FLTK 1.3.2
utf8.h
00001 /* $XFree86: xc/lib/X11/lcUniConv/utf8.h,v 1.3 2000/11/28 18:50:07 dawes Exp $ */
00002 
00003 /*
00004  * UTF-8
00005  */
00006 
00007 /* Specification: RFC 2279 */
00008 
00009 static int
00010 utf8_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)
00011 {
00012   unsigned char c = s[0];
00013 
00014   if (c < 0x80) {
00015     *pwc = c;
00016     return 1;
00017   } else if (c < 0xc2) {
00018     return RET_ILSEQ;
00019   } else if (c < 0xe0) {
00020     if (n < 2)
00021       return RET_TOOFEW(0);
00022     if (!((s[1] ^ 0x80) < 0x40))
00023       return RET_ILSEQ;
00024     *pwc = ((ucs4_t) (c & 0x1f) << 6)
00025            | (ucs4_t) (s[1] ^ 0x80);
00026     return 2;
00027   } else if (c < 0xf0) {
00028     if (n < 3)
00029       return RET_TOOFEW(0);
00030     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
00031           && (c >= 0xe1 || s[1] >= 0xa0)))
00032       return RET_ILSEQ;
00033     *pwc = ((ucs4_t) (c & 0x0f) << 12)
00034            | ((ucs4_t) (s[1] ^ 0x80) << 6)
00035            | (ucs4_t) (s[2] ^ 0x80);
00036     return 3;
00037   } else if (c < 0xf8) {
00038     if (n < 4)
00039       return RET_TOOFEW(0);
00040     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
00041           && (s[3] ^ 0x80) < 0x40
00042           && (c >= 0xf1 || s[1] >= 0x90)))
00043       return RET_ILSEQ;
00044     *pwc = ((ucs4_t) (c & 0x07) << 18)
00045            | ((ucs4_t) (s[1] ^ 0x80) << 12)
00046            | ((ucs4_t) (s[2] ^ 0x80) << 6)
00047            | (ucs4_t) (s[3] ^ 0x80);
00048     return 4;
00049   } else if (c < 0xfc) {
00050     if (n < 5)
00051       return RET_TOOFEW(0);
00052     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
00053           && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
00054           && (c >= 0xf9 || s[1] >= 0x88)))
00055       return RET_ILSEQ;
00056     *pwc = ((ucs4_t) (c & 0x03) << 24)
00057            | ((ucs4_t) (s[1] ^ 0x80) << 18)
00058            | ((ucs4_t) (s[2] ^ 0x80) << 12)
00059            | ((ucs4_t) (s[3] ^ 0x80) << 6)
00060            | (ucs4_t) (s[4] ^ 0x80);
00061     return 5;
00062   } else if (c < 0xfe) {
00063     if (n < 6)
00064       return RET_TOOFEW(0);
00065     if (!((s[1] ^ 0x80) < 0x40 && (s[2] ^ 0x80) < 0x40
00066           && (s[3] ^ 0x80) < 0x40 && (s[4] ^ 0x80) < 0x40
00067           && (s[5] ^ 0x80) < 0x40
00068           && (c >= 0xfd || s[1] >= 0x84)))
00069       return RET_ILSEQ;
00070     *pwc = ((ucs4_t) (c & 0x01) << 30)
00071            | ((ucs4_t) (s[1] ^ 0x80) << 24)
00072            | ((ucs4_t) (s[2] ^ 0x80) << 18)
00073            | ((ucs4_t) (s[3] ^ 0x80) << 12)
00074            | ((ucs4_t) (s[4] ^ 0x80) << 6)
00075            | (ucs4_t) (s[5] ^ 0x80);
00076     return 6;
00077   } else
00078     return RET_ILSEQ;
00079 }
00080 
00081 static int
00082 utf8_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n) /* n == 0 is acceptable */
00083 {
00084   int count;
00085   if (wc < 0x80)
00086     count = 1;
00087   else if (wc < 0x800)
00088     count = 2;
00089   else if (wc < 0x10000)
00090     count = 3;
00091   else if (wc < 0x200000)
00092     count = 4;
00093   else if (wc < 0x4000000)
00094     count = 5;
00095   else if (wc <= 0x7fffffff)
00096     count = 6;
00097   else
00098     return RET_ILSEQ;
00099   if (n < count)
00100     return RET_TOOSMALL;
00101   switch (count) { /* note: code falls through cases! */
00102     case 6: r[5] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x4000000;
00103     case 5: r[4] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x200000;
00104     case 4: r[3] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x10000;
00105     case 3: r[2] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0x800;
00106     case 2: r[1] = 0x80 | (wc & 0x3f); wc = wc >> 6; wc |= 0xc0;
00107     case 1: r[0] = wc;
00108   }
00109   return count;
00110 }