🧪
This library is experimental!
Its API is not stabilized yet, and writing tests is still a bit tedious. Use at your own risk. Looking forward to your feedback :)
What is vis-assert?
It's a Kotlin library to write visually appealing ASCII-art-like test assertions for math functions. For example, you can test that your (Float) -> Float
function describing a sine wave produces proper values. Or if you have a game where the player jumps, you can describe player's vertical position as a function of time - you could test this function to make sure that the jump movement is fluent and fast enough.
Under the hood, each such ASCII visualisation is translated into a collection of constraints, where each constraint looks at a single X value of the function and performs a certain check on its Y value at this point.
Installation
In your build.gradle
or build.gradle.kts
:
repositories {
mavenCentral()
}
dependencies {
testCompile("it.krzeminski.vis-assert:vis-assert:0.4.1-beta")
}
// or
kotlin {
sourceSets {
val ...Test by getting {
dependencies {
implementation("it.krzeminski.vis-assert:vis-assert:0.4.1-beta")
}
}
}
}
Examples
@Test
fun sineWaveFor2HzOnePeriod() {
assertFunctionConformsTo(
functionUnderTest = sineWave(frequency = 2.0f),
visualisation = {
row(1.0f, " IIXII ")
row( " III III ")
row( " I I ")
row( " II II ")
row( " I I ")
row(0.0f, "X I I")
row( " I I ")
row( " II II ")
row( " I I ")
row( " III III ")
row(-1.0f, " IIIII ")
xAxis {
markers("| | |")
values( 0.0f, 0.25f, 0.5f)
}
})
}
or for high-frequency function and higher sampling:
@Test
fun assertFunctionConformsToForHighFrequencyFunctionWhenAssertionsAreFulfilledAndSamplingHigherThan1IsUsed() {
assertFunctionConformsTo(
functionUnderTest = { x: Float -> (sin(100*x) * sin(x) * x * 0.3).toFloat() },
samplesPerCharacter = 100,
visualisation = {
row( 2.0f, " ")
row( " ")
row( " IIIIIIIIIIIIII ")
row( 1.0f, " IIIIIIIIIIIIIIIIIIIII ")
row( " IIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIII")
row( " IIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII")
row( 0.0f, "IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII")
row( " IIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIII")
row( " IIIIIIII IIIIIIIIIIIIIIIIIIIIIIIIIII")
row(-1.0f, " IIIIIIIIIIIIIIIIIIII ")
row( " IIIIIIIIIIIIII ")
row( " ")
row(-2.0f, " ")
xAxis {
markers("| | | | | | |")
values( 0.0f, 1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f)
}
}
)
}
Where:
I
characters mean that for a given X argument, the function's value can be in a certain range around a given Y value. Also, this constraint is "strict", which means that making it wider or narrower vertically would make the assertion fail. In this example, eachI
character has a tolerance of +/- 0.1. The tolerance is calculated based on the vertical axis description.X
characters mean that for a given X argument, the function's value has to exactly match the given Y value.
There's also i
constraint, which just checks that all values are in a certain range.
More examples can be found in unit tests for krzema12/fsynth - a project that vis-assert was created for.
Limitations
- the library performs sampling, as given by the
xAxis
description andsamplesPerCharacter
parameter. It means that if two subsequent X values are 0.2 and 0.3, and not enough sampling rate is given, the library may not check what happens for 0.25 or 0.20001. In most cases, such simple sampling is enough. - only
(Float) -> Float
functions are currently supported. Mitigation: it's possible to assert on any other function, as long as it can be presented as a(Float) -> Float
function. See this example for adapting an(Int) -> Float
function - when assertions fail, the current message just says about failed first (x, y) constraint, going from the left. It's thus quite time-consuming to write a test. Ideally, if the assertion fails, vis-assert should show how the ASCII visualisation could look like.