content [nfc]: Use non-capturing group in _emojiClassRegexp

We never look at the captured group, so it's more efficient to
not ask the regexp engine to capture it in the first place.

lib/model/content.dart

@@ -639,7 +639,7 @@ class _ZulipContentParser {
     return RegExp("^(?:$mentionClass(?: silent)?|silent $mentionClass)\$");
   }();
 
-  static final _emojiClassRegexp = RegExp(r"^emoji(-[0-9a-f]+)*$");
+  static final _emojiClassRegexp = RegExp(r"^emoji(?:-[0-9a-f]+)*$");
 
   InlineContentNode parseInlineContent(dom.Node node) {
     assert(_debugParserContext == _ParserContext.inline);