EscapeNonASCII.scala

def escapeNonASCII(str: String) : String = {
    val result = new java.lang.StringBuilder
    var i = 0
    while(i < str.length) {
        val cp: Int = Character.codePointAt(str, i)
        val n : Int = Character.charCount(cp)
        if(n > 1) {
            i += n-1
            if(i >= str.length) throw new IllegalArgumentException("truncated unexpectedly")
        }
        if(cp < 128)
            result.appendCodePoint(cp)
        else
            result.append("\\u%x".format(cp))
        i += 1
    }
    result.toString
}

Java (and by tradition, Scala) employs UTF-16 Unicode, which is a precursor of what we call today Unicode.
A quick approach to the matter is:

• Avoid methods employing char (or Char) data type.
• You cannot tell the length of a String simply counting how many code points it has.
• In general, 16 bits is sufficient for accomodating most languages. But not all languages. Not even all characters of a given language.
• Java Tutorial on Unicode here: https://docs.oracle.com/javase/tutorial/essential/regex/unicode.html
• More in-depth details on Java Unicode here: https://docs.oracle.com/javase/tutorial/i18n/text/unicode.html
• Glossary on Unicode here: http://unicode.org/glossary/

You can generate test data here: http://generator.lorem-ipsum.info/