chell-0.4.0.1: A simple and intuitive library for automated testing.

Safe HaskellNone
LanguageHaskell98

Test.Chell

Contents

Description

Chell is a simple and intuitive library for automated testing. It natively supports assertion-based testing, and can use companion libraries such as chell-quickcheck to support more complex testing strategies.

An example test suite, which verifies the behavior of artithmetic operators.

{-# LANGUAGE TemplateHaskell #-}

import Test.Chell

suite_Math :: Suite
suite_Math = suite "math"
    [ test_Addition
    , test_Subtraction
    ]

test_Addition :: Test
test_Addition = assertions "addition" $ do
    $expect (equal (2 + 1) 3)
    $expect (equal (1 + 2) 3)

test_Subtraction :: Test
test_Subtraction = assertions "subtraction" $ do
    $expect (equal (2 - 1) 1)
    $expect (equal (1 - 2) (-1))

main :: IO ()
main = defaultMain [suite_Math]
 
$ ghc --make chell-example.hs
$ ./chell-example
PASS: 2 tests run, 2 tests passed

Synopsis

Main

defaultMain :: [Suite] -> IO () #

A simple default main function, which runs a list of tests and logs statistics to stdout.

Test suites

data Suite #

A suite is a named collection of tests.

Note: earlier versions of Chell permitted arbitrary nesting of test suites. This feature proved too unwieldy, and was removed. A similar result can be achieved with suiteTests; see the documentation for suite.

Instances

suite :: String -> [Test] -> Suite #

Define a new Suite, with the given name and children.

Note: earlier versions of Chell permitted arbitrary nesting of test suites. This feature proved too unwieldy, and was removed. A similar result can be achieved with suiteTests:

test_Addition :: Test
test_Subtraction :: Test
test_Show :: Test

suite_Math :: Suite
suite_Math = suite "math"
    [ test_Addition
    , test_Subtraction
    ]

suite_Prelude :: Suite
suite_Prelude = suite "prelude"
    (
      [ test_Show
      ]
      ++ suiteTests suite_Math
    )

suiteName :: Suite -> String #

Get a suite's name. Suite names may be any string, but are typically plain ASCII so users can easily type them on the command line.

$ ghci chell-example.hs
Ok, modules loaded: Main.

*Main> suiteName tests_Math
"math"

suiteTests :: Suite -> [Test] #

Get the full list of tests contained within this Suite. Each test is given its full name within the test hierarchy, where names are separated by periods.

$ ghci chell-example.hs
Ok, modules loaded: Main.

*Main> suiteTests tests_Math
[Test "math.addition",Test "math.subtraction"]

Skipping some tests

class SuiteOrTest a #

Minimal complete definition

skipIf_, skipWhen_

Instances

skipIf :: SuiteOrTest a => Bool -> a -> a #

Conditionally skip tests. Use this to avoid commenting out tests which are currently broken, or do not work on the current platform.

tests :: Suite
tests = suite "tests"
    [ test_Foo
    , skipIf builtOnUnix test_WindowsSpecific
    , test_Bar
    ]

skipWhen :: SuiteOrTest a => IO Bool -> a -> a #

Conditionally skip tests, depending on the result of a runtime check. The predicate is checked before each test is started.

tests :: Suite
tests = suite "tests"
    [ test_Foo
    , skipWhen noNetwork test_PingGoogle
    , test_Bar
    ]

Basic testing library

data Assertions a #

See assertions.

Instances

assertions :: String -> Assertions a -> Test #

Convert a sequence of pass/fail assertions into a runnable test.

test_Equality :: Test
test_Equality = assertions "equality" $ do
    $assert (1 == 1)
    $assert (equal 1 1)

class IsAssertion a #

See assert and expect.

Minimal complete definition

runAssertion

Instances

data Assertion #

A single pass/fail assertion. Failed assertions include an explanatory message.

Instances

assertionPassed :: Assertion #

See Assertion.

assertionFailed :: String -> Assertion #

See Assertion.

assert :: Q Exp #

expect :: Q Exp #

die :: Q Exp #

trace :: Q Exp #

note :: String -> String -> Assertions () #

Attach a note to a test run. Notes will be printed to stdout and included in reports, even if the test fails or aborts. Notes are useful for debugging failing tests.

afterTest :: IO () -> Assertions () #

Register an IO action to be run after the test completes. This action will run even if the test failed or aborted.

requireLeft :: Q Exp #

requireRight :: Q Exp #

Built-in assertions

equal :: (Show a, Eq a) => a -> a -> Assertion #

Assert that two values are equal.

notEqual :: (Eq a, Show a) => a -> a -> Assertion #

Assert that two values are not equal.

equalWithin #

Arguments

:: (Real a, Show a) 
=> a 
-> a 
-> a

delta

-> Assertion 

Assert that two values are within some delta of each other.

just :: Maybe a -> Assertion #

Assert that some value is Just.

nothing :: Show a => Maybe a -> Assertion #

Assert that some value is Nothing.

left :: Show b => Either a b -> Assertion #

Assert that some value is Left.

right :: Show a => Either a b -> Assertion #

Assert that some value is Right.

