Based on my understanding, anonymous function in Kotlin allow you to specify return type. In addition to that, return statement inside anonymous will exit only the function block, while in lambda it will exit the enclosing function.
Still, I can't imagine what would be the real world use case of anonymous function in Kotlin that lambda syntax cannot provide?
The use case is that sometimes we may wish to be explicit about the return type. In those cases, we can use so called an anonymous function. Example:
fun(a: String, b: String): String = a + b
Or better return control like:
fun(): Int {
try {
// some code
return result
} catch (e: SomeException) {
// handler
return badResult
}
}
Anonymous functions (a.k.a function expressions) are very handy when you have to pass a huge lambda with complex logic and want early returns to work. For example if you write a dispatcher in spark-java:
get("/", fun(request, response) {
// Your web page here
// You can use `return` to interrupt the handler
})
get("/") { req, res -> if (sth) return@get; dosth() } works just fine