return (0);
}
+
+/*
+ * Get the combing class.
+ *
+ * Similar to CFUniCharGetCombiningPropertyForCharacter.
+ */
+static inline u_int8_t
+get_combining_class(u_int16_t character) {
+ const u_int8_t *bitmap = __CFUniCharCombiningPropertyBitmap;
+
+ u_int8_t value = bitmap[(character >> 8)];
+
+ if (value) {
+ bitmap = bitmap + (value * 256);
+ return bitmap[character % 256];
+ }
+ return (0);
+}
+
+
static int unicode_decompose(u_int16_t character, u_int16_t *convertedChars);
static u_int16_t unicode_combine(u_int16_t base, u_int16_t combining);
+static void priortysort(u_int16_t* characters, int count);
char utf_extrabytes[32] = {
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
{
u_int16_t* bufstart;
u_int16_t* bufend;
- u_int16_t ucs_ch;
- u_int8_t byte;
+ unsigned int ucs_ch;
+ unsigned int byte;
+ int combcharcnt = 0;
int result = 0;
int decompose, precompose, swapbytes;
/* check for ascii */
if (byte < 0x80) {
- ucs_ch = byte; /* 1st byte */
+ ucs_ch = byte; /* 1st byte */
} else {
u_int32_t ch;
int extrabytes = utf_extrabytes[byte >> 3];
utf8len -= extrabytes;
switch (extrabytes) {
- case 1: ch = byte; /* 1st byte */
- ch <<= 6;
- ch += *utf8p++; /* 2nd byte */
- ch -= 0x00003080UL;
- if (ch < 0x0080)
- goto invalid;
- ucs_ch = ch;
+ case 1:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x00003080UL;
+ if (ch < 0x0080)
+ goto invalid;
+ ucs_ch = ch;
break;
-
- case 2: ch = byte; /* 1st byte */
- ch <<= 6;
- ch += *utf8p++; /* 2nd byte */
- ch <<= 6;
- ch += *utf8p++; /* 3rd byte */
- ch -= 0x000E2080UL;
- if (ch < 0x0800)
+ case 2:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x000E2080UL;
+ if (ch < 0x0800)
+ goto invalid;
+ if (ch >= 0xD800) {
+ if (ch <= 0xDFFF)
+ goto invalid;
+ if (ch == 0xFFFE || ch == 0xFFFF)
goto invalid;
- ucs_ch = ch;
- break;
-
- case 3: ch = byte; /* 1st byte */
- ch <<= 6;
- ch += *utf8p++; /* 2nd byte */
- ch <<= 6;
- ch += *utf8p++; /* 3rd byte */
- ch <<= 6;
- ch += *utf8p++; /* 4th byte */
- ch -= 0x03C82080UL + SP_HALF_BASE;
- ucs_ch = (ch >> SP_HALF_SHIFT) + SP_HIGH_FIRST;
- *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
- if (ucsp >= bufend)
- goto toolong;
- ucs_ch = (ch & SP_HALF_MASK) + SP_LOW_FIRST;
- *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
+ }
+ ucs_ch = ch;
+ break;
+ case 3:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 4th byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x03C82080UL + SP_HALF_BASE;
+ ucs_ch = (ch >> SP_HALF_SHIFT) + SP_HIGH_FIRST;
+ if (ucs_ch < SP_HIGH_FIRST || ucs_ch > SP_HIGH_LAST)
+ goto invalid;
+ *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
+ if (ucsp >= bufend)
+ goto toolong;
+ ucs_ch = (ch & SP_HALF_MASK) + SP_LOW_FIRST;
+ if (ucs_ch < SP_LOW_FIRST || ucs_ch > SP_LOW_LAST)
+ goto invalid;
+ *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
continue;
-
default:
- goto invalid;
+ goto invalid;
}
if (decompose) {
if (unicode_decomposeable(ucs_ch)) {
if (ucsp >= bufend)
goto toolong;
}
+ combcharcnt += count - 1;
continue;
}
} else if (precompose && (ucsp != bufstart)) {
if (ucs_ch == altslash)
ucs_ch = '/';
+ /*
+ * Make multiple combining character sequences canonical
+ */
+ if (unicode_combinable(ucs_ch)) {
+ ++combcharcnt; /* start tracking a run */
+ } else if (combcharcnt) {
+ if (combcharcnt > 1) {
+ priortysort(ucsp - combcharcnt, combcharcnt);
+ }
+ combcharcnt = 0; /* start over */
+ }
*ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
}
+ /*
+ * Make a previous combining sequence canonical
+ */
+ if (combcharcnt > 1) {
+ priortysort(ucsp - combcharcnt, combcharcnt);
+ }
exit:
*ucslen = (u_int8_t*)ucsp - (u_int8_t*)bufstart;
}
+/*
+ * utf8_validatestr - Check for a valid UTF-8 string.
