/*
- * Copyright (c) 2000-2001 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2005 Apple Computer, Inc. All rights reserved.
*
- * @APPLE_LICENSE_HEADER_START@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
- * The contents of this file constitute Original Code as defined in and
- * are subject to the Apple Public Source License Version 1.1 (the
- * "License"). You may not use this file except in compliance with the
- * License. Please obtain a copy of the License at
- * http://www.apple.com/publicsource and read it before using this file.
+ * This file contains Original Code and/or Modifications of Original Code
+ * as defined in and that are subject to the Apple Public Source License
+ * Version 2.0 (the 'License'). You may not use this file except in
+ * compliance with the License. The rights granted to you under the License
+ * may not be used to create, or enable the creation or redistribution of,
+ * unlawful or unlicensed copies of an Apple operating system, or to
+ * circumvent, violate, or enable the circumvention or violation of, any
+ * terms of an Apple operating system software license agreement.
*
- * This Original Code and all software distributed under the License are
- * distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY KIND, EITHER
+ * Please obtain a copy of the License at
+ * http://www.opensource.apple.com/apsl/ and read it before using this file.
+ *
+ * The Original Code and all software distributed under the License are
+ * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
* EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
* INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT. Please see the
- * License for the specific language governing rights and limitations
- * under the License.
+ * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
+ * Please see the License for the specific language governing rights and
+ * limitations under the License.
*
- * @APPLE_LICENSE_HEADER_END@
+ * @APPLE_OSREFERENCE_LICENSE_HEADER_END@
+ */
+
+ /*
+ Includes Unicode 3.2 decomposition code derived from Core Foundation
*/
+
#include <sys/param.h>
#include <sys/utfconv.h>
#include <sys/errno.h>
-#include <architecture/byte_order.h>
+#include <sys/malloc.h>
+#include <libkern/OSByteOrder.h>
/*
* UTF-8 (Unicode Transformation Format)
#define SP_HIGH_FIRST 0xD800UL
#define SP_HIGH_LAST 0xDBFFUL
#define SP_LOW_FIRST 0xDC00UL
-#define SP_LOW_LAST 0xDFFFUL
+#define SP_LOW_LAST 0xDFFFUL
+
+
+#include "vfs_utfconvdata.h"
+
+
+/*
+ * Test for a combining character.
+ *
+ * Similar to __CFUniCharIsNonBaseCharacter except that
+ * unicode_combinable also includes Hangul Jamo characters.
+ */
+int
+unicode_combinable(u_int16_t character)
+{
+ const u_int8_t *bitmap = __CFUniCharCombiningBitmap;
+ u_int8_t value;
+
+ if (character < 0x0300)
+ return (0);
+
+ value = bitmap[(character >> 8) & 0xFF];
+
+ if (value == 0xFF) {
+ return (1);
+ } else if (value) {
+ bitmap = bitmap + ((value - 1) * 32) + 256;
+ return (bitmap[(character & 0xFF) / 8] & (1 << (character % 8)) ? 1 : 0);
+ }
+ return (0);
+}
+
+/*
+ * Test for a precomposed character.
+ *
+ * Similar to __CFUniCharIsDecomposableCharacter.
+ */
+int
+unicode_decomposeable(u_int16_t character) {
+ const u_int8_t *bitmap = __CFUniCharDecomposableBitmap;
+ u_int8_t value;
+
+ if (character < 0x00C0)
+ return (0);
+
+ value = bitmap[(character >> 8) & 0xFF];
+
+ if (value == 0xFF) {
+ return (1);
+ } else if (value) {
+ bitmap = bitmap + ((value - 1) * 32) + 256;
+ return (bitmap[(character & 0xFF) / 8] & (1 << (character % 8)) ? 1 : 0);
+ }
+ 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);
-static u_int16_t ucs_decompose(u_int16_t, u_int16_t *);
+static u_int16_t ucs_to_sfm(u_int16_t ucs_ch, int lastchar);
-static u_int16_t ucs_combine(u_int16_t base, u_int16_t comb);
+static u_int16_t sfm_to_ucs(u_int16_t ucs_ch);
char utf_extrabytes[32] = {
-1, -1, -1, -1, -1, -1, -1, -1, 1, 1, 1, 1, 2, 2, 3, -1
};
+const char hexdigits[16] = {
+ '0', '1', '2', '3', '4', '5', '6', '7',
+ '8', '9', 'A', 'B', 'C', 'D', 'E', 'F'
+};
/*
- * utf8_encodelen - Calculates the UTF-8 encoding length for a Unicode filename
+ * utf8_encodelen - Calculate the UTF-8 encoding length
*
- * NOTES:
- * If '/' chars are allowed on disk then an alternate
- * (replacement) char must be provided in altslash.
+ * This function takes a Unicode input string, ucsp, of ucslen bytes
+ * and calculates the size of the UTF-8 output in bytes (not including
+ * a NULL termination byte). The string must reside in kernel memory.
*
- * input flags:
- * UTF_REVERSE_ENDIAN: Unicode byteorder is opposite current runtime
+ * If '/' chars are possible in the Unicode input then an alternate
+ * (replacement) char should be provided in altslash.
+ *
+ * FLAGS
+ * UTF_REVERSE_ENDIAN: Unicode byte order is opposite current runtime
+ *
+ * UTF_BIG_ENDIAN: Unicode byte order is always big endian
+ *
+ * UTF_LITTLE_ENDIAN: Unicode byte order is always little endian
+ *
+ * UTF_DECOMPOSED: generate fully decomposed output
+ *
+ * UTF_PRECOMPOSED is ignored since utf8_encodestr doesn't support it
+ *
+ * ERRORS
+ * None
*/
size_t
-utf8_encodelen(const u_int16_t * ucsp, size_t ucslen, u_int16_t altslash,
- int flags)
+utf8_encodelen(const u_int16_t * ucsp, size_t ucslen, u_int16_t altslash, int flags)
{
u_int16_t ucs_ch;
+ u_int16_t * chp = NULL;
+ u_int16_t sequence[8];
+ int extra = 0;
int charcnt;
int swapbytes = (flags & UTF_REVERSE_ENDIAN);
+ int decompose = (flags & UTF_DECOMPOSED);
size_t len;
-
+
charcnt = ucslen / 2;
len = 0;
while (charcnt-- > 0) {
- ucs_ch = *ucsp++;
-
- if (swapbytes)
- ucs_ch = NXSwapShort(ucs_ch);
- if (ucs_ch == '/')
- ucs_ch = altslash ? altslash : '_';
- else if (ucs_ch == '\0')
- ucs_ch = UCS_ALT_NULL;
-
+ if (extra > 0) {
+ --extra;
+ ucs_ch = *chp++;
+ } else {
+ ucs_ch = *ucsp++;
+ if (swapbytes) {
+ ucs_ch = OSSwapInt16(ucs_ch);
+ }
+ if (ucs_ch == '/') {
+ ucs_ch = altslash ? altslash : '_';
+ } else if (ucs_ch == '\0') {
+ ucs_ch = UCS_ALT_NULL;
+ } else if (decompose && unicode_decomposeable(ucs_ch)) {
+ extra = unicode_decompose(ucs_ch, sequence) - 1;
+ charcnt += extra;
+ ucs_ch = sequence[0];
+ chp = &sequence[1];
+ }
+ }
len += UNICODE_TO_UTF8_LEN(ucs_ch);
}
*
* input flags:
* UTF_REVERSE_ENDIAN: Unicode byteorder is opposite current runtime
+ *
+ * UTF_BIG_ENDIAN: Unicode byte order is always big endian
+ *
+ * UTF_LITTLE_ENDIAN: Unicode byte order is always little endian
+ *
+ * UTF_DECOMPOSED: generate fully decomposed output
+ *
* UTF_NO_NULL_TERM: don't add NULL termination to UTF-8 output
*
* result:
* ENAMETOOLONG: Name didn't fit; only buflen bytes were encoded
+ *
* EINVAL: Illegal char found; char was replaced by an '_'.
