feat(perf): implement lazy loading and code splitting for faster star…

[sweetesty]

Closes #410

---

This PR implements lazy loading and code splitting across the app to reduce cold start time and bundle size.

What was done:

metro.config.js

• Metro code splitting configured to enable async bundle loading
• Chunking strategy defined: critical auth/home screens in the main bundle; secondary screens (settings, reports, details) in async chunks

src/navigation/AppNavigator.tsx

• Non-critical screens converted to React.lazy() imports
• Suspense boundaries added around lazy-loaded screen groups with lightweight fallback components
• Frequently used screens (e.g. home, dashboard) kept as eager imports to avoid perceptible loading delays on primary flows
• Bundle loading failures caught at the Suspense boundary level with an error fallback UI and retry option, preventing a blank screen on chunk load failure

src/screens/

• Secondary and infrequently accessed screens refactored for dynamic import compatibility
• No changes to screen component logic — only import strategy updated

Prefetching

• Critical secondary modules prefetched during app idle time using InteractionManager.runAfterInteractions so they are warm before the user navigates to them

Bundle size analysis

• Full bundle analysis report generated and included at docs/bundle-analysis.md
• Main bundle size reduction documented; per-chunk sizes listed

Performance CI check

• Performance regression check added to CI pipeline — build fails if main bundle size exceeds the defined threshold
• Cold start time measured and documented against the <2s target

Acceptance criteria met:

• ✅ Metro code splitting configuration
• ✅ Lazy-loaded screens with Suspense boundaries
• ✅ Bundle loading failure handling
• ✅ Eager vs lazy tradeoff applied — critical screens kept eager
• ✅ Prefetching critical modules on app idle
• ✅ Bundle size analysis report
• ✅ Performance regression CI check
• ✅ Startup time reduction targeting <2s cold start

[drips-wave]

@sweetesty Great news! 🎉 Based on an automated assessment of this PR, the linked Wave issue(s) no longer count against your application limits.

You can now already apply to more issues while waiting for a review of this PR. Keep up the great work! 🚀

Learn more about application limits