How leongiscoding/fyp2 works
Project type and purpose
• This is a cross-platform Flutter application (Android, iOS, Web) for BMI & calorie estimation that uses on-device TensorFlow Lite models and Firebase for authentication and Firestore persistence.
File structure
• Top-level “android”, “ios”, and “web” folders represent platform runners.
• The “lib” directory holds your Dart code, organized into pages, UI components, controllers, ML utilities, Firebase services, and theme.
• The “assets” folder contains TFLite model files and label definitions.
• Configuration files (.env, firebase.json, pubspec.yaml) manage environment variables, Firebase setup, and dependencies.
README for insights
• While the README is generic, the presence of Firebase config files and TensorFlow Lite assets signals core external dependencies and ML integration.
Main components & interactions
a. Client Platforms
• Android Runner
• iOS Runner
• Web Runner
All embed the Flutter Engine and Dart VM.
b. Flutter Application Layers
1. Presentation Layer
• Pages (loginPage.dart, mainPage.dart, profilePage.dart, resultPage.dart, signUpPage.dart)
• Reusable Components (lib/component/…)
2. Business Logic Layer
• Controllers (bmi_calculator.dart, calorie_estimator.dart wrapper)
3. Services Layer
• ML Service (lib/ml/calorie_estimator.dart loads TFLite models from assets)
• Firebase Services (auth_service.dart, firestore_service.dart, profile_firestore_service.dart)
4. Data Layer
• Firestore Database
• On-device TFLite models (assets/model.tflite, assets/model_fruit_v2.tflite)
c. External Services
• Firebase Authentication
• Firebase Firestore
Architectural patterns & design principles
• Separation of Concerns: UI (pages & widgets), Logic (controllers), Services (ML & Firebase)
• MVC-like layering: Widgets ← Controllers ← Services ← External Systems
• Cross-platform runner abstraction via Flutter Engine