[MBL-19469][Student] - Discussion Checkpoints Notifications E2E Test

[kdeakinstructure]

Implement Discussion checkpoint notifications E2E test.
Refactor notification assertion and click logic to handle contains.

refs: MBL-19469
affects: Student
release note: -

• Run E2E test suite

[claude]

Review: Discussion Checkpoints Notifications E2E Test

This PR adds an E2E test for Discussion Checkpoints notifications and refactors NotificationPage to support both exact and partial text matching. The test structure is solid and follows the existing patterns well.

Issues Found

• Breaking behavior change in NotificationPage (potential regression) — NotificationPage.kt lines 42–47 and 77–84
  The original assertNotificationDisplayed(title) and clickNotification(title) always used containsTextCaseInsensitive, which does substring + case-insensitive matching. With the new default (contains = false), all existing callers silently switch to withText(title) — exact, case-sensitive matching. Any existing test relying on the old partial/case-insensitive behavior will now fail without any compile-time warning. Consider inverting the default to preserve backward compatibility, or renaming the parameter to make the intent explicit:

  // Option A: preserve old default behavior
  fun assertNotificationDisplayed(title: String, exactMatch: Boolean = false) {
      val matcher = if (exactMatch) {
          allOf(withText(title), withId(R.id.title))
      } else {
          allOf(containsTextCaseInsensitive(title), withId(R.id.title))
      }

• Potential timezone-related test flakiness — NotificationsE2ETest.kt, assignmentDetailsDisplayFormat line
  getCustomDateCalendar(n) creates a Calendar in UTC (sets hour=10, minute=1, second=1 in UTC). However, assignmentDetailsDisplayFormat does not set a timezone, so SimpleDateFormat defaults to the device's local timezone when formatting. On a CI runner in a non-UTC timezone, the formatted time will differ from what the app displays (which uses UTC). Add an explicit timezone:

  val assignmentDetailsDisplayFormat = SimpleDateFormat("MMM d, yyyy h:mm a", Locale.US).apply {
      timeZone = TimeZone.getTimeZone("UTC")
  }

• Unused syllabusBody in seedData — NotificationsE2ETest.kt, testDiscussionCheckpointsNotificationsE2E
seedData(..., syllabusBody = "this is the syllabus body") seeds data that is never referenced by this test. Removing it reduces noise and avoids unnecessary server calls.

• Inconsistent retry block formatting — NotificationsE2ETest.kt, retry block
  The closing brace of catchBlock is placed on the same line as the call (refresh() }), while the main lambda opens on a new line. This looks inconsistent with the Kotlin style used elsewhere in the file:

  retryWithIncreasingDelay(times = 10, maxDelay = 3000, catchBlock = { refresh() }) {
      notificationPage.assertNotificationDisplayed(discussionWithCheckpointsTitle)
      notificationPage.assertNotificationDisplayed("Assignment Created - $discussionWithCheckpointsTitle", contains = true)
  }

Positive Notes

• Migrating from StudentTest to StudentComposeTest is the right call for a test that interacts with Compose-based pages like discussionDetailsPage.
• Using retryWithIncreasingDelay for polling notification appearance is a good pattern for flake-resistant E2E tests.
• The date formatting split (seed format vs. display format) is thorough and covers the format mismatch between the API and the UI.
• The contains: Boolean = false extension to the page object methods is a clean, backward-compatible API addition in concept — just needs the default value adjusted as noted above.

[github-actions]

🧪 Unit Test Results

✅ 📱 Student App

• Tests: 1252 total, 0 failed, 0 skipped
• Duration: 0.000s
• Success Rate: 100%

✅ 🌅 Horizon

• Tests: 790 total, 0 failed, 0 skipped
• Duration: 37.838s
• Success Rate: 100%

✅ 📦 Submodules

• Tests: 3348 total, 0 failed, 0 skipped
• Duration: 52.346s
• Success Rate: 100%

---

📊 Summary

• Total Tests: 5390
• Failed: 0
• Skipped: 0
• Status: ✅ All tests passed!

Last updated: Mon, 06 Apr 2026 22:19:09 GMT

[github-actions]

📊 Code Coverage Report

✅ Student

• PR Coverage: 42.65%
• Master Coverage: 42.65%
• Delta: +0.00%

✅ Teacher

• PR Coverage: 25.37%
• Master Coverage: 25.37%
• Delta: +0.00%

✅ Pandautils

• PR Coverage: 23.68%
• Master Coverage: 23.68%
• Delta: +0.00%

---

📈 Overall Average

• PR Coverage: 30.57%
• Master Coverage: 30.57%
• Delta: +0.00%

[github-actions]

Student Install Page