libdrawtext

annotate src/utf8.c @ 56:f9c8eaa9b8d6

find changesets by author, revision, files, or words in the commit message
done with the unicode sample
author John Tsiombikas <nuclear@mutantstargoat.com>
date Fri, 16 Sep 2011 08:31:06 +0300
parents fe0c54e574ae
children 095ff7ca4e74
rev   line source
nuclear@0 1 #include "drawtext.h"
nuclear@0 2
nuclear@0 3 #define U8_IS_FIRST(x) (((((x) >> 7) & 1) == 0) || ((((x) >> 6) & 3) == 3))
nuclear@0 4
nuclear@0 5 static const char first_mask[] = {
nuclear@0 6 0,
nuclear@0 7 0x7f, /* single byte, 7 bits valid */
nuclear@0 8 0x1f, /* two-bytes, 5 bits valid */
nuclear@0 9 0xf, /* three-bytes, 4 bits valid */
nuclear@0 10 0x7 /* four-bytes, 3 bits valid */
nuclear@0 11 };
nuclear@3 12 static const char first_shift[] = { 0, 7, 5, 4, 3 }; /* see above */
nuclear@0 13
nuclear@0 14 #define CONT_PREFIX 0x80
nuclear@0 15 #define CONT_MASK 0x3f
nuclear@0 16 #define CONT_SHIFT 6
nuclear@0 17
nuclear@0 18 /* last charcodes for 1, 2, 3 or 4-byte utf8 chars */
nuclear@0 19 static const int utf8_lastcode[] = { 0x7f, 0x7ff, 0xfff, 0x1fffff };
nuclear@0 20
nuclear@0 21 #define prefix_mask(x) (~first_mask[x])
nuclear@0 22 #define prefix(x) ((prefix_mask(x) << 1) & 0xff)
nuclear@0 23
nuclear@0 24
nuclear@0 25 char *dtx_utf8_next_char(char *str)
nuclear@0 26 {
nuclear@0 27 return str + dtx_utf8_nbytes(str);
nuclear@0 28 }
nuclear@0 29
nuclear@0 30 int dtx_utf8_char_code(const char *str)
nuclear@0 31 {
nuclear@0 32 int i, nbytes, shift, code = 0;
nuclear@3 33 int mask;
nuclear@0 34
nuclear@0 35 if(!U8_IS_FIRST(*str)) {
nuclear@0 36 return -1;
nuclear@0 37 }
nuclear@0 38
nuclear@0 39 nbytes = dtx_utf8_nbytes(str);
nuclear@0 40 mask = first_mask[nbytes];
nuclear@0 41 shift = 0;
nuclear@0 42
nuclear@0 43 for(i=0; i<nbytes; i++) {
nuclear@0 44 if(!*str) {
nuclear@0 45 break;
nuclear@0 46 }
nuclear@0 47
nuclear@0 48 code = (code << shift) | (*str++ & mask);
nuclear@0 49 mask = 0x3f;
nuclear@3 50 shift = 6;
nuclear@0 51 }
nuclear@0 52 return code;
nuclear@0 53 }
nuclear@0 54
nuclear@0 55 int dtx_utf8_nbytes(const char *str)
nuclear@0 56 {
nuclear@0 57 int i, numset = 0;
nuclear@0 58 int c = *str;
nuclear@0 59
nuclear@0 60 if(!U8_IS_FIRST(c)) {
nuclear@0 61 for(i=0; !U8_IS_FIRST(str[i]); i++);
nuclear@0 62 return i;
nuclear@0 63 }
nuclear@0 64
nuclear@0 65 /* count the leading 1s */
nuclear@0 66 for(i=0; i<4; i++) {
nuclear@0 67 if(((c >> (7 - i)) & 1) == 0) {
nuclear@0 68 break;
nuclear@0 69 }
nuclear@0 70 numset++;
nuclear@0 71 }
nuclear@0 72
nuclear@0 73 if(!numset) {
nuclear@0 74 return 1;
nuclear@0 75 }
nuclear@0 76 return numset;
nuclear@0 77 }
nuclear@0 78
nuclear@0 79 int dtx_utf8_char_count(const char *str)
nuclear@0 80 {
nuclear@0 81 int n = 0;
nuclear@0 82
nuclear@0 83 while(*str) {
nuclear@0 84 n++;
nuclear@0 85 str = dtx_utf8_next_char((char*)str);
nuclear@0 86 }
nuclear@0 87 return n;
nuclear@0 88 }
nuclear@0 89
nuclear@0 90 size_t dtx_utf8_from_char_code(int code, char *buf)
nuclear@0 91 {
nuclear@0 92 size_t nbytes = 0;
nuclear@0 93 int i;
nuclear@0 94
nuclear@0 95 for(i=0; i<4; i++) {
nuclear@0 96 if(code <= utf8_lastcode[i]) {
nuclear@0 97 nbytes = i + 1;
nuclear@0 98 break;
nuclear@0 99 }
nuclear@0 100 }
nuclear@0 101
nuclear@0 102 if(!nbytes && buf) {
nuclear@0 103 for(i=0; i<nbytes; i++) {
nuclear@0 104 int idx = nbytes - i - 1;
nuclear@0 105 int mask, shift, prefix;
nuclear@0 106
nuclear@0 107 if(idx > 0) {
nuclear@0 108 mask = CONT_MASK;
nuclear@0 109 shift = CONT_SHIFT;
nuclear@0 110 prefix = CONT_PREFIX;
nuclear@0 111 } else {
nuclear@0 112 mask = first_mask[nbytes];
nuclear@0 113 shift = first_shift[nbytes];
nuclear@0 114 prefix = prefix(nbytes);
nuclear@0 115 }
nuclear@0 116
nuclear@0 117 buf[idx] = (code & mask) | (prefix & ~mask);
nuclear@0 118 code >>= shift;
nuclear@0 119 }
nuclear@0 120 }
nuclear@0 121 return nbytes;
nuclear@0 122 }
nuclear@0 123
nuclear@0 124 size_t dtx_utf8_from_string(const wchar_t *str, char *buf)
nuclear@0 125 {
nuclear@0 126 size_t nbytes = 0;
nuclear@0 127 char *ptr = buf;
nuclear@0 128
nuclear@0 129 while(*str) {
nuclear@0 130 int cbytes = dtx_utf8_from_char_code(*str++, ptr);
nuclear@0 131 if(ptr) {
nuclear@0 132 ptr += cbytes;
nuclear@0 133 }
nuclear@0 134 nbytes += cbytes;
nuclear@0 135 }
nuclear@0 136 return nbytes;
nuclear@0 137 }