Android: sleep stage-timeline rows (iOS #988 port) + HC sleep import fixes by TheBoroer · Pull Request #240 · ryanbr/noop

TheBoroer · 2026-07-10T20:32:55Z

What this PR does

Ports the iOS #988 sleep stage-timeline rows to Android: replaces the hero card's StageLegend with per-stage StageTimeline rows (span extraction, displaySmoothed port, Metrics constants, interval reconstruction from hero weight pairs).

Also fixes three Health Connect / sleep-hero bugs found while validating against strap data:

HC importer coveredDays gate — sleep import now skips days covered by the active strap ids, so a flat HC night can no longer shadow a fully-staged strap night.
In-bed figure — hero in-bed now spans the full night group (min onset → max wake across fragments) instead of the main fragment only, matching iOS SleepView.mergeDay.
Cleanup migration — purges previously-written HC my-whoop sleep rows that shadow strap-computed nights.

Type of change

How it was tested

./gradlew testFullDebugUnitTest — full com.noop.ingest.* and com.noop.ui.* packages pass, including the new HealthConnectCoveredDaysTest (gate + purge behavior).
Room compile-time SQL verification covers the two new purge queries in WhoopDao.
No BLE path touched — importer, DAO/repository, and Compose UI only.
Timeline rows verified on-device against strap-staged nights and HC-only nights (fallback/empty states).

Checklist

Swift package tests pass for any package I touched (none touched — Android only)
Android unit tests pass if I touched android/ (./gradlew testFullDebugUnitTest)
No new build warnings introduced
UI changes use only StrandDesign tokens — no hardcoded colors, fonts, or spacing
No hardcoded hex frame bytes; protocol facts live in the schema / decoders
Follows the conventions in docs/CONTRIBUTING.md
I did not commit generated output (Strand.xcodeproj/) or any secrets/keystores

Related issues

Android port of iOS #988 (stage timeline rows).

…ows (ryanbr#988 port)

…nbr#988 port)

… rows

…port)

…rt order (iOS twin of #240) (#241) * fix: prefer richer stagesJSON over recency in sleep-session merge A re-imported session whose newer twin lacked stage data used to clobber the older stage-bearing row. Merge now keeps the candidate with non-empty stagesJSON regardless of recency; ties still resolve by recency. Covered by richness/tie/regression cases in SleepMergeTests. * fix: canonical import outranks computed siblings in sourceCandidates After a device re-add (active strap id != canonical "my-whoop"), the resolver tried the active strap's computed sibling before the canonical import, so fresh computed estimates shadowed richer imported history. Candidate order is now: active strap > canonical import > computed siblings > Apple fallback. Swift twin of the #240 precedence fix; covered by SourceCandidatesOrderTests (single-device path stays byte-identical). * fix: point StrandTests TEST_HOST at renamed "NOOP Staging" product The staging rename set the app's PRODUCT_NAME to "NOOP Staging" but TEST_HOST still referenced NOOP.app, so local `xcodebuild test` could not find the test host. Track the renamed product path (regenerate with xcodegen after pulling).

