First version

Just log things. No visualization yet
zenazn · Nov 29, 2011 · 4484648 · 4484648
1 parent 81f19a0
commit 4484648
Show file tree

Hide file tree

Showing 7 changed files with 129 additions and 0 deletions.
diff --git a/db.js b/db.js
@@ -0,0 +1,42 @@
+var db = openDatabase(
+  'referrer_log',
+  '',
+  "A log of referrer data you leak to third parties on the web",
+  64 * 1024 * 1024 // 64 megs
+);
+
+db.changeVersion('', '1', function(tx) {
+  tx.executeSql(
+    "CREATE TABLE requests (" +
+    "  id INTEGER PRIMARY KEY AUTOINCREMENT," +
+    "  created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP," +
+    "  req_url TEXT," +
+    "  req_domain TEXT," +
+    "  req_root TEXT," +
+    "  ref_url TEXT," +
+    "  ref_domain TEXT," +
+    "  ref_root TEXT" +
+    ")"
+  );
+  ["req_domain", "req_root", "ref_domain", "ref_root"].forEach(function(idx) {
+    tx.executeSql(
+      "CREATE INDEX " + idx + "_idx ON requests (" + idx + ")"
+    );
+  });
+});
+
+function db_insert_request(req, ref) {
+  db.transaction(function(tx) {
+    tx.executeSql(
+      "INSERT INTO requests (req_url, req_domain, req_root, ref_url, " +
+        "ref_domain, ref_root) VALUES (?, ?, ?, ?, ?, ?)",
+      [req.href, req.host, req.root, ref.href, ref.host, ref.root]
+    );
+  });
+}
+
+function db_reset() {
+  db.transaction(function(tx) {
+    tx.executeSql("DELETE FROM requests");
+  });
+}
diff --git a/images/icon.png b/images/icon.png
diff --git a/inject.js b/inject.js
@@ -0,0 +1,19 @@
+function hook(e) {
+  // We're only interested in http (and https)
+  if (window.location.protocol.indexOf("http") == -1) {
+    return;
+  }
+
+  var req = parse_url(e.url), ref = parse_url(document.location.href);
+
+  // We're only really interested in cross-domain requests. This is some sort of
+  // poor approximation of same-origin. ish.
+  if (req.root == ref.root) {
+    return;
+  }
+
+  // Log it!
+  chrome.extension.sendRequest({'msg': 'log', 'req': req, 'ref': ref});
+}
+
+document.addEventListener('beforeload', hook, true);
diff --git a/logger.html b/logger.html
@@ -0,0 +1,5 @@
+<html>
+  <script type="text/javascript" src="db.js"></script>
+  <script type="text/javascript" src="util.js"></script>
+  <script type="text/javascript" src="logger.js"></script>
+</html>
diff --git a/logger.js b/logger.js
@@ -0,0 +1,26 @@
+function log(msg, cb) {
+  // We actually do a bit more filtering on this end. We only want requests
+  // which have cookies attached (CDN URLs don't count, for instance), and we
+  // don't have access to the cookie API from the injected script
+  chrome.cookies.getAll({url: msg.req.href}, function(cookies) {
+    if (cookies && cookies.length > 0) {
+      // We're sending cookies. Potential spying. Rut roh!
+      // Turns out you can get a lot of data from a request even if explicit
+      // tracking cookies aren't involved. However, since that's considerably
+      // more complex (from the service provider's point of view) and since
+      // doing the traditional tracker cookie thing is so easy, we'll assume
+      // that's what everyone is doing. This is probably Mostly Correct (tm)
+      db_insert_request(msg.req, msg.ref);
+    }
+  });
+}
+
+var handlers = {
+  'log': log
+};
+
+chrome.extension.onRequest.addListener(function(req, sender, cb) {
+  // Simple dispatch. It's super effective!
+  handlers[req.msg](req, cb);
+});
+
diff --git a/manifest.json b/manifest.json
@@ -0,0 +1,21 @@
+{
+  "name": "Double Cross",
+  "version": "0.1.0",
+  "description": "Spy on the websites that spy on your web activity",
+  "browser_action": {
+    "default_icon": "images/icon.png"
+  },
+  "permissions": [
+    "<all_urls>",
+    "background",
+    "cookies",
+    "unlimitedStorage"
+  ],
+  "background_page": "logger.html",
+  "content_scripts": [{
+    "matches": ["http://*/*", "https://*/*"],
+    "js": ["util.js", "inject.js"],
+    "run_at": "document_start",
+    "all_frames": true
+  }]
+}
diff --git a/util.js b/util.js
@@ -0,0 +1,16 @@
+// This probably isn't particularly rigorous. We're trying to go from
+// google.com -> google.com
+// news.google.com -> google.com
+var root_re = /((?:[^.]+\.)?[^.]+)$/;
+
+function parse_url(url) {
+  // This is a hack. I don't want to parse the URL, so make a fake <a> instead
+  var a = document.createElement('a');
+  a.href = url;
+  return {
+    "href": a.href,
+    "host": a.host,
+    "root": root_re.exec(a.host)[1],
+    "protocol": a.protocol
+  };
+}