Properly fix #622 by escaping input into DOM filters

index.html

@@ -21,7 +21,7 @@
     <script src="//ajax.googleapis.com/ajax/libs/angularjs/1.3.14/angular-sanitize.min.js"></script>
     <script src="//ajax.googleapis.com/ajax/libs/angularjs/1.3.14/angular-touch.min.js"></script>
     <script src="//cdnjs.cloudflare.com/ajax/libs/underscore.js/1.7.0/underscore-min.js"></script>
-    <script src="//cdnjs.cloudflare.com/ajax/libs/emojione/1.3.0/lib/js/emojione.min.js"></script>
+    <script src="//cdnjs.cloudflare.com/ajax/libs/emojione/1.4.0/lib/js/emojione.min.js"></script>
     <script type="text/javascript" src="3rdparty/inflate.min.js"></script>
     <script type="text/javascript" src="js/localstorage.js"></script>
     <script type="text/javascript" src="js/weechat.js"></script>
@@ -284,7 +284,7 @@ $ openssl req -nodes -newkey rsa:4096 -keyout relay.pem -x509 -days 365 -out rel
               <td class="prefix"><a ng-click="addMention(bufferline.prefix)"><span class="hidden-bracket">&lt;</span><span ng-repeat="part in ::bufferline.prefix" ng-class="::part.classes" ng-bind="::part.text|prefixlimit:25"></span><span class="hidden-bracket">&gt;</span></a></td><!--
            --><td class="message"><!--
              --><div ng-repeat="metadata in ::bufferline.metadata" plugin data="::metadata"></div><!--
-             --><span ng-repeat="part in ::bufferline.content" class="text" ng-class="::part.classes.concat(['line-' + part.$$hashKey.replace(':','_')])" ng-bind-html="::part.text | escape | linky:'_blank' | DOMfilter:'irclinky' | DOMfilter:'emojify':settings.enableJSEmoji | DOMfilter:'inlinecolour' | DOMfilter:'mathjax':('.line-' + part.$$hashKey.replace(':','_')):settings.enableMathjax"></span>
+             --><span ng-repeat="part in ::bufferline.content" class="text" ng-class="::part.classes.concat(['line-' + part.$$hashKey.replace(':','_')])" ng-bind-html="::part.text | linky:'_blank' | DOMfilter:'irclinky' | DOMfilter:'emojify':settings.enableJSEmoji | DOMfilter:'inlinecolour' | DOMfilter:'mathjax':('.line-' + part.$$hashKey.replace(':','_')):settings.enableMathjax"></span>
               </td>
             </tr>
             <tr class="readmarker" ng-if="activeBuffer().lastSeen==$index">

js/filters.js

@@ -30,13 +30,6 @@ weechat.filter('irclinky', ['$filter', function($filter) {
             return text;
         }
 
-        // First, escape entities to prevent escaping issues because it's a bad idea
-        // to parse/modify HTML with regexes, which we do a couple of lines down...
-        var entities = {"<": "&lt;", ">": "&gt;", '"': '&quot;', "'": '&#39;', "&": "&amp;", "/": '&#x2F;'};
-        text = text.replace(/[<>"'&\/]/g, function (char) {
-            return entities[char];
-        });
-
         // This regex in no way matches all IRC channel names (they could also begin with &, + or an
         // exclamation mark followed by 5 alphanumeric characters, and are bounded in length by 50).
         // However, it matches all *common* IRC channels while trying to minimise false positives.
@@ -73,6 +66,15 @@ weechat.filter('DOMfilter', ['$filter', '$sce', function($filter, $sce) {
             return text;
         }
 
+        var escape_html = function(text) {
+            // First, escape entities to prevent escaping issues because it's a bad idea
+            // to parse/modify HTML with regexes, which we do a couple of lines down...
+            var entities = {"<": "&lt;", ">": "&gt;", '"': '&quot;', "'": '&#39;', "&": "&amp;", "/": '&#x2F;'};
+            return text.replace(/[<>"'&\/]/g, function (char) {
+                return entities[char];
+            });
+        };
+
         // hacky way to pass extra arguments without using .apply, which
         // would require assembling an argument array. PERFORMANCE!!!
         var extraArgument = (arguments.length > 2) ? arguments[2] : null;
@@ -85,8 +87,12 @@ weechat.filter('DOMfilter', ['$filter', '$sce', function($filter, $sce) {
         // Recursive DOM-walking function applying the filter to the text nodes
         var process = function(node) {
             if (node.nodeType === 3) { // text node
-                var value = filterFunction(node.nodeValue, extraArgument, thirdArgument);
+                // apply the filter to *escaped* HTML, and only commit changes if
-                if (value !== node.nodeValue) {
+                // it changed the escaped value. This is because setting the result
+                // as innerHTML causes it to be unescaped.
+                var input = escape_html(node.nodeValue);
+                var value = filterFunction(input, extraArgument, thirdArgument);
+                if (value !== input) {
                     // we changed something. create a new node to replace the current one
                     // we could also only add its children but that would probably incur
                     // more overhead than it would gain us
@@ -141,18 +147,6 @@ weechat.filter('getBufferQuickKeys', function () {
     };
 });
 
-weechat.filter('escape', ['$sanitize', function($sanitize) {
-    return function(text) {
-        // manual escaping because ng-sanitize is shit
-        return text
-            .replace(/&/g, "&amp;")
-            .replace(/</g, "&lt;")
-            .replace(/>/g, "&gt;")
-            .replace(/"/g, "&quot;")
-            .replace(/'/g, "&#039;");
-    };
-}]);
-
 // Emojifis the string using https://github.com/Ranks/emojione
 weechat.filter('emojify', function() {
     return function(text, enable_JS_Emoji) {