+ */
+int
+utf8_validatestr(const u_int8_t* utf8p, size_t utf8len)
+{
+ unsigned int byte;
+ u_int32_t ch;
+ unsigned int ucs_ch;
+ size_t extrabytes;
+
+ while (utf8len-- > 0 && (byte = *utf8p++) != '\0') {
+ if (byte < 0x80)
+ continue; /* plain ascii */
+
+ extrabytes = utf_extrabytes[byte >> 3];
+
+ if (utf8len < extrabytes)
+ goto invalid;
+ utf8len -= extrabytes;
+
+ switch (extrabytes) {
+ case 1:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x00003080UL;
+ if (ch < 0x0080)
+ goto invalid;
+ break;
+ case 2:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x000E2080UL;
+ if (ch < 0x0800)
+ goto invalid;
+ if (ch >= 0xD800) {
+ if (ch <= 0xDFFF)
+ goto invalid;
+ if (ch == 0xFFFE || ch == 0xFFFF)
+ goto invalid;
+ }
+ break;
+ case 3:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 4th byte */
+ if ((byte >> 6) != 2)
+ goto invalid;
+ ch += byte;
+ ch -= 0x03C82080UL + SP_HALF_BASE;
+ ucs_ch = (ch >> SP_HALF_SHIFT) + SP_HIGH_FIRST;
+ if (ucs_ch < SP_HIGH_FIRST || ucs_ch > SP_HIGH_LAST)
+ goto invalid;
+ ucs_ch = (ch & SP_HALF_MASK) + SP_LOW_FIRST;
+ if (ucs_ch < SP_LOW_FIRST || ucs_ch > SP_LOW_LAST)
+ goto invalid;
+ break;
+ default:
+ goto invalid;
+ }
+
+ }
+ return (0);
+invalid:
+ return (EINVAL);
+}
+
+
/*
* Unicode 3.2 decomposition code (derived from Core Foundation)
*/
return (value);
}
+
+/*
+ * priortysort - order combining chars into canonical order
+ *
+ * Similar to CFUniCharPrioritySort
+ */
+static void
+priortysort(u_int16_t* characters, int count)
+{
+ u_int32_t p1, p2;
+ u_int16_t *ch1, *ch2;
+ u_int16_t *end;
+ int changes = 1;
+
+ end = characters + count;
+ do {
+ changes = 0;
+ ch1 = characters;
+ ch2 = characters + 1;
+ p2 = get_combining_class(*ch1);
+ while (ch2 < end) {
+ p1 = p2;
+ p2 = get_combining_class(*ch2);
+ if (p1 > p2) {
+ u_int32_t tmp;
+
+ tmp = *ch1;
+ *ch1 = *ch2;
+ *ch2 = tmp;
+ changes = 1;
+ }
+ ++ch1;
+ ++ch2;
+ }
+ } while (changes);
+}
projects / apple / xnu.git / blobdiff