FastAPI LINE Bot with Gemini AI and Firebase Integration

This project is a FastAPI application that integrates with LINE Messaging API, Gemini AI, and Firebase to handle various types of messages and provide intelligent responses.

Features

• Health Check Endpoint: Simple endpoint to check if the service is running.
• LINE Webhook Handler: Handles incoming messages from LINE and responds accordingly.
• Gemini AI Integration: Uses Gemini AI to process and generate responses based on the content of the messages.
• Firebase Integration: Stores and retrieves chat history from Firebase.

Prerequisites

• Python 3.7+
• LINE Messaging API account
• Gemini AI API key
• Firebase project
• .env file with the following environment variables:
  • API_ENV
  • LINE_CHANNEL_SECRET
  • LINE_CHANNEL_ACCESS_TOKEN
  • LOG
  • FIREBASE_URL
  • GEMINI_API_KEY
  • OPEN_API_KEY

Installation

1. Clone the repository:

   git clone <repository_url>
   cd <repository_directory>

2. Create and activate a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install the dependencies:

   pip install -r requirements.txt

4. Create a .env file in the root directory and add the required environment variables.

Usage

1. Run the FastAPI application:

   uvicorn main:app --host 0.0.0.0 --port 8080 --reload

2. The application will start and listen for incoming requests on the specified port.

Endpoints

• GET /health: Health check endpoint to verify if the service is running.
• POST /webhooks/line: Webhook endpoint to handle incoming messages from LINE.

Environment Variables

• API_ENV: Set to production or develop.
• LINE_CHANNEL_SECRET: Your LINE channel secret.
• LINE_CHANNEL_ACCESS_TOKEN: Your LINE channel access token.
• LOG: Logging level (default is WARNING).
• FIREBASE_URL: Your Firebase database URL.
• GEMINI_API_KEY: Your Gemini AI API key.
• OPEN_API_KEY: Your Open Data API key.

Logging

The application uses Python's built-in logging module. The log level can be set using the LOG environment variable.

License

This project is licensed under the MIT License.