given ```enum class Bar { ONE, TWO, }``` and ```enum class Foo { A, B, C, ; val asBar ...

# getting-started

[y]

given

enum class Bar {
  ONE,
  TWO,
}

and

enum class Foo {
  A,
  B,
  C,
  ;
  val asBar get() = 
    when (this) {
      A, B -> Bar.ONE,
      C -> Bar.TWO,
    }
}

vs.

enum class Foo(val asBar: Bar) {
  A(Bar.ONE),
  B(Bar.ONE),
  C(Bar.TWO),
}

the second one compiles to something shorter, so is it the preferred way to write this?
specifically, is it also "zero sized" (whatever that term would mean in JVM land)?

[jamshedalamqaderi]

For me #2 source is good and more cleaner then source #1.

[Joffrey]

I wouldn't pay so much attention to the compiled code in general unless you want to investigate particular issues that you or your users experienced. Here I would just look at the source. The second snippet is better in terms of readability. The property name asBar might need to be revisited based on the real domain names, as this is usually a function name.

[ephemient]

"zero-sized" doesn't have an equivalent in JVM

[ephemient]

here, enum entries are objects and all constructed when the class is loaded, no matter which way you write it