Real-Time Chat System

A real-time chat backend built with Go for scalability, concurrency safety, and clean architecture.

Supports private and group chats, with WebSocket-based communication, MongoDB for messages, and PostgreSQL for room metadata.

✨ Features

• 🔒 Concurrency-safe message handling
• ⚡ Real-time communication via WebSocket
• 📬 Private chat flow (lazy room creation) → room would be created when first message sent
• 👥 Group chat flow → WhatsApp/Discord-like group creation & invites
• 📨 Async worker for background tasks (priority queue, message persistence)
• 💾 Hybrid storage → MongoDB for messages, SQL for metadata
• ⏱️ Queue-based architecture (priority queue + worker)

🏗️ Project Tree Architecture

.
├── cmd/                # Entry point (main.go)
├── config/             # App configuration (config.go)
├── internal/           # Core business logic
│   ├── dtos/           # Data transfer objects (request/response)
│   ├── entity/         # Domain entities (Message, Room, User, dll.)
│   ├── errors/         # Custom error handling
│   ├── handlers/       # HTTP/WebSocket handlers
│   ├── middleware/     # Middleware (auth, logging, dsb.)
│   ├── queue/          # Queue management (priority queue)
│   ├── repo/           # Repository layer (Mongo + SQL access)
│   ├── routers/        # Routing definitions
│   ├── use-case/       # Application/business logic
│   ├── utils/          # Utility functions
│   ├── websocket/      # WebSocket hub & client management
│   └── worker/         # Async workers (process tasks dari queue)
├── migrations/         # Database migrations
├── state/              # State initialization (Mongo, Postgres, Redis, Secret)
│   ├── initMongo.go
│   ├── initPostgres.go
│   ├── initRedis.go
│   ├── initSecret.go
│   └── initState.go
├── docker-compose.yaml # Docker setup (Postgres, Mongo, Redis, dsb.)
└── ...

💡 Private Chat Flow

1. User sent first message to subject -> system would checking if there any chat room respect to participant is already exists or not

• If yes -> reuse
• If not -> create new room + add chat participant

2. Store message content in MongoDB, metadata in SQL
3. Broadcast to subject user via WebSocket

💡 Group Chat Flow

• Working on it

🚀 Getting Started

Prerequisites

• Go >=1.21
• PostgreSQL / MySQL (just fit with your preference SQL)
• MongoDB
• Redis (for working queue) Run Locally

# clone repo
git clone https://github.com/Xenn-00/chat-system.git
cd chat-system

# install deps
go mod tidy

# run service
go run main.go

🛠️ Tech Stack

• Go (chi, goroutines, channels)
• WebSocket
• MongoDB
• PostgreSQL
• Redis
• Docker

License

This project is licensed under the MIT LICENSE