How imabu0/cse327-project-nbm works
Project type and purpose
• This is a full-stack, multi-client “Infinite Cloud” application for managing files across Google Drive, Dropbox, and your own server.
• It provides three client surfaces:
• A React/Vite web front-end (“client/”)
• An Android mobile app (“mobile-app/”)
• A bot interface (“bot/”, likely a chat or messaging bot)
• Backend is a Node.js/Express service (“server/”), backed by a SQL database (“database/db.sql”). It integrates with external storage APIs (Google, Dropbox) and an LLM utility.
File structure
• Top-level directories:
• bot/ – Node.js bot logic (bot.js, package.json)
• client/ – React SPA built with Vite (components, pages, Tailwind config)
• mobile-app/ – Android app in Kotlin using Retrofit, TokenManager, MVVM-style ui/
• server/ – Express server with layered MVC:
– config/ (db.js)
– models/ (auth, user, file, bucket, dropbox, google)
– middlewares/ (authMiddleware)
– routes/ (authRoutes, userRoutes, fileRoutes, dropboxRoutes, googleRoutes)
– utils/ (llm integration)
– server.js entry point
• database/ – SQL schema (db.sql)
README
a. Main components
• Web client (React SPA)
• Mobile client (Android app)
• Bot client (Node.js bot)
• Backend API server (Node.js/Express)
• Database (SQL)
• External storage services (Google Drive API, Dropbox API)
• LLM service (in-house or third-party)
b. Relationships & interactions
• All three clients send HTTP(S) requests (REST) to the API Server (auth, file, user, Google, Dropbox endpoints).
• The API server:
– Authenticates via authMiddleware
– Reads/writes metadata to SQL database
– Forwards file operations to Google or Dropbox APIs
– Uses llm.js for any AI-driven features
• The Android app uses RetrofitClient → TokenManager → API server.
• The bot uses bot.js to receive messages, then calls server APIs.
c. Architectural patterns & principles
• Layered MVC on the server
• Separation of concerns: distinct front-ends share a common REST API
• External API integration modules (Google, Dropbox) abstracted in models
• Middleware for cross-cutting concerns (authentication)
d. Key technologies & frameworks
• Server: Node.js, Express, JWT/authMiddleware, Sequelize or raw SQL via db.js
• Web: React, Vite, Tailwind CSS
• Mobile: Kotlin, Android MVVM, Retrofit
• Bot: Node.js
• Database: MySQL/Postgres (SQL schema in db.sql)
• External APIs: Google Drive, Dropbox
• AI: LLM utility (server/utils/llm.js)
Detailed capture of architecture
• Route Handlers → Middlewares → Models → External Integrations / DB
• Database schema entities (User, File, Bucket, Google, Dropbox) if space permits
• LLM utility as an external AI service or an internal micro-module