KMP phase 6D: core:data:database becomes a KMP module, on AndroidSQLiteDriver - #239
Merged
Conversation
… byte epochToIso becomes kotlin.time.Instant.fromEpochMilliseconds(ms).toString() — the zero-new-dependency replacement (stdlib, stable since Kotlin 2.3), taken now so WorkoutExportMapper is commonMain-eligible when the module converts. The rendering is pinned twice, in different currencies: a new literal-string test (whole-second, zero, three-digit fractional, negative epoch) and the existing DatabaseJsonExporterImplTest assertion, which keeps java.time on the test side deliberately — it is now a cross-library oracle, not a leftover. Mutation M-ISO1 (fromEpochSeconds) reds 4 tests including the cross-oracle; the epoch-zero pin alone survives, as 0s and 0ms name the same instant. Reverted, not committed. Module suite: 128 tests / 0 failures. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
Room.databaseBuilder(context, AppDatabase::class.java, name) becomes the KMP-shared Room.databaseBuilder<AppDatabase>(context, name), across the one production site, seven src/test, four androidTest, two database-test and one app:app androidTest. The sixteenth site — MigrationTestHelper's KClass argument — was already the Room 3 form and does not move. The module is still Android-only; the reified overload resolves the constructor reflectively until @ConstructedBy arrives with the conversion. One inference consequence, worth knowing before phase 7 repeats it: with the Class<T> parameter gone, a bare ApplicationProvider.getApplicationContext() argument no longer anchors either type parameter — the builder's factory overloads and the context accessor's own T deadlock inference. The three such sites pin it as getApplicationContext<Application>(). Suites: database 128/0, exercise 212/0 (RepositoryTestEnv consumers untouched); both androidTest APKs compile; detekt separate, 56/56. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
…ve before the module that needs them KmpLibraryConventionPlugin now calls withDeviceTest unconditionally — mirroring classic AGP, where every library carries an androidTest component whether or not sources exist — and registers two more aliases: assembleDebugAndroidTest → assembleAndroidDeviceTest (the vanish ConfigureInstrumentedSuiteGate pre-diagnosed) and connectedDebugAndroidTest → connectedAndroidDeviceTest. The second is a finding, not a plan item: ui_tests.yml invokes connectedDebugAndroidTest literally, so the assemble alias alone would build a device-test APK the weekly suite never runs — the same hole one workflow later. Proven both directions at the convention level: on the parent commit both repo-wide graphs contain 0 *AndroidDeviceTest tasks; with this commit, assembleDebugAndroidTest pulls every KMP module's device-test build and connectedDebugAndroidTest pulls connectedAndroidDeviceTest on all three. An empty device-test APK runs green on-device under the Regression annotation filter (the exact ui_tests.yml invocation), so componentizing modules without instrumented sources costs the weekly suite nothing. RoomLibraryConventionPlugin keeps one plugin id and branches: the KMP side declares the room bundle + paging-common in commonMain (paging-runtime-ktx is Android-only; PagingSource lives in paging-common), sqlite-framework in androidMain (the driver artifact is per-target by construction), room-compiler on kspAndroid + kspIosSimulatorArm64, room-testing on androidDeviceTest. Dead until the first KMP Room module lands in the next commit. The stale KmpLibraryConventionPlugin KDoc sentence pointing android.* needs at the phase-3-deleted core:core-android sibling is corrected to the androidMain + direct-Metro reality core:data:dataStore already proves. KMP modules re-verified green with the new component: core, dataStore, backup:api assemble + host tests, 140/140 executed. detekt separate, 56/56. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