*/
int
u_int8_t * bufstart;
u_int8_t * bufend;
u_int16_t ucs_ch;
- u_int16_t extra[2] = {0};
+ u_int16_t * chp = NULL;
+ u_int16_t sequence[8];
+ int extra = 0;
int charcnt;
int swapbytes = (flags & UTF_REVERSE_ENDIAN);
int nullterm = ((flags & UTF_NO_NULL_TERM) == 0);
int decompose = (flags & UTF_DECOMPOSED);
+ int sfmconv = (flags & UTF_SFM_CONVERSIONS);
int result = 0;
-
+
bufstart = utf8p;
bufend = bufstart + buflen;
if (nullterm)
charcnt = ucslen / 2;
while (charcnt-- > 0) {
- if (!decompose)
- ucs_ch = swapbytes ? NXSwapShort(*ucsp++) : *ucsp++;
- else if (extra[0]) {
- ucs_ch = extra[0]; extra[0] = 0;
- } else if (extra[1]) {
- ucs_ch = extra[1]; extra[1] = 0;
+ if (extra > 0) {
+ --extra;
+ ucs_ch = *chp++;
} else {
- ucs_ch = swapbytes ? NXSwapShort(*ucsp++) : *ucsp++;
- ucs_ch = ucs_decompose(ucs_ch, &extra[0]);
- if (extra[0])
- charcnt++;
- if (extra[1])
- charcnt++;
+ ucs_ch = swapbytes ? OSSwapInt16(*ucsp++) : *ucsp++;
+
+ if (decompose && unicode_decomposeable(ucs_ch)) {
+ extra = unicode_decompose(ucs_ch, sequence) - 1;
+ charcnt += extra;
+ ucs_ch = sequence[0];
+ chp = &sequence[1];
+ }
}
/* Slash and NULL are not permitted */
*utf8p++ = 0x80 | (0x3f & ucs_ch);
} else {
+ /* These chars never valid Unicode. */
+ if (ucs_ch == 0xFFFE || ucs_ch == 0xFFFF) {
+ result = EINVAL;
+ break;
+ }
+
/* Combine valid surrogate pairs */
if (ucs_ch >= SP_HIGH_FIRST && ucs_ch <= SP_HIGH_LAST
&& charcnt > 0) {
u_int16_t ch2;
u_int32_t pair;
- ch2 = swapbytes ? NXSwapShort(*ucsp) : *ucsp;
+ ch2 = swapbytes ? OSSwapInt16(*ucsp) : *ucsp;
if (ch2 >= SP_LOW_FIRST && ch2 <= SP_LOW_LAST) {
pair = ((ucs_ch - SP_HIGH_FIRST) << SP_HALF_SHIFT)
+ (ch2 - SP_LOW_FIRST) + SP_HALF_BASE;
*utf8p++ = 0x80 | (0x3f & pair);
continue;
}
+ } else if (sfmconv) {
+ ucs_ch = sfm_to_ucs(ucs_ch);
+ if (ucs_ch < 0x0080) {
+ if (utf8p >= bufend) {
+ result = ENAMETOOLONG;
+ break;
+ }
+ *utf8p++ = ucs_ch;
+ continue;
+ }
}
if ((utf8p + 2) >= bufend) {
result = ENAMETOOLONG;
* (replacement) char must be provided in altslash.
*
* input flags:
- * UTF_REV_ENDIAN: Unicode byteorder is oposite current runtime
- * UTF_DECOMPOSED: Unicode output string must be fully decompsed
+ * UTF_REV_ENDIAN: Unicode byte order is opposite current runtime
+ *
+ * UTF_BIG_ENDIAN: Unicode byte order is always big endian
+ *
+ * UTF_LITTLE_ENDIAN: Unicode byte order is always little endian
+ *
+ * UTF_DECOMPOSED: generate fully decomposed output (NFD)
+ *
+ * UTF_PRECOMPOSED: generate precomposed output (NFC)
+ *
+ * UTF_ESCAPE_ILLEGAL: percent escape any illegal UTF-8 input
*
* result:
* ENAMETOOLONG: Name didn't fit; only ucslen chars were decoded.
+ *
* EINVAL: Illegal UTF-8 sequence found.
*/
int
{
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;
+ int decompose, precompose, swapbytes, escaping;
+ int sfmconv;
+ int extrabytes;
- decompose = (flags & UTF_DECOMPOSED);
+ decompose = (flags & UTF_DECOMPOSED);
precompose = (flags & UTF_PRECOMPOSED);
- swapbytes = (flags & UTF_REVERSE_ENDIAN);
+ swapbytes = (flags & UTF_REVERSE_ENDIAN);
+ escaping = (flags & UTF_ESCAPE_ILLEGAL);
+ sfmconv = (flags & UTF_SFM_CONVERSIONS);
bufstart = ucsp;
bufend = (u_int16_t *)((u_int8_t *)ucsp + buflen);
/* check for ascii */
if (byte < 0x80) {
- ucs_ch = byte; /* 1st byte */
+ ucs_ch = sfmconv ? ucs_to_sfm(byte, utf8len == 0) : byte;
} else {
u_int32_t ch;
- int extrabytes = utf_extrabytes[byte >> 3];
- if (utf8len < extrabytes)
- goto invalid;
+ extrabytes = utf_extrabytes[byte >> 3];
+ if ((extrabytes < 0) || ((int)utf8len < extrabytes)) {
+ goto escape;
+ }
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 escape2;
+ ch += byte;
+ ch -= 0x00003080UL;
+ if (ch < 0x0080)
+ goto escape2;
+ 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)
- 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;
+ case 2:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto escape2;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto escape3;
+ ch += byte;
+ ch -= 0x000E2080UL;
+ if (ch < 0x0800)
+ goto escape3;
+ if (ch >= 0xD800) {
+ if (ch <= 0xDFFF)
+ goto escape3;
+ if (ch == 0xFFFE || ch == 0xFFFF)
+ goto escape3;
+ }
+ ucs_ch = ch;
+ break;
+ case 3:
+ ch = byte; ch <<= 6; /* 1st byte */
+ byte = *utf8p++; /* 2nd byte */
+ if ((byte >> 6) != 2)
+ goto escape2;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 3rd byte */
+ if ((byte >> 6) != 2)
+ goto escape3;
+ ch += byte; ch <<= 6;
+ byte = *utf8p++; /* 4th byte */
+ if ((byte >> 6) != 2)
+ goto escape4;
+ 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 escape4;
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)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) {
+ --ucsp;
+ goto escape4;
+ }
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)ucs_ch;
continue;
-
default:
- goto invalid;
+ result = EINVAL;
+ goto exit;
}
if (decompose) {
- u_int16_t comb_ch[2];
-
- ucs_ch = ucs_decompose(ucs_ch, &comb_ch[0]);
+ if (unicode_decomposeable(ucs_ch)) {
+ u_int16_t sequence[8];
+ int count, i;
+
+ /* Before decomposing a new unicode character, sort
+ * previous combining characters, if any, and reset
+ * the counter.
