Kotest is in the latest Thoughtworks Radar! A number of Thoughtworkers (including myself) use Kotest regularly and recommend using it for Kotlin development. Congratulations on an excellent testing framework! https://www.thoughtworks.com/radar/languages-and-frameworks?blipid=201911021

question @sam a co-worker of mine just ran into a problem like below

test("should print out failing inputs and shrinks when a function throws") {
      suspend fun functionThatUnexpectedlyThrows() { throw RuntimeException("aaaa") }
      checkAll(Arb.bigInt(4, 4)) { value ->
         functionThatUnexpectedlyThrows()
      }
   }

and then I realized that we don’t report the inputs or attempt any shrinking to the users when the exception is not kotest’s assertion errs. i.e.

Property failed after 1 attempts
java.lang.AssertionError: Property failed after 1 attempts

Repeat this test by using seed 320358835705021305

Caused by RuntimeException: aaaa

if that function was wrapped around

shouldNotThrowAny { .. }

then that’ll print out the inputs and properly shrunk it. it’s written here https://github.com/kotest/kotest/blob/aba6984d6a08530e931274f7a9f6e3b62a66b681/kot[…]perty/src/commonMain/kotlin/io/kotest/property/internal/test.kt is this by design? or should we fix that? if you reckon that’s a bug i can raise an issue and raise a fix PR?

Hey when generating an arbitrary domain model instance how can one lock some properties? Suppose I have a Student data class with an id, number, name and some other fields. If I have the following generator

fun Arb.Companion.student(): Arb<Student> = arbitrary { rs ->
    Student(
        id = Arb.uuid(V4,false).next(rs)
        name = Arb.string().next(rs),
        number = <http://Arb.int|Arb.int>().next(rs), ...
    )
}

What can I add here to lock the number I which my generated student have?

Arb.student().<here>?
// Something like
Arb.student().with(number, 123)

Anyway I can do this?

Hey guys! When using assertion

shouldBeFailure

, the error message I get is

java.lang.AssertionError: Result should be a failure but was kotlin.Unit

While I actually expected it to be more significant that it's a Success, like:

java.lang.AssertionError: Result should be a failure but was Success(kotlin.Unit)

This is due to the fact that

BeFailure

uses

getOrNull

in the

Result

instead of

toString

. I think it would be better to replace it, but I don't know. What do you think? I can open an issue to discuss this further in the Kotest repository

How do I use the SystemEnvironmentProjectListener (or really any externally defined project listener) in my project? I found a README that says to do this:

class ProjectConfig : AbstractProjectConfig() {

    override fun listeners(): List<TestListener> = listOf(SystemEnvironmentProjectListener("foo", "bar"))

}

but AbstractProjectConfig::listeners is deprecated and the deprecation message, “`Use extensions."` , isn’t the most helpful. I would like to define the listener once for the whole project. * Not a contribution *

When I run a single test of a spec that contains multiple tests, Kotest generates output for every ignored test. It's challenging to wade through all that logging when I'm looking for the output of print statements that I've added to the code under test in order to debug. Is there a way to turn off the logging? Otherwise, how could I have the output from print statements be associated with the test that was run instead of the spec containing all the tests?

anyone else having a problem with with

5.2.2

- i think it may be related to the new

1.6.1

coroutine release from yesterday?

We’re just migrating to Kotest 5.x, and I found some behavior that I can’t figure out if it’s a bug or not. I managed to trim it down to the following repro, and as far as I can see it’s triggered by two tests having the same name (as seen in the test). When I change one of the duplicated tests names, the test pass. Code and error details in 🧵

Anyone working with KMM or Multiplatform? It seems that Kotlin 1.6.20 broke Kotest references in Android Studio at least on my environment. This seems related to an old bug I filed 8 months ago:

Here’s the bug: https://youtrack.jetbrains.com/issue/KT-48148

Hi ya, is there an a way to run some “after all tests / specs” logic? At the moment i’m using

Runtime.getRuntime().addShutdownHook(…)

but was curious if there’s something exposed by kotest to more easily accomplish this?

Is there any logging capabilities embedded within the framework that I could use in my tests? I'd like to log some intermediate test outputs at debug level to toggle on when something goes wrong.

How do I properly use the same `withData`for multiple tests? 🧵

Perhaps this is more of a mockk question, but I've noticed that records from mocks and spys tend to bleed over between tests, even with

confirmVerified

or

unmockX()

between cases. 🧵

posted on stackoverflow but has anyone ran into similar issue here? I'm new to springboot and kotest https://stackoverflow.com/questions/71833886/springboot-test-fails-to-run

is there a plugin for continuous testing, as Ncrunch for C#? the auto-test function in Intellij is not as useful

Hi. I have a problem i’m struggling with for a few days. I want to create a DSL-like

