Channels
#stdlib
- j
jw
03/08/2016, 4:26 PMsimplify to your will - e
eshioji
03/08/2016, 4:26 PMPardon if there’s any documentation on this, but say you have a suggestion like this, do I create an issue on Youtrack? - j
jw
03/08/2016, 4:27 PMCopy codeinline fun <reified K : Enum<K>, V> enumMapOf(vararg pairs: Pair<K, V>) = EnumMap<K, V>(K::class.java).let { map -> pairs.forEach { map.put(it.first, it.second) } } - j
jw
03/08/2016, 4:27 PMcould probably do it as a
as wellfold - j
jw
03/08/2016, 4:34 PMyes, create a YouTrack - j
jw
03/08/2016, 4:35 PMand link it here so I can vote on it - e
eshioji
03/08/2016, 4:35 PMok cool, thanks! Will do. And thank you for the snippet, will be using that in my code simple smile - e
eshioji
03/08/2016, 4:40 PMoh, nvm - e
eshioji
03/08/2016, 4:40 PM - e
eshioji
03/08/2016, 4:40 PMThere already was a ticket - k
kirillrakhman
03/08/2016, 4:57 PM@jw your second version returnsUnit - j
jw
03/08/2016, 4:57 PMoops ¯\_(ツ)_/¯ - k
kirillrakhman
03/08/2016, 4:57 PMCopy codeinline fun <reified K : Enum<K>, V> enumMapOf(vararg pairs: Pair<K, V>) = EnumMap<K, V>(K::class.java).apply { pairs.forEach { put(it.first, it.second) } } - j
jw
03/08/2016, 4:59 PMthanks. was just messing around in a text editor, not IDE or anything to verify - k
kittinunf
03/08/2016, 5:40 PMCan we do a bit shorter with
?associateByToCopy codeinline fun <reified K : Enum<K>, V> enumMapOf(vararg pairs: Pair<K, V>) = pairs.associateByTo(EnumMap<K, V>(K::class.java), { it.first }, { it.second }) - y
yoavst
03/08/2016, 5:55 PMCopy code
It could be nicer ifinline fun <reified K : Enum<K>, V> enumMapOf(vararg pairs: Pair<K, V>) = pairs.associateByTo(EnumMap<K, V>(K::class.java), Pair<K, V>::first, Pair<K, V>::second)
would have worked.Pair::first - e
eshioji
03/08/2016, 6:43 PMHm looks like this approach isn’t completely bullet proof - e
eshioji
03/08/2016, 6:43 PM-.scala - e
eshioji
03/08/2016, 6:43 PMThe above doesn’t compile if
is implemented as followsenumMapOf - e
eshioji
03/08/2016, 6:44 PMUntitled.scala - e
eshioji
03/08/2016, 6:46 PMIt will however work if I do - e
eshioji
03/08/2016, 6:46 PM-.scala - e
eshioji
03/08/2016, 6:47 PMOr of course - e
eshioji
03/08/2016, 6:47 PM-.scala - e
eshioji
03/08/2016, 6:50 PMHard to tell what’s the right thing to do here…. - e
eshioji
03/08/2016, 6:56 PMI guess the last one is the safest choice, although a bit unfortunate - j
jw
03/08/2016, 7:12 PMthe first one is broken because K isn't reified through that class - i
ilya.gorbunov
03/09/2016, 2:16 PMhttps://kotlinlang.slack.com/archives/stdlib/p1457439208000312
represents a total order relation on a set. By redefiningcompareTo
to partial order you break its contract.compareTo - v
voddan
03/09/2016, 3:27 PM@ilya.gorbunov: is it a reason not to have the operators overloaded? The partial order would matter if I wanted to sort a collection of such items or something - v
voddan
03/09/2016, 3:28 PMBTW, someone was concerned that
should not throw:compareToe.compareTo(null) should throw a NullPointerException
from the docs: https://docs.oracle.com/javase/7/docs/api/java/lang/Comparable.html