+ */
+ if (combcharcnt > 1) {
+ priortysort(ucsp - combcharcnt, combcharcnt);
+ }
+ combcharcnt = 0;
- if (comb_ch[0]) {
- *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
- if (ucsp >= bufend)
- goto toolong;
- ucs_ch = comb_ch[0];
- if (comb_ch[1]) {
- *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : ucs_ch;
+ count = unicode_decompose(ucs_ch, sequence);
+ for (i = 0; i < count; ++i) {
+ ucs_ch = sequence[i];
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)ucs_ch;
if (ucsp >= bufend)
goto toolong;
- ucs_ch = comb_ch[1];
}
+ combcharcnt += count - 1;
+ continue;
}
} else if (precompose && (ucsp != bufstart)) {
u_int16_t composite, base;
- base = swapbytes ? NXSwapShort(*(ucsp - 1)) : *(ucsp - 1);
- composite = ucs_combine(base, ucs_ch);
- if (composite) {
- --ucsp;
- ucs_ch = composite;
+ if (unicode_combinable(ucs_ch)) {
+ base = swapbytes ? OSSwapInt16(*(ucsp - 1)) : *(ucsp - 1);
+ composite = unicode_combine(base, ucs_ch);
+ if (composite) {
+ --ucsp;
+ ucs_ch = composite;
+ }
}
}
if (ucs_ch == UCS_ALT_NULL)
if (ucs_ch == altslash)
ucs_ch = '/';
- *ucsp++ = swapbytes ? NXSwapShort(ucs_ch) : 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 ? OSSwapInt16(ucs_ch) : (u_int16_t)ucs_ch;
+ continue;
+
+ /*
+ * Escape illegal UTF-8 into something legal.
+ */
+escape4:
+ utf8p -= 3;
+ goto escape;
+escape3:
+ utf8p -= 2;
+ goto escape;
+escape2:
+ utf8p -= 1;
+escape:
+ if (!escaping) {
+ result = EINVAL;
+ goto exit;
+ }
+ if (extrabytes > 0)
+ utf8len += extrabytes;
+ byte = *(utf8p - 1);
+ if ((ucsp + 2) >= bufend)
+ goto toolong;
+
+ /* Make a previous combining sequence canonical. */
+ if (combcharcnt > 1) {
+ priortysort(ucsp - combcharcnt, combcharcnt);
+ }
+ combcharcnt = 0;
+
+ ucs_ch = '%';
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)ucs_ch;
+ ucs_ch = hexdigits[byte >> 4];
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)ucs_ch;
+ ucs_ch = hexdigits[byte & 0x0F];
+ *ucsp++ = swapbytes ? OSSwapInt16(ucs_ch) : (u_int16_t)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;
return (result);
-invalid:
- result = EINVAL;
- goto exit;
-
toolong:
result = ENAMETOOLONG;
goto exit;
/*
- * Lookup tables for Unicode chars 0x00C0 thru 0x00FF
- * primary_char yields first decomposed char. If this
- * char is an alpha char then get the combining char
- * from the combining_char table and add 0x0300 to it.
+ * 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;
-static unsigned char primary_char[8*36] = {
- 0x41, 0x41, 0x41, 0x41, 0x41, 0x41, 0x00, 0x43,
-
- 0x45, 0x45, 0x45, 0x45, 0x49, 0x49, 0x49, 0x49, /* CF */
-
- 0x00, 0x4E, 0x4F, 0x4F, 0x4F, 0x4F, 0x4F, 0x00,
-
- 0x00, 0x55, 0x55, 0x55, 0x55, 0x59, 0x00, 0x00, /* DF */
-
- 0x61, 0x61, 0x61, 0x61, 0x61, 0x61, 0x00, 0x63,
-
- 0x65, 0x65, 0x65, 0x65, 0x69, 0x69, 0x69, 0x69, /* EF */
-
- 0x00, 0x6E, 0x6F, 0x6F, 0x6F, 0x6F, 0x6F, 0x00,
-
- 0x00, 0x75, 0x75, 0x75, 0x75, 0x79, 0x00, 0x79, /* FF */
-
- 0x41, 0x61, 0x41, 0x61, 0x41, 0x61, 0x43, 0x63,
-
- 0x43, 0x63, 0x43, 0x63, 0x43, 0x63, 0x44, 0x64, /* 10F */
-
- 0x00, 0x00, 0x45, 0x65, 0x45, 0x65, 0x45, 0x65,
-
- 0x45, 0x65, 0x45, 0x65, 0x47, 0x67, 0x47, 0x67, /* 11F */
-
- 0x47, 0x67, 0x47, 0x67, 0x48, 0x68, 0x00, 0x00,
-
- 0x49, 0x69, 0x49, 0x69, 0x49, 0x69, 0x49, 0x69,
-
- 0x49, 0x00, 0x00, 0x00, 0x4A, 0x6A, 0x4B, 0x6B,
-
- 0x00, 0x4C, 0x6C, 0x4C, 0x6C, 0x4C, 0x6C, 0x00, /* 13F */
-
- 0x00, 0x00, 0x00, 0x4E, 0x6E, 0x4E, 0x6E, 0x4E,
-
- 0x6E, 0x00, 0x00, 0x00, 0x4F, 0x6F, 0x4F, 0x6F,
-
- 0x4F, 0x6F, 0x00, 0x00, 0x52, 0x72, 0x52, 0x72,
-
- 0x52, 0x72, 0x53, 0x73, 0x53, 0x73, 0x53, 0x73, /* 15F */
-
- 0x53, 0x73, 0x54, 0x74, 0x54, 0x74, 0x00, 0x00,
-
- 0x55, 0x75, 0x55, 0x75, 0x55, 0x75, 0x55, 0x75,
-
- 0x55, 0x75, 0x55, 0x75, 0x57, 0x77, 0x59, 0x79,
-
- 0x59, 0x5A, 0x7A, 0x5A, 0x7A, 0x5A, 0x7A, 0x00, /* 17F */
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 19F */
-
- 0x4F, 0x6F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x55,
-
- 0x75, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 1BF */
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
-
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x41, 0x61, 0x49,
-
- 0x69, 0x4F, 0x6F, 0x55, 0x75, 0xDC, 0xFC, 0xDC,
-
- 0xFC, 0xDC, 0xFC, 0xDC, 0xFC, 0x00, 0xC4, 0xE4 /* 1DF */
-
-};
-
-static unsigned char combining_char[8*36] = {
- 0x00, 0x01, 0x02, 0x03, 0x08, 0x0A, 0xFF, 0x27,
