☰

Link Untracker

Remove tracking elements from links

  1. // ==UserScript==
  2. // @name        Link Untracker
  3. // @namespace   IzzySoft
  4. // @description Remove tracking elements from links
  5. // @license     CC BY-NC-SA
  6. // @include     *
  7. // @exclude     *phpmyadmin*
  8. // @version     13
  9. // @run-at      document-idle
  10. // @grant       unsafeWindow
  11. // @homepage    https://codeberg.org/izzy/userscripts
  12. // ==/UserScript==
  13.  
  14. /* badp: we strip parameters starting with that. Some are taken from NetURL, see
  15.    https://github.com/Smile4ever/firefoxaddons/tree/master/Neat%20URL-webext
  16.    fb_: Facebook; ga_/_gl: Google; hmb_: HumbleBundle(?); utm_: Urchin Tracker; wt_: Webtrekk; WT.: Webtrends
  17. */
  18. var badp = ['fb_','ga_','_ga','_gl','hmb_','utm_',
  19.             'ei@google.','gws_rd@google.','sei@google.','ved@google.',
  20.             'pd_rd_r@amazon.','pd_rd_w@amazon.','pd_rd_wg@amazon.','psc@amazon.','_encoding@amazon.',
  21.             'wt_','WT.','yclid','referrer','clickfrom','pf_rd'
  22.            ];
  23. var anch = ''; /* variable to hold the "anchor part" of the URL, if any – i.e. "#anchor" */
  24. var replace_semicolon = true; /* some sites use ";" to separate URL params;
  25.                                  URLSearchParams can't deal with that and messes up.
  26.                                  If those sites don't work with "&", either set this
  27.                                  to "false" or add that site on the exclude list
  28.                                */
  29.  
  30. function getRealLinkFromGoogleUrl(a) { /* adapted from https://github.com/Rob--W/dont-track-me-google at 2018-01-03 */
  31.     if ((a.hostname === location.hostname || a.hostname === 'www.google.com') &&
  32.         /^\/(local_)?url$/.test(a.pathname)) {
  33.         // Google Maps / Dito (/local_url?q=<url>)
  34.         // Mobile (/url?q=<url>)
  35.         var url = /[?&](?:q|url)=((?:https?|ftp)[%:][^&]+)/.exec(a.search);
  36.         if (url) {
  37.             return decodeURIComponent(url[1]);
  38.         }
  39.         // Help pages, e.g. safe browsing (/url?...&q=%2Fsupport%2Fanswer...)
  40.         url = /[?&](?:q|url)=((?:%2[Ff]|\/)[^&]+)/.exec(a.search);
  41.         if (url) {
  42.             return a.origin + decodeURIComponent(url[1]);
  43.         }
  44.     }
  45.     return a.href;
  46. }
  47.  
  48. class UrlParams { /* adapted from https://stackoverflow.com/a/45516670 as the built-in URLSearchParams messes up parameters without values */
  49.   constructor(search) { /* extracts the query string by stripping the leading "?" and initializes properties */
  50.     this.qs = (search || location.search).substr(1);
  51.     this.params = {};
  52.     this.parseQuerstring();
  53.   }
  54.   parseQuerstring() { /* walks the query string and extracts the parameters as key-value-pairs */
  55.     this.qs.split('&').reduce((a, b) => {
  56.       let [key, val] = b.split('=');
  57.       a[key] = val;
  58.       return a;
  59.     }, this.params);
  60.   }
  61.   filterQuerystring(badparms) { /* eliminates all "bad parameters" (tracking parameters) from our key-value-pairs */
  62.     var keysToDelete = [];
  63.     for (var key of Object.keys(this.params)) {
  64.       for (let p of badparms) {
  65.         if (key.startsWith(p)) console.log('removing '+key+' from '+this.qs);
  66.         if (key.startsWith(p)) { keysToDelete.push(key); break; }
  67.       }
  68.     }
  69.     for(var key of keysToDelete) {
  70.       delete this.params[key];
  71.     }
  72.   }
  73.   getString() { /* concatenates all parameters and returns them as a single query string without the leading "?" */
  74.     var str = '';
  75.     for (var key of Object.keys(this.params)) {
  76.       str = str + '&' + key;
  77.       if (this.params[key] != undefined) str = str + '=' + this.params[key]; /* only add the "=value" part if the parameter has a value! */
  78.     }
  79.     return str.substring(1);
  80.   }
  81. }
  82.  
  83. // MetaGer.de uses different methods of obfuscation with tracking URLs. This is catching most of them (hopefully):
  84. function metager1(elem) {
  85.   if ( elem.hostname == 'api.smartredirect.de' ) {
  86.     var purl = new UrlParams(elem.search);
  87.     elem.href = decodeURIComponent(purl.params['url']);
  88.   } else if ( elem.href.startsWith ('https://metager.de/partner/r')) {
  89.     var purl = new UrlParams(elem.search);
  90.     elem.href = decodeURIComponent(purl.params['link']);
  91.   } else if ( elem.href.startsWith ('https://metager.de/r/metager')) {
  92.     if (elem.text.trim().startsWith('http:') || elem.text.trim().startsWith('https:')) {
  93.       elem.href = decodeURI(elem.text.trim());
  94.     } else {
  95.       elem.href = 'https://'+decodeURI(elem.text.trim());
  96.     }
  97.   }
  98. }
  99.  
  100. function parseLinks(k) {
  101.   for(var i = k; i < document.links.length; i++) { // for some reason stops at 100 loops
  102.     if (i - k > 98) return i;
  103.     var elem = document.links[i];
  104.     if (window.location.hostname == 'metager.de' && elem.className.includes('result-link')) { metager1(elem); continue; }
  105.     if (elem.search == '') continue;
  106.     if (replace_semicolon) var purl = new UrlParams(elem.search.replace(new RegExp(';','g'),'&'));
  107.     else var purl = new UrlParams(elem.search);
  108.     purl.filterQuerystring(badp);
  109.     if (elem.href.indexOf('#') > 0) anch = '#' + elem.href.split('#')[1];
  110.     else anch = '';
  111.     elem.href = elem.href.split('?')[0] + '?' + purl.getString() + anch;
  112.     elem.href = getRealLinkFromGoogleUrl(elem);
  113.   }
  114.   return i;
  115. }
  116.  
  117. console.log(document.links.length+' links to parse');
  118. k = 0; while ( k < document.links.length) {
  119.   console.log('parsing from '+k);
  120.   k = parseLinks(k+1);
  121.   console.log('returned '+k)
  122. }
  123. console.log('DONE');