…ing a byte of schema core:data:database converts to convention.kmpLibrary + the Room convention's KMP branch: 51 files to commonMain, 4 to androidMain (buildAppDatabase and the snapshot/ package — Context, framework SQLite, java.io.File), 26 test classes to androidHostTest, 5 to src/androidDeviceTest/kotlin. AppDatabase gains @ConstructedBy(AppDatabaseConstructor) — a bare expect object; Room KSP generates the actual per target, and NoActualForExpectSuppressionRule is why it carries no NO_ACTUAL_FOR_EXPECT suppression: the metadata compilation never demands an actual, and if codegen stops, the platform compile reds instead of green-over-nothing. The schema did not move, proven the way §6 prescribes: schemaInput redirected to an empty directory out-of-tree, the processor exported a fresh 6.json, byte-identical, identityHash 9a848458f02ef4c4ace9841b9f2e3d42. The mechanism that made the probe necessary is itself the finding — Room 3 exports write-on-diff against schemaInput (= the committed schemas/), so a production build cannot silently rewrite the directory. What the first instrumented conversion surfaced, each fixed at the root: - Room-KMP puts no schemas on the device-test APK; all 7 MigrationTestHelper tests failed loud (Missing file: …/5.json). The Room convention's KMP branch now wires schemas/ as a static device-test asset via the variant API — which requires androidResources.enable = true first, because with it off the component's sources.assets is null (measured): the asset pipeline does not exist. - The first KMP device APK depending on an Android-convention AAR fails checkAndroidDeviceTestAarMetadata over desugaring; the KMP convention now mirrors configureKotlinAndroid's repo-wide core-library desugaring. - Lint does not classify KSP output as generated on KMP: a consumer with checkDependencies analyzed the generated Room code and red 1087 RestrictedApi + 6 SyntheticAccessor — all under build/generated, none in handwritten code. lint.xml gains per-issue **/build/generated/** ignores; mutation M-LINT1 (break the path) brings back exactly 1093, reverted. - src/androidTest was never full-detekt-checked on classic modules (default sources are src/main + src/test); the KMP source-set model put the device tests under the full ruleset for the first time and surfaced two pre-existing Wrapping violations in AtomicRollbackDeviceTest — fixed. - detekt.yml: every list carrying the androidHostTest glob gains **/androidDeviceTest/**, and the header note now points at the convention (configureDetektSources) instead of the pre-05b75a4c per-module wiring. Gates, all --rerun-tasks --no-build-cache --no-configuration-cache: full forced gate 3248/3248 tasks 0 failed; detekt separate 56/56; unit tests 2230/230 (testDebugUnitTest scope) + 165/33 (testAndroidHostTest scope) = 2395, accounting exactly as 2354 baseline + 4 new epochToIso pins + 37 KMP-module tests the old testDebugUnitTest-glob methodology never counted (core:core 32, dataStore 5 — execution was never affected, only counting; report both scopes from here on). Device: Starting 28 tests on Pixel_6_API_34, 28/28 green via connectedDebugAndroidTest alias under the exact ui_tests.yml Regression annotation filter. Suite gate prints ":core:data:database: 5 instrumented source files, 206 classpath entries scanned, 0 missing" over androidDeviceTestRuntimeClasspath. iOS: klib compiled + kspKotlinIosSimulatorArm64 in the same repo-wide graph. Goldens 456, 0 touched. schemas/ untouched by git. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
… of its own claims New §9 carries the conversion's findings for phase 7's ~30 modules: the connectedDebugAndroidTest fifth vanish, the Room-KMP device-asset gap behind androidResources.enable, write-on-diff schema export, KSP2 config spellings, the forbidden-and-unneeded NO_ACTUAL_FOR_EXPECT suppression, the desugaring AAR-metadata check, lint's blindness to generated code on KMP, and detekt seeing device tests under the full ruleset for the first time. §3's setQueryCoroutineContext and 5-file-androidMain claims are dated in place, not deleted. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
Contributor
Unit Test Results2 066 tests 2 066 ✅ 9m 55s ⏱️ Results for commit c857cb6. ♻️ This comment has been updated with latest results. |
