Replace \\ found at a particular position in a String in java (not all occurrences) with \\\\

I am trying to replace \ with double \ in java. Thing is sometimes the \ belongs to a unicode character (eg:\u00E9) and in this case I don’t want to do the escape.
Currently I am stuck at the actual replacement of . This is just an example that will be a bit refactored :

static boolean isUnicodeChar(String string) {
        final String regex = "(?<!\\\\)(\\\\\\\\)*\\\\u[A-Fa-f\\d]{4}";
        final Pattern pattern = Pattern.compile(regex, Pattern.MULTILINE | Pattern.UNICODE_CASE);
        final Matcher matcher = pattern.matcher(string);
        return matcher.find();
    }

    public static void main(String[] args) {

        String string = "Muir-Torr \\ \\u00E9 syndrome \\u1234 skd just some \\uabcd arbitrary text \\ s";
        System.out.println("Old String: " + string);

        for (int startPosition = string.indexOf('\\'); startPosition >= 0; startPosition = string.indexOf('\\', startPosition + 1)) {
            int strLength = string.length();
//            int endPosition = startPosition + 5;
            int endPosition = startPosition + 6;
            System.out.println("\\ found at position: " + startPosition);

            if (endPosition <= strLength - 1) {
                String stringToCheck = string.substring(startPosition, endPosition);
                System.out.println("Checking : " + stringToCheck);
                if (!isUnicodeChar(stringToCheck)) {
                  //here I should replace the char that is found at start position with "\\\\"
                    System.out.println("New String: " + string);

                } else {
                    System.out.println("No replacement needed, is unicode char: " + stringToCheck);

                }
            } else {
              //here I should replace the char that is found at start position with "\\\\"
                System.out.println("New String: " + string);
            }
        }

    }

Replaced string should be:
"Muir-Torr \\\\ \\u00E9 syndrome \\u1234 skd just some \\uabcd arbitrary text \\\\ s";

>Solution :

First let’s just replace all backslashes with 2. The joy of Java escaping…

String replaced = string.replaceAll("\\\\", "\\\\\\\\");

Now we need to tell the regex engine "don’t do that if the slash is followed by u and four hexadecimal digits". We can use a negative lookahead: (?!u[0-9a-fA-F]{4})

String replaced = string.replaceAll("\\\\(?!u[0-9a-fA-F]{4})", "\\\\\\\\");

Result:

Muir-Torr \\ \u00E9 syndrome \u1234 skd just some \uabcd arbitrary text \\ s

You can repeat step one to get your desired result.