Harold is a twitter bot which can manage tweeting from feeds at specified intervals, keeping your timeline groomed by removing tweets after a certain amount of time, of even removing all of your tweets or DMs entirely.
Harold takes a config file in TOML format (with -config command line flag), and whether or not you want to run in botmode, deletedm, or deletetimeline.
harold -botmode -config=/path/to/config.toml
You will need to create a new Twitter application and generate API keys. Harold assumes the following environment variables are set:
TWITTER_CONSUMER_KEY
TWITTER_CONSUMER_SECRET
TWITTER_ACCESS_TOKEN
TWITTER_ACCESS_TOKEN_SECRET
One way to do this would be to get the API keys and then put them into a file, which you can then source before running Harold.
export TWITTER_CONSUMER_KEY=xxxxx
.
.
.
Then just source keys.env or whatever you have named the file containing the keys, or you will get a response from Twitter saying you are unauthorized.
The config file has a few different parameters, and an example is below.
cleanTimeline = true
maxTweetAge = "24h"
cleanTimelineEvery = "1h"
[feeds]
[feeds.aadrakecom]
url = "https://aadrake.com/index.xml"
hashTags = "#tech #datascience #ai #ml #techto"
maxTweetInterval = "12h"
[feeds.datamultireddit]
url = "https://www.reddit.com/user/adrake/m/data/.rss"
hashTags = "#data #bigdata #ai #ml"
maxTweetInterval = "12h"
[feeds.datatau]
url = "http://www.datatau.com/rss"
hashTags = "#data #bigdata #ai #ml #datascience"
maxTweetInterval = "12h"
[feeds.cryptocurrencynews]
url = "https://cryptocurrencynews.com/feed/"
hashTags = "#cryptocurrency #blockchain #btc #eth #xrp #xrb"
maxTweetInterval = "12h"
cleanTimeline determines whether or not Harold will remove tweets older than maxTweetAge from your timeline when running in botmode, which will be done at frequency cleanTimelineEvery.
Each entry in the feeds section takes a url, a list of hashtags in the hashTags parameter which you want appended to tweets from that feed, and a maxTweetInterval, which is the longest time between tweets for that feed. The frequency of tweets from a given feed is uniformly distributed between now and now + maxTweetInterval.