bot_kette

A simple Discord bot that logs all messages to the console and can respond to specific trigger words.

Features

• Logs all messages from Discord servers to the console
• Logs attachments with their URLs
• Responds to configured trigger words with custom responses
• Supports different word matching modes (whole word, starts with, ends with)
• Containerized with Docker for easy deployment
• GitHub Actions workflow for automatic container publishing

Requirements

• Node.js 22.0.0 or higher
• Discord.js v14
• A Discord bot token

Setup

Local Development

1. Clone this repository
2. Install dependencies:

   npm install
   

3. Create a .env file with your Discord bot token:

   DISCORD_TOKEN=your_token_here
   

4. Start the bot:

   npm start
   

Using Docker

Option 1: Pull from GitHub Container Registry

# Pull the latest image
docker pull ghcr.io/Flyy-y/bot_kette:latest

# Run with your Discord token
docker run -d --name bot_kette -e DISCORD_TOKEN=your_token_here ghcr.io/Flyy-y/bot_kette:latest

Option 2: Using Docker Compose

1. Clone this repository
2. Update the Discord token in compose.yml
3. Run with Docker Compose:

   docker compose up -d
   

Example compose.yml:

services:
  bot:
    image: ghcr.io/Flyy-y/bot_kette:latest
    # Alternatively, build locally:
    # build: .
    container_name: bot_kette
    restart: unless-stopped
    environment:
      - DISCORD_TOKEN=your_token_here
    volumes:
      # Mount custom answerMap.json (optional)
      - ./answerMap.json:/app/answerMap.json:ro

Option 3: Build Locally

1. Clone this repository
2. Build the Docker image:

   docker build -t bot_kette .
   

3. Run the container:

   docker run -d --name bot_kette -e DISCORD_TOKEN=your_token_here bot_kette
   

Configuration

The bot uses an answerMap.json file to configure trigger words and responses. The format is:

{
  "trigger1": {"answer": "response1", "on": "endsWith"},
  "trigger2": {"answer": "response2", "on": "startsWith"},
  "trigger3": {"answer": "response3", "on": "always"},
  "trigger4": {"answer": ["response4a", "response4b"], "on": "always", "secondaryMatches": ["alt1", "alt2"]}
}

Response Types

The bot supports both single responses and arrays of possible responses:

• Single Response: Provide a string as the answer value

  • Example: "trigger": {"answer": "response", "on": "always"}

• Multiple Responses: Provide an array of strings as the answer value

  • Example: "trigger": {"answer": ["response1", "response2", "response3"], "on": "always"}
  • When triggered, the bot will randomly select one of the responses

Secondary Matches

You can define alternative words that will also trigger the same response:

• Use the secondaryMatches key with an array of alternative trigger words
• Example: "ca": {"answer": "lope", "on": "always", "secondaryMatches": ["sa", "ça"]}
• This will trigger the same response for "ca", "sa", or "ça"

Multiple Matches Handling

When a message contains multiple trigger words:

• The bot will collect all matching responses
• Responses are kept in the same order as they appear in the original message
• Responses are joined with " + "
• If there are 3 or more matches, "+ ratio" is added at the end
• Example: For a message containing "oui c'est quoi ça", the response would be "stiti + feur + lope + ratio"

Matching Modes

The bot supports three different matching modes:

• endsWith: Triggers when the message ends with the specified word

  • Example: With "oui": {"answer": "stiti", "on": "endsWith"}, the message "ça va oui" will trigger the response

• startsWith: Triggers when the message starts with the specified word

  • Example: With "hello": {"answer": "world", "on": "startsWith"}, the message "hello everyone" will trigger the response

• always: Triggers when the word appears anywhere in the message as a whole word (not as part of another word)

  • Example: With "thanks": {"answer": "you're welcome", "on": "always"}, the message "thanks for your help" will trigger the response, but "thanksgiving" will not

The bot only checks for whole words, so partial matches like "jouir" will not trigger the "oui" response.

Accent Handling

The bot automatically removes accents from words before matching:

• Accented characters are normalized to their non-accented equivalents
• For example, "ça" will match "ca" and vice versa
• This works for all matching modes (always, startsWith, endsWith)
• Example: With "ca": {"answer": "lope", "on": "always"}, the message "ça va bien" will trigger the response

Testing

This project includes a comprehensive test suite using Jest. The tests cover:

• Utility functions for string matching (containsWholeWord, startsWithWord, endsWithWord)
• answerMap loading functionality
• Message handling and response logic

Running Tests

Local Development

# Install dependencies including dev dependencies
npm install

# Run tests
npm test

# Run tests with coverage report
npm run test:coverage

Using Docker

# Build and run tests using Docker
docker build -t bot_kette_test -f Dockerfile.test .
docker run bot_kette_test

Test Status

The current test coverage is:

File	% Statements	% Branches	% Functions	% Lines
All files	85.18	87.5	50	84.9
index.js	82.97	83.33	20	82.6
utils.js	100	100	100	100

GitHub Actions

This repository includes GitHub Actions workflows that automatically:

1. Run Tests Workflow

• Builds a Docker test image
• Runs all tests inside the Docker container
• Generates and uploads test coverage reports
• Automatically comments on PRs with test results if tests fail
• Prevents Docker image building if tests fail
• Runs on every push to main and add-unit-tests branches, pull requests to main, and manual triggering

2. Docker Publish Workflow

• Builds the Docker image
• Tags it with the version from package.json
• Publishes it to GitHub Container Registry (ghcr.io)
• Runs on every push, version tags (v*..), pull requests to main, and manual triggering

The image will be available at ghcr.io/Flyy-y/bot_kette:latest and ghcr.io/Flyy-y/bot_kette:1.0.0 (or whatever version is in package.json).

License

This project is licensed under the MIT License - see the LICENSE file for details.