Improve Chart E2E Tests – Time Range Selection Causes Slow and Flaky UI Interaction

[tenthe]

Body

The current E2E tests for chart components need improvement. At the moment, selecting a custom time range through the UI is unnecessarily time-consuming and likely introduces flakiness.

The problem lies in how the time range is selected. The test implementation navigates month-by-month in the calendar widget, often clicking “previous month” multiple times to reach historical dates.

StreamPipes Committer

I acknowledge that I am a maintainer/committer of the Apache StreamPipes project.

[alisha-1000]

Thanks for raising this — I agree that navigating month-by-month in the date picker is both slow and prone to flakiness in E2E tests.
Instead of interacting with the calendar UI step-by-step, we could consider the following improvements:

1. Direct Input Manipulation (Preferred)

   • Programmatically set the start/end date via the underlying input field.
   • Trigger the appropriate change events afterward.
   • This keeps the test focused on chart behavior rather than date-picker navigation.

2. API-Level Time Injection

   • If possible, mock or override the time range at the request/interceptor level.
   • This avoids UI interaction entirely and reduces test runtime significantly.

3. Test Utility for Time Selection

   • Introduce a helper like:

     selectTimeRange({ from: '2023-01-01', to: '2023-01-31' })

   • Internally this could:

     • Either manipulate the input directly, or
     • Jump to a target year/month instead of clicking “previous” repeatedly.

4. Use Fixed Test Data Window

   • Ensure test datasets use recent timestamps so that the default calendar month already contains the target range.
   • This avoids historical navigation altogether.

5. Stability Improvements

   • Wait explicitly for chart re-render or data fetch completion.
   • Avoid relying on arbitrary timeouts.

This should:

• Reduce test execution time
• Improve reliability
• Decouple chart validation from date-picker behavior

If desired, I can also help draft a concrete Cypress/Playwright refactoring example tailored to the current test setup.