fix(app): add non-creating child-store peek for sidebar timestamps (closes #475) by Astro-Han · Pull Request #552 · Astro-Han/pawwork

Astro-Han · 2026-05-11T11:07:11Z

Summary

Add peekExisting(directory) as a cache-only child-store accessor.
Switch sidebar timestamp message and part reads to peekExisting so missing directory caches are not materialized during sidebar row building.
Add child-store and layout source-level tests that lock the non-creating behavior.

Why

Sidebar timestamp reads only need cached messages and parts when a child store already exists. Calling globalSync.child(..., { bootstrap: false, pin: false }) still creates a child store and its persisted vcs/project/icon caches just to discover absence.

Related Issue

Closes #475.

Human Review Status

Pending. A human should make the final merge decision after reviewing the final diff and verification evidence.

Review Focus

peekExisting must stay read-only: validate the directory, then return children[directory] without calling ensureChild, pinning, bootstrapping, marking, or evicting.
Existing child() and peek() semantics should remain unchanged.
Sidebar timestamp messagesForSession and partsForMessage should both use the non-creating accessor.

peekExisting semantics

peekExisting(directory) returns the existing ChildStoreTuple if present in children, otherwise undefined. It does not call ensureChild, does not pin for the current owner, does not bootstrap loading, and does not run eviction. It is intentionally a non-reactive read: if the child store does not yet exist, peekExisting does not subscribe to its eventual creation. Sidebar timestamp reads use this to avoid materializing per-directory caches just to discover absence.

If a future caller needs absent-to-present reactivity, that is a separate reactive-registry design and out of scope for this PR.

Risk Notes

Low. The new accessor is only used by sidebar timestamp cache reads. Missing caches now return undefined, which preserves the existing row-builder fallback to session creation time. Existing child() and peek() callers keep their current ensure/create behavior.

How To Verify

Targeted app tests: 1048 passed, 0 failed
Command: bun --cwd packages/app test --preload ./happydom.ts src/context/global-sync/child-store.test.ts src/pages/layout/pawwork-sidebar-session-rows.test.ts

App typecheck: passed
Command: bun --cwd packages/app typecheck

Diff check: passed
Command: git diff --check

Screenshots or Recordings

Not applicable. This is a cache access behavior change with no intended visible UI change.

Checklist

Summary by CodeRabbit

New Features
- Added optimized data access method for existing application state, allowing more efficient queries without triggering unnecessary initialization.
Improvements
- Enhanced sidebar rendering performance by streamlining how session data is accessed and displayed.

Astro-Han

Opus second-pass review (product layer / three questions / sidebar timestamp path)

Conclusion: approve

Diff matches the brief shape; both call sites are switched; tests lock the contract.

Three questions

Could this be even less? peekExisting is 5 lines of implementation, the type alias is a 1-liner extracted from the previously inline tuple shape, plus 2 child-store tests and 1 source-level layout invariant. Skipping the type alias would force [Store<State>, SetStoreFunction<State>] to be repeated inline; not worth the saving. Check.

Is what remains good enough?

peekExisting does exactly validateChildStoreDirectory + return children[directory] — no ensure, no pin, no bootstrap, no mark, no eviction.
Both messagesForSession and partsForMessage switched; the per-call tuple?.[0].message[session.id] keeps the existing fallback contract intact.
Source-level invariant test locks layout.tsx against drifting back to globalSync.child(session.directory, { bootstrap: false, pin: false }).
Check.

Is it reassuring? Sidebar timestamp user path traced:

First boot, no cache yet: peekExisting(dir) returns undefined → tuple?.[0]... short-circuits to undefined → buildPawworkSidebarSessionRows falls back to session.time.created (covered by existing row-builder tests at line 138-143 in the same test file).
After directory has been visited: peekExisting(dir) returns the existing tuple → cache hit, identical to the old child(dir, { bootstrap: false, pin: false }) behavior except we no longer materialize vcs/meta/icon caches on the side.
Disk persistence layer untouched; no new I/O introduced.
Check.

