth component from a record

We can use Groovy’s normal positional indexing to obtain a particular component in a record like so:

record Point(int x, int y, String color) { }

def p = new Point(100, 200, 'green')
assert p[1] == 200

You can use @RecordOptions(getAt=false) to disable this feature.

6.3. Optional Groovy features

6.3.1. Copying

It can be useful to make a copy of a record with some components changed. This can be done using an optional copyWith method which takes named arguments. Record components are set from the supplied arguments. For components not mentioned, a (shallow) copy of the original record component is used. Here is how you might use copyWith for the Fruit record:

@RecordOptions(copyWith=true)
record Fruit(String name, double price) {}
def apple = new Fruit('Apple', 11.6)
assert 'Apple' == apple.name()
assert 11.6 == apple.price()

def orange = apple.copyWith(name: 'Orange')
assert orange.toString() == 'Fruit[name=Orange, price=11.6]'

The copyWith functionality can be disabled by setting the RecordOptions#copyWith annotation attribute to false.

6.3.2. Deep immutability

As with Java, records by default offer shallow immutability. Groovy’s @Immutable transform performs defensive copying for a range of mutable data types. Records can make use of this defensive copying to gain deep immutability as follows:

@ImmutableProperties
record Shopping(List items) {}

def items = ['bread', 'milk']
def shop = new Shopping(items)
items << 'chocolate'
assert shop.items() == ['bread', 'milk']

These examples illustrate the principal behind Groovy’s record feature offering three levels of convenience:

  • Using the record keyword for maximum succinctness

  • Supporting low-ceremony customization using declarative annotations

  • Allowing normal method implementations when full control is required

6.3.3. Obtaining the components of a record as a typed tuple

You can obtain the components of a record as a typed tuple:

import groovy.transform.*

@RecordOptions(components=true)
record Point(int x, int y, String color) { }

@CompileStatic
def method() {
    def p1 = new Point(100, 200, 'green')
    def (int x1, int y1, String c1) = p1.components()
    assert x1 == 100
    assert y1 == 200
    assert c1 == 'green'

    def p2 = new Point(10, 20, 'blue')
    def (x2, y2, c2) = p2.components()
    assert x2 * 10 == 100
    assert y2 ** 2 == 400
    assert c2.toUpperCase() == 'BLUE'

    def p3 = new Point(1, 2, 'red')
    assert p3.components() instanceof Tuple3
}

method()

Groovy has a limited number of TupleN classes. If you have a large number of components in your record, you might not be able to use this feature.

6.4. Other differences to Java

Groovy supports creating record-like classes as well as native records. Record-like classes don’t extend Java’s Record class and such classes won’t be seen by Java as records but will otherwise have similar properties.

The @RecordOptions annotation (part of @RecordType) supports a mode annotation attribute which can take one of three values (with AUTO being the default):

NATIVE

Produces a class similar to what Java would do. Produces an error when compiling on JDKs earlier than JDK16.

EMULATE

Produces a record-like class for all JDK versions.

AUTO

Produces a native record for JDK16+ and emulates the record otherwise.

Whether you use the record keyword or the @RecordType annotation is independent of the mode.

7. Sealed hierarchies (incubating)

Sealed classes, interfaces and traits restrict which subclasses can extend/implement them. Prior to sealed classes, class hierarchy designers had two main options:

  • Make a class final to allow no extension.

  • Make the class public and non-final to allow extension by anyone.

Sealed classes provide a middle-ground compared to these all or nothing choices.

Sealed classes are also more flexible than other tricks previously used to try to achieve a middle-ground. For example, for class hierarchies, access modifiers like protected and package-private give some ability to restrict inheritance hierarchies but often at the expense of flexible use of those hierarchies.

Sealed hierarchies provide full inheritance within a known hierarchy of classes, interfaces and traits but disable or only provide controlled inheritance outside the hierarchy.

As an example, suppose we want to create a shape hierarchy containing only circles and squares. We also want a shape interface to be able to refer to instances in our hierarchy. We can create the hierarchy as follows:

sealed interface ShapeI permits Circle,Square { }
final class Circle implements ShapeI { }
final class Square implements ShapeI { }

Groovy also supports an alternative annotation syntax. We think the keyword style is nicer but you might choose the annotation style if your editor doesn’t yet have Groovy 4 support.

@Sealed(permittedSubclasses=[Circle,Square]) interface ShapeI { }
final class Circle implements ShapeI { }
final class Square implements ShapeI { }

We can have a reference of type ShapeI which, thanks to the permits clause, can point to either a Circle or Square and, since our classes are final, we know no additional classes will be added to our hierarchy in the future. At least not without changing the permits clause and recompiling.