-
- 0x00, 0x01, 0x02, 0x08, 0x00, 0x01, 0x02, 0x08, /* CF */
-
- 0xFF, 0x03, 0x00, 0x01, 0x02, 0x03, 0x08, 0xFF,
-
- 0xFF, 0x00, 0x01, 0x02, 0x08, 0x01, 0xFF, 0xFF, /* DF */
-
- 0x00, 0x01, 0x02, 0x03, 0x08, 0x0A, 0xFF, 0x27,
-
- 0x00, 0x01, 0x02, 0x08, 0x00, 0x01, 0x02, 0x08, /* EF */
-
- 0xFF, 0x03, 0x00, 0x01, 0x02, 0x03, 0x08, 0xFF,
-
- 0xFF, 0x00, 0x01, 0x02, 0x08, 0x01, 0xFF, 0x08, /* FF */
-
- 0x04, 0x04, 0x06, 0x06, 0x28, 0x28, 0x01, 0x01,
-
- 0x02, 0x02, 0x07, 0x07, 0x0C, 0x0C, 0x0C, 0x0C,
-
- 0x00, 0x00, 0x04, 0x04, 0x06, 0x06, 0x07, 0x07,
-
- 0x28, 0x28, 0x0C, 0x0C, 0x02, 0x02, 0x06, 0x06,
-
- 0x07, 0x07, 0x27, 0x27, 0x02, 0x02, 0x00, 0x00,
-
- 0x03, 0x03, 0x04, 0x04, 0x06, 0x06, 0x28, 0x28,
-
- 0x07, 0x00, 0x00, 0x00, 0x02, 0x02, 0x27, 0x27,
-
- 0x00, 0x01, 0x01, 0x27, 0x27, 0x0C, 0x0C, 0x00, /* 13F */
-
- 0x00, 0x00, 0x00, 0x01, 0x01, 0x27, 0x27, 0x0C,
-
- 0x0C, 0x00, 0x00, 0x00, 0x04, 0x04, 0x06, 0x06,
+ while (utf8len-- > 0 && (byte = *utf8p++) != '\0') {
+ if (byte < 0x80)
+ continue; /* plain ascii */
- 0x0B, 0x0B, 0x00, 0x00, 0x01, 0x01, 0x27, 0x27,
+ extrabytes = utf_extrabytes[byte >> 3];
- 0x0C, 0x0C, 0x01, 0x01, 0x02, 0x02, 0x27, 0x27,
+ if (utf8len < extrabytes)
+ goto invalid;
+ utf8len -= extrabytes;
- 0x0C, 0x0C, 0x27, 0x27, 0x0C, 0x0C, 0x00, 0x00,
+ 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);
+}
- 0x03, 0x03, 0x04, 0x04, 0x06, 0x06, 0x0A, 0x0A, /* 16F */
+/*
+ * utf8_normalizestr - Normalize a UTF-8 string (NFC or NFD)
+ *
+ * This function takes an UTF-8 input string, instr, of inlen bytes
+ * and produces normalized UTF-8 output into a buffer of buflen bytes
+ * pointed to by outstr. The size of the output in bytes (not including
+ * a NULL termination byte) is returned in outlen. In-place conversions
+ * are not supported (i.e. instr != outstr).]
+
+ * FLAGS
+ * UTF_DECOMPOSED: output string will be fully decomposed (NFD)
+ *
+ * UTF_PRECOMPOSED: output string will be precomposed (NFC)
+ *
+ * UTF_NO_NULL_TERM: do not add null termination to output string
+ *
+ * UTF_ESCAPE_ILLEGAL: percent escape any illegal UTF-8 input
+ *
+ * ERRORS
+ * ENAMETOOLONG: output did not fit or input exceeded MAXPATHLEN bytes
+ *
+ * EINVAL: illegal UTF-8 sequence encountered or invalid flags
+ */
+int
+utf8_normalizestr(const u_int8_t* instr, size_t inlen, u_int8_t* outstr,
+ size_t *outlen, size_t buflen, int flags)
+{
+ u_int16_t unicodebuf[32];
+ u_int16_t* unistr = NULL;
+ size_t unicode_bytes;
+ size_t uft8_bytes;
+ size_t inbuflen;
+ u_int8_t *outbufstart, *outbufend;
+ const u_int8_t *inbufstart;
+ unsigned int byte;
+ int decompose, precompose;
+ int result = 0;
- 0x0B, 0x0B, 0x28, 0x28, 0x02, 0x02, 0x02, 0x02,
+ if (flags & ~(UTF_DECOMPOSED | UTF_PRECOMPOSED | UTF_NO_NULL_TERM | UTF_ESCAPE_ILLEGAL)) {
+ return (EINVAL);
+ }
+ decompose = (flags & UTF_DECOMPOSED);
+ precompose = (flags & UTF_PRECOMPOSED);
+ if ((decompose && precompose) || (!decompose && !precompose)) {
+ return (EINVAL);
+ }
+ outbufstart = outstr;
+ outbufend = outbufstart + buflen;
+ inbufstart = instr;
+ inbuflen = inlen;
+
+ while (inlen-- > 0 && (byte = *instr++) != '\0') {
+ if (outstr >= outbufend) {
+ result = ENAMETOOLONG;
+ goto exit;
+ }
+ if (byte >= 0x80) {
+ goto nonASCII;
+ }
+ /* ASCII is already normalized. */
+ *outstr++ = byte;
+ }
+exit:
+ *outlen = outstr - outbufstart;
+ if (((flags & UTF_NO_NULL_TERM) == 0)) {
+ if (outstr < outbufend)
+ *outstr++ = '\0';
+ else
+ result = ENAMETOOLONG;
+ }
+ return (result);
- 0x08, 0x01, 0x01, 0x07, 0x07, 0x0C, 0x0C, 0x00,
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 17F */
+ /*
+ * Non-ASCII uses the existing utf8_encodestr/utf8_decodestr
+ * functions to perform the normalization. Since this will
+ * presumably be used to normalize filenames in the back-end
+ * (on disk or over-the-wire), it should be fast enough.
+ */
+nonASCII:
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+ /* Make sure the input size is reasonable. */
+ if (inbuflen > MAXPATHLEN) {
+ result = ENAMETOOLONG;
+ goto exit;
+ }
+ /*
+ * Compute worst case Unicode buffer size.
+ *
+ * For pre-composed output, every UTF-8 input byte will be at
+ * most 2 Unicode bytes. For decomposed output, 2 UTF-8 bytes
+ * (smallest composite char sequence) may yield 6 Unicode bytes
+ * (1 base char + 2 combining chars).