Findings

No P0/P1/P2.

P3 note (not blocking): child() and peek() are now indistinguishable in semantics (both ensure + optional bootstrap; only child pins). That equivalence pre-dates this PR and is out of scope, but worth a follow-up issue if the team wants the manager surface to be smaller.

Approve

coderabbitai · 2026-05-11T11:08:44Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Pro Plus

Run ID: dd19fd31-7559-45e8-9849-c8572ed335c9

📥 Commits

Reviewing files that changed from the base of the PR and between 7ac9d47 and 7f97886.

📒 Files selected for processing (6)

packages/app/src/context/global-sync.tsx
packages/app/src/context/global-sync/child-store.test.ts
packages/app/src/context/global-sync/child-store.ts
packages/app/src/context/global-sync/types.ts
packages/app/src/pages/layout.tsx
packages/app/src/pages/layout/pawwork-sidebar-session-rows.test.ts

📝 Walkthrough

Walkthrough

This PR adds a non-creating child-store peek accessor (peekExisting) to enable cache-only reads without creating, bootstrapping, or pinning directory stores. The method is wired through global-sync context to sidebar timestamp reads, reducing lifecycle side effects for cache lookups.

Changes

Non-creating Child Store Accessor

Layer / File(s)	Summary
Type Definition `packages/app/src/context/global-sync/types.ts`	`ChildStoreTuple` type alias added to represent a SolidJS store and setter tuple pair.
Child Store Manager `packages/app/src/context/global-sync/child-store.ts`	Imports and uses `ChildStoreTuple` for internal registry; adds `peekExisting(directory)` method returning existing child store tuple or `undefined` without calling `ensureChild(...)`; exposes `peekExisting` in public API.
Global Sync Context `packages/app/src/context/global-sync.tsx`	Exposes `peekExisting` on `GlobalSyncProvider` context value by forwarding to `children.peekExisting`.
Sidebar Cache Reads `packages/app/src/pages/layout.tsx`	Updates `buildPawworkSidebarSessionRows` to read message/part data via `globalSync.peekExisting(session.directory)` instead of `globalSync.child(..., { bootstrap: false, pin: false })`.
Child Store Tests `packages/app/src/context/global-sync/child-store.test.ts`	Enhances `createManager` test helper to accept `overrides` parameter; adds test verifying `peekExisting(missingDir)` returns `undefined` without creating caches or triggering bootstrap; adds test verifying `peekExisting(existingDir)` returns the created store tuple.
Layout Sidebar Integration Test `packages/app/src/pages/layout/pawwork-sidebar-session-rows.test.ts`	Adds test reading `layout.tsx` source and asserting cache reads use `globalSync.peekExisting(...)` without calling `globalSync.child(...)`.

🎯 2 (Simple) | ⏱️ ~12 minutes

Possibly related PRs

Astro-Han/pawwork#161: Both modify createChildStoreManager and its tests; the main PR adds peekExisting API and tuple type while this PR adds directory validation and persisted-cache error handling.
Astro-Han/pawwork#101: Both address child-store access patterns and lifecycle concerns; the main PR adds non-creating peekExisting accessor while this PR changes how current child is resolved to avoid provider-owned caching.

Suggested labels

bug, ui

Poem

🐰 A peek without the crashing sound,
No bootstrap stores to spin around,
Just read the cache that's waiting there,
Sidebar stats with lighter care!

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and specifically describes the main change: adding a non-creating child-store peek for sidebar timestamps, with issue reference.
Description check	✅ Passed	The description is comprehensive and follows the template structure with all key sections completed: summary, why, related issue, review focus, risks, verification steps, and completed checklist.
Linked Issues check	✅ Passed	The PR implements all coding requirements from `#475`: adds peekExisting accessor with correct semantics (no ensureChild/bootstrap/pin), integrates into sidebar timestamp reads, and includes focused tests validating non-creating behavior.
Out of Scope Changes check	✅ Passed	All changes are scoped to `#475` requirements: new accessor implementation, type definitions, sidebar integration, and targeted tests. No unrelated refactors, dependency changes, or broad call-site replacements introduced.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

Create stacked PR
Commit on current branch

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch fix/peek-existing-child-store

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

Astro-Han · 2026-05-11T11:09:43Z

GPT-X engineering final review

Result: pass. I found no P0/P1/P2 issues.

