Skip to content

Commit 24a9c18

Browse files
cursoragentwilliscool
cursoragent
and
williscool
committed
docs: Add detailed test suite consolidation plan for Issue #174
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>
1 parent 323f37d commit 24a9c18

1 file changed

Lines changed: 201 additions & 0 deletions

File tree

‎docs/test-consolidation-plan.md‎

Lines changed: 201 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,201 @@
1+
# Test Suite Consolidation Plan
2+
3+
**Issue:** [#174 - Test suite getting bloated](https://github.com/williscool/CalendarNotification/issues/174)
4+
5+
**Coverage Baseline:** See `docs/coverage-baseline/BASELINE-SUMMARY.md`
6+
7+
## Current State
8+
9+
| Category | Files | Lines of Code |
10+
|----------|-------|---------------|
11+
| Unit Tests (Robolectric) | ~50 | ~18,582 |
12+
| Integration Tests (Instrumented) | ~45 | ~19,743 |
13+
| **Total** | ~95 | **~38,325** |
14+
15+
## Analysis Summary
16+
17+
### Coverage Baseline (Pre-Consolidation)
18+
19+
| Metric | Unit Tests | Integration Tests |
20+
|--------|------------|-------------------|
21+
| Instruction Coverage | 25.97% | 43.79% |
22+
| Branch Coverage | 21.80% | 31.07% |
23+
| Line Coverage | 24.84% | 45.34% |
24+
| Method Coverage | 34.30% | 46.07% |
25+
26+
### Key Findings
27+
28+
1. **Massive Duplication**: ~20 test classes exist as BOTH Robolectric (unit) AND instrumented (integration) versions with nearly identical tests
29+
2. **Deprecated Tests**: ~5,000 lines in `deprecated_raw_calendarmonitor/` folder that should be removed or archived
30+
3. **Testing Overly Simple Logic**: Some tests verify trivial functionality that could be simplified
31+
4. **Large Monolithic Tests**: Several test files exceed 700+ lines and could be split or simplified
32+
33+
---
34+
35+
## Consolidation Plan
36+
37+
### Phase 1: Remove Deprecated Tests (~5,000 lines reduction)
38+
39+
**Target Directory:** `app/src/androidTest/java/.../deprecated_raw_calendarmonitor/`
40+
41+
| File | Lines | Action |
42+
|------|-------|--------|
43+
| `CalendarMonitorServiceTest.kt` | 2,001 | DELETE - replaced by `FixturedCalendarMonitorServiceTest` |
44+
| `CalendarMonitorServiceTestFirstScanEver.kt` | 2,034 | DELETE - covered by newer tests |
45+
| `CalendarMonitorServiceEventReminderTest.kt` | 966 | DELETE - functionality tested elsewhere |
46+
47+
**Verification:** Run integration tests after removal to ensure no functionality gaps.
48+
49+
---
50+
51+
### Phase 2: Consolidate Duplicate Test Pairs (~5,000-7,000 lines reduction)
52+
53+
For tests that exist in both unit AND integration form with nearly identical logic, consolidate to ONE version (preferring Robolectric for speed, keeping instrumented only where Android runtime is essential).
54+
55+
#### Priority 1 - Direct Duplicates (Pure Logic Tests - Keep Robolectric Only)
56+
57+
These tests verify pure Kotlin/Java logic that doesn't require real Android runtime:
58+
59+
| Test Pair | Unit Lines | Int Lines | Action |
60+
|-----------|------------|-----------|--------|
61+
| `TagsManager*Test.kt` | 206 | 205 | **DELETE integration** - pure string parsing |
62+
| `AlarmScheduler*Test.kt` | 282 | 289 | **KEEP BOTH** - integration verifies real AlarmManager |
63+
| `CalendarReloadManager*Test.kt` | 292 | 283 | **DELETE integration** - logic can be tested with mocks |
64+
| `SnoozeTest.kt` / `SnoozeRobolectricTest.kt` | 334 | 380 | **DELETE integration** - snooze logic testable with mocks |
65+
| `ComponentIsolation*Test.kt` | 319 | 297 | **DELETE integration** - isolation tests work in Robolectric |
66+
67+
**Estimated Reduction:** ~1,400 lines
68+
69+
#### Priority 2 - UI Tests (Keep Integration, Delete/Reduce Robolectric)
70+
71+
UI tests benefit from real Android runtime. Robolectric UI tests are often flaky or limited:
72+
73+
| Test Pair | Unit Lines | Int Lines | Action |
74+
|-----------|------------|-----------|--------|
75+
| `MainActivityModernTest.kt` | 775 | 873 | **REVIEW** - consider keeping integration only |
76+
| `MainActivityTest.kt` | 367 | 460 | **REVIEW** - may consolidate with Modern |
77+
| `ViewEventActivityTest.kt` | 383 | 306 | **REVIEW** - prefer integration for UI |
78+
| `SnoozeAllActivityTest.kt` | N/A | 279 | **KEEP** - no Robolectric version |
79+
80+
**Estimated Reduction:** ~500-1,000 lines (after review)
81+
82+
#### Priority 3 - CalendarProvider Tests (Keep Integration)
83+
84+
CalendarProvider tests require real ContentProvider access:
85+
86+
| Test Pair | Unit Lines | Int Lines | Action |
87+
|-----------|------------|-----------|--------|
88+
| `CalendarProviderBasicTest.kt` | 349 | 270 | **DELETE Robolectric** - needs real ContentProvider |
89+
