@Documented @Retention(value=SOURCE) @Target(value=TYPE) public @interface InheritConstructors
TupleConstructor
.
@InheritConstructors
saves you typing some boilerplate code.
Example usage:
class Person {
String first, last
Person(String first, String last) {
this.first = first
this.last = last.toUpperCase()
}
}
@groovy.transform.InheritConstructors
class PersonAge extends Person {
int age
}
def js = new PersonAge('John', 'Smith')
js.age = 25
assert "$js.last, $js.first is $js.age years old" == 'SMITH, John is 25 years old'
for this case, the PersonAge
class will be
equivalent to the following code:
class PersonAge extends Person { PersonAge(String first, String last) { super(first, last) } int age }You may add additional constructors in addition to inherited ones. If the argument types of a supplied constructor exactly match those of a parent constructor, then that constructor won't be inherited.
Style note: Don't go overboard using this annotation.
Typical Groovy style is to use named-arg constructors when possible.
This is easy to do for Groovy objects or any objects following JavaBean
conventions. In other cases, inheriting the constructors may be useful.
However, sub-classes often introduce new properties and these are often best
set in a constructor; especially if that matches the style adopted
in parent classes. So, even for the example above, it may have been
better style to define an explicit constructor for PersonAge
that also set the age
property. Sometimes, consistent
style is much more important than saving a few keystrokes.
As another example, this:
@InheritConstructors
class CustomException extends RuntimeException { }
is equivalent to this:
class CustomException extends RuntimeException { CustomException() { super() } CustomException(String message) { super(message) } CustomException(String message, Throwable cause) { super(message, cause) } CustomException(Throwable cause) { super(cause) } }Advanced note:If you create Groovy constructors with optional arguments this leads to multiple constructors created in the byte code. The expansion to multiple constructors occurs in a later phase to this AST transformation. This means that you can't override (i.e. not inherit) the constructors with signatures that Groovy adds later. If you get it wrong you will get a compile-time error about the duplication.
More examples:
//-------------------------------------------------------------------------- import groovy.transform.InheritConstructors @InheritConstructors class MyException extends Exception { } def e = new MyException() def e1 = new MyException('message') // Other constructors are available. assert 'message' == e1.message
//-------------------------------------------------------------------------- import groovy.transform.InheritConstructors class Person { String name Person(String name) { this.name = name } } @InheritConstructors class Child extends Person {} def child = new Child('Liam') assert 'Liam' == child.name
Modifier and Type | Optional Element and Description |
---|---|
boolean |
constructorAnnotations
Whether to carry over annotations on the copied constructors.
|
boolean |
parameterAnnotations
Whether to carry over parameter annotations on the copied constructors.
|