CFUnicodePrecomposition.c

   1 /*
   2  * Copyright (c) 2014 Apple Inc. All rights reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License. Please obtain a copy of the License at
  10  * http://www.opensource.apple.com/apsl/ and read it before using this
  11  * file.
  12  *
  13  * The Original Code and all software distributed under the License are
  14  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  15  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  16  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  17  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  18  * Please see the License for the specific language governing rights and
  19  * limitations under the License.
  20  *
  21  * @APPLE_LICENSE_HEADER_END@
  22  */
  23
  24 /*      CFUnicodePrecomposition.c
  25         Copyright (c) 1999-2014, Apple Inc. All rights reserved.
  26         Responsibility: Aki Inoue
  27 */
  28
  29 #include <string.h>
  30 #include <CoreFoundation/CFBase.h>
  31 #include <CoreFoundation/CFCharacterSet.h>
  32 #include "CFUniChar.h"
  33 #include "CFUnicodePrecomposition.h"
  34 #include "CFInternal.h"
  35 #include "CFUniCharPriv.h"
  36
  37 // Canonical Precomposition
  38 static UTF32Char *__CFUniCharPrecompSourceTable = NULL;
  39 static uint32_t __CFUniCharPrecompositionTableLength = 0;
  40 static uint16_t *__CFUniCharBMPPrecompDestinationTable = NULL;
  41 static uint32_t *__CFUniCharNonBMPPrecompDestinationTable = NULL;
  42
  43 static const uint8_t *__CFUniCharNonBaseBitmapForBMP_P = NULL; // Adding _P so the symbol name is different from the one in CFUnicodeDecomposition.c
  44 static const uint8_t *__CFUniCharCombiningClassForBMP = NULL;
  45
  46 static CFLock_t __CFUniCharPrecompositionTableLock = CFLockInit;
  47
  48 static void __CFUniCharLoadPrecompositionTable(void) {
  49
  50     __CFLock(&__CFUniCharPrecompositionTableLock);
  51
  52     if (NULL == __CFUniCharPrecompSourceTable) {
  53         const uint32_t *bytes = (const uint32_t *)CFUniCharGetMappingData(kCFUniCharCanonicalPrecompMapping);
  54         uint32_t bmpMappingLength;
  55
  56         if (NULL == bytes) {
  57             __CFUnlock(&__CFUniCharPrecompositionTableLock);
  58             return;
  59         }
  60
  61         __CFUniCharPrecompositionTableLength = *(bytes++);
  62         bmpMappingLength = *(bytes++);
  63         __CFUniCharPrecompSourceTable = (UTF32Char *)bytes;
  64         __CFUniCharBMPPrecompDestinationTable = (uint16_t *)((intptr_t)bytes + (__CFUniCharPrecompositionTableLength * sizeof(UTF32Char) * 2));
  65         __CFUniCharNonBMPPrecompDestinationTable = (uint32_t *)(((intptr_t)__CFUniCharBMPPrecompDestinationTable) + bmpMappingLength);
  66
  67         __CFUniCharNonBaseBitmapForBMP_P = CFUniCharGetBitmapPtrForPlane(kCFUniCharNonBaseCharacterSet, 0);
  68         __CFUniCharCombiningClassForBMP = (const uint8_t *)CFUniCharGetUnicodePropertyDataForPlane(kCFUniCharCombiningProperty, 0);
  69     }
  70
  71     __CFUnlock(&__CFUniCharPrecompositionTableLock);
  72 }
  73
  74  // Adding _P so the symbol name is different from the one in CFUnicodeDecomposition.c
  75 #define __CFUniCharIsNonBaseCharacter   __CFUniCharIsNonBaseCharacter_P
  76 CF_INLINE bool __CFUniCharIsNonBaseCharacter(UTF32Char character) {
  77     return CFUniCharIsMemberOfBitmap(character, (character < 0x10000 ? __CFUniCharNonBaseBitmapForBMP_P : CFUniCharGetBitmapPtrForPlane(kCFUniCharNonBaseCharacterSet, ((character >> 16) & 0xFF))));
  78 }
  79
  80 typedef struct {
  81     UTF16Char _key;
  82     UTF16Char _value;
  83 } __CFUniCharPrecomposeBMPMappings;
  84
  85 static UTF16Char __CFUniCharGetMappedBMPValue(const __CFUniCharPrecomposeBMPMappings *theTable, uint32_t numElem, UTF16Char character) {
  86     const __CFUniCharPrecomposeBMPMappings *p, *q, *divider;
  87
  88     if ((character < theTable[0]._key) || (character > theTable[numElem-1]._key)) {
