JustBDD.Core 1.0.0-beta-161

JustBDD

Credits

This library is an enhanced copy of a BDD framework that was originally developed Darren Hall @Darren166. The original design of a BDD syntax that is in the core of JustBDD belongs to him, my implementation just extends it a bit for a more convenient and extendable usage.

Description

This library is intended to solve two main problems:

• tests readability. In most of the cases, tests written by developers can be read only by developers and only right after they were written. Also, the test case description in most of the cases is just a test method name. This library allows to make test body as readable as possible for everybody and document a test case properly.
• test steps reusability. When there are too many tests, inevitably there will be a custom solution to group test steps somehow. This lib suggest GWT-style structure as a possible solution.

Comparison to other libraries:

• Specflow. Specflow is a great BDD library to use for complex specification-first development scenarios. JustBDD is intended to be used when there is no BDD specification available at the beginnging.
• LightBDD. LightBDD is also a great library, but it does not facilitate the structure of the tests as well does not facilitate code reusability.
• BDDfy. Almost the same as LightBDD.
• to be added...

Pros:

• Simple. Subjectively, it is way easier to write human-readable scenarios compared to Specflow
• Lightweight. It can be fully forked and redesigned for a specific project if needed. There is not too much code in there.
• Maintainability. The library forces to have a structure for test step methods.
• Code reusability. Allows ultimate reusability of tests steps which facilitates TDD.
• Readability. Test body looks like human readable text, so test can be used for onboarding new people on the project.
• Flexibility. Yes, it facilitates the structure of the tests, but everything else can be adjusted as needed for specific project and domain language.

Cons:

• No parallelization. There are some static things that block parallelization support. Solution: there is no solution for now, however - the library use case is high-level API tests, so parallelization there most-likely may cause random tests failures or complex data setup.
• Awaitable methods look a bit ugly. Solution: no solution here for now, but in the future integration with LightBDD may help to resolve this issue.

Future plans:

• Provide more samples
• Provide integration with LightBDD for prettier output
• Provide more out-of-box helpers, e.g. logging, authentication, etc.

Test example:

See samples folder for an example of usage. The example is not a guidelines, but just one of many possible ways to use the library.

    public class GetAllCategoriesApi_HappyPath_Should : TestFixtureBase
    {
        [Test]
        public void ReturnAllCategories()
        {
            var existingCategory = new CategoryBuilder()
                .WithRandomValues()
                .Build();

            var expectedResponse = new CategoryResponseItemBuilder()
                .FromDomainModel(existingCategory)
                .Build();

            Given.IHave.LoggedInAs.AValidUser().And.DatabaseHas.Category(existingCategory);
            When.ICall.TheCategoriesApi.GetAllEndpoint();
            Then.TheCall.WillSucceed().And.TheCall.WillHaveAResponseEqualTo(new[] { expectedResponse });
        }
    }