Reconcile package selection when a window size rule hides the selected package

[joshdholtz]

Stacked on #4193. Closes the selection gap flagged in review: a package hidden by a window size rule could stay selected and purchasable.

What

• reconcileSelectionForWindowSize on PaywallState.Loaded.Components: when the measured paywall bounds change and the selected package no longer resolves visible, selection moves to the authored default if it renders, else the first visible package. A visible selection is never moved; if nothing is visible, selection stays put.
• Runs from MeasurePaywallBounds, keyed on the measured bounds — also covers the initial selection, which is resolved before the bounds are known.
• Tab switches and sheet-dismiss resets restore selections that were resolved without the bounds, so those paths reconcile too.
• resolvesVisible and the default-selection helpers take an optional measured size; the initial-selection path is unchanged (window unknown).

Design note

When nothing resolves visible at the current size, the hidden selection is kept — matching iOS. Rules that hide every package at some size are an authoring error; the plan is to guard against that in the dashboard (validation when window rules can leave zero visible packages) rather than guess at runtime.

Scope note for reviewers

Size-class (screen condition) rules have the same gap and keep their existing pinned-to-COMPACT behavior — reconciling those changes behavior of already-shipped paywalls, so it's left as a deliberate follow-up decision. Mirrors iOS reconcileSelection (window-size scope).

Tests

• WindowSizeSelectionReconcileTests: moves off a hidden package, keeps a visible one, no-op on unknown size, keeps selection when nothing is visible, prefers a visible authored default, no flapping on resize-back, height rules.
• PaywallStateLoadedComponentsPackageSelectionTests: tab switch onto a hidden remembered default, sheet-dismiss reset restoring a hidden default, user-selected package hidden by a rule.

🤖 Generated with Claude Code

---

Note

Medium Risk
Changes which package is selected for purchase across resize, tabs, and sheet dismiss, but behavior is guarded by extensive tests and explicit iOS parity rules.

Overview
Fixes a gap where a package hidden by window width/height or size-class visibility rules could remain selected and purchasable after bounds change, tab switch, or sheet reset—initial selection still runs before measure.

reconcileSelectionForWindowSize on PaywallState.Loaded.Components re-checks whether the current selection would render visible (using measured paywall bounds for dimension rules and windowScreenCondition from the app window for compact/medium/expanded). If not, selection moves to the tab’s authored default, then other visible fallbacks; if nothing is visible, selection stays put (iOS parity). Replacement candidates are evaluated unselected; the keep-check uses selected view state so selected-condition rules don’t oscillate.

MeasurePaywallBounds now writes windowScreenCondition and runs reconciliation in a LaunchedEffect keyed on bounds, screen condition, and state list (including prewarmed workflow states). Tab update, resetToDefaultPackage, and peekDefaultPackageUniqueIdAfterSheetDismiss use the same visibility-aware helpers so analytics and dismiss match reset behavior.

resolvesVisible / default-selection helpers accept optional measured size and screen condition; FakePaywallState stops synthesizing extra always-visible package components that broke visibility tests. New unit and Compose window-resize tests cover resize, tabs, medium/selected rules, and late-added states.

^{Reviewed by Cursor Bugbot for commit 1763d7b. Bugbot is set up for automated code reviews on this repo. Configure here.}

[emerge-tools]

📸 Snapshot Test

632 unchanged

Name	Added	Removed	Modified	Renamed	Unchanged	Errored	Approval
TestPurchasesUIAndroidCompatibility Paparazzi com.revenuecat.testpurchasesuiandroidcompatibility.paparazzi	0	0	0	0	285	0	N/A
TestPurchasesUIAndroidCompatibility com.revenuecat.testpurchasesuiandroidcompatibility	0	0	0	0	347	0	N/A

---

🛸 Powered by Emerge Tools

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 83.70%. Comparing base (3172730) to head (1763d7b).

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #4198   +/-   ##
=======================================
  Coverage   83.70%   83.70%           
=======================================
  Files         456      456           
  Lines       18636    18636           
  Branches     2813     2813           
=======================================
  Hits        15599    15599           
  Misses       2134     2134           
  Partials      903      903           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[cursor]

Cursor Bugbot has reviewed your changes and found 3 potential issues.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 301466a. Configure here.}