Consolidate push/PR test runs into single dotnet test invocation for one HTML report

[Copilot]

Two separate dotnet test steps on push/PR each triggered [BeforeTestRun], causing ExtentReportHelper to initialize twice and produce two timestamped HTML artifacts instead of one consolidated report.

Workflow (tests.yml)

• Replaced the two push/PR steps (filtered + intentional-fail) with a single unfiltered dotnet test step with continue-on-error: true — one process, one [BeforeTestRun], one report

ExtentReportHelper

• Added LogKnownFailure(string) — logs with Warning status and [EXPECTED FAILURE] prefix
• Added MarkTestAsKnownFailure(string) — logs a ⚠️ [KNOWN / EXPECTED FAILURE] warning on the test node itself, which drives ExtentReports to render the test as orange (Warning) rather than red (Fail)

Hooks

• Extracted "intentional-fail" to private const string IntentionalFailTag
• AfterStep: detects @intentional-fail → calls LogWarning instead of LogFail for the failing step
• AfterScenario: detects @intentional-fail → uses "EXPECTED FAIL" screenshot label and calls MarkTestAsKnownFailure instead of treating the error as a genuine failure

// AfterScenario — intentional-fail treated as Warning, not Fail
bool isKnownFailure = _scenarioContext.ScenarioInfo.Tags.Contains(IntentionalFailTag);
var screenshotLabel = isKnownFailure ? "EXPECTED FAIL" : "FAIL";
// ...
if (isKnownFailure)
    ExtentReportHelper.Instance.MarkTestAsKnownFailure("Deliberately incorrect assertion...");

Net result: one HTML report per CI run; regular tests ✅ green, intentional-fail ⚠️ orange with clear label, pipeline stays green.

Original prompt

Problem

When the workflow runs on push or pull_request, two separate dotnet test steps are executed:

1. Run tests excluding intentional-fail (push/PR) — runs all tests except @intentional-fail
2. Run intentional-fail test (push/PR - expected to fail) — runs only @intentional-fail

Because Reqnroll fires [BeforeTestRun] once per process, each dotnet test invocation initializes ExtentReportHelper separately and creates its own timestamped HTML report. This results in two uploaded artifacts — one per test run — which is unprofessional.

Goal

Produce one single consolidated HTML report per CI run where:

• All regular tests show as ✅ Pass
• The intentional-fail scenario shows as ⚠️ Warning (Expected Failure) with a clear [EXPECTED FAILURE] label — not as a red Fail
• The workflow still exits successfully (pipeline is green)

Changes Required

1. .github/workflows/tests.yml

Replace the two separate push/PR test steps with a single dotnet test step that runs all tests together with continue-on-error: true.

The current two steps to replace:

      - name: Run tests excluding intentional-fail (push/PR)
        if: github.event_name != 'workflow_dispatch'
        env:
          BROWSER_HEADLESS: 'true'
        run: |
          dotnet test NotchContactFormTests/NotchContactFormTests/NotchContactFormTests.csproj \
            --no-build \
            --filter "Category!=intentional-fail"

      - name: Run intentional-fail test (push/PR - expected to fail)
        if: github.event_name != 'workflow_dispatch'
        env:
          BROWSER_HEADLESS: 'true'
        run: |
          dotnet test NotchContactFormTests/NotchContactFormTests/NotchContactFormTests.csproj \
            --no-build \
            --filter "Category=intentional-fail"
        continue-on-error: true

Replace with a single step:

      - name: Run all tests (push/PR)
        if: github.event_name != 'workflow_dispatch'
        env:
          BROWSER_HEADLESS: 'true'
        run: |
          dotnet test NotchContactFormTests/NotchContactFormTests/NotchContactFormTests.csproj \
            --no-build
        continue-on-error: true

2. NotchContactFormTests/NotchContactFormTests/Helpers/ExtentReportHelper.cs

Add a LogKnownFailure(string message) method that logs with Warning status:

public void LogKnownFailure(string message)
    => _currentTest?.Warning($"[EXPECTED FAILURE] {message}");

Also add a MarkTestAsKnownFailure(string reason) method that marks the current test node itself as Warning status with a clear label. In ExtentReports, a test's final status is determined by the highest-severity log entry — so logging a Warning (instead of Fail) will make the test show as Warning/orange rather than red Fail.

Add:

public void MarkTestAsKnownFailure(string reason)
{
    _currentTest?.Warning($"⚠️ [KNOWN / EXPECTED FAILURE] This test is intentionally failing to demonstrate failure reporting.\nReason: {reason}");
}

3. NotchContactFormTests/NotchContactFormTests/Hooks/Hooks.cs

In the AfterScenario hook, detect whether the failing scenario has the @intentional-fail tag. If it does, call MarkTestAsKnownFailure instead of LogFail, so the report shows it as Warning (orange) instead of Fail (red).

Current AfterScenario:

[AfterScenario]
public void AfterScenario()
{
    if (_scenarioContext.TestError != null)
    {
        try
        {
            var screenshotHelper = _container.Resolve<ScreenshotHelper>();
            var base64 = screenshotHelper.CaptureScreenshotAsBase64(
                _scenarioContext.ScenarioInfo.Title, "FAIL");

            if (!string.IsNullOrEmpty(base64))
                ExtentReportHelper.Instance.AttachScreenshot(base64, "Screenshot at failure");
        }
        catch { /*screenshot failure must not mask the original error */ }
    }

    _driverContext?.Dispose();
}

Update to:

[AfterScenario]
public void AfterScenario()
{
    if (_scenarioContext.TestError != null)
    {
        bool isKnownFailure = _scenarioContext.ScenarioInfo.Tags.Contains("intentional-fail");

        try
        {
            var screenshotHelper = _container.Resolve<ScreenshotHelper>();
            var screenshotLabel = isKnownFailure ? "EXPECTED FAIL" : "FAIL";
            var base64 = screenshotHelper.CaptureScreenshotAsBase64(
                _scenarioContext.ScenarioInfo.Title, screenshotLabel);

            if (!string.IsNullOrEmpty(base64))
                ExtentReportHelper.Instance.AttachScreenshot(base64, $"Screenshot at {screenshotLabel}");
        }
        catch { /* screenshot failure must not mask the original error */ }

        if (isKnownFailure)
        {
            ExtentReportHelper.Instance.MarkTestAsKnownFailure(
                "Assertion uses a deliberately incorrect expected value to demonstrate failure reporting in ExtentReports.");
        }
    }

...

</details>



<!-- START COPILOT CODING AGENT SUFFIX -->

*This pull request was created from Copilot chat.*
>

<!-- START COPILOT CODING AGENT TIPS -->
---

📍 Connect Copilot coding agent with [Jira](https://gh.io/cca-jira-docs), [Azure Boards](https://gh.io/cca-azure-boards-docs) or [Linear](https://gh.io/cca-linear-docs) to delegate work to Copilot in one click without leaving your project management tool.