Overloadshape

Overloadshape is a concept used in software engineering to describe the arrangement and characteristics of a set of overloaded function or method signatures within an interface. The term is used to discuss how the available overloads form a “shape” that affects discoverability, maintainability, and dispatch performance.

Definition and scope: An overload shape comprises the arity (number of parameters), the types of each parameter, and sometimes qualifiers such as constness or reference types, that together distinguish each overload. The return type is typically not part of the resolution by the compiler, but it can influence API design decisions. The shape can also be influenced by default arguments and the presence of implicit conversions.

Role in design: The shape influences overload resolution rules and how easy it is for a caller to select the intended overload. A well-formed shape uses unique discriminators to minimize ambiguity and reduce the need for explicit casts. In languages with type inference or implicit conversions, overlapping overloads can complicate resolution and affect performance.

Examples: Consider an API with draw(int w, int h) and draw(Size s). The overload shape is defined by the parameter types and arity. If Size has a constructor from a pair of ints, calls can be ambiguous without disambiguation, illustrating how shape interacts with conversions and resolution.

Impact and considerations: When designing an API, teams may manage overload shape to balance expressiveness and simplicity. Alternatives to complex shapes include separate method names, variadic arguments, or a single generic method with runtime type checks. Critics warn that overly rich shapes can lead to ambiguity and maintenance challenges.

See also: function overloading, method resolution, API design, type inference.