Skip to main content
Version: 2.0.x

Built-in Generators

In the companion object of the Gen data type, there are tons of generators for various data types.

Primitive Types Generators

ZIO Test provides generators for primitive types such as, Gen.string, Gen.boolean, Gen.float, Gen.double, Gen.bigInt, Gen.byte, Gen.bigdecimal, Gen.long, Gen.char, and Gen.short.

Let's create an Int generator:

import zio._
import zio.test._

val intGen: Gen[Any, Int] =

Character Generators

In addition to Gen.char, ZIO Test offers a variety of specialized character generators:

  • Gen.alphaChar — e.g. Z, z, A, t, o, e, K, E, y, N
  • Gen.alphaNumericChar — e.g. b, O, X, B, 4, M, k, 9, a, p
  • Gen.asciiChar — e.g. , >, , , , 2, k, , , 
  • Gen.unicodeChar — e.g. 了, , 옷, 嗀, , 뮲, ﹓, 癮, , ᜣ)
  • Gen.numericChar — e.g. 1, 0, 1, 5, 6, 9, 4, 4, 5, 2
  • Gen.printableChar — e.g. H, J, (, Q, n, g, 4, G, 9, l
  • Gen.whitespaceChars — e.g. , ,  , , ,  ,  , ,  ,
  • Gen.hexChar — e.g. 3, F, b, 5, 9, e, 2, 8, b, e
  • Gen.hexCharLower — e.g. f, c, 4, 4, c, 2, 5, 4, f, 3
  • Gen.hexCharUpper — e.g. 4, 8, 9, 8, C, 9, F, A, E, C

String Generators

Besides the primitive string generator, Gen.string, ZIO Test also provides the following specialized generators:

  1. Gen.stringBounded — A generator of strings whose size falls within the specified bounds:

    Gen.stringBounded(1, 5)(Gen.alphaChar)
    // Sample Output: List(b, YJXzY, Aro, y, WMPbj, Abxt, kJep, LKN, kUtr, xJ)
  2. Gen.stringN — A generator of strings of fixed size:

    // Sample Output: List(BuywQ, tXCEy, twZli, ffLwI, BPEbz, OKYTi, xeDJW, iDUVn, cuMCr, keQAA)
  3. Gen.string1 — A generator of strings of at least one character.

  4. Gen.alphaNumericString — A generator of alphanumeric characters.

  5. Gen.alphaNumericStringBounded — A generator of alphanumeric strings whose size falls within the specified bounds.

  6. Gen.iso_8859_1 — A generator of strings that can be encoded in the ISO-8859-1 character set.

  7. Gen.asciiString — A generator of US-ASCII characters.

Generating Fixed Values

  1. Gen.const — A constant generator of the specified value.

    // Output: List(true, true, true, true, true)
  2. Gen.constSample — A constant generator of the specified sample:

    // Output: List(true, true, true, true, true)
  3. Gen.unit — A constant generator of the unit value.

  4. Gen.throwable — A generator of throwables.

Note that there is an empty generator called Gen.empty, which generates no values and returns nothing. We can think of that as a generator of empty stream, Gen(Stream.empty).

Generating from Fixed Values

  1. Gen.elements — Constructs a non-deterministic generator that only generates randomly from the fixed values:
import java.time._

  1. Gen.fromIterable — Constructs a deterministic generator that only generates the specified fixed values:
Gen.fromIterable(List("red", "green", "blue"))

Collection Generators

ZIO Test has generators for collection data types such as sets, lists, vectors, chunks, and maps. These data types share similar APIs. The following example illustrates how the generator of sets works:

// A sized generator of sets
// Sample Output: Set(Y, M, c), Set(), Set(g, x, Q), Set(s), Set(f, J, b, R)

// A sized generator of non-empty sets
// Sample Output: Set(Y), Set(L, S), Set(i), Set(H), Set(r, Z, z)

// A generator of sets whose size falls within the specified bounds.
Gen.setOfBounded(1, 3)(Gen.alphaChar)
// Sample Output: Set(Q), Set(q, J), Set(V, t, h), Set(c), Set(X, O)

// A generator of sets of the specified size.
// Sample Output: Set(J, u), Set(u, p), Set(i, m), Set(b, N), Set(B, Z)

Bounded Generator

The Gen.bounded constructor is a generator whose size falls within the specified bounds:

Gen.bounded(2, 5)(Gen.stringN(_)(Gen.alphaChar))

Suspended Generator

The Gen.suspend constructs a generator lazily. This is useful to avoid infinite recursion when creating generators that refer to themselves.

