Yet Another Discord Music Bot

I love using Discord, so why not make another generic music bot for personal use? I got really interested in the way the other bots out there handled commands, queries, and all sorts of interactions, so I got to researching and creating my own.

Description

A pet project coded in my free time using TypeScript. Uses ts-node as runtime and the discord.js v13 API for Discord v9. Integrates slash commands. User playlists are saved in Firebase.

This was originally written in JavaScript, but I got really fed up with loose typing (coming from a Java background). This project was then migrated over from JavaScript.

Credits

Credits to Zechariah Tan for coming up with a 'framework' for the bot.

Main Libraries

• discord.js
• @discordjs/voice
• @discordjs/builders
• [@discordjs/rest]
• [@discordjs/opus]
• [axios]
• ytdl-core
• [youtube-music-api]
• [youtube-dl-exec]
• [spotify-web-api-node]
• [firebase-admin]
• ts-node
• [express]

Features

• Play Youtube URLs, Spotify songs/albums/playlists or search YouTube with queries (in one command!)
• Loop, remove songs, swap songs, replace songs (all that jazz!)
• Adjustable audio quality (tested by an audiophile but no guaranteed change in quality!)
• Create, play and delete your own playlists - max 25 (admittedly useless but cool!)

Notes

Creating a track's audio resource can be extremely buggy when working with ytdl-core or ytdl. Most of the time, the track cuts out halfway and throws an error. This has been a known issue on ytdl's part and discordjs came up with a solution.

Here's an extract from a music bot example that they thankfully provided on GitHub:

export default class Track {
    /**
     * Creates an AudioResource from this Track.
     */
    public createAudioResource(): Promise<AudioResource<Track>> {
        return new Promise((resolve, reject) => {
            const process = ytdl(
                this.url,
                {
                    o: '-',
                    q: '',
                    f: 'bestaudio[ext=webm+acodec=opus+asr=48000]/bestaudio',
                    r: '100K',
                },
                { stdio: ['ignore', 'pipe', 'ignore'] },
            );
            if (!process.stdout) {
                reject(new Error('No stdout'));
                return;
            }
            const stream = process.stdout;
            const onError = (error: Error) => {
                if (!process.killed) process.kill();
                stream.resume();
                reject(error);
            };
            process
                .once('spawn', () => {
                    demuxProbe(stream)
                        .then(probe => resolve(createAudioResource(probe.stream, {
                            metadata: this,
                            inputType: probe.type
                        })))
                        .catch(onError);
                })
                .catch(onError);
        });
    }
}

From what I know, audio quality is significantly improved and the midway crashing is gone, so I greatly recommend using this.