Type signature
|
|
A type signature defines the inputs and outputs for a function or method. Type signatures may appear to be extremely restrictive but they allow for effortless code re-use, in that given a function that performs a certain task, it can reliably be swapped with any other code that matches this type signature and the overall functionality of the entire program can be assured (given that both functions operate properly). A type signature includes the function name and the number of its parameters. In some programming languages, it may also specify the function's return type or the types of its parameters.
Haskell
A type signature in Haskell is written, generally, in the following format:
functionName :: arg1Type -> arg2Type -> ... -> argNType
Notice that the final output can be regarded as an argument. This is a consequence of currying. That is, given a function that had one argument supplied, but takes in two inputs, the function is "curried" and becomes a function of one argument -- the one that is not supplied.
The actual type specifications can consist of an actual type, such as Integer, or a type variable that is used in parametric polymorphic functions, such as "a", or "b", or "anyType". So we can write something like:
functionName :: a -> a -> ... -> a
Since Haskell supports higher-order functions, functions can be passed as arguments, and this is written as:
functionName :: (a -> a) -> a
This function takes in a function with type signature a -> a, and returns data of type "a" out.
Java
In the Java virtual machine, so-called internal type signatures are used to identify methods and classes at the level of the virtual machine code.
Example: The method String String.substring(int, int) is represented as java/lang/String/substring(II)Ljava/lang/String;