Take for example the following pseudo-code,
Assert.assertNotNull(cut.func())
Where cut is a non-null class under test, and func() is some function that returns void.
Furthermore, assuming func() does not throw any errors this will result in a valid assertion.
Since it isn't possible to set an object to be null from within itself in Java / Kotlin, (i.e. func() cannot set cut to null), why is it valid to call void functions within non-null assertions?
Is it simply because void is not the same as null?
Java's void return type is mapped to the Unit object in Kotlin, and Unit is not null.
val x = System.out.println("") // void Java method
println(x == null) // "false"
println(x == Unit) // "true"