fix: restore analysis on reload instead of rendering a blank page

[bhavik-mangla]

Addressed Issues:

Fixes #204

What was wrong

The analytical model lived only in React state. Any fresh page load — a reload, a bookmark, a shared link — started with model === null, and nothing restored it. Two different failures followed:

Route	Behaviour before
/overview, /repositories, /contributors, /analytics, /governance	if (!model) return null → blank page between navbar and footer
/network	crash — TypeError: Cannot read properties of null (reading 'allRepos') at NetworkPage.jsx:257

NetworkPage was simply missing the null guard its five sibling pages already had.

Reproducible on main today: npm run dev, then open http://localhost:5173/overview directly — <main> renders with 0 children.

The fix

Two parts, matching the approach suggested in #204:

1. Persist and restore the analysis (src/services/cache.js)

Cache the model in IndexedDB with a one-hour TTL — the same "1HR intelligent cache" the landing page already advertises — and restore it on startup. A reload now costs zero API calls, which matters against a 60 req/hr unauthenticated budget.

IndexedDB rather than localStorage because a single model runs to several megabytes for a large org (239 repos for one the size of Vercel), well past the ~5MB quota. Every cache operation is best-effort, so a private window or a browser with storage disabled degrades to the old behaviour rather than breaking.

Audit and pull-request results are cached too — they are the most expensive data to refetch.

2. Guard the routes that need a model (src/components/RequireAnalysis.jsx)

For the case where there is genuinely nothing to restore (no cache, or an expired one), redirect to the organization picker — the only place an analysis can be started — instead of rendering an empty page.

Two details worth review:

• It waits on hydrating, so a reload is not redirected away a fraction of a second before its own cached data arrives.
• It keeps children mounted while explore() refetches, because explore() clears the model first and each page shows its own skeleton during that window.

/settings and /support-us are deliberately left unguarded — Settings is where a PAT is added, so it has to work with no analysis loaded.

One subtlety handled: the write-back skips the save that would immediately follow a restore. Without that, every page load would stamp a fresh savedAt and the entry would never reach its TTL.

Verification

Check	Result
npm test	✅ 44/44 passing (40 before, +4 new for RequireAnalysis)
npm run build	✅ passes
Reload with a valid cache	✅ analysis restored, stays on route, 0 GitHub API calls
Reload with an expired cache	✅ stale entry purged, redirected to the picker
TTL not refreshed by a restore	✅ entry still reports its original age after reload
/network with no analysis	✅ redirects — was 4 console TypeErrors
/overview with no analysis	✅ redirects — was 0 children in <main>

Verified in a real browser against real IndexedDB, seeding the cache directly so the restore path could be exercised without spending API quota.

Additional Notes:

• Independent of fix: serve SPA shell for deep links on GitHub Pages #203 (the GitHub Pages deep-link 404 fix) and mergeable in either order. They are complementary: fix: serve SPA shell for deep links on GitHub Pages #203 makes a reload reach the app at all on Pages, this makes the app render data once it gets there.
• Unrelated pre-existing warning noticed while testing: OverviewPage's language-distribution list logs Each child in a list should have a unique "key" prop. Left alone to keep this diff focused — happy to fix separately.

Checklist

• My code follows the project's code style and conventions
• I have made corresponding changes to the documentation
• My changes generate no new warnings or errors
• I have joined the Discord server and I will share a link to this PR with the project maintainers there
• I have read the Contributing Guidelines

⚠️ AI Notice - Important!

AI assistance was used to diagnose and implement this change. Both failure modes were reproduced on a clean main checkout before any code was written, the fix was verified in a real browser against real IndexedDB (including TTL expiry and the no-API-calls claim), and the full build and test suite were run before submitting. New unit tests cover all four states of the route guard.

Summary by CodeRabbit

• New Features

  • Restores recent analysis results automatically when returning to the app.
  • Saves analysis data locally for up to one hour, including analytics and audit results.
  • Shows analysis pages only when analysis data is available.

• Bug Fixes

  • Prevents analysis pages from displaying blank or broken states during loading and restoration.
  • Fixes errors when opening the Network page without loaded analysis data.

[Copilot]

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

[coderabbitai]

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 9c6120f8-426f-4360-b735-b82c0bd39f0c

📥 Commits

Reviewing files that changed from the base of the PR and between 4ce2822 and 4ab8c0c.

📒 Files selected for processing (6)

• src/App.jsx
• src/components/RequireAnalysis.jsx
• src/components/RequireAnalysis.test.jsx
• src/context/AppContext.jsx
• src/pages/NetworkPage.jsx
• src/services/cache.js

---

Walkthrough

The app now persists the latest analysis in IndexedDB, restores it during startup, exposes hydration state, and guards analysis routes. Tests cover restoration, loading, missing-model, and rendered-model states.

Changes

Analysis state flow

Layer / File(s)	Summary
IndexedDB analysis cache src/services/cache.js	Adds best-effort IndexedDB storage with a one-hour TTL for the latest analysis.
App state hydration and persistence src/context/AppContext.jsx	Restores cached analysis during startup, exposes hydrating, and saves later analysis changes without refreshing the cache timestamp during restoration.
Analysis route guard and validation src/components/RequireAnalysis.jsx, src/components/RequireAnalysis.test.jsx, src/App.jsx, src/pages/NetworkPage.jsx	Guards analysis routes, preserves pages during loading, redirects when analysis is unavailable, tests state combinations, and prevents null model access in NetworkPage.

Estimated code review effort: 3 (Moderate) | ~20 minutes

Sequence Diagram(s)

sequenceDiagram
  participant Browser
  participant AppProvider
  participant IndexedDB
  participant RequireAnalysis
  participant AnalysisRoute
  Browser->>AppProvider: start application
  AppProvider->>IndexedDB: loadAnalysis()
  IndexedDB-->>AppProvider: cached analysis or null
  AppProvider->>RequireAnalysis: expose model, loading, hydrating
  RequireAnalysis->>AnalysisRoute: render when analysis is available
  RequireAnalysis-->>Browser: redirect to / when no model and not loading

Loading

Suggested labels: Typescript Lang

Suggested reviewers: rahul-vyas-dev, ri1tik

Poem

A rabbit watched the cache store bright,
IndexedDB held data tight.
Hydration crossed the startup stream,
Routes waited safely for the model’s gleam.
No null fields caused a fright,
The dashboard hopped back into sight.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

• X
• Mastodon
• Reddit
• LinkedIn

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