The Quote Command for Botyo uses Markov chains to generate quotes based on messages sent by the participants of a chat thread.
#quote [ <person> | me | all [ on <subject> ] ]
For example:
#quote Alice
- Generates a quote by Alice.#quote me
or simply#quote
- Quotes the sender of the message.#quote all
or#quote *
- Builds a Markov chain based on all but the bot's messages in the chat, and generates an anonymous quote.#quote me on JavaScript
- Generates a quote by the sender starting with 'JavaScript'.#quote Alice on cats
- Generates a quote by Alice starting with 'cats'.#quote all on the bot
- Generates a quote based on all messages, starting with 'the bot'.
- Persistence Bundle - must be configured and enabled to store the messages in the database, since the Quote Command generates sentences based on messages in DB in order to avoid downloading the entire chat history on every command invocation.
Step 1. Install the module from npm.
npm install --save botyo-command-quote
Step 2. Register the module.
import Botyo from "botyo";
import QuoteCommand from "botyo-command-quote"
Botyo.builder()
...
.registerModule(QuoteCommand)
...
.build()
.start();
The configuration of the module has sensible defaults and doesn't need to be explicitly configured. However, you can still override the defaults in your configuration file or upon module registration.
modules:
QuoteCommand:
enable: true
markov:
vom: true # whether to build a variable-order Markov model
order: 2 # order of the Markov model
maxWordCount: 20 # limit for the generated sentence's word count
censorship:
enable: false # whether to censor (i.e. not send) sentences matching the following regex, and generate new ones instead
regex: "/badword|worseword/gi" # JavaScript regex
maxRetries: 20 # how many times to retry generating new sentences until finding one that doesn't get censored