+ */
+ unicode_bytes = precompose ? (inbuflen * 2) : (inbuflen * 3);
+
+ if (unicode_bytes <= sizeof(unicodebuf))
+ unistr = &unicodebuf[0];
+ else
+ MALLOC(unistr, u_int16_t *, unicode_bytes, M_TEMP, M_WAITOK);
+
+ /* Normalize the string. */
+ result = utf8_decodestr(inbufstart, inbuflen, unistr, &unicode_bytes,
+ unicode_bytes, 0, flags & ~UTF_NO_NULL_TERM);
+ if (result == 0) {
+ /* Put results back into UTF-8. */
+ result = utf8_encodestr(unistr, unicode_bytes, outbufstart,
+ &uft8_bytes, buflen, 0, UTF_NO_NULL_TERM);
+ outstr = outbufstart + uft8_bytes;
+ }
+ if (unistr && unistr != &unicodebuf[0]) {
+ FREE(unistr, M_TEMP);
+ }
+ goto exit;
+}
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 19F */
+ /*
+ * Unicode 3.2 decomposition code (derived from Core Foundation)
+ */
- 0x1B, 0x1B, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+typedef struct {
+ u_int32_t _key;
+ u_int32_t _value;
+} unicode_mappings32;
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x1B,
+static inline u_int32_t
+getmappedvalue32(const unicode_mappings32 *theTable, u_int32_t numElem,
+ u_int16_t character)
+{
+ const unicode_mappings32 *p, *q, *divider;
- 0x1B, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+ if ((character < theTable[0]._key) || (character > theTable[numElem-1]._key))
+ return (0);
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+ p = theTable;
+ q = p + (numElem-1);
+ while (p <= q) {
+ divider = p + ((q - p) >> 1); /* divide by 2 */
+ if (character < divider->_key) { q = divider - 1; }
+ else if (character > divider->_key) { p = divider + 1; }
+ else { return (divider->_value); }
+ }
+ return (0);
+}
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
+#define RECURSIVE_DECOMPOSITION (1 << 15)
+#define EXTRACT_COUNT(value) (((value) >> 12) & 0x0007)
- 0x00, 0x00, 0x00, 0x00, 0x00, 0x0C, 0x0C, 0x0C, /* 1CF */
+typedef struct {
+ u_int16_t _key;
+ u_int16_t _value;
+} unicode_mappings16;
- 0x0C, 0x0C, 0x0C, 0x0C, 0x0C, 0x04, 0x04, 0x01,
+static inline u_int16_t
+getmappedvalue16(const unicode_mappings16 *theTable, u_int32_t numElem,
+ u_int16_t character)
+{
+ const unicode_mappings16 *p, *q, *divider;
- 0x01, 0x0C, 0x0C, 0x00, 0x00, 0x00, 0x04, 0x04 /* 1DF */
-};
+ if ((character < theTable[0]._key) || (character > theTable[numElem-1]._key))
+ return (0);
+ p = theTable;
+ q = p + (numElem-1);
+ while (p <= q) {
+ divider = p + ((q - p) >> 1); /* divide by 2 */
+ if (character < divider->_key)
+ q = divider - 1;
+ else if (character > divider->_key)
+ p = divider + 1;
+ else
+ return (divider->_value);
+ }
+ return (0);
+}
-/* CYRILLIC codepoints 0x0400 ~ 0x04FF */
-static const unsigned long __CyrillicDecompBitmap[] = {
- 0x510A0040, 0x00000040, 0x0000510A, 0x00000000, /* 0x0400 */
- 0x00000000, 0x00000000, 0x00000000, 0x00000000, /* 0x0480 */
-};
-/* CJK codepoints 0x3000 ~ 0x30FF */
-static const unsigned long __CJKDecompBitmap[] = {
- 0x00000000, 0x00000000, 0x000AAAAA, 0xA540DB6C, /* 0x3000 */
- 0x00000802, 0x000AAAAA, 0xA540DB6C, 0x000009E2, /* 0x3080 */
-};
-#define IS_DECOMPOSABLE(table,unicodeVal) \
- (table[(unicodeVal) / 32] & (1 << (31 - ((unicodeVal) % 32))))
+static u_int32_t
+unicode_recursive_decompose(u_int16_t character, u_int16_t *convertedChars)
+{
+ u_int16_t value;
+ u_int32_t length;
+ u_int16_t firstChar;
+ u_int16_t theChar;
+ const u_int16_t *bmpMappings;
+ u_int32_t usedLength;
+
+ value = getmappedvalue16(
+ (const unicode_mappings16 *)__CFUniCharDecompositionTable,
+ __UniCharDecompositionTableLength, character);
+ length = EXTRACT_COUNT(value);
+ firstChar = value & 0x0FFF;
+ theChar = firstChar;
+ bmpMappings = (length == 1 ? &theChar : __CFUniCharMultipleDecompositionTable + firstChar);
+ usedLength = 0;
+
+ if (value & RECURSIVE_DECOMPOSITION) {
+ usedLength = unicode_recursive_decompose((u_int16_t)*bmpMappings, convertedChars);
+
+ --length; /* Decrement for the first char */
+ if (!usedLength)
+ return 0;
+ ++bmpMappings;
+ convertedChars += usedLength;
+ }
+
+ usedLength += length;
+
+ while (length--)
+ *(convertedChars++) = *(bmpMappings++);
+
+ return (usedLength);
+}
+
+#define HANGUL_SBASE 0xAC00
+#define HANGUL_LBASE 0x1100
+#define HANGUL_VBASE 0x1161
+#define HANGUL_TBASE 0x11A7
+
+#define HANGUL_SCOUNT 11172
+#define HANGUL_LCOUNT 19
+#define HANGUL_VCOUNT 21
+#define HANGUL_TCOUNT 28
+#define HANGUL_NCOUNT (HANGUL_VCOUNT * HANGUL_TCOUNT)
/*
- * ucs_decompose - decompose a composed Unicode char
+ * unicode_decompose - decompose a composed Unicode char
*
* Composed Unicode characters are forbidden on
* HFS Plus volumes. ucs_decompose will convert a
* composed character into its correct decomposed
* sequence.
*
- * Currently only Tier-1 and Tier-2 languages
- * are handled. Other composed characters are
- * passed unchanged.
+ * Similar to CFUniCharDecomposeCharacter
+ */
+static int
+unicode_decompose(u_int16_t character, u_int16_t *convertedChars)
+{
+ if ((character >= HANGUL_SBASE) &&
+ (character <= (HANGUL_SBASE + HANGUL_SCOUNT))) {
+ u_int32_t length;
+
+ character -= HANGUL_SBASE;
+ length = (character % HANGUL_TCOUNT ? 3 : 2);
+
+ *(convertedChars++) =
+ character / HANGUL_NCOUNT + HANGUL_LBASE;
+ *(convertedChars++) =
+ (character % HANGUL_NCOUNT) / HANGUL_TCOUNT + HANGUL_VBASE;
+ if (length > 2)
+ *convertedChars = (character % HANGUL_TCOUNT) + HANGUL_TBASE;
+ return (length);
+ } else {
+ return (unicode_recursive_decompose(character, convertedChars));
+ }
+}
+
+/*
+ * unicode_combine - generate a precomposed Unicode char
+ *
+ * Precomposed Unicode characters are required for some volume
+ * formats and network protocols. unicode_combine will combine
+ * a decomposed character sequence into a single precomposed
+ * (composite) character.
+ *
+ * Similar toCFUniCharPrecomposeCharacter but unicode_combine
+ * also handles Hangul Jamo characters.