Unfold Generator

The unfoldGen takes the initial state and depending on the previous state, it determines what will be the next generated value:

def unfoldGen[R, S, A](s: S)(f: S => Gen[R, (S, A)]): Gen[R, List[A]]

Assume we want to test the built-in scala stack (scala.collection.mutable.Stack). One way to do that is to create an acceptable series of push and pop commands, and then check that the stack doesn't throw any exception by executing these commands:

sealed trait Command
case object Pop extends Command
final case class Push(value: Char) extends Command

val genPop: Gen[Any, Command] = Gen.const(Pop)
def genPush: Gen[Any, Command] =

val genCommands: Gen[Any, List[Command]] =
Gen.unfoldGen(0) { n =>
if (n <= 0) => (n + 1, command))
Gen.oneOf( => (n - 1, command)), => (n + 1, command))

We are now ready to test the generated list of commands:

import zio.test.{ test, _ }

test("unfoldGen") {
check(genCommands) { commands =>
val stack = scala.collection.mutable.Stack.empty[Int]
commands.foreach {
case Pop => stack.pop()
case Push(value) => stack.push(value)

From a ZIO Effect

  1. Gen.fromZIO

    val gen: Gen[Any, Int] = Gen.fromZIO(Random.nextInt) 
  2. Gen.fromZIOSample

    val gen: Gen[Any, Int] =

From a Random Effect

  1. Gen.fromRandom — Constructs a generator from a function that uses randomness:

    val gen: Gen[Any, Int] = Gen.fromRandom(_.nextInt) 
  2. Gen.fromRandomSample — Constructs a generator from a function that uses randomness to produce a sample:

    val gen: Gen[Any, Int] =

Uniform and Non-uniform Generators

  1. Gen.uniform — A generator of uniformly distributed doubles between [0, 1].

  2. Gen.weighted — A generator which chooses one of the given generators according to their weights. For example, the following generator will generate 90% true and 10% false values:

    val trueFalse = Gen.weighted((Gen.const(true), 9), (Gen.const(false), 1))
    // Sample Output: List(false, false, false, false, false, false, false, false, true, false)
  3. Gen.exponential — A generator of exponentially distributed doubles with mean 1: => math.round(x * 100) / 100.0)
    // Sample Output: List(0.22, 3.02, 1.96, 1.13, 0.81, 0.92, 1.7, 1.47, 1.55, 0.46)

Generating Date/Time Types

Date/Time TypesGenerators

Function Generators

To test some properties, we need to generate functions. There are two types of function generators:

  1. Gen.function — It takes a generator of type B and produces a generator of functions from A to B:

    def function[R, A, B](gen: Gen[R, B]): Gen[R, A => B]

Two A values will be considered to be equal, and thus will be guaranteed to generate the same B value, if they have the same hashCode.

  1. Gen.functionWith — It takes a generator of type B and also a hash function for A values, and produces a generator of functions from A to B:

    def functionWith[R, A, B](gen: Gen[R, B])(hash: A => Int): Gen[R, A => B]

Two A values will be considered to be equal, and thus will be guaranteed to generate the same B value, if they have the same hash. This is useful when A does not implement hashCode in a way that is consistent with equality.

Accordingly, ZIO Test provides a variety of function generators for Function2, Function3, ..., and also the PartialFunction:

  • Gen.function2 — Gen[R, C] => Gen[R, (A, B) => C]
  • Gen.functionWith2 — Gen[R, B] => ((A, B) => Int) => Gen[R, (A, B) => C]
  • Gen.partialFunction — Gen[R, B] => Gen[R, PartialFunction[A, B]]
  • Gen.partialFunctionWith — Gen[R, B] => (A => Int) => Gen[R, PartialFunction[A, B]]

Let's write a test for ZIO.foldLeft operator. This operator has the following signature:

def foldLeft[R, E, S, A](in: => Iterable[A])(zero: => S)(f: (S, A) => ZIO[R, E, S]): ZIO[R, E, S]

We want to test the following property:

(in, zero, f) => ZIO.foldLeft(in)(zero)(f) == ZIO(List.foldLeft(in)(zero)(f))

To test this property, we have an input of type (Int, Int) => Int. So we need a Function2 generator of integers:

val func2: Gen[Any, (Int, Int) => Int] = Gen.function2(

Now we can test this property:

import zio._
import zio.test.{test, _}

test("ZIO.foldLeft should have the same result with List.foldLeft") {
check(Gen.listOf(,, func2) { case (in, zero, f) =>
ZIO.foldLeft(in)(zero)((s, a) => ZIO.attempt(f(s, a)))
in.foldLeft(zero)((s, a) => f(s, a)))

