https://kotlinlang.org logo
Channels
Powered by
#stdlib
  • i

    ilya.gorbunov

    11/03/2015, 2:42 PM
  • m

    mikehearn

    11/03/2015, 2:42 PM
    ah ha, great, thanks
  • m

    mikehearn

    11/03/2015, 2:42 PM
    you guys are way ahead of me
  • m

    mikehearn

    11/03/2015, 2:43 PM
    though i guess by now the stdlib is frozen, right? so it must wait until post 1.0
  • o

    orangy

    11/03/2015, 2:43 PM
    no, stdlib is not as frozen as ABI, there will be some changes before 1.0
  • o

    orangy

    11/03/2015, 2:46 PM
    @mikehearn: do you mean something like this?
    Copy code
    fun check(inputs: Set<String>, outputs: Set<String>, link: String.() -> Set<String>): Boolean = inputs.flatMapTo(hashSetOf(),link) == outputs
  • m

    mikehearn

    11/03/2015, 2:48 PM
    hm, something of that form, yes. not sure my code maps exactly
  • m

    mikehearn

    11/03/2015, 2:48 PM
    but neat construct
  • m

    mikehearn

    11/03/2015, 5:21 PM
    List.get(i, j) -> subList …. yes or no?
  • m

    mplatvoet

    11/03/2015, 5:29 PM
    List.get(i..j) ?
  • m

    mikehearn

    11/03/2015, 5:30 PM
    as a range? it’d create an extra object, no?
  • m

    mplatvoet

    11/03/2015, 5:30 PM
    Yes
  • m

    mplatvoet

    11/03/2015, 5:31 PM
    I find it a bit more idiomatic though
  • m

    mikehearn

    11/03/2015, 5:32 PM
    the android guys would be happier with a lower-garbage solution, though
  • m

    mplatvoet

    11/03/2015, 5:34 PM
    That's true, yet if they are really concerned they aren't sublisting at all
  • o

    orangy

    11/03/2015, 5:53 PM
    We made it possible to add slices post-1.0 (smth like 
    list[0:1]
    ) by not using 
    :
    in important places. So we will probably not add some intermediate functions to stdlib as they live forever.
  • m

    mikehearn

    11/03/2015, 5:55 PM
    makes sense
  • m

    mplatvoet

    11/03/2015, 5:56 PM
    Nice!
  • e

    evanchooly

    11/03/2015, 7:00 PM
    slices++
  • v

    voddan

    11/03/2015, 7:24 PM
    @orangy: New syntax 
    list[a:b]
    just to skip an object allocation? Wow Can't it be solved by 
    parameter inlining
    for 
    inline
    functions? Or is that it?
  • v

    voddan

    11/03/2015, 7:25 PM
    There is a similar case with overhead in creating maps: 
    mapOf(1 to 2)
    . Will 
    1:2
    syntax solve that?
  • o

    orangy

    11/03/2015, 7:26 PM
    No design yet, just overall target idea and syntax availability post-1.0
  • e

    Eugenio

    11/03/2015, 8:56 PM
    love it, thank you so much for thinking about Android
  • j

    jkwatson

    11/09/2015, 4:30 AM
    Would love a 
    plusIfAbsent
    on immutable Maps.
  • k

    kirillrakhman

    11/19/2015, 12:56 PM
    any particular reason why 
    measureTimeMillis
    is not inlined?
  • i

    ilya.gorbunov

    11/19/2015, 1:43 PM
    They were written in "before-inline" era, and have never been revisited since then.
  • d

    damian

    11/19/2015, 2:42 PM
    re: slices/inling - fwiw, it looks from the bytecode that unassigned ranges, used for something like an 
    in
    -check, are already inlined
  • d

    damian

    11/19/2015, 2:42 PM
    e.g. 
    1 in 0..n
    is already inlined
  • d

    damian

    11/19/2015, 3:19 PM
    Also: is there currently anything in the kotlin stdlib to replace 
    System.arraycopy
    ?
  • i

    ilya.gorbunov

    11/19/2015, 3:24 PM
    Some cases are covered with Array.copyOfRange, but not the case when you need to copy elements from one to another. What is your case and how do you expect stdlib should cover it in a convenient way?
1...252627...149Latest