TestFactory

builder, that would nest further steps declared by user in a lambda as the last parameter of my test factory. In my specific case that would look like this:

class MySpec : FreeSpec({
    include(replicatorSpec(...) { // this: ReplicatorTestFactoryConfiguration extending TestFactoryConfiguration(), FreeSpecRootScope
        parent(...) { // this: ReplicatorTestFactoryConfiguration.AfterParent extending TestFactoryConfiguration(), FreeSpecRootScope
            create(...) { // this: ReplicatorTestFactoryConfiguration.AfterCreate extending TestFactoryConfiguration(), FreeSpecRootScope
                update(...)
            }
        }
    )}
})

inline fun RootScope.replicatorSpec(
    ...,
    crossinline configure: ReplicatorTestFactoryConfiguration.() -> Unit = { }
) = freeSpec {
    <http://logger.info|logger.info> { "replicatorSpec" }

    beforeSpec {...}

    val configuration = ReplicatorTestFactoryConfiguration(...)

    configuration.configure()

    afterSpec {...}
}

inline fun ReplicatorTestFactoryConfiguration.parent(
    ... ,
    crossinline then: ReplicatorTestFactoryConfiguration.AfterParent<P>.() -> Unit = {}
) { ... }

inline fun ReplicatorTestFactoryConfiguration.AfterParent.create(
    ... ,
    crossinline then: ReplicatorTestFactoryConfiguration.AfterParent<P>.AfterCreate<P, COUT>.() -> Unit = { }
) { ... }

inline fun ReplicatorTestFactoryConfiguration.AfterParent.AfterCreate.update(...) { ... }

But when running tests and putting debug FreeSpec test scopes (

"debug" - { }

) i see that execution gets only to the

replicatorSpec(...)

and the FreeSpecScopes defined in there, but anything further like the code defined in lambda. How can I create such a

Is there an "or" matcher. I would like to test that a json key is absent or false.

shouldContainJsonKeyValue("key", false) or shouldNotContainJsonKeyValue("key")

Hi there! I am having problems with jsTest and jsBrowserTest in a multiplatform project. On my machine tests are no longer executed for JS. When I create a test like

"FAIL!" {
   fail("BOOOM!")
}

I still see all tests being successful. The test suite also worked last week and the JS and Common tests where executed properly and suddenly stopped doing so. Does anyone have an idea what is wrong here. I am running on Ubuntu 20.04 and Kotest 5.2.3, Kotlin 1.6.20. I also know that my browsers where updated last week... So yeah am I lost here. Any help, ideas or pointers is very much appreciated!

Hi! I was wondering if there are any recommended formatter settings in conjunction with kotest. when using the standard kotlin formatter settings from intellij there is really much whitespace in front of all lines...

Wrote a quick blog on avoiding mocks by using Test Containers and Kotest https://medium.com/@sam_57450/unitigration-testing-with-kotest-and-test-containers-55b7ad28f615

Is there a way to twek the kotest plugin so i can navigate to the failing assertion by double clicking it in the results view (as i am used to do from plain junit tests)?

what does Arb stand for?

Hey suppose I have a Student like

data class Student(val id: Int, val name: String)

and I’m trying to build a Arb of Student like

fun Arb.Companion.student(
    id: Arb<Int> = positiveInt(),
    name: Arb<String> = string()
): Arb<Student> = arbitrary { rs ->
    Student(
        id.next(rs),
        name.next(rs) // or name.bind() and ignore rs?
    )
}

Should I call

bind

or

next

with the current randomSource?

Ever since my upgrade to 2022.1 (Ultimate Edition) I am seeing “Invalid value: -1” errors trying to run Kotest tests from the gutter (using the play button). My project uses gradle and I suspect that could be a cause as well. This is making is very difficult to run tests locally 😞 ** Not a contribution

Hi, does anyone have experience of testing an OpenAPI spec using Kotest? I'm asking because I'm very tempted to start creating a Kotest version of Schemathesis, which I've used before to great effect. I'd like a Kotlin version though! I think a Kotest 'OpenAPI' module would work really well. It would parse an OpenAPI spec (Swagger provides a Java parser), and generate random requests, using Arbs. So if a specific response isn't documented, or a parameter marked as 'optional' actually isn't, then the test would fail. Stateful testing (e.g. make sure that after a successful

POST /user/create

that

GET /user/{id}

works) is possible too. The benefits of this would be • the OpenAPI spec drives the testing, and would be completely separate from the codebase (no need for spec or code generation, which is always limiting) • there's no limitation or restriction on how to implement the API (for example Hikaku doesn't handle many custom implementations) • because the spec can be tested, it would always be up-to-date. I've attached a mock code example. I might work on a proof-of-concept...

Are there plans for kotest to support nested javascript tests (mybe runTest allows this)? I hava a huge library with 500 kotest-tests and wanted to enabled JS without loosing advantages of nested tests... (apart from all the work to refactor 500 nested tests to not nested tests).

Hi all, I'm trying to use Kotest + TestContainers (neo4J database) and I cannot make it work. The error is:

Not enough information to infer type variable T

io.kotest.extensions.testcontainers.TestContainerExtension public constructor TestContainerExtension<T : GenericContainer<out T>>(
    container: GenericContainer<out T>,
    lifecycleMode: LifecycleMode
)

Adding the test code in the thread 👇 Thanks for any help!!

Is there a form of

Map<K, V>.shouldContainAll(expected: Map<K, V>)

which recurses into any maps in the top level structure?