The JLSv2 section 3.6 only defines five characters as whitespace (space, tab, form feed, newline, carriage return). In particular, none of the non-ASCII Unicode space characters are defined as white space. The method Character.isWhiteSpace claims it "Determines if the specified character is white space according to Java." This mehtod admits many Unicode characters. Other portions of the lexical structure portion of the JLS have been updated to refer to new, more Unicode-aware, methods in Character. It seems the white space definition should be similarly updated.