tweetback Twitter Archive

Take ownership of your Twitter data. First talked about at Jamstack Conf 2019 and in this blog post.

• Referenced on Delete Twitter!

Demos

• https://www.zachleat.com/twitter/
• https://twitter.11ty.dev/

Features

• Built with Eleventy
• Each tweet has its own independent URL (with backwards/forwards threading!)
• Uses @tweetback/canonical to resolve other Twitter archives URLs (internal links stay in the archive and don’t link out to Twitter).
• t.co links are bypassed and original hyperlinks URLs are used.
• Links to users, tweets, non-truncated URLs.
• Nicer link formatting for links-to-tweets: @username/:id.
• Support some markdown: I sometimes use backtick markdown notation for code in my tweet text. This translates to <code> properly.
• Analytics:
  • See your most popular tweets
  • Who you retweet the most
  • Who you reply to the most
  • Frequently used swear words
  • Top emoji
  • Top hashtags

Usage

• Clone/download this repository
• In your terminal, cd to the folder of the project
• Install Node.js (there are many ways to do this depending on your system and preferences, but assuming you have homebrew, an easy installation method on Mac is to run brew install node)
• Run npm install

Populate the database from your Twitter Archive zip

1. Copy ./data/tweets.js (or ./data/tweet.js) from your Twitter Archive zip file into the ./database directory of this project.
   • Rename window.YTD.tweets.part0 (or window.YTD.tweet.part0) in tweets.js to module.exports
2. If you want to exclude Twitter Circles tweets (these are included in the archive, why 😭): copy ./data/twitter-circle-tweet.js from your Twitter Archive zip file into the ./database directory of this project.
   • Rename window.YTD.tweets.part0 or (window.YTD.tweet.part0) in twitter-circle-tweet.js to module.exports
3. Run npm run import or npm run import-without-circles

Build the web site

1. Edit the _data/metadata.js file to add metadata information.
   • username is your Twitter username (no @ symbol)
   • homeLabel and homeUrl are the the name and URL of a homepage or other link to include at the top of the main archive page
   • data.avatar should link to a 16x16 pixel profile picture; the default will use the favicon from the website at homeUrl
2. Run npm run build (will just create the proper files) or npm start (will run a server to look at them in your browser).
   • Optional: If you want the web site to live in a subdirectory (e.g. /twitter/), use Eleventy’s Path Prefix feature via the command line --pathprefix=twitter or via a return object in your configuration file. Careful: this is an option to Eleventy and not npm, so it needs to live after a -- separator (for instance, npm run build -- --pathprefix=twitter).

⚠️ Warning: the first build may take quite a long time (depending on the size of your archive), as remote media is fetched/downloaded into your project locally. There is currently no progress bar so be patient! Repeat builds will be much faster.

Fetch additional tweets from the API (optional)

Note: Twitter no longer offers free API access

If you want to fetch additional tweets from the API and put them into your sqlite database:

1. You will need a twitter developer token an a TWITTER_BEARER_TOKEN environment variable (from the Twitter API v2). Read more about App-only Bearer Tokens.
2. Run npm run fetch-new-data

Add your production URL to @tweetback/canonical (optional)

https://github.com/tweetback/tweetback-canonical has a mapping.js file that stores the existing twitter username → canonical URL mappings. These will be transformed automatically to point to other archives in all tweetback instances.

Publish your archive (optional)

• To GitHub Pages