Test: Consolidation Plan #205

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Open

williscool wants to merge 5 commits into master from cursor/jacoco-coverage-baseline-02cc

Owner

williscool commented Jan 18, 2026 •

edited by cursor bot

Loading

Note

Introduces documentation for current test coverage and a roadmap to reduce test bloat.

Adds coverage baseline under docs/coverage-baseline/ with BASELINE-SUMMARY.md, unit-test-coverage.csv, and integration-test-coverage.csv (Unit: 25.97% lines, Integration: 45.34% lines; 334 classes analyzed)
Adds docs/test-consolidation-plan.md outlining phased deletions and merges of duplicate/legacy tests, fixture consolidation, and refactors targeting ~10.9k LOC reduction (~28%) while maintaining coverage at or above baseline

^{Written by Cursor Bugbot for commit 24a9c18. This will update automatically on new commits. Configure here.}

cursoragent and others added 2 commits

January 18, 2026 00:20


          docs: Add JaCoCo coverage baseline from cursor/unit-test-jacoco-csv-66a5

323f37d

Baseline coverage captured from workflow run 21102796115 (2026-01-17):

Unit Test Coverage:
- Instructions: 25.97% (20,646 / 79,496)
- Branches: 21.80% (1,235 / 5,664)
- Lines: 24.84% (3,688 / 14,849)
- Methods: 34.30% (959 / 2,796)

Integration Test Coverage:
- Instructions: 43.79% (34,814 / 79,496)
- Branches: 31.07% (1,760 / 5,664)
- Lines: 45.34% (6,732 / 14,849)
- Methods: 46.07% (1,288 / 2,796)

This establishes a baseline for Issue #174 work on test suite consolidation.

Co-authored-by: wharris <wharris@upscalews.com>


          docs: Add detailed test suite consolidation plan for Issue #174

24a9c18

Analysis of ~38,000 lines of test code identified:
- ~5,000 lines in deprecated_raw_calendarmonitor/ to remove
- ~20 duplicate test pairs (Robolectric + instrumented) to consolidate
- Large tests (700+ lines) to simplify

Estimated reduction: ~10,900 lines (28% reduction)
Target: ~27,000 lines while maintaining coverage baseline

Co-authored-by: wharris <wharris@upscalews.com>

github-actions bot commented Jan 18, 2026

Build artifacts for PR #205 (commit bf9dd6f) are available:

You can download these artifacts from the "Artifacts" section of the workflow run.

github-actions bot commented Jan 18, 2026 •

edited

Loading

Code Coverage Report

Overall Project	34.55%	🍏

There is no coverage information present for the Files changed

github-actions bot commented Jan 18, 2026

📊 Code Coverage Summary

Coverage Type	Coverage
Overall	34.5
Changed Files	100

View detailed coverage report

williscool commented

View reviewed changes

docs/test-consolidation-plan.md Outdated

    
              ### Key Findings

              1. **Massive Duplication**: ~20 test classes exist as BOTH Robolectric (unit) AND instrumented (integration) versions with nearly identical tests

Owner Author

williscool Jan 18, 2026

This duplication is intentional. There are some things that roboelectric cannot test for instance anything to do with sqlite because

we use a version that has custom extensions
Even if we didn't, I heard that there. Their sequel light support is rudimentary at best so given how important dealing with storage is in this application, I don't trust testing it there

But robo electric is great for quick sanity checks for application functionality overall.

The way I have it set up is only the roboelectric test block PRS.

But releases don't if the Android instrumentation tests fail

williscool commented

View reviewed changes

docs/test-consolidation-plan.md Outdated

    
              ### Key Findings

              1. **Massive Duplication**: ~20 test classes exist as BOTH Robolectric (unit) AND instrumented (integration) versions with nearly identical tests

              2. **Deprecated Tests**: ~5,000 lines in `deprecated_raw_calendarmonitor/` folder that should be removed or archived

Owner Author

williscool Jan 18, 2026

I'm really paranoid here. At this point the fixtures have been around for almost a year so I trust them a lot but it's a little bit harder to follow exactly what's going on in them.

In the deprecated test, it's pretty straightforward to follow exactly what's happening.

And they test the app core application functionality (that it that we're watching the calendar properly to get events. If this breaks nothing else works)

Maybe there's a refactoring of the fixtures that we could do that would make them easier to follow. Maybe there's some kind of fixture framework we should be using or something like that then I would feel more comfortable getting rid of them

williscool commented

View reviewed changes

docs/test-consolidation-plan.md Outdated

    
              1. **Massive Duplication**: ~20 test classes exist as BOTH Robolectric (unit) AND instrumented (integration) versions with nearly identical tests

              2. **Deprecated Tests**: ~5,000 lines in `deprecated_raw_calendarmonitor/` folder that should be removed or archived

              3. **Testing Overly Simple Logic**: Some tests verify trivial functionality that could be simplified

