This website is automatically built for my personal interest only. The basic ideas:
- Search Twitter messages through predefined keywords using the R package rtweet on Travis CI on a daily basis;
- Convert the results to Markdown files and push to Github;
- Build the website using Hugo automatically on Netlify.
Then I can subscribe to the RSS feed through an RSS reader, and know if anybody was trying to praise me in a corner yesterday.
You are welcome to fork this project on Github, and use it by yourself. Some technical notes in case you want to do that:
- Learn a bit about rtweet, especially how to get your own access token;
- After you save the token to an
twitter_token.rds, follow the Travis guide to encrypt it and tweak your
- Tweak the keywords I defined for myself;
- Define an environment variable
GITHUB_PATon Travis, which is a Github personal access token that should have write access to your repo, because the Markdown files will need to be pushed to your repo from Travis;
- Connect your Github repo with Netlify, so that whenever there are new commits in your repo, Netlify will automatically rebuild your site. The build command can be
hugo_0.19, and the publish dir is
Then sit back and relax. Travis and Netlify will do the hard work for you. You may have a good or bad day tomorrow after you read the tweets. Who knows. Social network is a fantasy, filled with isolated and lonely people. You’d better not spend too much time there.
Well, you may ask, “Hey, why did you build this website then?” Good question. I just wanted to learn a bit about the rtweet package. I have never tried to scrape data from Twitter before. Since I’m working on blogdown recently, I thought it might be a good exercise to build a website automatically.