*/
static u_int16_t
-ucs_decompose(register u_int16_t ch, u_int16_t *cmb)
+unicode_combine(u_int16_t base, u_int16_t combining)
{
- u_int16_t base;
+ u_int32_t value;
+
+ /* Check HANGUL */
+ if ((combining >= HANGUL_VBASE) && (combining < (HANGUL_TBASE + HANGUL_TCOUNT))) {
+ /* 2 char Hangul sequences */
+ if ((combining < (HANGUL_VBASE + HANGUL_VCOUNT)) &&
+ (base >= HANGUL_LBASE && base < (HANGUL_LBASE + HANGUL_LCOUNT))) {
+ return (HANGUL_SBASE +
+ ((base - HANGUL_LBASE)*(HANGUL_VCOUNT*HANGUL_TCOUNT)) +
+ ((combining - HANGUL_VBASE)*HANGUL_TCOUNT));
+ }
- cmb[0] = 0;
- cmb[1] = 0;
+ /* 3 char Hangul sequences */
+ if ((combining > HANGUL_TBASE) &&
+ (base >= HANGUL_SBASE && base < (HANGUL_SBASE + HANGUL_SCOUNT))) {
+ if ((base - HANGUL_SBASE) % HANGUL_TCOUNT)
+ return (0);
+ else
+ return (base + (combining - HANGUL_TBASE));
+ }
+ }
- if (ch < 0x00C0) {
- base = ch;
- } else if (ch <= 0x01DF) {
-
- base = (u_int16_t) primary_char[ch - 0x00C0];
+ value = getmappedvalue32(
+ (const unicode_mappings32 *)__CFUniCharPrecompSourceTable,
+ __CFUniCharPrecompositionTableLength, combining);
- if (base == 0)
- base = ch;
- else {
- if ((base < 0x00C0) || (primary_char[base - 0x00C0] == 0))
- cmb[0] = (u_int16_t)0x0300 + (u_int16_t)combining_char[ch - 0x00C0];
- else {
- u_int16_t tch = base;
-
- base = (u_int16_t)primary_char[tch - 0x00C0];
- cmb[0] = (u_int16_t)0x0300 + (u_int16_t)combining_char[tch - 0x00C0];
- cmb[1] = (u_int16_t)0x0300 + (u_int16_t)combining_char[ch - 0x00C0];
+ if (value) {
+ value = getmappedvalue16(
+ (const unicode_mappings16 *)
+ ((const u_int32_t *)__CFUniCharBMPPrecompDestinationTable + (value & 0xFFFF)),
+ (value >> 16), base);
+ }
+ 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 = 0;
+
+ 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 && p2 != 0) {
+ u_int32_t tmp;
+
+ tmp = *ch1;
+ *ch1 = *ch2;
+ *ch2 = tmp;
+ changes = 1;
+
+ /*
+ * Make sure that p2 contains the combining class for the
+ * character now stored at *ch2. This isn't required for
+ * correctness, but it will be more efficient if a character
+ * with a large combining class has to "bubble past" several
+ * characters with lower combining classes.
+ */
+ p2 = p1;
}
+ ++ch1;
+ ++ch2;
}
- } else if ((ch >= 0x0400) && (ch <= 0x04FF) &&
- IS_DECOMPOSABLE(__CyrillicDecompBitmap, ch - 0x0400)) {
-
- /* Handle CYRILLIC LETTERs */
- switch(ch) {
- case 0x0401: base = 0x0415; cmb[0] = 0x0308; break; /* */
- case 0x0403: base = 0x0413; cmb[0] = 0x0301; break; /* */
- case 0x0407: base = 0x0406; cmb[0] = 0x0308; break; /* */
- case 0x040C: base = 0x041A; cmb[0] = 0x0301; break; /* */
- case 0x040E: base = 0x0423; cmb[0] = 0x0306; break; /* */
- case 0x0419: base = 0x0418; cmb[0] = 0x0306; break; /* */
- case 0x0439: base = 0x0438; cmb[0] = 0x0306; break; /* */
- case 0x0451: base = 0x0435; cmb[0] = 0x0308; break; /* */
- case 0x0453: base = 0x0433; cmb[0] = 0x0301; break; /* */
- case 0x0457: base = 0x0456; cmb[0] = 0x0308; break; /* */
- case 0x045C: base = 0x043A; cmb[0] = 0x0301; break; /* */
- case 0x045E: base = 0x0443; cmb[0] = 0x0306; break; /* */
-
- default:
- /* Should not be hit from bit map table */
- base = ch;
- }
- } else if (ch == 0x1E3F) {
- base = 0x006D; cmb[0] = 0x0301; /* LATIN SMALL LETTER M WITH ACUTE */
- } else if ((ch > 0x3000) && (ch < 0x3100) &&
- IS_DECOMPOSABLE(__CJKDecompBitmap, ch - 0x3000)) {
-
- /* Handle HIRAGANA LETTERs */
- switch(ch) {
- case 0x3071: base = 0x306F; cmb[0] = 0x309A; break; /* PA */
- case 0x3074: base = 0x3072; cmb[0] = 0x309A; break; /* PI */
- case 0x3077: base = 0x3075; cmb[0] = 0x309A; break; /* PU */
- case 0x307A: base = 0x3078; cmb[0] = 0x309A; break; /* PE */
-
- case 0x307D: base = 0x307B; cmb[0] = 0x309A; break; /* PO */
- case 0x3094: base = 0x3046; cmb[0] = 0x3099; break; /* VU */
- case 0x30D1: base = 0x30CF; cmb[0] = 0x309A; break; /* PA */
- case 0x30D4: base = 0x30D2; cmb[0] = 0x309A; break; /* PI */
-
- case 0x30D7: base = 0x30D5; cmb[0] = 0x309A; break; /* PU */
- case 0x30DA: base = 0x30D8; cmb[0] = 0x309A; break; /* PE */
- case 0x30DD: base = 0x30DB; cmb[0] = 0x309A; break; /* PO */
- case 0x30F4: base = 0x30A6; cmb[0] = 0x3099; break; /* VU */
-
- case 0x30F7: base = 0x30EF; cmb[0] = 0x3099; break; /* VA */
- case 0x30F8: base = 0x30F0; cmb[0] = 0x3099; break; /* VI */
- case 0x30F9: base = 0x30F1; cmb[0] = 0x3099; break; /* VE */
- case 0x30FA: base = 0x30F2; cmb[0] = 0x3099; break; /* VO */
-
- default:
- /* the rest (41 of them) have a simple conversion */
- base = ch - 1;
- cmb[0] = 0x3099;
- }
- } else if ((ch >= 0xAC00) && (ch < 0xD7A4)) {
- /* Hangul */
- ch -= 0xAC00;
- base = 0x1100 + (ch / (21*28));
- cmb[0] = 0x1161 + (ch % (21*28)) / 28;
-
- if (ch % 28)
- cmb[1] = 0x11A7 + (ch % 28);
- } else {
- base = ch;
- }
-
- return (base);
+ } while (changes);
}
-static const short diacrit_tbl[8*6] = {
- /* 300 - 307 */ 0, 58, 116, 174, 232, -1, 290, 348,
- /* 308 - 30F */ 406, -1, 464, 522, 580, -1, -1, -1,
- /* 310 - 317 */ -1, -1, -1, -1, -1, -1, -1, -1,
- /* 318 - 31F */ -1, -1, -1, 638, -1, -1, -1, -1,
- /* 320 - 327 */ -1, -1, -1, -1, -1, -1, -1, 696,
- /* 328 - 32F */ 754, -1, -1, -1, -1, -1, -1, -1
-};
+/*
+ * Invalid NTFS filename characters are encodeded using the
+ * SFM (Services for Macintosh) private use Unicode characters.
+ *
+ * These should only be used for SMB, MSDOS or NTFS.
