utf8.c

   1 /**
   2  * UTF-8 utility functions
   3  *
   4  * (c) 2010 Steve Bennett <steveb@workware.net.au>
   5  *
   6  * See LICENCE for licence details.
   7  */
   8
   9 #include <ctype.h>
  10 #include <stdlib.h>
  11 #include <string.h>
  12 #include <stdio.h>
  13 #include <assert.h>
  14 #include "utf8.h"
  15
  16 /* This one is always implemented */
  17 int utf8_fromunicode(char *p, unsigned uc)
  18 {
  19     if (uc <= 0x7f) {
  20         *p = uc;
  21         return 1;
  22     }
  23     else if (uc <= 0x7ff) {
  24         *p++ = 0xc0 | ((uc & 0x7c0) >> 6);
  25         *p = 0x80 | (uc & 0x3f);
  26         return 2;
  27     }
  28     else if (uc <= 0xffff) {
  29         *p++ = 0xe0 | ((uc & 0xf000) >> 12);
  30         *p++ = 0x80 | ((uc & 0xfc0) >> 6);
  31         *p = 0x80 | (uc & 0x3f);
  32         return 3;
  33     }
  34     /* Note: We silently truncate to 21 bits here: 0x1fffff */
  35     else {
  36         *p++ = 0xf0 | ((uc & 0x1c0000) >> 18);
  37         *p++ = 0x80 | ((uc & 0x3f000) >> 12);
  38         *p++ = 0x80 | ((uc & 0xfc0) >> 6);
  39         *p = 0x80 | (uc & 0x3f);
  40         return 4;
  41     }
  42 }
  43
  44 #if defined(JIM_UTF8) && !defined(JIM_BOOTSTRAP)
  45 int utf8_charlen(int c)
  46 {
  47     if ((c & 0x80) == 0) {
  48         return 1;
  49     }
  50     if ((c & 0xe0) == 0xc0) {
  51         return 2;
  52     }
  53     if ((c & 0xf0) == 0xe0) {
  54         return 3;
  55     }
  56     if ((c & 0xf8) == 0xf0) {
  57         return 4;
  58     }
  59     /* Invalid sequence */
  60     return -1;
  61 }
  62
  63 int utf8_strlen(const char *str, int bytelen)
  64 {
  65     int charlen = 0;
  66     if (bytelen < 0) {
  67         bytelen = strlen(str);
  68     }
  69     while (bytelen) {
  70         int c;
  71         int l = utf8_tounicode(str, &c);
  72         charlen++;
  73         str += l;
  74         bytelen -= l;
  75     }
  76     return charlen;
  77 }
  78
  79 int utf8_index(const char *str, int index)
  80 {
  81     const char *s = str;
  82     while (index--) {
  83         int c;
  84         s += utf8_tounicode(s, &c);
  85     }
  86     return s - str;
  87 }
  88
  89 int utf8_charequal(const char *s1, const char *s2)
  90 {
  91     int c1, c2;
  92
  93     utf8_tounicode(s1, &c1);
  94     utf8_tounicode(s2, &c2);
  95
  96     return c1 == c2;
  97 }
  98
  99 int utf8_prev_len(const char *str, int len)
 100 {
 101     int n = 1;
 102
 103     assert(len > 0);
 104
 105     /* Look up to len chars backward for a start-of-char byte */
 106     while (--len) {
 107         if ((str[-n] & 0x80) == 0) {
 108             /* Start of a 1-byte char */
 109             break;
 110         }
 111         if ((str[-n] & 0xc0) == 0xc0) {
 112             /* Start of a multi-byte char */
 113             break;
 114         }
 115         n++;
 116     }
 117     return n;
 118 }
 119
 120 int utf8_tounicode(const char *str, int *uc)
 121 {
 122     unsigned const char *s = (unsigned const char *)str;
 123
 124     if (s[0] < 0xc0) {
 125         *uc = s[0];
 126         return 1;
 127     }
 128     if (s[0] < 0xe0) {
 129         if ((s[1] & 0xc0) == 0x80) {
 130             *uc = ((s[0] & ~0xc0) << 6) | (s[1] & ~0x80);
 131             return 2;
 132         }
 133     }
 134     else if (s[0] < 0xf0) {
 135         if (((str[1] & 0xc0) == 0x80) && ((str[2] & 0xc0) == 0x80)) {
 136             *uc = ((s[0] & ~0xe0) << 12) | ((s[1] & ~0x80) << 6) | (s[2] & ~0x80);
 137             return 3;
 138         }
 139     }
 140     else if (s[0] < 0xf8) {
 141         if (((str[1] & 0xc0) == 0x80) && ((str[2] & 0xc0) == 0x80) && ((str[3] & 0xc0) == 0x80)) {
 142             *uc = ((s[0] & ~0xf0) << 18) | ((s[1] & ~0x80) << 12) | ((s[2] & ~0x80) << 6) | (s[3] & ~0x80);
 143             return 4;
 144         }
 145     }
 146
 147     /* Invalid sequence, so just return the byte */
 148     *uc = *s;
 149     return 1;
 150 }
 151
 152 struct casemap {
 153     unsigned short code;        /* code point */
 154     unsigned short altcode;     /* alternate case code point */
 155 };
 156
 157 /* Generated mapping tables */
 158 #include "_unicode_mapping.c"
 159
 160 #define ARRAYSIZE(A) sizeof(A) / sizeof(*(A))
 161
 162 static int cmp_casemap(const void *key, const void *cm)
 163 {
 164     return *(int *)key - (int)((const struct casemap *)cm)->code;
 165 }
 166
 167 static int utf8_map_case(const struct casemap *mapping, int num, int ch)
 168 {
 169     /* We only support 16 bit case mapping */
 170     if (ch <= 0xffff) {
 171         const struct casemap *cm =
 172             bsearch(&ch, mapping, num, sizeof(*mapping), cmp_casemap);
 173
 174         if (cm) {
 175             return cm->altcode;
 176         }
 177     }
 178     return ch;
 179 }
 180
 181 /* Some platforms don't have isascii */
 182 #ifndef isascii
 183 #define isascii(C) (!((C) & ~0x7f))
 184 #endif
 185
 186 int utf8_upper(int ch)
 187 {
 188     if (isascii(ch)) {
 189         return toupper(ch);
 190     }
 191     return utf8_map_case(unicode_case_mapping_upper, ARRAYSIZE(unicode_case_mapping_upper), ch);
 192 }
 193
 194 int utf8_lower(int ch)
 195 {
 196     if (isascii(ch)) {
 197         return tolower(ch);
 198     }
 199     return utf8_map_case(unicode_case_mapping_lower, ARRAYSIZE(unicode_case_mapping_lower), ch);
 200 }
 201
 202 int utf8_title(int ch)
 203 {
 204     int newch = utf8_map_case(unicode_case_mapping_title, ARRAYSIZE(unicode_case_mapping_title), ch);
 205     if (newch != ch) {
 206         return newch ? newch : ch;
 207     }
 208     return utf8_upper(ch);
 209 }
 210
 211 #endif /* JIM_BOOTSTRAP */