Why does the library contain both Stream.of(T t) and Stream.of(T... values)?

Question

Since the method

Stream.of(T... values)

includes the case of a single element, why is the method:

Stream.of(T t)

also included in the API? What's its purpose? When is it invoked? Does it do anything different from the varargs version?

For the same reason as http://stackoverflow.com/questions/31924249/why-does-enumset-have-many-overloaded-of-methods — Tunaki, Aug 16 '16 at 07:55

score 7 · Answer 1 · answered Aug 15 '16 at 23:14

7

What's its purpose?

A small improvement in efficiency for a common use case.

When is it invoked?

When you have one argument.

Does it do anything different from the varargs version?

The out come is the same, but it creates less garbage.

answered Aug 15 '16 at 23:14

Peter Lawrey

This also explains `Stream.empty()`, though that doesn't work as implicitly as an overload. – shmosel Aug 16 '16 at 00:04
3

@shmosel And `empty()` is much clearer semantically than what looks like a missing argument. – chrylis -cautiouslyoptimistic- Aug 16 '16 at 00:13
@chrylis Guava doesn't think so. – shmosel Aug 16 '16 at 00:13
2

@chrylis You can still use `Stream.of()` but I agree that `empty()` is a clearer choice. – Peter Lawrey Aug 16 '16 at 00:28

1 Answers1