Fix #556

`\uFEFF` (ZERO WIDTH NO-BREAK SPACE) is removed when parsing, but was
un-escaped for the output when `ascii_only` was false.

When using
UglifyJS multiple times (creating packages from minified sources, for
example), this would lead to problems because the byte was removed when
parsing for the second time.

Author: rvanvelzen

lib/output.js

@@ -86,7 +86,7 @@ function OutputStream(options) {
 
     function make_string(str) {
         var dq = 0, sq = 0;
-        str = str.replace(/[\\\b\f\n\r\t\x22\x27\u2028\u2029\0]/g, function(s){
+        str = str.replace(/[\\\b\f\n\r\t\x22\x27\u2028\u2029\0\ufeff]/g, function(s){
             switch (s) {
               case "\\": return "\\\\";
               case "\b": return "\\b";
@@ -98,6 +98,7 @@ function OutputStream(options) {
               case '"': ++dq; return '"';
               case "'": ++sq; return "'";
               case "\0": return "\\x00";
+              case "\ufeff": return "\\ufeff";
             }
             return s;
         });