In general, we might want to have some parts of our class hierarchy immediately locked down like we have here, where we marked the subclasses as final but other times we might want to allow further controlled inheritance.

sealed class Shape permits Circle,Polygon,Rectangle { }

final class Circle extends Shape { }

class Polygon extends Shape { }
non-sealed class RegularPolygon extends Polygon { }
final class Hexagon extends Polygon { }

sealed class Rectangle extends Shape permits Square{ }
final class Square extends Rectangle { }
<Click to see the alternate annotations syntax> 
@Sealed(permittedSubclasses=[Circle,Polygon,Rectangle]) class Shape { }

final class Circle extends Shape { }

class Polygon extends Shape { }
@NonSealed class RegularPolygon extends Polygon { }
final class Hexagon extends Polygon { }

@Sealed(permittedSubclasses=Square) class Rectangle extends Shape { }
final class Square extends Rectangle { }

 
In this example, our permitted subclasses for Shape are Circle, Polygon, and Rectangle. Circle is final and hence that part of the hierarchy cannot be extended. Polygon is implicitly non-sealed and RegularPolygon is explicitly marked as non-sealed. That means our hierarchy is open to any further extension by subclassing, as seen with Polygon → RegularPolygon and RegularPolygon → Hexagon. Rectangle is itself sealed which means that part of the hierarchy can be extended but only in a controlled way (only Square is permitted).

Sealed classes are useful for creating enum-like related classes which need to contain instance specific data. For instance, we might have the following enum:

enum Weather { Rainy, Cloudy, Sunny }
def forecast = [Weather.Rainy, Weather.Sunny, Weather.Cloudy]
assert forecast.toString() == '[Rainy, Sunny, Cloudy]'

but we now wish to also add weather specific instance data to weather forecasts. We can alter our abstraction as follows:

sealed abstract class Weather { }
@Immutable(includeNames=true) class Rainy extends Weather { Integer expectedRainfall }
@Immutable(includeNames=true) class Sunny extends Weather { Integer expectedTemp }
@Immutable(includeNames=true) class Cloudy extends Weather { Integer expectedUV }
def forecast = [new Rainy(12), new Sunny(35), new Cloudy(6)]
assert forecast.toString() == '[Rainy(expectedRainfall:12), Sunny(expectedTemp:35), Cloudy(expectedUV:6)]'

Sealed hierarchies are also useful when specifying Algebraic or Abstract Data Types (ADTs) as shown in the following example:

import groovy.transform.*

sealed interface Tree<T> {}
@Singleton final class Empty implements Tree {
    String toString() { 'Empty' }
}
@Canonical final class Node<T> implements Tree<T> {
    T value
    Tree<T> left, right
}

Tree<Integer> tree = new Node<>(42, new Node<>(0, Empty.instance, Empty.instance), Empty.instance)
assert tree.toString() == 'Node(42, Node(0, Empty, Empty), Empty)'

Sealed hierarchies work well with records as shown in the following example:

sealed interface Expr {}
record ConstExpr(int i) implements Expr {}
record PlusExpr(Expr e1, Expr e2) implements Expr {}
record MinusExpr(Expr e1, Expr e2) implements Expr {}
record NegExpr(Expr e) implements Expr {}

def threePlusNegOne = new PlusExpr(new ConstExpr(3), new NegExpr(new ConstExpr(1)))
assert threePlusNegOne.toString() == 'PlusExpr[e1=ConstExpr[i=3], e2=NegExpr[e=ConstExpr[i=1]]]'

7.1. Differences to Java

  • Java provides no default modifier for subclasses of sealed classes and requires that one of final, sealed or non-sealed be specified. Groovy defaults to non-sealed but you can still use non-sealed/@NonSealed if you wish. We anticipate the style checking tool CodeNarc will eventually have a rule that looks for the presence of non-sealed so developers wanting that stricter style will be able to use CodeNarc and that rule if they want.

  • Currently, Groovy doesn’t check that all classes mentioned in permittedSubclasses are available at compile-time and compiled along with the base sealed class. This may change in a future version of Groovy.

Groovy supports annotating classes as sealed as well as "native" sealed classes.

The @SealedOptions annotation supports a mode annotation attribute which can take one of three values (with AUTO being the default):

NATIVE

Produces a class similar to what Java would do. Produces an error when compiling on JDKs earlier than JDK17.

EMULATE

Indicates the class is sealed using the @Sealed annotation. This mechanism works with the Groovy compiler for JDK8+ but is not recognised by the Java compiler.

AUTO

Produces a native record for JDK17+ and emulates the record otherwise.

Whether you use the sealed keyword or the @Sealed annotation is independent of the mode.