tryhing whitliesting

assets/js/trackers.js

@@ -101,14 +101,24 @@ function generateTrackerPatternsArray() {
 
 
 // Actually strip out the tracking codes/parameters from a URL and return the cleansed URL
-function removeTrackersFromUrl(url) {
-  const urlPieces = url.split('?');
+function removeTrackersFromUrl(urlString) {
+
+  const url = new URL(urlString);
+
+  const {
+    search: searchString = '',
+    searchParams
+  } = url;
+
+
 
   // If no params, nothing to modify
-  if (urlPieces.length === 1) {
-    return url;
+  if (searchString === '') {
+    return urlString;
   }
 
+  const hostname = cleanHostname(url.hostname);
+
   // Go through all the pattern roots
   for (let root in TRACKER_REGEXES_BY_ROOT) {
     // If we see the root in the params part, then we should probably try to do some replacements
@@ -126,6 +136,22 @@ function removeTrackersFromUrl(url) {
   return urlPieces[1] ? urlPieces.join('?') : urlPieces[0];
 }
 
+// Clean up the hostname
+function cleanHostname(hostname) {
+  if (hostname.startsWith('www.')) {
+    return hostname.replace('www.', '');
+  }
+  return hostname;
+}
+
+function isWhitelistedDomain(host) {
+  for (whiteListedHost of whiteListedHosts) {
+    if ()
+  }
+
+  return false;
+}
+
 