…computed stages (parity with #241) (#247) * Android: mergeSleep richness exception — stage-less import yields to computed stages (parity with #241) Re-review of the #241/#246 merge surfaced a missed twin. Android's WhoopRepository.mergeSleep is the direct twin of Swift WhoopStore.SleepMerge.merge (its own comment says "Mirrors WhoopStore.SleepMerge"), but it still had the plain imported-wins-per-day logic — #241 added the richness exception to the Swift side only, so the shared-logic twins diverged. Port the exception: a sparse import (no stage data on ANY of its sessions that day) no longer clobbers a computed day that HAS stage data; days where the import carries stages, or where neither side does, keep the imported-wins rule unchanged. Adds hasStages() (null/""/"[]"/blank = no stages), byte-parity with Swift's hasStages. Not covered by #240: that's a Health-Connect-specific write-gate + heal-purge; it never touches mergeSleep and doesn't protect non-HC sparse imports (Apple/CSV). This is the mergeSleep-level twin. Mirrors Swift SleepMergeTests' 5 richness cases in MergeSleepLocalDayTest; compileFullDebugKotlin + MergeSleepLocalDayTest pass (8/0). * Android: route the Sleep screen through the richness merge too (re-review) Re-review of the first commit found it incomplete: mergeSleep feeds only CSV export + HealthConnectWriter. The Sleep SCREEN builds its night list with its OWN inline imported-wins merge (SleepScreen.kt, two sites: the main loader + the post-nap-edit reload), which had no richness exception — so a stage-less import would still blank stages on the Android sleep screen, while iOS gets richness via repo.sleeps (WhoopStore.SleepMerge). That's the exact user-visible surface the fix is meant to protect. Extract mergeSleepRichness (the imported-wins + #241 richness rule, returned UNSORTED) and route both SleepScreen sites through it, keeping their effectiveStartTs sort (#395) and the #814 union inputs unchanged. mergeSleep is now that helper keyed by wake-day + sorted by startTs, so all three consumers (browse, CSV, screen) share ONE richness implementation. compileFullDebugKotlin + MergeSleepLocalDayTest pass (9/0, +1 mergeSleepRichness case).

…rt order (iOS twin of #240) (#241) * fix: prefer richer stagesJSON over recency in sleep-session merge A re-imported session whose newer twin lacked stage data used to clobber the older stage-bearing row. Merge now keeps the candidate with non-empty stagesJSON regardless of recency; ties still resolve by recency. Covered by richness/tie/regression cases in SleepMergeTests. * fix: canonical import outranks computed siblings in sourceCandidates After a device re-add (active strap id != canonical "my-whoop"), the resolver tried the active strap's computed sibling before the canonical import, so fresh computed estimates shadowed richer imported history. Candidate order is now: active strap > canonical import > computed siblings > Apple fallback. Swift twin of the ryanbr/noop#240 precedence fix; covered by SourceCandidatesOrderTests (single-device path stays byte-identical). * fix: point StrandTests TEST_HOST at renamed "NOOP Staging" product The staging rename set the app's PRODUCT_NAME to "NOOP Staging" but TEST_HOST still referenced NOOP.app, so local `xcodebuild test` could not find the test host. Track the renamed product path (regenerate with xcodegen after pulling).

… not return raw fragments (#240 follow-up) (#316) The Android port of Swift Hypnogram.displaySmoothed added an early return `if (intervals.size <= 2 || minDurationSec <= 0.0) return intervals`. The `|| minDurationSec <= 0.0` short-circuit isn't in Swift, which guards ONLY on count and lets the "raw" (minDuration 0) case fall through to coalesce() — so Swift returns the coalesced timeline, while Android returned the 60-100 un-merged epoch fragments (the "comb" the smoothing exists to remove). Everything else in the port is faithful. Production is unaffected (the only caller passes STAGE_ROW_SMOOTH_SEC = 90.0), so this is a latent port-fidelity fix, not a shipping bug — but it makes the "straight port" claim true and protects a future 0-caller. Also strengthens `zeroFloorReturnsCoalescedInputUnsmoothed`: its old [light, deep, rem] fixture had no adjacent same-stage runs, so coalesce was a no-op and it passed even with the short-circuit. New fixture has a light/light seam and asserts the coalesced result — verified it FAILS against the old short-circuit and passes with the fix. Android-only (Health Connect / SleepScreen); testFullDebugUnitTest green.