This was referenced Aug 20, 2026
stslex
marked this pull request as ready for review
August 20, 2026 20:07
|
You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard. |
Owner
Author
|
@claude review |
Owner
Author
|
@codex review |
|
Claude finished @stslex's task in 2m 17s —— View job ReviewWent through the full diff (
Two minor, non-blocking notes:
No correctness issues found in the conversion itself. The gate table in the PR description (2395 unit tests, 28/28 device suite, byte-identical schema/identityHash, both alias directions verified) is consistent with what the diff implements. Nice work isolating this from the |
|
Codex Review: Didn't find any major issues. Can't wait for the next one! Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
Owner
Author
|
@claude review |
Owner
Author
|
@codex review |
|
Claude finished @stslex's task in 1m 30s —— View job Review
Went through Confirmed correct / no new issues found:
Still open from the prior automated review, both non-blocking:
No correctness issues found in the KMP conversion itself. The port is well-isolated from the |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: fea9edd127
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
Owner
Author
|
Review classification before the next push:
|
Owner
Author
|
Additional classification before the same push:
|
Owner
Author
|
@claude review |
Owner
Author
|
@codex review |
|
Claude finished @stslex's task in 53s —— View job Review
Re-reviewed at HEAD ( Both P1 fixes verified correct:
All three anchors resolve to real, matching section headers in Other changes in this commit ( Housekeeping note: the working tree in this review checkout ( No correctness issues found. Both outstanding P1 comments from Codex are resolved as claimed, and the fix commit is scoped exactly to what those comments asked for — no unrelated changes riding along. |
|
Codex Review: Didn't find any major issues. You're on a roll. Reviewed commit: ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
If Codex has suggestions, it will comment; otherwise it will react with 👍. Codex can also answer questions or update the PR. Try commenting "@codex address that feedback". |
stslex
added a commit
that referenced
this pull request
Aug 21, 2026
) **KMP phase 6, the driver flip: Android moves to `BundledSQLiteDriver`. Stacked on #239** — review that first; this is one commit on top, deliberately separate so a SQLite-behaviour regression and a source-set regression stay distinguishable under bisect, and so this half reverts alone. #241 (core:data:exercise → KMP) stacks on top of this PR. ## What flips, what doesn't, and why the line is where it is | Surface | Driver after this PR | Why | |---|---|---| | Production (`buildAppDatabase`) | **Bundled** (SQLite 3.50.x) | The point: one SQLite build on every device instead of the per-OEM system one | | The 5 device-test classes (migration, rollback, invalidation, round-trip, recently-trained) | **Bundled** | The migration suite is the flip's gate, so it must run the flipped driver | | `InMemoryDatabaseProvider` (app:app androidTest via `MetroTestRule`) | **Bundled** | Device-side; the whole instrumented app suite now exercises the production driver | | `RepositoryTestEnv` + host tests (`BaseDatabaseTest`, snapshot test) | **Framework, pinned with a comment** | **Measured, not assumed**: the bundled android variant ships Android-ABI natives only — under Robolectric on a desktop JVM it dies with `UnsatisfiedLinkError`. Robolectric is not an admissible driver oracle in this repo anyway (`AtomicRollbackDeviceTest` KDoc: two false negatives) | The Room convention's KMP branch now declares `sqlite-bundled` for androidMain; the database module pins `sqlite-framework` on its host tests explicitly, next to the reason. ## The gate — the migration suite on a real device, under the flipped driver - **28/28 device tests green under `BundledSQLiteDriver`** on Pixel_6_API_34 (arm64), collected count printed (`Starting 28 tests`), via the `connectedDebugAndroidTest` alias under the exact ui_tests.yml Regression