throws :: Exception err => (err -> Bool) -> IO a -> IO Assertion #

Assert that some computation throws an exception matching the provided predicate. This is mostly useful for exception types which do not have an instance for Eq, such as ErrorCall.

throwsEq :: (Eq err, Exception err, Show err) => err -> IO a -> IO Assertion #

Assert that some computation throws an exception equal to the given exception. This is better than just checking that the correct type was thrown, because the test can also verify the exception contains the correct information.

greater :: (Ord a, Show a) => a -> a -> Assertion #

Assert a value is greater than another.

greaterEqual :: (Ord a, Show a) => a -> a -> Assertion #

Assert a value is greater than or equal to another.

lesser :: (Ord a, Show a) => a -> a -> Assertion #

Assert a value is less than another.

lesserEqual :: (Ord a, Show a) => a -> a -> Assertion #

Assert a value is less than or equal to another.

sameItems :: (Foldable container, Show item, Ord item) => container item -> container item -> Assertion #

Assert that two containers have the same items, in any order.

equalItems :: (Foldable container, Show item, Ord item) => container item -> container item -> Assertion #

Assert that two containers have the same items, in the same order.

class IsText a #

Class for types which can be treated as text; see equalLines.

Minimal complete definition

toLines, unpack

Instances

IsText String # 

Methods

toLines :: String -> [String]

unpack :: String -> String

IsText ByteString #

Uses Data.ByteString.Lazy.Char8

IsText ByteString #

Uses Data.ByteString.Char8

IsText Text # 

Methods

toLines :: Text -> [Text]

unpack :: Text -> String

IsText Text # 

Methods

toLines :: Text -> [Text]

unpack :: Text -> String

equalLines :: (Ord a, IsText a) => a -> a -> Assertion #

Assert that two pieces of text are equal. This uses a diff algorithm to check line-by-line, so the error message will be easier to read on large inputs.

equalLinesWith :: Ord a => (a -> [String]) -> a -> a -> Assertion #

Variant of equalLines which allows a user-specified line-splitting predicate.

Custom test types

data Test #

A Test is, essentially, an IO action that returns a TestResult. Tests are aggregated into suites (see Suite).

Instances

Show Test # 

Methods

showsPrec :: Int -> Test -> ShowS #

show :: Test -> String #

showList :: [Test] -> ShowS #

SuiteOrTest Test # 

Methods

skipIf_ :: Bool -> Test -> Test

skipWhen_ :: IO Bool -> Test -> Test

test :: String -> (TestOptions -> IO TestResult) -> Test #

Define a test, with the given name and implementation.

testName :: Test -> String #

Get the name a test was given when it was defined; see test.

runTest :: Test -> TestOptions -> IO TestResult #

Run a test, wrapped in error handlers. This will return TestAborted if the test throws an exception or times out.

Test results

data TestResult #

The result of running a test.

To support future extensions to the testing API, any users of this module who pattern-match against the TestResult constructors should include a default case. If no default case is provided, a warning will be issued.

Constructors

TestPassed [(String, String)]

The test passed, and generated the given notes.

TestSkipped

The test did not run, because it was skipped with skipIf or skipWhen.

TestFailed [(String, String)] [Failure]

The test failed, generating the given notes and failures.

TestAborted [(String, String)] String

The test aborted with an error message, and generated the given notes.

Instances

Failures

data Failure #

Contains details about a test failure.

Instances

failure :: Failure #

An empty Failure; use the field accessors to populate this value.

failureLocation :: Failure -> Maybe Location #

If given, the location of the failing assertion, expectation, etc.

failureLocation is a field accessor, and can be used to update a Failure value.

failureMessage :: Failure -> String #

If given, a message which explains why the test failed.

failureMessage is a field accessor, and can be used to update a Failure value.

Failure locations

data Location #

Contains details about a location in the test source file.

Instances

location :: Location #

An empty Location; use the field accessors to populate this value.

locationFile :: Location -> String #

A path to a source file, or empty if not provided.

locationFile is a field accessor, and can be used to update a Location value.

locationModule :: Location -> String #

A Haskell module name, or empty if not provided.

locationModule is a field accessor, and can be used to update a Location value.

locationLine :: Location -> Maybe Integer #

A line number, or Nothing if not provided.

locationLine is a field accessor, and can be used to update a Location value.

Test options

data TestOptions #

Test options are passed to each test, and control details about how the test should be run.

Instances

defaultTestOptions :: TestOptions #

Default test options.

$ ghci
Prelude> import Test.Chell

Test.Chell> testOptionSeed defaultTestOptions
0

Test.Chell> testOptionTimeout defaultTestOptions
Nothing

testOptionSeed :: TestOptions -> Int #

Get the RNG seed for this test run. The seed is generated once, in defaultMain, and used for all tests. It is also logged to reports using a note.

When using defaultMain, users may specify a seed using the --seed command-line option.

testOptionSeed is a field accessor, and can be used to update a TestOptions value.

testOptionTimeout :: TestOptions -> Maybe Int #

An optional timeout, in millseconds. Tests which run longer than this timeout will be aborted.

When using defaultMain, users may specify a timeout using the --timeout command-line option.

testOptionTimeout is a field accessor, and can be used to update a TestOptions value.