How langochungdev/instar works
Project type and purpose
• This is a full-stack “Instagram clone” application:
• Backend: Spring Boot monolith exposing RESTful APIs, WebSocket chat, WebRTC signaling, JWT-based security.
• Frontend: VueJS SPA + Bootstrap, communicating via Axios, WebSocket, WebRTC.
• Database: SQL Server, with Redis for token blacklisting/possibly caching.
File structure
• Top-level directories:
• backend/ – Maven project with layered packages (config, common, feature/*, entity, repository, service, controller).
• frontend/ – Vue project organized by features (auth, device, init), plus global layouts, router, store, utils.
• TaiNguyen/ – miscellaneous docs, SQL scripts (instar.sql), design notes.
• Key config files:
• backend/src/main/resources/application.properties (DB, Redis, JWT settings)
• backend/src/main/java/.../config/*.java (WebMvcConfig, SecurityConfig, RedisConfig, WebSocketConfig)
• frontend/.env, vite.config.js, src/utils/firebase.js, axios.js.
• Deployment artifacts: Dockerfiles likely added in deploy scripts (not shown), CI/CD defined in project readme.
README for additional insight
• REST API endpoints across core domains: User, Post, Comment, Like, Follow, Save, Message, Notification, UserBehavior.
• Architectural notes: layered backend (Controller → Service → Repository → Entity), JWT & Spring Security, WebSocket for real-time, SQL Server schema, CI/CD on Ubuntu + Nginx.
• Frontend: Vue components organized by feature, real-time integrations.
Essential elements to capture
• All client interactions (login, feed, chat, call) mapped to API endpoints.
• Real-time flows separately drawn (WebSocket, WebRTC).
• Data persistence flows (CRUD ➔ SQL Server).
• Token lifecycle (Login → JWT issued → blacklist check via Redis on logout/refresh).
• CI/CD and deployment can be shown as an auxiliary pipeline if needed.