| `CalendarProviderEventTest.kt` | 314 | 276 | **DELETE Robolectric** - needs real ContentProvider |
90+
| `CalendarProviderReminderTest.kt` | 340 | 315 | **DELETE Robolectric** - needs real ContentProvider |
91+
| `CalendarBackupRestoreTest.kt` | 289 | 333 | **DELETE Robolectric** - needs real storage |
92+
93+
**Estimated Reduction:** ~1,300 lines
94+
95+
#### Priority 4 - Storage/Database Tests (Keep Integration)
96+
97+
Storage tests benefit from real SQLite/Room:
98+
99+
| Test Pair | Unit Lines | Int Lines | Action |
100+
|-----------|------------|-----------|--------|
101+
| `EventDismissTest.kt` | 726 | 735 | **DELETE Robolectric** - needs real Room |
102+
| `OriginalEventDismiss*Test.kt` | 318 | ~300 | **DELETE Robolectric** - needs real Room |
103+
104+
**Estimated Reduction:** ~1,000 lines
105+
106+
#### Priority 5 - Broadcast Receiver Tests (Keep Both, Reduce Overlap)
107+
108+
Some broadcast tests need real Android, others can be unit tested:
109+
110+
| Test Pair | Unit Lines | Int Lines | Action |
111+
|-----------|------------|-----------|--------|
112+
| `BroadcastReceiverTest.kt` | 356 | 315 | **MERGE** - combine unique tests, remove duplicates |
113+
| `ReminderAlarmTest.kt` | 422 | 305 | **MERGE** - keep real alarm behavior in integration |
114+
115+
**Estimated Reduction:** ~300 lines
116+
117+
---
118+
119+
### Phase 3: Simplify Large Tests
120+
121+
#### Tests to Review for Simplification
122+
123+
| File | Lines | Opportunity |
124+
|------|-------|-------------|
125+
| `EventNotificationManagerRobolectricTest.kt` | 1,480 | Extract helper methods, reduce setup boilerplate |
126+
| `NotificationContextInvariantTest.kt` | 999 | Review if all cases are necessary |
127+
| `EventFormatterRobolectricTest.kt` | 956 | Consider parameterized tests |
128+
| `MainActivityModernTest.kt` | 873 | Split into focused feature tests |
129+
| `SettingsBackupManagerRobolectricTest.kt` | 760 | Use shared fixtures |
130+
| `EventDismissTest.kt` | 735 | Consolidate similar test cases |
131+
| `EventsStorageMigrationTest.kt` | 637 | Consider if all migration paths need testing |
132+
133+
**Estimated Reduction:** ~500-1,500 lines through refactoring
134+
135+
---
136+
137+
### Phase 4: Consolidate Test Fixtures
138+
139+
| Fixture Pair | Lines | Action |
140+
|--------------|-------|--------|
141+
| `UITestFixture.kt` | 978 | **MERGE** - combine shared utilities |
142+
| `UITestFixtureRobolectric.kt` | 551 | Into single fixture with platform-specific overrides |
143+
| `CalendarProviderTestFixture.kt` | 414 | **KEEP** - integration only |
144+
| `CalendarMonitorTestFixture.kt` | 398 | **KEEP** - integration only |
145+
| `BaseCalendarTestFixture.kt` | 531 | **KEEP** - base class |
146+
147+
**Estimated Reduction:** ~400 lines
148+
149+
---
150+
151+
## Implementation Order
152+
153+
### Batch 1: Low-Risk Deletions (No coverage loss expected)
154+
1. Delete `deprecated_raw_calendarmonitor/` folder (~5,000 lines)
155+
2. Delete `TagsManagerTest.kt` (integration) - keep Robolectric version
156+
3. Delete CalendarProvider Robolectric tests (keep integration versions)
157+
158+
### Batch 2: Consolidations Requiring Verification
159+
4. Merge BroadcastReceiver test pairs
160+
5. Consolidate EventDismiss test pairs
161+
6. Review and consolidate SnoozeTest pairs
162+
163+
### Batch 3: UI Test Consolidation
164+
7. Review MainActivity test strategy
165+
8. Consolidate UI test fixtures
166+
167+
### Batch 4: Refactoring Large Tests
168+
9. Apply shared fixtures to large test files
169+
10. Use parameterized tests where applicable
170+
171+
---
172+
173+
## Success Criteria
174+
175+
1. **No Coverage Regression**: Coverage percentages should remain at or above baseline
176+
2. **Test Count Reduction**: Target 20-30% reduction in total test lines
177+
3. **Test Speed**: Unit test suite should run faster (fewer duplicate runs)
178+
4. **Maintainability**: Clear separation of unit vs integration test responsibilities
179+
180+
---
181+
182+
## Estimated Total Reduction
183+
184+
| Phase | Lines Removed |
185+
|-------|---------------|
186+
| Phase 1: Deprecated | ~5,000 |
187+
| Phase 2: Duplicates | ~4,500 |
188+
| Phase 3: Simplify | ~1,000 |
189+
| Phase 4: Fixtures | ~400 |
190+
| **Total** | **~10,900 lines (28% reduction)** |
191+
192+
**Post-Consolidation Target:** ~27,000 lines (down from ~38,000)
193+
194+
---
195+
196+
## Notes
197+
198+
- Always verify coverage after each batch
199+
- Run full test suite before and after changes
200+
- Document any intentional coverage reductions with justification
201+
- Keep integration tests for anything that truly requires Android runtime

0 commit comments

Comments
 (0)