  89         return 0;
  90     }
  91     p = theTable;
  92     q = p + (numElem-1);
  93     while (p <= q) {
  94         divider = p + ((q - p) >> 1);   /* divide by 2 */
  95         if (character < divider->_key) { q = divider - 1; }
  96         else if (character > divider->_key) { p = divider + 1; }
  97         else { return divider->_value; }
  98     }
  99     return 0;
 100 }
 101
 102 typedef struct {
 103     UTF32Char _key;
 104     uint32_t _value;
 105 } __CFUniCharPrecomposeMappings;
 106
 107 static uint32_t __CFUniCharGetMappedValue_P(const __CFUniCharPrecomposeMappings *theTable, uint32_t numElem, UTF32Char character) {
 108     const __CFUniCharPrecomposeMappings *p, *q, *divider;
 109
 110     if ((character < theTable[0]._key) || (character > theTable[numElem-1]._key)) {
 111         return 0;
 112     }
 113     p = theTable;
 114     q = p + (numElem-1);
 115     while (p <= q) {
 116         divider = p + ((q - p) >> 1);   /* divide by 2 */
 117         if (character < divider->_key) { q = divider - 1; }
 118         else if (character > divider->_key) { p = divider + 1; }
 119         else { return divider->_value; }
 120     }
 121     return 0;
 122 }
 123
 124 CF_PRIVATE
 125 UTF32Char CFUniCharPrecomposeCharacter(UTF32Char base, UTF32Char combining) {
 126     uint32_t value;
 127
 128     if (NULL == __CFUniCharPrecompSourceTable) __CFUniCharLoadPrecompositionTable();
 129
 130     if (!(value = __CFUniCharGetMappedValue_P((const __CFUniCharPrecomposeMappings *)__CFUniCharPrecompSourceTable, __CFUniCharPrecompositionTableLength, combining))) return 0xFFFD;
 131
 132     // We don't have precomposition in non-BMP
 133     if (value & kCFUniCharNonBmpFlag) {
 134         value = __CFUniCharGetMappedValue_P((const __CFUniCharPrecomposeMappings *)((uint32_t *)__CFUniCharNonBMPPrecompDestinationTable + (value & 0xFFFF)), (value >> 16) & 0x7FFF, base);
 135     } else {
 136         value = __CFUniCharGetMappedBMPValue((const __CFUniCharPrecomposeBMPMappings *)((uint32_t *)__CFUniCharBMPPrecompDestinationTable + (value & 0xFFFF)), (value >> 16), base);
 137     }
 138     return (value ? value : 0xFFFD);
 139 }
 140
 141 #define HANGUL_SBASE 0xAC00
 142 #define HANGUL_LBASE 0x1100
 143 #define HANGUL_VBASE 0x1161
 144 #define HANGUL_TBASE 0x11A7
 145 #define HANGUL_SCOUNT 11172
 146 #define HANGUL_LCOUNT 19
 147 #define HANGUL_VCOUNT 21
 148 #define HANGUL_TCOUNT 28
 149 #define HANGUL_NCOUNT (HANGUL_VCOUNT * HANGUL_TCOUNT)
 150
 151 CF_INLINE void __CFUniCharMoveBufferFromEnd0(UTF16Char *convertedChars, CFIndex length, CFIndex delta) {
 152     const UTF16Char *limit = convertedChars;
 153     UTF16Char *dstP;
 154
 155     convertedChars += length;
 156     dstP = convertedChars + delta;
 157
 158     while (convertedChars > limit) *(--dstP) = *(--convertedChars);
 159 }
 160
 161 bool CFUniCharPrecompose(const UTF16Char *characters, CFIndex length, CFIndex *consumedLength, UTF16Char *precomposed, CFIndex maxLength, CFIndex *filledLength) {
 162     UTF32Char currentChar = 0, lastChar = 0, precomposedChar = 0xFFFD;
 163     CFIndex originalLength = length, usedLength = 0;
 164     UTF16Char *currentBase = precomposed;
 165     uint8_t currentClass, lastClass = 0;
 166     bool currentBaseIsBMP = true;
 167     bool isPrecomposed;
 168
 169     if (NULL == __CFUniCharPrecompSourceTable) __CFUniCharLoadPrecompositionTable();
 170
 171     while (length > 0) {
 172         currentChar = *(characters++);
 173         --length;
 174
 175         if (CFUniCharIsSurrogateHighCharacter(currentChar) && (length > 0) && CFUniCharIsSurrogateLowCharacter(*characters)) {
 176             currentChar = CFUniCharGetLongCharacterForSurrogatePair(currentChar, *(characters++));
 177             --length;
 178         }
 179
 180         if (lastChar && __CFUniCharIsNonBaseCharacter(currentChar)) {