Generating ZIO Values

  1. Successful effects (Gen.successes):

    val gen: Gen[Any, UIO[Int]] = Gen.successes(, 10))
  2. Failed effects (Gen.failures):

    val gen: Gen[Any, IO[String, Nothing]] = Gen.failures(Gen.string)
  3. Died effects (Gen.died):

    val gen: Gen[Any, UIO[Nothing]] = Gen.died(Gen.throwable)
  4. Cause values (Gen.causes):

    val causes: Gen[Any, Cause[String]] = 
    Gen.causes(Gen.string, Gen.throwable)
  5. Chained effects (Gen.chined, Gen.chainedN): A generator of effects that are the result of chaining the specified effect with itself a random number of times.

Let's see some example of chained ZIO effects:

import zio._
val effect1 = ZIO(2).flatMap(x => ZIO(x * 2))
val effect2 = ZIO(1) *> ZIO(2)

By using Gen.chaned or Gen.chanedN generator, we can create generators of chained effects:

val chained : Gen[Any, ZIO[Any, Nothing, Int]] = 

val chainedN: Gen[Any, ZIO[Any, Nothing, Int]] =
  1. Concurrent effects (Gen.concurrent): A generator of effects that are the result of applying concurrency combinators to the specified effect that are guaranteed not to change its value.

    val random  : Gen[Any, UIO[Int]] = Gen.successes(
    val constant: Gen[Any, UIO[Int]] = Gen.concurrent(ZIO(3))
  2. Parallel effects (Gen.parallel): A generator of effects that are the result of applying parallelism combinators to the specified effect that are guaranteed not to change its value.

    val random: Gen[Any, UIO[String]] =

    val constant: Gen[Any, UIO[String]] =

Generating Compound Types

  1. tuples — We can combine generators using for-comprehension syntax and tuples:

    val tuples: Gen[Any, (Int, Double)] =
    for {
    a <-
    b <- Gen.double
    } yield (a, b)
  2. Gen.oneOf — It takes variable number of generators and select one of them:

    sealed trait Color
    case object Red extends Color
    case object Blue extends Color
    case object Green extends Color

    Gen.oneOf(Gen.const(Red), Gen.const(Blue), Gen.const(Green))
    // Sample Output: Green, Green, Red, Green, Red
  3. Gen.option — A generator of optional values:

    val intOptions: Gen[Any, Option[Int]] = Gen.option(
    val someInts: Gen[Any, Option[Int]] = Gen.some(
    val nons: Gen[Any, Option[Nothing]] = Gen.none
  4. Gen.either — A generator of either values:

    val char: Gen[Any, Either[Char, Char]] =
    Gen.either(Gen.numericChar, Gen.alphaChar)
  5. Gen.collectAll — Composes the specified generators to create a cartesian product of elements with the specified function:

    val gen: ZIO[Any, Nothing, List[List[Int]]] =
    Gen.fromIterable(List(1, 2)),
    Gen.fromIterable(List(4, 5))
    // Output:
    // List(
    // List(1, 3, 4),
    // List(1, 3, 5),
    // List(2, 3, 4),
    // List(2, 3, 5)
  6. Gen.concatAll — Combines the specified deterministic generators to return a new deterministic generator that generates all the values generated by the specified generators:

    val gen: ZIO[Any, Nothing, List[Int]] =
    Gen.fromIterable(List(1, 2)),
    Gen.fromIterable(List(4, 5))
    // Output: List(1, 2, 3, 4, 5)

Sized Generators

  1. Gen.sized — A sized generator takes a function from Int to Gen[R, A] and creates a generator by applying a size to that function:

    Gen.sized(, _))
    // Sample Output: List(5, 4, 1, 2, 0, 4, 2, 0, 1, 2)
  2. Gen.size — A generator which accesses the size from the environment and generates that:

    // Output: List(100, 100, 100, 100, 100)

There are also three sized generators, named small, medium and large, that use an exponential distribution of size values:

  1. Gen.small — The values generated will be strongly concentrated towards the lower end of the range but a few larger values will still be generated:

    // Output: List(6, 39, 73, 3, 57, 51, 40, 12, 110, 46)
  2. Gen.medium — The majority of sizes will be towards the lower end of the range but some larger sizes will be generated as well:

    // Output: List(93, 42, 58, 228, 42, 5, 12, 214, 106, 79)
  3. Gen.large — Uses a uniform distribution of size values. A large number of larger sizes will be generated:

    // Output: List(797, 218, 596, 278, 301, 779, 165, 486, 695, 788)