Sansofé Project Server

Overview

The Sansofe Project Server is a Node.js backend application built with Express.js and MongoDB. It provides a RESTful API for the Sansofe platform, handling authentication and data storage.

Features

• RESTful API architecture
• JWT-based authentication and authorization
• MongoDB database integration with Mongoose ODM
• Secure password handling with bcrypt
• Environment configuration with dotenv
• Cross-Origin Resource Sharing (CORS) support
• Request logging with Morgan

Technology Stack

• Node.js: Runtime environment
• Express.js: v4.21.2 - Web server and API framework
• MongoDB: Database (via Mongoose v8.11.0)
• JWT: v9.0.2 - For authentication
• bcrypt: v5.1.1 - For password hashing
• cors: v2.8.5 - For CORS support
• morgan: v1.10.0 - For HTTP request logging

Getting Started

Prerequisites

• Node.js (v18 or later recommended)
• MongoDB (local installation or MongoDB Atlas account)
• npm package manager

Installation

1. Clone the repository:

   git clone https://github.com/Jojocg/Sansofe-Project-backend.git
   cd Sansofe-Project-backend
   

2. Install dependencies:

   npm install
   

3. Create a .env file in the root directory with the following variables as examples:

   PORT=5005
   MONGODB_URI=mongodb://127.0.0.1:27017/sansofe_db
   TOKEN_SECRET=your_jwt_secret_key
   ORIGIN=http://localhost:3000
   

4. Start the development server:

   npm run dev
   

5. The API will be available at http://localhost:5005

Project Structure

sansofe-project-server/
├── config/             # Configuration files for frontend connection
├── controllers/        # Request controllers
├── db/                 # Database connection setup
├── middleware/         # Custom middleware
├── models/             # Mongoose models
├── routes/             # API routes
├── .env                # Environment variables
├── package.json        # Dependencies and scripts
└── server.js           # Entry point

AI Integration

Overview

The project integrates with Google's Gemini AI to provide intelligent responses about local markets in Gran Canaria. This integration enhances user experience by offering:

• Natural language interactions about markets
• Context-aware responses based on specific markets or towns
• Intelligent filtering of market information

Features

• 🤖 Smart Assistant: Provides detailed information about local markets
• 🔒 Rate Limiting: Protects the AI service with 20 requests per IP per 15 minutes
• ⚡ Response Caching: Improves performance for repeated queries
• 🔍 Context-Aware: Tailored responses based on specific market or town context

Technical Implementation

AI Request Flow

1. Request Validation

   POST /api/ai/assistant
   {
     "query": "string",     // Required: User's question
     "marketId": "string",  // Optional: Specific market
     "townId": "string"     // Optional: Specific town
   }

2. Response Format

   {
     "response": "string"   // AI-generated response
   }

3. Error Handling

   {
     "error": true,
     "type": "ErrorType",
     "message": "Error description",
     "details": {}
   }

Security Measures

• Request validation and sanitization
• Rate limiting protection
• Input length restrictions
• Context-based response filtering

Dependencies Added

• @google/generative-ai: Gemini AI integration
• express-rate-limit: API request limiting
• express-validator: Request validation

Environment Variables

Add to your .env:

GEMINI_API_KEY=your_gemini_api_key

Usage Limits

• 20 AI requests per IP address per 15 minutes
• Maximum query length: 500 characters
• Response caching for queries under 5000 characters

Additional Documentation

For more details about the AI integration, check:

• Google Gemini AI Documentation
• Express Rate Limit
• Express Validator