16. State Normalization (Redux / Large Apps)
Problem
Large Flutter apps often suffer from:
- Deeply nested objects
- Hard-to-update UI
- Excessive rebuilds
Solution → Normalize State
Use Maps instead of deep trees.
Bad
User {
List<Post> posts;
}
Good
Map<String, User> users;
Map<String, Post> posts;
Why?
- O(1) access
- Easier updates
- Smaller rebuild scope
Data Structure Used: HashMap
17. Event-Driven Architecture (Streams as Queues)
Flutter is Event-Based
- User taps
- API responses
- WebSocket messages
All of these form an event queue.
Stream = Observer + Queue
final controller = StreamController<Event>();
Internally:
[Event1] → [Event2] → [Event3] → UI
Used in:
- BLoC
- Riverpod StreamProvider
- WebSocket listeners
18. Background Tasks & Isolates (Message Queues)
Problem
Heavy computation blocks UI.
Solution → Isolates
Isolates communicate via message queues.
ReceivePort receivePort = ReceivePort();
await Isolate.spawn(worker, receivePort.sendPort);
Data Structures
- Queue (messages)
- Immutable data (safe transfer)
19. Image & Asset Caching (LRU in Practice)
Flutter Image Cache
Flutter internally uses:
- HashMap
- LRU eviction
You should mirror this for:
- API images
- WebView content
- Offline files
Custom Cache Layer
final imageCache = LRUCache<String, Uint8List>(100);
Rules:
- Fixed capacity
- Evict least-used
- Prevent memory leaks
20. Offline-First Architecture
Data Flow
UI → Cache → API → Cache → UI
Data Structures
| Layer | Structure |
|---|---|
| Local DB | B-Tree (SQLite) |
| Memory Cache | HashMap |
| Sync Queue | Queue |
| Conflict tracking | Set |
Sync Queue Example
Queue<SyncTask> pendingTasks = Queue();
- Retry failed requests
- Process sequentially
- Resume on app restart
21. Pagination at Scale
Wrong Approach
- Load everything into
List - Rebuild entire widget tree
Correct Approach
- Windowed lists
- Lazy loading
- Indexed access
Structures Used
| Purpose | DS |
|---|---|
| Visible items | List |
| Prefetch buffer | Queue |
| Loaded pages | Map<int, Page> |
22. Flutter Rendering Pipeline
Internals
| Stage | Data Structure |
|---|---|
| Widget Tree | Tree |
| Element Tree | Tree |
| Render Tree | Tree |
| Layer Tree | Tree |
Understanding this helps you:
- Reduce rebuilds
- Avoid layout thrashing
- Optimize animations
23. Dependency Injection
Providers Form a Graph
ProviderA → ProviderB → ProviderC
- Circular dependencies = graph cycle
- Riverpod prevents invalid cycles
Data Structure: Directed Acyclic Graph (DAG)
24. Feature Flags & Permissions
Efficient Access
Set<String> enabledFeatures;
Why Set?
- O(1) lookup
- No duplicates
- Clean semantics
Used for:
- A/B testing
- Remote config
- Role-based UI
25. Error Handling Pipelines
Error Flow
Error → Queue → Logger → UI
Use:
- Queue for retries
- Stack for call traces
- Map for error metadata
26. Large Forms & Validation Trees
Form as Tree
Form
├── Section
│ ├── Field
│ └── Field
Traversal:
- DFS → validate all
- BFS → step-by-step wizard
27. Performance Checklist (DS-Focused)
| Symptom | Likely DS Issue |
|---|---|
| UI jank | Large List rebuild |
| Memory leak | Unbounded cache |
| Slow search | Linear scan instead of Map/Trie |
| Crashes on back | Stack overflow |
| API spam | Missing cache |
28. How Senior Flutter Devs Think
They don’t ask:
“Which package should I use?”
They ask:
“What data structure models this problem best?”
29. Architecture Mapping Summary
| Flutter Concern | Data Structure |
|---|---|
| Navigation | Stack |
| State | Map |
| UI | Tree |
| Cache | LRU |
| Search | Trie |
| Events | Queue |
| Dependencies | Graph |
| Offline sync | Queue + Set |
30. Final Takeaway
Flutter is not about widgets.
It is about data flowing through well-chosen structures.
If you master this:
- Your apps scale
- Your UI stays smooth
- Your architecture stays clean