filter. That includes all 7 `AppDatabaseMigrationTest` cases — a real v5 file seeded on disk, `Migration6` applied, schema validated against `6.json` — plus the atomicity-rollback oracle (both production transaction shapes), invalidation across connections, and the on-disk round-trip. - **The same 28 ran green under `AndroidSQLiteDriver` in #239** — the spec's "under both drivers" requirement, satisfied across the two PRs by construction. - **41/41 app:app device Regression tests green** with the flipped in-memory provider (41 = the 35 pinned in tech-debt plus #232's six DataStore singleton tests). ## Risk register (from the pre-flip SQL audit of all 105 `@Query` sites + every raw-SQL site) - **Zero double-quoted string literals** repo-wide (the classic `SQLITE_DQS=0` breakage): every SQL literal is single-quoted. Zero version-sensitive constructs in use — no window functions, CTEs, upserts (`@Insert(REPLACE/IGNORE)` compile to the pre-upsert forms, frozen since 3.6.19), date/time/json/math functions, `RETURNING`, or `IIF`. - **One query with version-floating output, pre-existing**: `SessionDao.getRecentSessionsForExercise` projects a bare column under `GROUP BY` with no aggregate — legal and arbitrary in *both* SQLite builds; a planner change may pick a different in-group row when one session holds the same exercise twice. Not a flip regression; noted for the record. - **Cross-library file interop is deliberate and stays**: the `snapshot/` package opens the live DB via framework SQLite (peek + pre-migration checkpoint) while Room holds bundled connections. Main-db and WAL formats are frozen, so reads are safe by format; the concurrent-open path leans on POSIX locking shared by both libraries. This is the residual risk the device-restore pass (Ilya's gate, per the recovery spec) covers end-to-end; flagged here rather than silently carried. - Improvements that come free: host-variable limit 999 → 32766, and `ROW_NUMBER()` becomes available — `SessionDao`'s PR-batch comment now records the rewrite as *available but not taken* (this PR is behaviour-neutral by design). ## Gates | Gate | Result | |---|---| | device suite under bundled | ✅ 28/28, `Starting 28 tests on Pixel_6_API_34(AVD)` | | app:app device Regression | ✅ 41/41 | | host tests (framework pin) | ✅ database 128/26, exercise 212/19, 0 failures | | full forced gate (assembleDebug testDebugUnitTest verifyPaparazziDebug lintDebug assembleDebugAndroidTest) | ✅ `3248 actionable tasks: 3248 executed`, 0 failed | | detekt (separate invocation) | ✅ `56 actionable tasks: 56 executed` | | unit tests | ✅ 2230/230 + 165/33 = 2395, 0 failures — identical to #239, as a behaviour-neutral flip must be | | goldens | ✅ 456, 0 touched | 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
stslex
added a commit
that referenced
this pull request
Aug 21, 2026
…roidMain (#241) **KMP phase 6, increment E: `core:data:exercise` becomes a KMP module — with zero androidMain. Stacked on #240** → #239. Review in that order. The last code increment of phase 6. Four commits: the one genuine behaviour change (red-first, on device), the wall-clock swap, the pure-mechanics conversion, and the hook-deployment fix the conversion's own near-miss surfaced. ## 1. The behaviour change the spec predicted, observed live before it was fixed Spec §5 called `catch (SQLiteConstraintException)` in `ExerciseRepositoryImpl.saveItem` *control flow, not a type name*: under a Room 3 driver the constraint violation arrives as a different type, the catch silently stops matching, and the exception escapes — with no compile error. That is exactly what happened. A new on-device duplicate-name test (this module's first device test, `@Regression`, in-memory DB under the flipped `BundledSQLiteDriver`) ran **red against the unfixed repository**, escaping with: ``` android.database.SQLException: Error code: 2067, message: UNIQUE constraint failed: exercise_table.name ``` The fix is smaller than the spec's own sketch, because of a fact worth keeping: **on Android, `androidx.sqlite.SQLiteException` is `actual typealias … = android.database.SQLException`** (read from the 2.7.0 sources jar). So one common catch covers the framework driver's `SQLiteConstraintException` (a subtype), the bundled driver's own throw (measured above), and iOS's native actual — no expect/actual seam needed at all. The constraint check gates on the one string both engines emit (`UNIQUE constraint failed`), because the common type has no code field and the framework subtype is not nameable from common code. Proven in all four directions: device red on the old catch → green on the new; the Robolectric twin (framework driver) green throughout; mutation **M-DUP1** (never-matching string) reds **both** legs — 1 of 54 host tests, 1 of 1 device — reverted. ## 2. Mechanics - **Eight `System.currentTimeMillis()` sites** → `kotlin.time.Clock.System.now().toEpochMilliseconds()` (stdlib, same wall clock; its own commit so the conversion stays pure moves). - **The conversion**: all 42 production files → commonMain. **No androidMain** — after the catch fix, the module had no platform type left; the paging surface lives in `paging-common` (verified `iosSimulatorArm64` published). 19 test classes + seed helper → `androidHostTest`, the device test → `src/androidDeviceTest/kotlin`. The dead `compose-bom`/`compose-runtime`/`room3-runtime` declarations (zero usages, spec §5) do not return; `paging-runtime-ktx` → `paging-common`; the new `androidx-sqlite` (base) alias supplies the exception type the database module's `implementation` deps rightly don't leak. - The whole PR-D convention machinery ran unchanged for the second module: device-test component, both CI aliases, schema-asset wiring (n/a here), desugaring, detekt device-test globs. **Nothing new was needed** — the first conversion paid; the second rode. ## 3. A process finding, in the open `setup-hooks.sh` **copies** `.githooks/pre-commit` into `.git/hooks/` — it does not set `core.hooksPath`, though CLAUDE.md says it does. Consequences, both real: a clone that never ran the script has **no hook at all** (this session's clone — caught when a red detekt run and a passing commit coexisted; the separate forced detekt invocations in this arc's process were the actual gate all along), and a clone that ran it **before #235 still carries the ACM filter** the fixture-move PR proved blind to renames. Fixed here: the script now sets `core.hooksPath` to the tracked `.githooks/`, so hook fixes deploy with `git pull` instead of with a ritual — which is what CLAUDE.md already claimed. Reported rather than silently corrected. ## 4. The §7 riders, re-priced with numbers (closing the phase brief's ask) - **`java.time`**: production `core/data` is now clean — `WorkoutExportMapper`'s line was taken in #239. What remains: 1 test file (`DatabaseJsonExporterImplTest`, deliberately kept as a cross-library oracle) and 14 `feature:exercise-chart` files, which are Phase 7's by the spec's own split. **This rider is discharged for phase 6.** - **`java.io.File` → `okio.Path`**: the dependency half is now free (okio rides the graph via `datastore-preferences-core` since #233; using it in signatures needs only a toml alias). The remaining cost is unchanged in kind: a signature ripple across 7 modules (backup:api, google-drive, worker, database, recovery, settings, app:app). And its urgency **dropped** with this phase: after #239 every `File`-typed surface sits in an androidMain of a KMP module — nothing common is blocked by it. Still correctly deferred until a real iOS backup implementation makes the right shape observable; `android.net.Uri` still has no common analogue (SAF). ## Gates All `--rerun-tasks --no-build-cache --no-configuration-cache`. | Gate | Result | |---|---| | device: duplicate-name under bundled | ✅ red on old catch (escape above) → 1/1 green via the `connectedDebugAndroidTest` alias under the ui_tests.yml Regression filter | | host tests | ✅ exercise 212/19 (now `testAndroidHostTest` scope), feature:exercise 113/9, exercise-chart 66/8, 0 failures | | full forced gate (assembleDebug testDebugUnitTest verifyPaparazziDebug lintDebug assembleDebugAndroidTest) | ✅ `3265 