https://kotlinlang.org logo
Channels
Powered by
#spek
  • d

    dimsuz

    08/15/2016, 12:17 PM
    you helped us uncover a really annoying performance problem, which when fixed can improve life for almost every Kotlin developer
    @orangy: this might be a bit late reaction, but is there any YT issue I can track? Currently I am arranging my code a bit differently as a workaround and I'd like to stick a comment which describes when this workaround can be removed 🙂
  • o

    orangy

    08/15/2016, 1:13 PM
    @nk: do you have issue#?
  • d

    dimsuz

    08/15/2016, 4:28 PM
    @artem_zin: did you by chance have any issue with 
    RxJavaHooks.reset(); RxJavaHooks.onIOScheduler(Schedulers.immediate())
    ? I try to call it in several places: a) in 
    describe()
    body b) in 
    Spek({ beforeEach() { HERE } describe()...
    and I have a strange result: my tests which rely on sync-behaviour of io-scheduler always pass on my local machine, but always fail on CI server? o_O Didn't have time to investigate deeply yet, thought that maybe this is some known symptom...
  • a

    artem_zin

    08/15/2016, 5:24 PM
    @dimsuz yes, but it's not related to Spek, just DI schedulers
  • d

    dimsuz

    08/15/2016, 5:34 PM
    I see. I wonder what is a Spek's equivalent to 
    @BeforeClass
  • k

    kirillrakhman

    08/15/2016, 5:40 PM
    companion object?
  • d

    dimsuz

    08/15/2016, 5:59 PM
    hmm. didn't think of it.
  • a

    artem_zin

    08/15/2016, 7:59 PM
    Why would you need it? Just use @afterEach on top level
  • d

    dimsuz

    08/16/2016, 9:28 AM
    Why would you need it? Just use @afterEach on top level
    for some reason injecting io scheduler this way doesn't work on my CI server (tests fail) while it does work everywhere else (tests pass). It even works on CI server when I run tests by hand with 
    gradlew test
    . But when CI agent runs the build it constantly fails. And if I remove explicity 
    subscribeOn(<http://Schedulers.io|Schedulers.io>())
    it gets fixed immediately... I'm puzzled (for now).
  • a

    artem_zin

    08/16/2016, 9:32 AM
    @dimsuz: we’ve faced this in two projects with like 2-3 months difference (though we’re building them on same CI with same JDK/etc) Since we were using hooks for legacy reasons we just migrated to DI for schedulers everywhere
  • d

    dimsuz

    08/16/2016, 10:01 AM
    thanks for advice, I'll try to also do that once I have some time!
  • n

    nk

    08/16/2016, 1:16 PM
    @nk pinned a message to this channel.
  • n

    nk

    08/16/2016, 1:47 PM
    I have just collected all known information into https://youtrack.jetbrains.com/issue/KT-13474
  • d

    dimsuz

    08/16/2016, 2:35 PM
    thank you!
  • h

    hhariri

    08/16/2016, 2:54 PM
    Well we’re on Maven (albeit a test deploy), but we’ll have the release on Maven for sure
  • a

    artem_zin

    08/18/2016, 10:26 AM
    Anyone solved problems with using Spek in Android Instrumentation tests? 1. Adding nodes to the Spek tree breaks tests: we have to use flat hierarchy without additional 
    on/describe/context/given
    2. Non-clean run of tests fails even though app under the test have clean state each time 3. Negative test run duration in reports (
    -248 ms
    , etc)
  • h

    hhariri

    08/18/2016, 10:50 AM
    @raniejade: ^^^
  • r

    raniejade

    08/18/2016, 10:53 AM
    @artem_zin: are you using the new 1.0?
  • a

    artem_zin

    08/18/2016, 10:53 AM
    @raniejade: no, we’re on 
    1.0.25
  • a

    artem_zin

    08/18/2016, 10:53 AM
    is new 
    1.0
    stable enough to try it?
  • r

    raniejade

    08/18/2016, 10:57 AM
    Yeah, it is. But I don't think it will work for android instrumentation tests.
  • r

    raniejade

    08/18/2016, 10:58 AM
    I still need to write some docs on setting up.
  • a

    artem_zin

    08/18/2016, 11:06 AM
    Ok, I’ll try to apply it for our unit tests first
  • j

    john.shelley

    08/18/2016, 2:08 PM
    confused.. is 
    1.0
    higher then 
    1.0.25
    somehow? what does new 
    1.0
    mean?
  • h

    hhariri

    08/18/2016, 2:10 PM
    @john.shelley: we’ve been publishing Snapshots of 1.0.x to Artifactory and are preparing for a release. You can see all builds here: https://teamcity.jetbrains.com/viewType.html?buildTypeId=Spek_BuildAndTests So it’s not 1.0, but 1.0.X which is currently at .72
    👍 1
  • a

    adambl4

    08/23/2016, 9:54 AM
    Can I do 
    beforeEach
    before 
    on
    in the Spek snapshot?
  • s

    simon.vergauwen

    08/26/2016, 10:26 AM
    Any1 here experience with PowerMock in combintation with Spek?
  • d

    dmitriy.m

    08/26/2016, 11:40 AM
    i’m @simon.vergauwen
  • s

    simon.vergauwen

    08/26/2016, 11:41 AM
    I’m not able to get mockStatic working.
  • s

    simon.vergauwen

    08/26/2016, 11:41 AM
    Copy code
    PowerMockito.mockStatic(Uri::class.java)
`when`(Uri.parse(parseUrlString)).thenReturn(uri)
1...383940...112Latest