…rt order (iOS twin of ryanbr#240) (ryanbr#241) * fix: prefer richer stagesJSON over recency in sleep-session merge A re-imported session whose newer twin lacked stage data used to clobber the older stage-bearing row. Merge now keeps the candidate with non-empty stagesJSON regardless of recency; ties still resolve by recency. Covered by richness/tie/regression cases in SleepMergeTests. * fix: canonical import outranks computed siblings in sourceCandidates After a device re-add (active strap id != canonical "my-whoop"), the resolver tried the active strap's computed sibling before the canonical import, so fresh computed estimates shadowed richer imported history. Candidate order is now: active strap > canonical import > computed siblings > Apple fallback. Swift twin of the ryanbr#240 precedence fix; covered by SourceCandidatesOrderTests (single-device path stays byte-identical). * fix: point StrandTests TEST_HOST at renamed "NOOP Staging" product The staging rename set the app's PRODUCT_NAME to "NOOP Staging" but TEST_HOST still referenced NOOP.app, so local `xcodebuild test` could not find the test host. Track the renamed product path (regenerate with xcodegen after pulling).

…computed stages (parity with ryanbr#241) (ryanbr#247) * Android: mergeSleep richness exception — stage-less import yields to computed stages (parity with ryanbr#241) Re-review of the ryanbr#241/ryanbr#246 merge surfaced a missed twin. Android's WhoopRepository.mergeSleep is the direct twin of Swift WhoopStore.SleepMerge.merge (its own comment says "Mirrors WhoopStore.SleepMerge"), but it still had the plain imported-wins-per-day logic — ryanbr#241 added the richness exception to the Swift side only, so the shared-logic twins diverged. Port the exception: a sparse import (no stage data on ANY of its sessions that day) no longer clobbers a computed day that HAS stage data; days where the import carries stages, or where neither side does, keep the imported-wins rule unchanged. Adds hasStages() (null/""/"[]"/blank = no stages), byte-parity with Swift's hasStages. Not covered by ryanbr#240: that's a Health-Connect-specific write-gate + heal-purge; it never touches mergeSleep and doesn't protect non-HC sparse imports (Apple/CSV). This is the mergeSleep-level twin. Mirrors Swift SleepMergeTests' 5 richness cases in MergeSleepLocalDayTest; compileFullDebugKotlin + MergeSleepLocalDayTest pass (8/0). * Android: route the Sleep screen through the richness merge too (re-review) Re-review of the first commit found it incomplete: mergeSleep feeds only CSV export + HealthConnectWriter. The Sleep SCREEN builds its night list with its OWN inline imported-wins merge (SleepScreen.kt, two sites: the main loader + the post-nap-edit reload), which had no richness exception — so a stage-less import would still blank stages on the Android sleep screen, while iOS gets richness via repo.sleeps (WhoopStore.SleepMerge). That's the exact user-visible surface the fix is meant to protect. Extract mergeSleepRichness (the imported-wins + ryanbr#241 richness rule, returned UNSORTED) and route both SleepScreen sites through it, keeping their effectiveStartTs sort (ryanbr#395) and the ryanbr#814 union inputs unchanged. mergeSleep is now that helper keyed by wake-day + sorted by startTs, so all three consumers (browse, CSV, screen) share ONE richness implementation. compileFullDebugKotlin + MergeSleepLocalDayTest pass (9/0, +1 mergeSleepRichness case).

…mport fixes (ryanbr#240) * docs: design + implementation plan for Android sleep stage-timeline rows (ryanbr#988 port) * feat: sleep stage-interval reconstruction from hero weight pairs (ryanbr#988 port) * docs: plan uses full-flavor gradle task names (app has product flavors) * feat: port Hypnogram.displaySmoothed to Kotlin for the sleep timeline rows * feat: per-stage row span extraction for the sleep timeline rows * feat: Metrics constants for the sleep stage-timeline rows * feat: StageTimeline row composables for the sleep hero (iOS ryanbr#988 port) * feat: wire StageTimeline into the sleep hero card, drop StageLegend * fix: gate HC sleep import on days covered by the active strap * fix: in-bed figure spans the full night group, not the main fragment * fix: purge HC my-whoop sleep rows shadowing strap-computed nights