actionable tasks: 3265 executed`, 0 failed | | detekt (separate invocation) | ✅ `57 actionable tasks: 57 executed` — the 57th is the suite gate over this module's new device test, previously NO-SOURCE | | repo-wide unit tests | ✅ **2018 / 211** (`testDebugUnitTest` scope) + **377 / 52** (`testAndroidHostTest` scope) = **2395**, 0 failures, 0 skipped — identical total to #239/#240; exercise's 212 moved scope, not existence | | iOS | ✅ exercise klib + `kspKotlinIosSimulatorArm64` in the repo-wide `assembleDebug` graph | | goldens | ✅ 456, 0 touched | 🤖 Generated with [Claude Code](https://claude.com/claude-code) https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
KMP phase 6, increment D:
core:data:databasebecomes a KMP module, onAndroidSQLiteDriver.The centrepiece of phase 6's second half (#232 carried the spec; #235 was this PR's step 1). The port only — the
BundledSQLiteDriverflip is deliberately NOT here; it follows as its own PR with its own gate, so a SQLite regression and a source-set regression stay distinguishable under bisect. The schema did not move: the regenerated6.jsonis byte-identical,identityHash 9a848458f02ef4c4ace9841b9f2e3d42.The shape — five commits, bisect-green each
epochToIsoleavesjava.time—kotlin.time.Instant.fromEpochMilliseconds(ms).toString(), the zero-new-dependency stdlib replacement, measured byte-identical againstjava.timefor whole-second / zero / fractional / negative epochs. Pinned twice: new literal-string tests, andDatabaseJsonExporterImplTestkeepsjava.timeon the test side deliberately as a cross-library oracle. Mutation M-ISO1 (fromEpochSeconds) reds 4 tests — the epoch-zero pin alone survives, since 0 s and 0 ms name the same instant; reverted. This makesWorkoutExportMappercommonMain-eligible: androidMain ends at 4 files, not the spec's 5.Room.databaseBuilder<AppDatabase>(context, name)etc., while the module is still Android-only (the reified overload falls back to reflection until@ConstructedByexists — so this commit is pure mechanics, bisectable on its own). The 16th site was already the Room 3 KClass form; the spec's total said 15, measured 16. Phase-7 note: withClass<T>gone, a bareApplicationProvider.getApplicationContext()argument deadlocks inference against the builder's factory overloads — pin asgetApplicationContext<Application>().withDeviceTestunconditional in the KMP convention (classic-AGP parity: every library carries the component whether or not sources exist) plusassembleDebugAndroidTest → assembleAndroidDeviceTestandconnectedDebugAndroidTest → connectedAndroidDeviceTest. The second alias is a finding, not a plan item:ui_tests.ymlinvokesconnectedDebugAndroidTestliterally, so the pre-diagnosed assemble alias alone would build a device-test APK the weekly suite never runs — the fifth silent-vanish instance of this arc. Proven both directions: on the parent commit both repo-wide graphs contain 0*AndroidDeviceTesttasks; with the commit, both are fully wired for every KMP module, and an empty KMP device-test APK runs green on-device under the Regression annotation filter (the exact ui_tests.yml invocation), so the three source-less KMP modules cost the weekly suite nothing.RoomLibraryConventionPluginkeeps one plugin id and branches (KMP: room bundle +paging-commonin commonMain,sqlite-frameworkin androidMain,kspAndroid/kspIosSimulatorArm64, room-testing on the device suite).buildAppDatabase,snapshot/×3: Context, framework SQLite,java.io.File), 26 test classes →androidHostTest, 5 →src/androidDeviceTest/kotlin;@ConstructedBy+ a bareexpect object AppDatabaseConstructor(see below); detekt.yml gains**/androidDeviceTest/**beside every host-test glob, in this commit because that is when the sources exist (spec §3.4).kmp-phase-6-data-layer.md; two of the spec's own §3 claims dated in place.What the first instrumented conversion surfaced — each measured, each fixed at the root
