Skip to content

A small .NET Standard library for writing well-structured, clear-cut 3A tests.

License

Notifications You must be signed in to change notification settings

balazs-kis/test-lite

Repository files navigation

TestLite

Test Lite is a small library for writing well-structured, clear-cut 3A tests.

Build Status Coverage Status Nuget License: MIT pull requests: welcome

Purpose

Test Lite helps to enforce the 3A pattern for unit tests. Each test can be written in the following format: Test.Arrange().Act().Assert(). The fluent API makes it possible to define test methods as expression-bodied members.

Usage

The Arrange method expects a function which returns the unit under test, and optionally some additional parameter(s). If additional parameters are present, the return type should be a value tuple, e.g. return (underTest, parameter);. If no arrangement is needed, the Arrange method can be called without any parameters.

The Act method expects a function which takes the arguments determined by the previous Arrange method, and returns a result. This result, again, can be a value tuple if multiple return values are needed. Any exception thrown be the function passed to Act will be caught and can be validated later.

The Assert method has two overloads. One of them takes a validator function, which can be used to declare assertions. In this case the input type is an ActResult class. The other variant is parameterless and can be used to chain validation methods. A generic Validate method can also be chained to run custom validations on the result. In this case the input is not an ActResult instance, but the value passed from the Act method.

Examples

A simple test case which only validates if the Act section did not threw any exception (using Fluent Assertions):

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod())
    .Assert(result => result.IsSuccess.Should().BeTrue("This basic operation should succeed"));

The same test case with chained assertion:

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod())
    .Assert().IsSuccess("This basic operation should succeed");

A test case checking if the operation threw the expected exception (using Fluent Assertions):

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod(10))
    .Assert(result => 
    {
        result.IsFailure.Should()
            .BeTrue("The array should be smaller than 10");

        result.Exception.Should()
            .BeAssignableTo<IndexOutOfRangeException>("The array should be smaller than 10");
    });

The same test case with chained assertion:

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod(10))
    .Assert().ThrewException<IndexOutOfRangeException>("The array should be smaller than 10");

A test case checking if the operation returned the expected value (using Fluent Assertions):

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod(42))
    .Assert(result => 
    {
        result.IsSuccess.Should()
            .BeTrue("This basic operation should succeed");

        result.Value.Should()
            .Be(42, "The output should be equal to the input");
    });

The same test case with chained assertion (using Fluent Assertions):

[TestMethod]
MyTestCase() => Test
    .Arrange(() => new MyClass())
    .Act(underTest => underTest.InstanceMethod(42))
    .Assert()
        .Validate(result => result.Should().Be(42, "The output should be equal to the input"));