Fix issue caused by pretokenising of UTF8 strings (which get converted after the pretokenise step and cause the length to be wrong)

Even if length were correct, pretokenising the string stops the scan for UTF8 char patterns so the string wouldn't be treated correctly

Author: gfwilliams

plugins/pretokenise.js

@@ -196,7 +196,8 @@
         resultCode += "\n";
       if (tok.str==")" || tok.str=="}" || tok.str=="]") brackets--;
       // if we have a token for something, use that - else use the string
-      if (pretokeniseStrings && tok.type == "STRING") {
+      if (pretokeniseStrings && tok.type == "STRING" &&
+          tok.value.split("").reduce((r,ch) => r&&(ch.charCodeAt(0)<256), true)/*all 8 bit*/) {
         let str = tok.value;  // get string value
         lastIdx = tok.endIdx; // get next token
         lastlastTok = lastTok;