DJALLIN2

A mildly configurable twitch noisemaker capable of responding to chat and point redemptions.

Available through pip (djallin2) or as a standalone executable for macOS and Windows.

Note for Windows users, your antivirus will probably think it's a trojan. It's not, but it kinda looks like one to your antivirus. If you get the option to report a false positive, please do so.

Quickstart

Want !sound to just work?

• Download the latest build from the releases tab
• Download the quick config file, rename it to config.txt
• Place your sounds in a folder called sounds, next to the app and config
• Run it, your browser will open to handle the initial Twitch connection
  • This will happen every ~3 months as necessary

Configuration

config.txt holds your token and configures chat and point responders.

Take a look at the example configuration for examples, and a detailed explanation for all settings.

The following block defines a basic chat listener:

[chat.sound]
    badges = ['moderator', 'broadcaster', 'vip']
    target = './sounds/'
    command = '!sound '

Mods, VIPs, and the broadcaster can use !sound xxx to play ./sounds/xxx.mp3

If badge and name filters are not supplied, anyone can use the command. The random setting enables random playback. A value of 2 selects a random file from the directory.

[chat.wisdom]
    target = './wisdom/'
    command = '!wisdom'
    random = 2

Points rewards are similarly configured. Set the name to the name configured in your twitch rewards. A random value of 1 enables the random command where a random file will be selected if a user enters random.

[points.sound]
    name = 'play a sound'
    target = './sounds/'
    random = 1

Alternatively, you can duplicate configurations using links. Setting applied in the block with a link statement override settings in the linked responder.

[points.sound]
    link = 'chat.sound'
    random = 1

command_mode can be used to define different match modes for the command. contains allows you to play a single file or randomly select from a directory when a message contains the command phrase. Chat listeners are checked in alphabetical order and when one plays a sound, checks stop.

[chat.z_cowboy]
    command = 'cowboy'
    command_mode = 'contains'
    target = './sound/yeehaw.mp3'

Both point and chat blocks support custom code actions, check the example configuration for details.

Playback statistics can be recorded to a sqlite3 database of your choice. Simply add stats = true to your configuration block.

[chat.sound]
    target = './sounds/'
    command = '!sound '
    stats = true

The name of the database can be changed by setting stats_db = 'database.sqlite' at the top of your configuration file.

Known issues

If a red screen shows when registering your twitch token, there may be a problem with your connection. Close the soundbot and try again.

TODO:

1. This
2. Let exit on fire be a configurable option?
3. HTTP command receiver (lioran?)
4. custom hook code - lioran example? obs websocket?
5. Bits?
6. Does linux support audio these days?
7. https://inloop.github.io/sqlite-viewer/