+ *
+ * Illegal NTFS Char SFM Unicode Char
+ * ----------------------------------------
+ * 0x01-0x1f 0xf001-0xf01f
+ * '"' 0xf020
+ * '*' 0xf021
+ * '/' 0xf022
+ * '<' 0xf023
+ * '>' 0xf024
+ * '?' 0xf025
+ * '\' 0xf026
+ * '|' 0xf027
+ * ' ' 0xf028 (Only if last char of the name)
+ * '.' 0xf029 (Only if last char of the name)
+ * ----------------------------------------
+ *
+ * Reference: http://support.microsoft.com/kb/q117258/
+ */
-static const u_int16_t composite_tbl[58*14] = {
- /*
- * A B C D E F G H I J K L M
- * N O P Q R S T U V W X Y Z
- * [ \ ] ^ _ `
- * a b c d e f g h i j k l m
- * n o p q r s t u v w x y z
- */
+#define MAX_SFM2MAC 0x29
+#define SFMCODE_PREFIX_MASK 0xf000
- /*
- * 0x300 - grave accent
- */
- 0x0C0, 0, 0, 0,0x0C8, 0, 0, 0,0x0CC, 0, 0, 0, 0,
- 0,0x0D2, 0, 0, 0, 0, 0,0x0D9, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x0E0, 0, 0, 0,0x0E8, 0, 0, 0,0x0EC, 0, 0, 0, 0,
- 0,0x0F2, 0, 0, 0, 0, 0,0x0F9, 0, 0, 0, 0, 0,
- /*
- * 0x301 - acute accent
- */
- 0x0C1, 0,0x106, 0,0x0C9, 0, 0, 0,0x0CD, 0, 0,0x139, 0,
- 0x143,0x0D3, 0, 0,0x154,0x15A, 0,0x0DA, 0, 0, 0,0x0DD,0x179,
- 0, 0, 0, 0, 0, 0,
- 0x0E1, 0,0x107, 0,0x0E9, 0, 0, 0,0x0ED, 0, 0,0x13A,0x1E3F,
- 0x144,0x0F3, 0, 0,0x155,0x15B, 0,0x0FA, 0, 0, 0,0x0FD,0x17A,
- /*
- * 0x302 - circumflex accent
- */
- 0x0C2, 0,0x108, 0,0x0CA, 0,0x11C,0x124,0x0CE,0x134, 0, 0, 0,
- 0,0x0D4, 0, 0, 0,0x15C, 0,0x0DB, 0,0x174, 0,0x176, 0,
- 0, 0, 0, 0, 0, 0,
- 0x0E2, 0,0x109, 0,0x0EA, 0,0x11D,0x125,0x0EE,0x135, 0, 0, 0,
- 0,0x0F4, 0, 0, 0,0x15D, 0,0x0FB, 0,0x175, 0,0x177, 0,
- /*
- * 0x303 - tilde
- */
- 0x0C3, 0, 0, 0, 0, 0, 0, 0,0x128, 0, 0, 0, 0,
- 0x0D1,0x0D5, 0, 0, 0, 0, 0,0x168, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x0E3, 0, 0, 0, 0, 0, 0, 0,0x129, 0, 0, 0, 0,
- 0x0F1,0x0F5, 0, 0, 0, 0, 0,0x169, 0, 0, 0, 0, 0,
- /*
- * 0x304 - macron
- */
- 0x100, 0, 0, 0,0x112, 0, 0, 0,0x12A, 0, 0, 0, 0,
- 0,0x14C, 0, 0, 0, 0, 0,0x16A, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x101, 0, 0, 0,0x113, 0, 0, 0,0x12B, 0, 0, 0, 0,
- 0,0x14D, 0, 0, 0, 0, 0,0x16B, 0, 0, 0, 0, 0,
- /*
- * 0x306 - breve
- */
- 0x102, 0, 0, 0,0x114, 0,0x11E, 0,0x12C, 0, 0, 0, 0,
- 0,0x14E, 0, 0, 0, 0, 0,0x16C, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x103, 0, 0, 0,0x115, 0,0x11F, 0,0x12D, 0, 0, 0, 0,
- 0,0x14F, 0, 0, 0, 0, 0,0x16D, 0, 0, 0, 0, 0,
- /*
- * 0x307 - dot above
- */
- 0, 0,0x10A, 0,0x116, 0,0x120, 0,0x130, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0x17B,
- 0, 0, 0, 0, 0, 0,
- 0, 0,0x10B, 0,0x117, 0,0x121, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0x17C,
- /*
- * 0x308 - diaeresis
- */
- 0x0C4, 0, 0, 0,0x0CB, 0, 0, 0,0x0CF, 0, 0, 0, 0,
- 0,0x0D6, 0, 0, 0, 0, 0,0x0DC, 0, 0, 0,0x178, 0,
- 0, 0, 0, 0, 0, 0,
- 0x0E4, 0, 0, 0,0x0EB, 0, 0, 0,0x0EF, 0, 0, 0, 0,
- 0,0x0F6, 0, 0, 0, 0, 0,0x0FC, 0, 0, 0,0x0FF, 0,
- /*
- * 0x30A - ring above
- */
- 0x0C5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0,0x16E, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x0E5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0,0x16F, 0, 0, 0, 0, 0,
- /*
- * 0x30B - double aute accent
- */
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0,0x150, 0, 0, 0, 0, 0,0x170, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0,0x151, 0, 0, 0, 0, 0,0x171, 0, 0, 0, 0, 0,
- /*
- * 0x30C - caron
- */
- 0x1CD, 0,0x10C,0x10E,0x11A, 0, 0, 0,0x1CF, 0, 0,0x13D, 0,
- 0x147,0x1D1, 0, 0,0x158,0x160,0x164,0x1D3, 0, 0, 0, 0,0x17D,
- 0, 0, 0, 0, 0, 0,
- 0x1CE, 0,0x10D,0x10F,0x11B, 0, 0, 0,0x1D0, 0, 0,0x13E, 0,
- 0x148,0x1D2, 0, 0,0x159,0x161,0x165,0x1D4, 0, 0, 0, 0,0x17E,
- /*
- * 0x31B - horn
- */
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0,0x1A0, 0, 0, 0, 0, 0,0x1AF, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
- 0,0x1A1, 0, 0, 0, 0, 0,0x1B0, 0, 0, 0, 0, 0,
- /*
- * 0x327 - cedilla
- */
- 0, 0,0x0C7, 0, 0, 0,0x122, 0, 0, 0,0x136,0x13B, 0,
- 0x145, 0, 0, 0,0x156,0x15E,0x162, 0, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0, 0,0x0E7, 0, 0, 0,0x123, 0, 0, 0,0x137,0x13C, 0,
- 0x146, 0, 0, 0,0x157,0x15F,0x163, 0, 0, 0, 0, 0, 0,
- /*
- * 0x328 - ogonek
- */
- 0x104, 0, 0, 0,0x118, 0, 0, 0,0x12E, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0,0x172, 0, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0,
- 0x105, 0, 0, 0,0x119, 0, 0, 0,0x12F, 0, 0, 0, 0,
- 0, 0, 0, 0, 0, 0, 0,0x173, 0, 0, 0, 0, 0,
+/*
+ * In the Mac OS 9 days the colon was illegal in a file name. For that reason
+ * SFM had no conversion for the colon. There is a conversion for the
+ * slash. In Mac OS X the slash is illegal in a file name. So for us the colon
+ * is a slash and a slash is a colon. So we can just replace the slash with the
+ * colon in our tables and everything will just work.