 module.exports = {
   TRACKERS_BY_ROOT,

foo.js

@@ -0,0 +1,214 @@
+'use strict';
+
+const {URL} = require('url');
+
+'utm_source, *@*.google.com';
+
+function parseRules(rules) {
+
+  rules = rules.split(',');
+
+  for (let rule of rules) {
+    rule = rule.trim();
+
+    let split = rule.split('@');
+
+    const tracker = split[0];
+    const domain = split[1];
+
+    if (!(tracker || domain)) {
+      throw new Error('must have a tracker or a domain');
+    }
+
+    if (tracker == '*') {
+      if (!domain) {
+        throw new Error('must have a domain for a whitelist');
+      }
+
+      let startWith, endsWith, contains, equals;
+      if (domain.startWith('*.')) {
+        endsWith = true;
+      }
+      if (domain.endsWith('.*')) {
+        if (endsWith) {
+          startsWith = false;
+          endsWith = false;
+          contains = true;
+        }
+        else {
+
+        }
+      }
+    }
+  }
+}
+
+
+
+const TRACKERS = [];
+
+class DomainMatcher {
+  constructor(hostname, options) {
+
+    this.hostname = hostname;
+
+    ['startWith', 'endsWith', 'contains', 'equals'].forEach(type => {
+      this[type] = options[type] === true;
+    });
+
+    if (this.contains) {
+      this.containsPattern = new RegExp("^([\\w\\-]*\\.)*" + hostname.replace('.', '\.') + "(\\.[\\w\\-]+)*$");
+    }
+  }
+
+  matches(hostname) {
+    // console.log('testing', hostname);
+    if (this.startWith && hostname.startsWith(this.hostname)) {
+      return true;
+    }
+    if (this.equals || this.endsWith) {
+      if (this.hostname == hostname) return true;
+      if (hostname.endsWith('.' + this.hostname)) return true;
+    }
+    if (this.contains) {
+      console.log('does it contain?');
+      // if (hostname.test(this.containsPattern)) return true;
+      if (this.containsPattern.test(hostname)) return true;
+    }
+    return false;
+  }
+}
+
+const whiteListedDomainsByRoot = {
+  'google.com': new DomainMatcher('google.com', {startWith: true, endsWith: true, contains: true, equals: true})
+};
+
+const blackListedDomainsByTracker = {
+  'utm_source': [
+    new DomainMatcher('google.com', {startWith: true, endsWith: true, contains: true, equals: true})
+  ]
+};
+
+
+// Go through all the trackers by their root and turn them into a big regex...
+const TRACKER_REGEXES_BY_ROOT = {};
+for (let root in TRACKERS_BY_ROOT) {
+  // Old way, matching at the end 1 or unlimited times.
+  // TRACKER_REGEXES_BY_ROOT[root] = new RegExp("((^|&)" + root + "(" + TRACKERS_BY_ROOT[root].join('|') + ")=[^&#]+)", "ig");
+  // New way, matching at the end 0 or unlimited times. Hope this doesn't come back to be a problem.
+  TRACKER_REGEXES_BY_ROOT[root] = new RegExp("((^|&)" + root + "(" + TRACKERS_BY_ROOT[root].join('|') + ")=[^&#]*)", "ig");
+
+  TRACKERS_BY_ROOT[root].forEach(suffix => {
+    TRACKERS.push(root + suffix);
+  });
+}
+
+
+// Generate the URL patterns used for webRequest filtering
+// https://developer.chrome.com/extensions/match_patterns
+function generateTrackerPatternsArray() {
+  const array = [];
+  for (let root in TRACKERS_BY_ROOT) {
+    for (let i=0; i < TRACKERS_BY_ROOT[root].length; i++) {
+      array.push( "*://*/*?*" + root + TRACKERS_BY_ROOT[root][i] + "=*" );
+    }
+  }
+
+  return array;
+}
+
+
+// Actually strip out the tracking codes/parameters from a URL and return the cleansed URL
+function removeTrackersFromUrl(urlString) {
+
+  const url = new URL(urlString);
+
+  const {
+    search: searchString = '',
+    searchParams
+  } = url;
+
+
+
+  // If no params, nothing to modify
+  if (searchString === '') {
+    return urlString;
+  }
+
+  const hostname = cleanHostname(url.hostname);
+
+  const isWhite = isWhitelistedDomain(hostname);
+
+  console.log({
+    searchString,
+    searchParams,
+    hostname,
+    isWhite
+  });
+
+  if (isWhite) {
+    return urlString;
+  }
+
+  // return urlString;
+
+  // Go through all the pattern roots
+  for (let root in TRACKER_REGEXES_BY_ROOT) {
+    // If we see the root in the params part, then we should probably try to do some replacements
+    if (urlPieces[1].indexOf(root) !== -1) {
+      urlPieces[1] = urlPieces[1].replace(TRACKER_REGEXES_BY_ROOT[root], '');
+    }
+  }
+
+  // If we've collapsed the URL to the point where there's an '&' against the '?'
+  // then we need to get rid of that.
+  while (urlPieces[1].charAt(0) === '&') {
+    urlPieces[1] = urlPieces[1].substr(1);
+  }
+
+  return urlPieces[1] ? urlPieces.join('?') : urlPieces[0];
+}
+
+// Clean up the hostname
+function cleanHostname(hostname) {
+  if (hostname.startsWith('www.')) {
+    return hostname.replace('www.', '');
+  }
+  return hostname;
+}
+
+function isWhitelistedDomain(host) {
+  for (let root in whiteListedDomainsByRoot) {
+    if(whiteListedDomainsByRoot[root].matches(host)) return true;
+  }
+
+  return false;
+}
+
+const urls = [
+  'https://www.foo.com/path/to/nowhere.html?foo=bar&bar=baz',
+  'https://www.google.com/blah?foo=bar',
+  'https://mail.google.com/blah?foo=bar',
+  'https://mail.notgoogle.com/blah?foo=bar'
+];
+
+urls.forEach(url => {
+  console.log({url});
+  console.log(removeTrackersFromUrl(url));
+});
+
+// const regeExDomainChars = '[\\w\\-]';
+
+
+
+
+
+
+// const d = new DomainMatcher('google.com', {startWith: true, endsWith: true, contains: true, equals: true});
+
+// console.log({d});
+
+// ['google.com', 'mail.google.com', 'google.com.uk', 'mail.google.com.uk', 'mail.google.com..'].forEach(host => {
+//   console.log(host, d.matches(host));
+// });
+