The entity or concept being tested.

Define an IO-based test.

Should typically be written to be read as full sentences in traditional BDD style: https://en.wikipedia.org/wiki/Behavior-driven_development.

describe "User" $ do
  it "can be checked for equality" $ do
    user1 shouldBe user1

Define a property test.

describe "User" $ do
  prop "decode . encode === Just" $ do
    let genUser = ...
    (decode . encode) P.=== Just `shouldSatisfy` P.isoWith genUser

Mark the given spec as expected to fail with the given description. Fails tests if they unexpectedly pass.

Can be selected with the marker @xfail

Skip all tests in the given spec with the given description.

Can be selected with the marker @skip

If at least one test is focused, skip all unfocused tests.

This definition includes a WARNING so that CI errors if it's accidentally committed (assuming CI runs with -Wall -Werror).

Since: 0.3.4

Mark tests as tests that should only be run when explicitly specified on the command line.

Adds the given names as plain markers to all tests in the given spec.

See getMarkerName.

Adds the given marker to all the tests in the given spec.

Useful for selecting tests from the command line or identifying tests in hooks

Assert that the given input should match the given value. Equivalent to actual shouldSatisfy P.eq expected

Assert that the given input should not match the given value. Equivalent to actual shouldNotSatisfy P.eq expected

Assert that the given input should satisfy the given predicate.

Assert that the given input should not satisfy the given predicate.

Assert that the given input should return the given value. Equivalent to actual shouldSatisfy (P.returns . P.eq) expected @since 0.3.3

Add any context to display if the test fails.

>>> (code, stdout) <- runCommand ...
>>> context stdout $ code `shouldBe` ExitSuccess

Unconditionally fail the test with the given message.

Request a CallStack.

NOTE: The implicit parameter ?callStack :: CallStack is an implementation detail and should not be considered part of the CallStack API, we may decide to change the implementation in the future.

Since: base-4.9.0.0

Like skip, except allows skipping tests at runtime.

Since: 0.4.0

A property to run, with optional configuration settings specified up front.

Settings should be specified before any other Property calls; any settings specified afterwards are ignored.

Generator for random values of a.

Load a fixture, initializing it if it hasn't been cached already.

A helper for specifying no cleanup.

A helper for defining the cleanup function in-line.

A fixture that provides a temporary directory that can be used in a test.

Fixture for inspecting the captured output.

Intended to be used with OverloadedRecordDot:

output <- getFixture @FixtureCapturedOutput

-- Read all of stdout/stderr so far
stdout <- output.getStdout
stderr <- output.getStderr

-- Read everything in stdout/stderr since the last read
stdout_chunk <- output.readStdout
stderr_chunk <- output.readStderr

Register a CLI flag.

Usage:

{- MyFixture.hs -}
import Skeletest

newtype MyFlag = MyFlag String
instance IsFlag MyFlag where
  flagName = "my-flag"
  flagHelp = "The value for MyFixture"
  flagSpec =
    OptionalFlag
      { default_ = "foo"
      , parse = case
          "illegal" -> Left "invalid flag value"
          s -> Right (MyFlag s)
      }

instance Fixture MyFixture where
  fixtureAction = do
    MyFlag val <- getFlag
    ...

{- Main.hs -}
import MyFixture

cliFlags =
  [ flag @MyFlag
  ]