Data and Reporting Model

AppraiseJS separates authored test design from execution telemetry.

Authored model (design-time)

• Structure chain: Module -> TestSuite -> TestCase -> TestCaseStep.
• Reusable behavior: TemplateStepGroup -> TemplateStep -> TemplateStepParameter.
• Selector layer: LocatorGroup -> Locator.
• Target context: Environment.
• Filtering/grouping: Tag with FILTER and IDENTIFIER types.

Runtime model (execution-time)

• TestRun: one execution instance with runId, environment, browser, workers, lifecycle status.
• TestRunTestCase: per-case status/result within the run.
• TestRunLog: persisted normalized log stream keyed by runId.
• tracePath and reportPath: artifact pointers for debugging and ingestion.

Report ingestion model

storeReportFromFile() parses Cucumber JSON into:

1. Report
2. ReportFeature + ReportFeatureTag
3. ReportScenario + ReportScenarioTag
4. ReportStep and ReportHook
5. ReportTestCase links back to TestRunTestCase

Scenario-to-test-case matching

Report ingestion matches scenario outcomes using two strategies:

1. Scenario name prefix extraction from [TestCaseTitle].
2. Fallback matching via identifier tags between scenario tags and test-case tags.

Metrics model

• TestCaseMetrics: failure rate, consecutive failures, flaky/repeated-failure flags.
• TestSuiteMetrics: last executed timestamp.
• DashboardMetrics: aggregate operational counters used by dashboard cards/drawers.

Current thresholds in metric calculator:

• Recent window: 7 days.
• Repeated failure: 3+ consecutive failures.
• Flaky: failure rate > 5% and < 100%.