Semprini / blortbot

Twitchbot which implements natural landuage processing with wikipedia pages as corpus

blortbot project

Twitchbot which implements natural landuage processing with wikipedia pages as corpus

Docker

docker build -t blortbot .
docker run -e TWITCH_OAUTH_TOKEN='your token heere' --name blortbot_1 blortbot

You can also override these environment variables using more -e args when running:

• BOT_NAME (default = blortbot)
• BOT_CHANNEL (default = beginbot)

Code has 2 bots:

BlortBot

Extends BaseBot (see below) by overriding handle_direct_message and adding more commands:

• !blortbot: List all the magical things blortbot knows
• !qod: Quote of the day grabbed from quotes.rest
• !cookie: Get cookie monsters opinion
• !learn: Swap knowledge to new subject

Once a subject has been learned then you can message @blortbot to get answers.

For Example in twitch chat:

 > !learn kung fu
blortbot: I know kung fu
 > @blortbot who was bruce lee's teacher
blortbot: yip man was the teacher of [[bruce lee]].

BlortBot does have dependencies so use:

pip install -r requirements.txt

And needs an environment variable:

• TWITCH_OAUTH_TOKEN # Get the value for this here: https://twitchapps.com/tmi/

BlortBot uses natural language processing and cosine similarity to respond with a vaguely intellegible answer.

More info: https://scikit-learn.org/stable/modules/metrics.html#cosine-similarity

BaseBot

An extensible bot which recognises !hello messages and responds with hello {user}. Oo'd from https://gitlab.com/beginbot/python_twitch_bot_no_deps

Uses only standard library so no need to install dependencies

To run you need to set 3 environment variables:

• BOT_NAME
• TWITCH_OAUTH_TOKEN
• BOT_CHANNEL

Then run:

python basebot.py

To add commands you annotate a function with the command decorator:

@command('hello', 'responds with hello')
def command_hello(bot, user, msg):
    ...