User script to resolve t.co links on twitter

resolve_tco_links.user.js

// ==UserScript==
// @name        Resolve t.co links on twitter
// @namespace   Violentmonkey Scripts
// @match       *://twitter.com/*
// @grant       GM_xmlhttpRequest
// @version     1.1
// @author      @jix_
// @homepage    https://gist.github.com/jix/e67c127820954e1d1571c1f6f800ac2c
// @description Tries to resolve all annoying t.co links on twitter and replaces them with the
//              target URL.
// ==/UserScript==


// 2021-02-23: 1.1: Also handle http not only https. Handle initial DOM nodes. Slightly improved
//                  error handling/logging.
// 2021-02-17: 1.0: Initial version

const inFlight = {}; // Have just one request in flight per URL
const cached = {}; // Cache results until page reload


// Takes a 't.co' url, strips '?amp=1' if present and resolves it by making a GET request. With curl
// on the command line you can also use a HEAD request, but I didn't manage to make it work in here.
const resolveTcoUrl = function (url, callback) {
  url = url.replace('?amp=1', '');
  if (cached.hasOwnProperty(url)) {
    callback(cached[url]);
    return;
  }
  console.log(`resolving ${url}`);

  if (inFlight.hasOwnProperty(url)) {
    // Request already in flight, just push this callback to it
    inFlight[url].push(callback);
    return;
  } else {
    inFlight[url] = [callback];
  }

  GM_xmlhttpRequest({
    method: "GET",
    url: url,
    anonymous: true,
    onload: function (response) {
      if (response.status != 200) {
        console.log(`failed resolve for ${url}`);
        // We reset the callbacks so if the url appears again we can retry... untested though
        delete inFlight[url];
        return;
      }

      let target;

      try {
        // The resulting url is in a meta http eqiv tag, easy to parse with browser APIs
        const parsed = new DOMParser().parseFromString(response.responseText, 'text/html');
        const meta = parsed.querySelector('META[http-equiv="refresh"]');
        target = meta.content.replace(/^0;URL=/, '');

        cached[url] = target; // Cache until page reload
      } catch (e) {
        console.log([`parsing response for ${url} failed`, e]);
        delete inFlight[url];
        return;
      }

      const callbacks = inFlight[url]; // Get and reset all pending callbacks
      delete inFlight[url];
      for (const callback of callbacks) {
        try {
          callback(target); // And notify them
        } catch (e) {
          console.log([`callback after resolving ${url} failed`, e]);
        }
      }
    }
  });
};

// Resolve all <a href=...> contained in the passed DOM node
const replaceContainedLinks = function (subtree) {
  for (const linkNode of subtree.querySelectorAll('a')) {
    if (!linkNode.href.startsWith('https://t.co/') && !linkNode.href.startsWith('http://t.co/')) continue;
    resolveTcoUrl(linkNode.href, function (resolved) {
      linkNode.href = resolved;
    });
  }
}

// This gets called whenever twitter javascript adds new subtrees to the DOM which we'll scan for
// links.
const mutationCallback = function (mutationsList, observer) {
  // We scan all added subtrees for all <a> tags and check whether we need to rsolve them
  for (const mutation of mutationsList) {
    for (const addedSubtree of mutation.addedNodes) {
      replaceContainedLinks(addedSubtree)
    }
  }
};

// Install the mutation callback
new MutationObserver(mutationCallback).observe(document, { childList: true, subtree: true });

// Replace links in static content
replaceContainedLinks(document);