… not return raw fragments (ryanbr#240 follow-up) (ryanbr#316) The Android port of Swift Hypnogram.displaySmoothed added an early return `if (intervals.size <= 2 || minDurationSec <= 0.0) return intervals`. The `|| minDurationSec <= 0.0` short-circuit isn't in Swift, which guards ONLY on count and lets the "raw" (minDuration 0) case fall through to coalesce() — so Swift returns the coalesced timeline, while Android returned the 60-100 un-merged epoch fragments (the "comb" the smoothing exists to remove). Everything else in the port is faithful. Production is unaffected (the only caller passes STAGE_ROW_SMOOTH_SEC = 90.0), so this is a latent port-fidelity fix, not a shipping bug — but it makes the "straight port" claim true and protects a future 0-caller. Also strengthens `zeroFloorReturnsCoalescedInputUnsmoothed`: its old [light, deep, rem] fixture had no adjacent same-stage runs, so coalesce was a no-op and it passed even with the short-circuit. New fixture has a light/light seam and asserts the coalesced result — verified it FAILS against the old short-circuit and passes with the fix. Android-only (Health Connect / SleepScreen); testFullDebugUnitTest green.

TheBoroer added 11 commits July 10, 2026 15:04

docs: design + implementation plan for Android sleep stage-timeline r…

6f6e60b

…ows (ryanbr#988 port)

feat: sleep stage-interval reconstruction from hero weight pairs (rya…

55d9dbd

…nbr#988 port)

docs: plan uses full-flavor gradle task names (app has product flavors)

f098874

feat: port Hypnogram.displaySmoothed to Kotlin for the sleep timeline…

4e003e1

… rows

feat: per-stage row span extraction for the sleep timeline rows

ff44719

feat: Metrics constants for the sleep stage-timeline rows

fd743d3

feat: StageTimeline row composables for the sleep hero (iOS ryanbr#988 …

83c0928

…port)

feat: wire StageTimeline into the sleep hero card, drop StageLegend

19bb236

fix: gate HC sleep import on days covered by the active strap

bceae02

fix: in-bed figure spans the full night group, not the main fragment

07bc1d9

fix: purge HC my-whoop sleep rows shadowing strap-computed nights

f25e90c

TheBoroer marked this pull request as ready for review July 10, 2026 20:49

TheBoroer mentioned this pull request Jul 10, 2026

Fix sleep-merge richness precedence + sourceCandidates canonical-import order (iOS twin of #240) #241

Merged

12 tasks

ryanbr mentioned this pull request Jul 11, 2026

Android: canonical import outranks computed sibling in sourceCandidates (parity with #241) #246

Merged

ryanbr mentioned this pull request Jul 11, 2026

Android: mergeSleep richness exception — stage-less import yields to computed stages (parity with #241) #247

Merged

ryanbr added enhancement New feature or request Major Release (TODO) Major Release Version labels Jul 11, 2026

ryanbr merged commit 5e8f0e8 into ryanbr:main Jul 12, 2026

ryanbr mentioned this pull request Jul 12, 2026

fix(sleep): displaySmoothed(minDuration<=0) coalesces like Swift instead of returning raw fragments (#240 follow-up) #316

Merged

This was referenced Jul 12, 2026

Oura (experimental): sanctioned path to display a ring-provided sleep stage timeline — without HR/RR ? #323

Closed

feat(oura): reconstruct BLE sleep-stage hypnogram #325

Closed

Navigation Menu

Android: sleep stage-timeline rows (iOS #988 port) + HC sleep import fixes - #240

Android: sleep stage-timeline rows (iOS #988 port) + HC sleep import fixes#240
ryanbr merged 11 commits into
ryanbr:mainfrom
TheBoroer:sleep-stage-timeline

TheBoroer commented Jul 10, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

TheBoroer commented Jul 10, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What this PR does

Type of change

How it was tested

Checklist

Related issues

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

TheBoroer commented Jul 10, 2026 •

edited

Loading