+ */
+static u_int8_t
+sfm2mac[42] = {
+ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, /* 00 - 07 */
+ 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, /* 08 - 0F */
+ 0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, /* 10 - 17 */
+ 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f, /* 18 - 1F */
+ 0x22, 0x2a, 0x3a, 0x3c, 0x3e, 0x3f, 0x5c, 0x7c, /* 20 - 27 */
+ 0x20, 0x2e /* 28 - 29 */
};
-
-/* CJK codepoints 0x3000 ~ 0x30FF */
-static const unsigned long __CJKCombBitmap[] = {
- 0x00000000, 0x00000000, 0x02155555, 0x4A812490, /* 0x3000 */
- 0x00000004, 0x02155555, 0x4A812490, 0x0001E004, /* 0x3080 */
+static u_int8_t
+mac2sfm[112] = {
+ 0x20, 0x21, 0x20, 0x23, 0x24, 0x25, 0x26, 0x27, /* 20 - 27 */
+ 0x28, 0x29, 0x21, 0x2b, 0x2c, 0x2d, 0x2e, 0x22, /* 28 - 2f */
+ 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, /* 30 - 37 */
+ 0x38, 0x39, 0x22, 0x3b, 0x23, 0x3d, 0x24, 0x25, /* 38 - 3f */
+ 0x40, 0x41, 0x42, 0x43, 0x44, 0x45, 0x46, 0x47, /* 40 - 47 */
+ 0x48, 0x49, 0x4a, 0x4b, 0x4c, 0x4d, 0x4e, 0x4f, /* 48 - 4f */
+ 0x50, 0x51, 0x52, 0x53, 0x54, 0x55, 0x56, 0x57, /* 50 - 57 */
+ 0x58, 0x59, 0x5a, 0x5b, 0x26, 0x5d, 0x5e, 0x5f, /* 58 - 5f */
+ 0x60, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, /* 60 - 67 */
+ 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f, /* 68 - 6f */
+ 0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, /* 70 - 77 */
+ 0x78, 0x79, 0x7a, 0x7b, 0x27, 0x7d, 0x7e, 0x7f /* 78 - 7f */
};
-#define CAN_COMBINE(table,unicodeVal) \
- (table[(unicodeVal) / 32] & (1 << (31 - ((unicodeVal) % 32))))
/*
- * ucs_combine - generate a precomposed Unicode char
+ * Encode illegal NTFS filename characters into SFM Private Unicode characters
*
- * Precomposed Unicode characters are required for some volume
- * formats and network protocols. ucs_combine will combine a
- * decomposed character sequence into a single precomposed
- * (composite) character.
- *
- * Currently only decomcomposed sequences from Apple's Tier 1
- * and Tier 2 languages are handled.
- *
- * INPUT:
- * base - base character
- * comb - combining character
- * OUTPUT:
- * result - precomposed char or zero if not combinable
+ * Assumes non-zero ASCII input.
*/
static u_int16_t
-ucs_combine(u_int16_t base, u_int16_t comb)
+ucs_to_sfm(u_int16_t ucs_ch, int lastchar)
{
- /* Get out early if we can */
- if (comb < 0x0300)
- return (0);
-
- /* Try ordinary diacritics (0x300 - 0x32F) */
- if (comb <= 0x032F) {
- int index;
-
- if (base >= 'A' && base <= 'z') {
- index = diacrit_tbl[comb - 0x0300];
- if (index < 0 ) return (0);
-
- return (composite_tbl[index + (base - 'A')]);
- }
-
- /* Handle Cyrillic and some 3 char latin sequences */
- switch (comb) {
- case 0x0300:
- switch (base) {
- case 0x00DC: return (0x01DB);
- case 0x00FC: return (0x01DC);
- } break;
- case 0x0301:
- switch (base) {
- case 0x00DC: return (0x01D7);
- case 0x00FC: return (0x01D8);
- case 0x0413: return (0x0403);
- case 0x041A: return (0x040C);
- case 0x0433: return (0x0453);
- case 0x043A: return (0x045C);
- } break;
- case 0x0304:
- switch (base) {
- case 0x00DC: return (0x01D5);
- case 0x00FC: return (0x01D6);
- case 0x00C4: return (0x01DE);
- case 0x00E4: return (0x01DF);
- } break;
- case 0x0306:
- switch (base) {
- case 0x0418: return (0x0419);
- case 0x0423: return (0x040E);
- case 0x0438: return (0x0439);
- case 0x0443: return (0x045E);
- } break;
- case 0x0308:
- switch (base) {
- case 0x0406: return (0x0407);
- case 0x0415: return (0x0401);
- case 0x0435: return (0x0451);
- case 0x0456: return (0x0457);
- } break;
- case 0x030C:
- switch (base) {
- case 0x00DC: return (0x01D9);
- case 0x00FC: return (0x01DA);
- } break;
- }
- return (0);
+ /* The last character of filename cannot be a space or period. */
+ if (lastchar) {
+ if (ucs_ch == 0x20)
+ return (0xf028);
+ else if (ucs_ch == 0x2e)
+ return (0xf029);
}
-
- /* Now try HANGUL */
- if (comb < 0x1161)
- return (0);
-
- /* 2 char Hangul sequences */
- if ((comb <= 0x1175) && (base >= 0x1100 && base <= 0x1112))
- return (0xAC00 + ((base - 0x1100)*(21*28)) + ((comb - 0x1161)*28));
-
- /* 3 char Hangul sequences */
- if ((comb >= 0x11A8 && comb <= 0x11C2) &&
- (base >= 0xAC00 && base <= 0xD788)) {
- if ((base - 0xAC00) % 28)
- return (0);
- else
- return (base + (comb - 0x11A7));
+ /* 0x01 - 0x1f is simple transformation. */
+ if (ucs_ch <= 0x1f) {
+ return (ucs_ch | 0xf000);
+ } else /* 0x20 - 0x7f */ {
+ u_int16_t lsb;
+
+ lsb = mac2sfm[ucs_ch - 0x0020];
+ if (lsb != ucs_ch)
+ return(0xf000 | lsb);
}
+ return (ucs_ch);
+}
- /* Now try HIRAGANA and KATAKANA */
- if ((comb == 0x3099 || comb == 0x309A) &&
- (base > 0x3000 && base < 0x3100) &&
- CAN_COMBINE(__CJKCombBitmap, base - 0x3000)) {
- if (comb == 0x309A) {
- switch(base) {
- case 0x306F: return (0x3071); /* PA */
- case 0x3072: return (0x3074); /* PI */
- case 0x3075: return (0x3077); /* PU */
- case 0x3078: return (0x307A); /* PE */
- case 0x307B: return (0x307D); /* PO */
- case 0x30CF: return (0x30D1); /* PA */
- case 0x30D2: return (0x30D4); /* PI */
- case 0x30D5: return (0x30D7); /* PU */
- case 0x30D8: return (0x30DA); /* PE */
- case 0x30DB: return (0x30DD); /* PO */
- default: return (0);
- }
- } else /* 0x3099 */ {
- switch (base) {
- case 0x3046: return (0x3094); /* VU */
- case 0x30A6: return (0x30F4); /* VU */
- case 0x30EF: return (0x30F7); /* VA */
- case 0x30F0: return (0x30F8); /* VI */
- case 0x30F1: return (0x30F9); /* VE */
- case 0x30F2: return (0x30FA); /* VO */
- default: return (base + 1); /* 41 code points here */
- }
- }
+/*
+ * Decode any SFM Private Unicode characters
+ */
+static u_int16_t
+sfm_to_ucs(u_int16_t ucs_ch)
+{
+ if (((ucs_ch & 0xffC0) == SFMCODE_PREFIX_MASK) &&
+ ((ucs_ch & 0x003f) <= MAX_SFM2MAC)) {
+ ucs_ch = sfm2mac[ucs_ch & 0x003f];
}
-
- return (0);
+ return (ucs_ch);
}
+
projects / apple / xnu.git / blobdiff