String lenght of a multy-byte string

Question

Hi Friends,
I am working with multy-byte(Korea) Strings. What I am doing is I am looking for the presence of an english substring in a multybyte string, find out the position of this substring using java indexof ()method.I am facing a problem here. For the purpose of comparing I am doing a toUppercase() of the multybyte string . I have noticed that this toUppercase() changes the string length and the position of the substring identified by the indexof() method becomes wrong. I could not find out a way to apply indexof() method in an "ignore-case" mode. Is there any java methods to resolve this. Please tell me a way out of this .
This is a bit urgent. Thanks in advance!!!

Answer 1

Well, I'm not sure what doing toUpperCase does to non-ASCII characters. If you want to know if it's
English characters, you should probably do something else...
java.lang.Character
    public static char toUpperCase(char ch) {
        char mapChar = ch;
        int val = A[Y[((X[ch>>5]&0xFF)<<4)|((ch>>1)&0xF)]|(ch&0x1)];
        if ((val & 0x00010000) != 0) {
            if ((val & 0x07FC0000) == 0x07FC0000) {
                switch(ch) {
                    // map chars with overflow offsets
                    case '\u00B5' : mapChar = '\u039C'; break;
                    case '\u017F' : mapChar = '\u0053'; break;
                    case '\u1FBE' : mapChar = '\u0399'; break;
                    // map char that have both a 1:1 and 1:M map
                    case '\u1F80' : mapChar = '\u1F88'; break;
                    case '\u1F81' : mapChar = '\u1F89'; break;
                    case '\u1F82' : mapChar = '\u1F8A'; break;
                    case '\u1F83' : mapChar = '\u1F8B'; break;
                    case '\u1F84' : mapChar = '\u1F8C'; break;
                    case '\u1F85' : mapChar = '\u1F8D'; break;
                    case '\u1F86' : mapChar = '\u1F8E'; break;
                    case '\u1F87' : mapChar = '\u1F8F'; break;
                    case '\u1F90' : mapChar = '\u1F98'; break;
                    case '\u1F91' : mapChar = '\u1F99'; break;
                    case '\u1F92' : mapChar = '\u1F9A'; break;
                    case '\u1F93' : mapChar = '\u1F9B'; break;
                    case '\u1F94' : mapChar = '\u1F9C'; break;
                    case '\u1F95' : mapChar = '\u1F9D'; break;
                    case '\u1F96' : mapChar = '\u1F9E'; break;
                    case '\u1F97' : mapChar = '\u1F9F'; break;
                    case '\u1FA0' : mapChar = '\u1FA8'; break;
                    case '\u1FA1' : mapChar = '\u1FA9'; break;
                    case '\u1FA2' : mapChar = '\u1FAA'; break;
                    case '\u1FA3' : mapChar = '\u1FAB'; break;
                    case '\u1FA4' : mapChar = '\u1FAC'; break;
                    case '\u1FA5' : mapChar = '\u1FAD'; break;
                    case '\u1FA6' : mapChar = '\u1FAE'; break;
                    case '\u1FA7' : mapChar = '\u1FAF'; break;
                    case '\u1FB3' : mapChar = '\u1FBC'; break;
                    case '\u1FC3' : mapChar = '\u1FCC'; break;
                    case '\u1FF3' : mapChar = '\u1FFC'; break;
                    // ch must have a 1:M case mapping, but we
                    // can't handle it here. Return ch.
                    // since mapChar is already set, no need
                    // to redo it here.
                    //default       : mapChar = ch;
            else {
                int offset = val  << 5 >> (5+18);
                mapChar =  (char)(ch - offset);
        return mapChar;
    }