denice_twitter

a port of denice's chatter algorithm to twitter's api

behavior

denice will read the content of your twitter timeline, and create a markov-chain based dictionary from it. then, denice can post dumb reconstructed tweets based on the dictionary.

you will need to make a twitter api key (and client credentials) for the account denice will be using.

a demo instance is running as @denice_wetfish.

requirements

lua >=5.1
luasql (mysql)
luaoauth

config

edit denice.lua, and just fill in the sql and twitter credentials at the top of the file.

running

just call lua denice.lua <mode>.
choose from these 4 modes: fetch, post, test, and stats.

fetch will parse the latest tweets from your timeline into the database dictionary
post will post a tweet based on the content of the dictionary
test will print a candidate tweet to stdout
stats will print information about the content of the database

you can set up a cronjob to call 'fetch' and 'post' at whatever intervals you want, e.g. fetch every 5 minutes and post every 15 minutes.

contact