Engineering checks

peekExisting(directory) is a true non-creating accessor: it validates the directory and returns children[directory]; it does not call ensureChild, mark, pin, pinForOwner, onBootstrap, or runEviction.
Existing child() and peek() semantics are unchanged. This keeps retainDirectory and the route-resolution call site working with their current tuple-return assumptions.
Both sidebar timestamp cache reads were switched: messagesForSession and partsForMessage now use globalSync.peekExisting(session.directory).
The tests cover the important invariant: missing child stores return undefined without creating children, vcsCache, metaCache, or iconCache, and an existing child tuple is still returned after normal creation.
The layout source-level test locks the sidebar timestamp path against regressing back to globalSync.child(session.directory, { bootstrap: false, pin: false }).

Binary check

v1 would not crash without this PR, but the previous sidebar timestamp path could still materialize per-directory child stores and persisted caches during row building. This is a focused memory/lifecycle cleanup at the correct layer: cache-only reads get a cache-only accessor, while normal create/bootstrap/pin callers keep their existing APIs.

Notes

peekExisting is intentionally non-reactive. If a future caller needs absent-to-present reactivity, that should be designed separately as a reactive registry rather than added to this cache peek API.

gemini-code-assist

Code Review

This pull request introduces a peekExisting method to the globalSync context and childStoreManager, allowing for the retrieval of existing child stores without triggering new store creation or side effects. The Layout component was updated to utilize this method for session data lookups, and new tests were added to ensure correct behavior and prevent regressions. I have no feedback to provide.

fix(app): add non-creating child-store peek

7f97886

Astro-Han added enhancement New feature or request P3 Low priority app Application behavior and product flows task Narrow execution, audit, spike, migration, tracking, or upstream follow-up work labels May 11, 2026

Astro-Han commented May 11, 2026

View reviewed changes

gemini-code-assist Bot reviewed May 11, 2026

View reviewed changes

Astro-Han merged commit 3c15b37 into dev May 11, 2026
23 checks passed

Astro-Han deleted the fix/peek-existing-child-store branch May 11, 2026 11:20

Astro-Han mentioned this pull request May 11, 2026

[Task] Windows advisory: session loop tests time out at 3000ms #555

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix(app): add non-creating child-store peek for sidebar timestamps (closes #475)#552

fix(app): add non-creating child-store peek for sidebar timestamps (closes #475)#552
Astro-Han merged 1 commit into
devfrom
fix/peek-existing-child-store

Astro-Han commented May 11, 2026 •

edited by coderabbitai Bot

Loading

Uh oh!

Astro-Han left a comment

Uh oh!

coderabbitai Bot commented May 11, 2026 •

edited

Loading

Walkthrough

Changes

Possibly related PRs

Suggested labels

Poem

❌ Failed checks (1 warning)

Uh oh!

Astro-Han commented May 11, 2026

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

Astro-Han commented May 11, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Why

Related Issue

Human Review Status

Review Focus

peekExisting semantics

Risk Notes

How To Verify

Screenshots or Recordings

Checklist

Summary by CodeRabbit

Uh oh!

Astro-Han left a comment

Choose a reason for hiding this comment

Opus second-pass review (product layer / three questions / sidebar timestamp path)

Conclusion: approve

Three questions

Findings

Approve

Uh oh!

coderabbitai Bot commented May 11, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Possibly related PRs

Suggested labels

Poem

❌ Failed checks (1 warning)

Uh oh!

Astro-Han commented May 11, 2026

GPT-X engineering final review

Engineering checks

Binary check

Notes

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Astro-Han commented May 11, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented May 11, 2026 •

edited

Loading