Randomizer

The Randomizer tool provides methods to create instances populated with random data, as well as injection behavior.

Example

Typically, many tests will verify behavior based upon specific values on an object. Randomizing an object then specifying the value to test keeps the scope of the test clearly defined and ensures none of the other values matter as expected:

/// <summary>Verifies the method can't find a child on null.</summary>
[Fact]
public void FindFirstChild_NullChildrenThrows()
{
    DataSample item = Tools.Randomizer.Create<DataSample>();
    item.Children = null;

    Tools.Asserter.Throws<ArgumentNullException>(() => item.FindFirstChild());
}

Injection

Using the inject functionality, the Randomizer can create a type using provided values and falling back to randomization for any necessary data not provided. When utilizing the CreateFor functionality via attributes, fakes are automatically injected into objects created afterwards. For example:

[Theory, RandomData]
public void Test([Fake] Child a, Parent b) { }

If Parent requires an instance of Child in the constructor, a will be used to create b.

Creation & Customization

The Randomizer requires a Faker to provide stubbing and an IRandom instance which controls the core value randomization. Like other tools, custom CreateHint instances can be passed in to control behavior for any specific types.

By passing a SeededRandom during construction, the Randomizer can make tests random but deterministic so that results can be repeated to help track down issues. See the IRandom page for more details.