Owner Author

williscool Jan 18, 2026

100% agree here

williscool commented

View reviewed changes

docs/test-consolidation-plan.md Outdated

    
              1. **Massive Duplication**: ~20 test classes exist as BOTH Robolectric (unit) AND instrumented (integration) versions with nearly identical tests

              2. **Deprecated Tests**: ~5,000 lines in `deprecated_raw_calendarmonitor/` folder that should be removed or archived

              3. **Testing Overly Simple Logic**: Some tests verify trivial functionality that could be simplified

              4. **Large Monolithic Tests**: Several test files exceed 700+ lines and could be split or simplified

Owner Author

williscool Jan 18, 2026

100% agree here also

williscool commented

View reviewed changes

docs/test-consolidation-plan.md Outdated

    
              ---

              ### Phase 2: Consolidate Duplicate Test Pairs (~5,000-7,000 lines reduction)

Owner Author

williscool Jan 18, 2026

Kind of paranoid here too. Don't really want to see the coverage number go down in the instrumentation test, especially since we don't block PRS on them


          docs: Revise consolidation plan based on PR feedback

d74d4cf

Key changes based on @williscool's review:

- KEEP Robolectric + Instrumented pairs (intentional design)
  - Robolectric: quick PR-blocking sanity checks
  - Instrumentation: real Android verification for releases
  - SQLite/Room MUST use instrumentation (custom extensions)

- DEFER deprecated test removal
  - Tests are clearer than fixtures
  - Test core calendar monitoring (critical path)
  - Prerequisite: improve fixture clarity first

- FOCUS on safe improvements:
  - Remove overly simple tests
  - Refactor large monolithic tests
  - Improve fixture documentation

Revised estimate: ~1,000-2,500 lines (3-7%) vs original ~10,900 (28%)
Quality > quantity for test consolidation

Co-authored-by: wharris <wharris@upscalews.com>

github-actions bot commented Jan 18, 2026

Build artifacts for PR #205 (commit d62c33b) are available:

You can download these artifacts from the "Artifacts" section of the workflow run.


          docs: Add Android testing best practices research

af99f6f

Research comparing current test setup against industry best practices:

Current Setup:
- JUnit 4, MockK, Robolectric 4.16, Ultron, Espresso
- Custom fixtures (BaseCalendarTestFixture, MockCalendarProvider, etc.)

Quick Wins Identified:
- Add Truth assertions for better failure messages
- Use ShadowAlarmManager instead of manual mocking
- Create Object Mother classes for test data
- Verify using PAUSED looper mode in Robolectric

Medium Effort:
- Formalize Robot Pattern for UI tests
- Extract Test Data Builders from fixtures
- Document fixture usage with examples

Investigated but not recommended:
- Hilt Testing (requires architecture changes)
- JUnit 5 (limited Android support)
- Moving SQLite tests to Robolectric (custom extensions need real DB)

Includes sample Object Mother implementation for EventAlertRecord.

Co-authored-by: wharris <wharris@upscalews.com>

github-actions bot commented Jan 18, 2026

📊 Code Coverage Summary

Coverage Type	Coverage
Overall	34.51
Changed Files	100

View detailed coverage report

github-actions bot commented Jan 18, 2026

Build artifacts for PR #205 (commit 2e0e008) are available:

You can download these artifacts from the "Artifacts" section of the workflow run.

github-actions bot commented Jan 18, 2026

📊 Code Coverage Summary

Coverage Type	Coverage
Overall	25.35
Changed Files	100

View detailed coverage report


          feat: Implement testing quick wins and add future improvements to plan

f7d90e0

Quick Wins Implemented:

1. Truth assertions (com.google.truth:truth:1.4.2)
   - Added for both unit and instrumented tests
   - Better failure messages and fluent API

2. Object Mother pattern for test data
   - EventMother.kt - EventAlertRecord factory methods
   - MonitorAlertMother.kt - MonitorEventAlertEntry factory methods
   - Reduces boilerplate, centralizes test data creation

3. Robolectric PAUSED looper mode
   - Created robolectric.properties with explicit config
   - looperMode=PAUSED for realistic async testing
   - sdk=34 for Android 14 behavior

4. ShadowAlarmManager helper
   - AlarmManagerTestHelper.kt for cleaner alarm testing
   - Provides introspection into scheduled alarms
   - Alternative to MockK mocking for AlarmManager

Updated consolidation plan with:
- Documentation of implemented quick wins
- Future improvements section (Robot Pattern, Gradle Managed Devices, etc.)
- Not recommended items with reasons

Co-authored-by: wharris <wharris@upscalews.com>

github-actions bot commented Jan 18, 2026

Build artifacts for PR #205 (commit c21be17) are available:

You can download these artifacts from the "Artifacts" section of the workflow run.

github-actions bot commented Jan 18, 2026

📊 Code Coverage Summary

Coverage Type	Coverage
Overall	34.55
Changed Files	100

View detailed coverage report

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet