In software engineering, the delegation pattern is an object-oriented design pattern that allows object composition to achieve the same code reuse as inheritance.
In delegation, an object handles a request by delegating to a second object (the delegate). The delegate is a helper object, but with the original context. With language-level support for delegation, this is done implicitly by having [[self (computer science)|self]] in the delegate refer to the original (sending) object, not the delegate (receiving object). In the delegate pattern, this is instead accomplished by explicitly passing the original object to the delegate, as an argument to a method.[1] "Delegation" is often used loosely to refer to the distinct concept of forwarding, where the sending object simply uses the corresponding member on the receiving object, evaluated in the context of the receiving object, not the original object.
This article uses "sending object/receiving object" for the two objects, rather than "receiving object/delegate", emphasizing which objects send and receive the delegation call, not the original call.
In the Introduction to Gamma et al. 1994, delegation is defined as:
In the example below (using the Kotlin programming language), the class Window delegates the area call to its internal Rectangle object (its delegate).
class Window(val bounds: Rectangle)
Some languages have special support for delegation built in. For example, in the Kotlin programming language the by keyword[2] delegates to another object's interface:
class Rectangle(val width: Int, val height: Int) : ClosedShape
// The ClosedShape implementation of Window delegates to that of the Rectangle that is boundsclass Window(private val bounds: Rectangle) : ClosedShape by bounds