Hey! This seems inconsistent. Does anyone know why...

# announcements

[Michael Böiers]

Hey! This seems inconsistent. Does anyone know why? @thana

import java.util.*

fun compiles() {
    val map: AbstractMap<String, String> = TreeMap()
    map["key"] = null // non-null not enforced (platform type)
}

fun `does not compile` () {
    val map = TreeMap<String, String>()
    map["key"] = null // non-null enforced? Why? Isn't this also a platform type?
}

[Michael Böiers]

More specifically:

import java.util.*

fun compiles() {
    val map = TreeMap<String, String>() as AbstractMap<String, String>
    map["key"] = null // non-null not enforced (platform type)
}

fun `doesn't compile` () {
    val map = TreeMap<String, String>()
    map["key"] = null // non-null enforced in TreeMap, but not AbstractMap?
}

[nanodeath]

does the IntelliJ popup show anything significantly different for put for the two calls?

[Michael Böiers]

Which popup do you mean? This one? It’s identical for the two set calls.

[nanodeath]

yeah...that's the one. hmm.

[Michael Böiers]

Nice trick that Jetbrains used here: In order to add the assignment syntax for these platform types they just defined a set method as an operator extension function. Sweet. It still doesn’t work when I use put explicitly, and there the only difference in the popups is TreeMap vs. AbstractMap.

[Michael Böiers]

Incidentally, I always found it odd that Map uses put in Java, but Set uses set.

[Roukanken]

Okay this is some black magic

[nanodeath]

I'm seeing something different from you it looks like

[nanodeath]

correction, they're both extension method calls

[nanodeath]

I think Kotlin must have some tricks for identifying what is actually a MutableMap, and AbstractMap isn't included

[nanodeath]

well, that's not quite right either...in any case agreed, magic

[Alexey Belkov [JB]]

@Michael Böiers Good catch. Can you please report this to http://kotl.in/issue. I'm not sure if this is a bug, but it would be nice to keep track of the problem.

[Michael Böiers]

https://youtrack.jetbrains.com/issue/KT-45953

[thana]

@Michael Böiers thank you very much for investigating on that issue <3