Skill: Testing Guide

Write or review xUnit tests for a Syncfusion Toolkit for .NET MAUI control change.

When to Use This Skill

• ✅ Writing new tests after a bug fix or feature
• ✅ Reviewing whether existing tests cover a change adequately
• ✅ Called from another agent to handle the test-writing step

Inputs Required

The caller (agent or user) must provide:

Step 1: Locate the Test File

maui/tests/Syncfusion.Maui.Toolkit.UnitTest/{ControlArea}/

find maui/tests/Syncfusion.Maui.Toolkit.UnitTest/{ControlArea}/ -name "*.cs" | head -20

If no test file exists for the control area, create one following the naming pattern of adjacent test files.

Step 2: What to Test

Add tests for all scenarios that are possible given the control's design:

Naming convention: MethodOrProperty_Scenario_ExpectedBehavior

[Fact]
public void SomeProperty_WhenSetToValue_ShouldUpdateBehavior()
{
	// Arrange
	var control = new SomeControl();

	// Act
	control.SomeProperty = expectedValue;

	// Assert
	Assert.Equal(expectedValue, control.InternalState);
}

Use hard tabs for indentation. Do NOT use the private keyword.

Step 3: What NOT to Test (Impossible Scenarios)

Do NOT write tests for scenarios that are architecturally impossible.

A scenario is impossible if:

Step 4: Document Skipped Scenarios

For each impossible scenario, add a row to ## Skipped Tests in the caller's state file (or report directly to the user if invoked standalone):

## Skipped Tests
| Scenario | Reason Impossible |
|----------|-------------------|
| `Items_WhenNull_ShouldNotThrow` | `Items` is initialized in constructor; can never be null |
| `OnPlatformRenderer_WhenDetached` | Requires native platform runtime; not testable in xUnit |

Step 5: Build and Run Tests

# Build first to catch compile errors
dotnet build ./Syncfusion.Maui.Toolkit.sln

# Run all unit tests
dotnet test maui/tests/Syncfusion.Maui.Toolkit.UnitTest/

# Run tests filtered to this control area
dotnet test maui/tests/Syncfusion.Maui.Toolkit.UnitTest/ --filter "FullyQualifiedName~{ControlArea}"

Report results to the caller or user: tests passed ✅, failed ❌, or skipped with reason.