MigrationTestHelpertests failed on device, loudly:FileNotFoundException … Missing file: …/5.json. The classic integration's androidTest-assets copy has no KMP counterpart, so the Room convention's KMP branch wiresschemas/as a static device-test asset via the variant API — which is null untilandroidResources.enable = true: AGP-KMP defaults the entire asset pipeline off (measured:sources.assets == nullbefore, six schemas inside the APK after).checkAndroidDeviceTestAarMetadatafails on the first KMP device APK that depends on an Android-convention AAR ("requires core library desugaring"). The KMP convention now mirrorsconfigureKotlinAndroid's repo-wide desugaring (flag +desugar_jdk_libs).checkGeneratedSources = falseis already set and honored on classic modules; here a consumer withcheckDependencies = truered 1087 RestrictedApi + 6 SyntheticAccessor — 100% underbuild/generated, zero in handwritten code. Encoded inlint.xmlas per-issue**/build/generated/**ignores. Mutation M-LINT1 (break the path) brings back exactly 1093; reverted.NoActualForExpectSuppressionRulerejected theNO_ACTUAL_FOR_EXPECTsuppression I first wrote on the expect object — and it is right: the metadata compilation never demands an actual, the platform compilations get Room's generated ones, and without the suppression a codegen failure reds the build instead of vanishing. And detekt, now seeing device tests through the KMP source-set model, surfaced two pre-existingWrappingviolations inAtomicRollbackDeviceTest—src/androidTestwas never full-detekt-checked on classic modules (default sources aresrc/main+src/test), the same shape as KMP phase 6D step 1: the DB fixtures leave testFixtures, and the pre-commit hook stops ignoring moves #235's testFixtures hole.schemaOutputonly when the generated schema differs fromschemaInput(= the committedschemas/), so a production build can never silently rewrite the schema directory. The §6 gate ran as prescribed: schemaInput redirected to an empty dir via an out-of-tree init script, fresh6.jsonexported, byte-identical.setQueryCoroutineContextdeliberately NOT added (deviation from spec §3, argued): the Android builder runs without it today, adding it would change which context Room queries use — not "the port only" — and nothing constructs a database off-Android until phase 7's iOS composition root, which is where it belongs, next to iOS's driver choice.String.toByteArray(Charsets.UTF_8)is JVM-only →encodeToByteArray()(one site).A counting correction, in the open
The repo-wide unit-test invariant number (2352 at #235) was produced by a
test-results/testDebugUnitTestglob — which never saw KMP modules'testAndroidHostTestresults (core:core 32, dataStore 5). Execution was never affected (CI's alias runs them); only the counting was blind, on both sides of every comparison, so past deltas stayed honest. From this PR on the number is reported in both scopes.Gates
All
--rerun-tasks --no-build-cache --no-configuration-cache; baseline measured on dev@c81c4490 the same way.3248 actionable tasks: 3248 executed, 0 failed (dev baseline: 3080/3080; the +168 is the KMP module's task set + device-test components)56 actionable tasks: 56 executedtestDebugUnitTestscope) + 165 / 33 (testAndroidHostTestscope) = 2395, 0 failures, 0 skipped, 0 stale XMLs. Accounting vs 2354 baseline: +4 new epochToIso pins, +37 previously uncounted KMP host tests, −0 lost; database's 128 moved scope, not existenceStarting 28 tests on Pixel_6_API_34(AVD), 28/28 green — via theconnectedDebugAndroidTestalias under the exact ui_tests.yml Regression annotation filter; red direction observed live (7 migration failures while schemas were missing from assets):core:data:database: 5 instrumented source files, 206 classpath entries scanned, 0 missing— resolved throughandroidDeviceTestRuntimeClasspath6.jsonbyte-identical to committed;git status schemas/emptycompileKotlinIosSimulatorArm64+kspKotlinIosSimulatorArm64+ klib in the same repo-wide graph (theassembleDebugalias pulls them)*AndroidDeviceTesttasks in either repo-wide graph; this branch: fully wiredStacked work: #240 (the
BundledSQLiteDriverflip) sits on top of this branch; PR E (core:data:exercise→ KMP) will stack on #240. Waiting on review here does not block that work.🤖 Generated with Claude Code
https://claude.ai/code/session_011kNU4XxzXp6NpCnVUQA6yH