Explicit parameters:
{ parameterName: ParameterType, otherParameterName: OtherParameterType -> anExpression() }
Inferred parameters:
val addition: (Int, Int) -> Int = { a, b -> a + b }
Single parameter it
shorthand
val square: (Int) -> Int = { it*it }
Signature:
() -> ResultType
(InputType) -> ResultType
(InputType1, InputType2) -> ResultType
Input type parameters can be left out when they can be left out when they can be inferred from the context. For example say you have a function on a class that takes a function:
data class User(val fistName: String, val lastName: String) {
fun username(userNameGenerator: (String, String) -> String) =
userNameGenerator(firstName, secondName)
}
You can use this function by passing a lambda, and since the parameters are already specified in the function signature there's no need to re-declare them in the lambda expression:
val user = User("foo", "bar")
println(user.userName { firstName, secondName ->
"${firstName.toUppercase}"_"${secondName.toUppercase}"
}) // prints FOO_BAR
This also applies when you are assigning a lambda to a variable:
//valid:
val addition: (Int, Int) = { a, b -> a + b }
//valid:
val addition = { a: Int, b: Int -> a + b }
//error (type inference failure):
val addition = { a, b -> a + b }
When the lambda takes one parameter, and the type can be inferred from the context, you can refer to the parameter by it
.
listOf(1,2,3).map { it * 2 } // [2,4,6]