/* by Jinsuk Kim, http://www.jinsuk.pe.kr */

#include <stdio.h>
#include <HanjaHangulMap.h>


int Hanja2Hangul(char *s)	/* input should be UTF-8 encoded stream */
{
    unsigned char c;
    unsigned short unicode = 0x00000000;

    int i, j, len;

    len = strlen(s);
    for (i=0; i<len;) {
	c = s[i]&0xe0;
	if (c < 0x80) {
	    i++;
	    continue; /* no need to process chars in this area */
	}
	else if (c < 0xe0) {
	    i+=2;
	    continue; /* no need to process chars in this area */
	}
	else if (c < 0xf0) {
	    unicode = (unsigned short) s[i]&0x0f;
	    i++;
	    unicode = unicode << 6;
	    unicode = unicode | ((unsigned short) s[i] & 0x3f);
	    i++;
	    unicode = unicode << 6;
	    unicode = unicode | ((unsigned short) s[i] & 0x3f);
	    i++;
	} /* from UTF-8 to UCS-2 */

	if (HjHgMap[unicode] != unicode ) { /* different ? hanja: non-CJK */
	    unsigned char byte[4];
	    /* full checking
	    int nbytes;
	    unicode = HjHgMap[unicode];
	    if (unicode < 0x80) {
		nbytes = 1;
		byte[0] = unicode;
	    } else if (unicode < 0x800) {
		nbytes = 2;
		byte[1] = (unicode & 0x3f) | 0x80;
		byte[0] = ((unicode << 2) & 0xcf00 | 0xc000) >> 8;
	    } else {
		nbytes = 3;
		byte[2] = (unicode & 0x3f) | 0x80;
		byte[1] = ((unicode << 2) & 0x3f00 | 0x8000) >> 8;
		byte[0] = ((unicode << 4) & 0x3f0000 | 0xe00000) >> 16;
	    }
	    */
	    /* entered here, we guarantee unicode is greater than 0x0800 */
	    unicode = HjHgMap[unicode];
	    byte[2] = (unicode & 0x3f) | 0x80;
	    byte[1] = ((unicode << 2) & 0x3f00 | 0x8000) >> 8;
	    byte[0] = ((unicode << 4) & 0x3f0000 | 0xe00000) >> 16;
	    for (j=0; j<3; j++) {
		s[i-3+j] = byte[j];
	    }
	    continue;
	}
    }
    return 0;
}