Discord Bot

A Spring Boot bot that bridges a Discord server and a Telegram chat. It plays audio into voice channels, speaks text out loud, records what happens in a channel and forwards the recording to Telegram — and it can be driven from either side: Discord commands or Telegram commands.

---

What it does

• Two-way Discord ↔ Telegram bridge. A Telegram command can make the bot speak or play audio in a Discord voice channel; a Discord command can push a recording or a message into Telegram.
• Soundboard. Any .mp3 in the audio folder becomes a command automatically — drop csgo.mp3 in and !csgo plays it. No code change, no restart of the command registry.
• Text-to-speech. !speak <channel> <text> synthesises speech with gTTS and streams it into the voice channel through LavaPlayer.
• Voice recording. !record captures a voice channel to a file, !stop ends the session and delivers the audio to the configured Telegram chat.
• Voice channel moderation. Removes users from a channel after playing a message, with a persisted block list of users and roles.
• Web admin panel. A Thymeleaf UI at /admin to trigger playback, send TTS, and manage blocked users and roles without touching Discord.

---

Commands

Discord

Command	Description	Example
!<filename>	Plays <filename>.mp3 from the audio folder in your current or a named voice channel	!csgo · !alo General
!speak <channel> <text>	Converts text to speech and plays it in the channel	!speak General Hello, how are you?
!record <channel>	Starts recording the voice channel	!record General
!stop	Stops recording and sends the file to Telegram	!stop
!delete <channel>	Plays a message, then removes users from the channel	!delete General
!case	Posts "Who will go for the case?" to both Discord and Telegram	!case

Unknown !command names fall through to the soundboard: the bot looks for command.mp3 and reports back if the file is missing.

Telegram

Command	Description
/vcinfo	Lists who is currently sitting in which voice channel
/sendvoice <channel>	Sends a voice message into the given Discord channel
/case	Posts "Who will go for the case?" to both sides

---

Tech stack

Area	Choice
Runtime	Java 17, Spring Boot 3.4, Gradle
Discord	JDA 5 + LavaPlayer for audio
Telegram	Long-polling client (TelegramBotPoller)
Speech	gtts4j (Google Text-to-Speech)
Web	Spring MVC + Thymeleaf admin UI, Spring Security, springdoc OpenAPI
Storage	Spring Data JPA over H2 — blocked users and roles
Media	ffmpeg (installed in the runtime image)

---

Architecture

src/main/java/com/alakey/discordbot/
├── discordbot/
│   ├── config/JdaConfig.java           # JDA client bootstrap
│   ├── command/                        # one class per command, dispatched by CommandManager
│   │   ├── CommandManager.java         #   routes "!x" → command, else → soundboard lookup
│   │   ├── SpeakCommand.java
│   │   ├── RecordCommand.java / SimpleAudioRecorder.java
│   │   ├── StopCommand.java
│   │   ├── PlayFileCommand.java
│   │   └── CaseCommand.java
│   ├── audio/AudioPlayerSendHandler.java  # LavaPlayer → JDA audio bridge
│   ├── listener/VoiceChannelListener.java
│   └── session/RecordingSession.java
├── telegrambot/TelegramBotPoller.java  # Telegram long polling, mirrors commands
├── service/                            # AudioService, TextToSpeechService, GuildService,
│                                       # VoiceChannelInfoService, BlockedEntityService, cache
├── controller/AdminController.java     # web admin panel
├── model/ + repository/                # BlockedUser, BlockedRole (JPA)
└── config/SecurityConfig.java

CommandManager is the single dispatch point: explicit commands are registered in a map, and anything unmatched is resolved as a soundboard file. That is why adding a sound requires no code.

---

Configuration

The application reads everything from environment variables — it will not start without them:

Variable	Purpose
DISCORD_BOT_TOKEN	Discord bot token
DISCORD_GUILD_BY_ID	ID of the guild (server) the bot serves
TELEGRAM_BOT_TOKEN	Telegram bot token
TELEGRAM_CHAT_ID	Chat that receives recordings and messages
AUDIO_FOLDER	Path to the .mp3 soundboard folder
USERNAME / PASSWORD	Credentials for the web admin panel

---

Running

Docker (recommended)

docker pull ialakey/discordbot:latest

docker run -p 8080:8080 \
  -e DISCORD_BOT_TOKEN=... \
  -e DISCORD_GUILD_BY_ID=... \
  -e TELEGRAM_BOT_TOKEN=... \
  -e TELEGRAM_CHAT_ID=... \
  -e AUDIO_FOLDER=/app/audio \
  -e USERNAME=admin \
  -e PASSWORD=secret \
  ialakey/discordbot:latest

The image is a two-stage build: Gradle produces the boot jar, then a slim JDK 17 image adds ffmpeg and the bundled audio folder.

Building it yourself

git clone https://github.com/ialakey/discordbot.git
cd discordbot
docker build -t discordbot .

From source

./gradlew bootRun

Requires a local ffmpeg installation.

---

Web interface

URL	What
http://localhost:8080/admin	Admin panel — play audio, send TTS, kick from a channel, manage blocked users and roles
http://localhost:8080/swagger-ui/index.html	OpenAPI documentation

The panel is protected by Spring Security using USERNAME / PASSWORD.

---

Author

Ilia Alakov — Telegram · LinkedIn