Currently the language detection method expects a case sensitive BCP47 string. But reality is less perfect and this it should handle it as a case insensitive string.