 181             isPrecomposed = (precomposedChar == 0xFFFD ? false : true);
 182             if (isPrecomposed) lastChar = precomposedChar;
 183
 184             currentClass = (currentChar > 0xFFFF ? CFUniCharGetUnicodeProperty(currentChar, kCFUniCharCombiningProperty) : CFUniCharGetCombiningPropertyForCharacter(currentChar, __CFUniCharCombiningClassForBMP));
 185
 186             if ((lastClass == 0) || (currentClass > lastClass)) {
 187                 if ((precomposedChar = CFUniCharPrecomposeCharacter(lastChar, currentChar)) == 0xFFFD) {
 188                     if (isPrecomposed) precomposedChar = lastChar;
 189                     lastClass = currentClass;
 190                 } else {
 191                     continue;
 192                 }
 193             }
 194             if (currentChar > 0xFFFF) { // Non-BMP
 195                 usedLength += 2;
 196                 if (usedLength > maxLength) break;
 197                 currentChar -= 0x10000;
 198                 *(precomposed++) = (UTF16Char)((currentChar >> 10) + 0xD800UL);
 199                 *(precomposed++) = (UTF16Char)((currentChar & 0x3FF) + 0xDC00UL);
 200             } else {
 201                 ++usedLength;
 202                 if (usedLength > maxLength) break;
 203                 *(precomposed++) = (UTF16Char)currentChar;
 204             }
 205         } else {
 206             if ((currentChar >= HANGUL_LBASE) && (currentChar < (HANGUL_LBASE + 0xFF))) { // Hangul Jamo
 207                 int8_t lIndex = currentChar - HANGUL_LBASE;
 208
 209                 if ((length > 0) && (0 <= lIndex) && (lIndex <= HANGUL_LCOUNT)) {
 210                     int16_t vIndex = *characters - HANGUL_VBASE;
 211
 212                     if ((vIndex >= 0) && (vIndex <= HANGUL_VCOUNT)) {
 213                         int16_t tIndex = 0;
 214
 215                         ++characters; --length;
 216
 217                         if (length > 0) {
 218                             tIndex = *characters - HANGUL_TBASE;
 219                             if ((tIndex < 0) || (tIndex > HANGUL_TCOUNT)) {
 220                                 tIndex = 0;
 221                             } else {
 222                                 ++characters; --length;
 223                             }
 224                         }
 225                         currentChar = (lIndex * HANGUL_VCOUNT + vIndex) * HANGUL_TCOUNT + tIndex + HANGUL_SBASE;
 226                     }
 227                 }
 228             }
 229
 230             if (precomposedChar != 0xFFFD) {
 231                 if (currentBaseIsBMP) { // Non-BMP
 232                     if (lastChar > 0xFFFF) { // Last char was Non-BMP
 233                         --usedLength;
 234                         memmove(currentBase + 1, currentBase + 2, (precomposed - (currentBase + 2)) * sizeof(UTF16Char));
 235                     }
 236                     *(currentBase) = (UTF16Char)precomposedChar;
 237                 } else {
 238                     if (lastChar < 0x10000) { // Last char was BMP
 239                         ++usedLength;
 240                         if (usedLength > maxLength) break;
 241                         __CFUniCharMoveBufferFromEnd0(currentBase + 1, precomposed - (currentBase + 1), 1);
 242                     }
 243                     precomposedChar -= 0x10000;
 244                     *currentBase = (UTF16Char)((precomposedChar >> 10) + 0xD800UL);
 245                     *(currentBase + 1) = (UTF16Char)((precomposedChar & 0x3FF) + 0xDC00UL);
 246                 }
 247                 precomposedChar = 0xFFFD;
 248             }
 249             currentBase = precomposed;
 250
 251             lastChar = currentChar;
 252             lastClass = 0;
 253
 254             if (currentChar > 0xFFFF) { // Non-BMP
 255                 usedLength += 2;
 256                 if (usedLength > maxLength) break;
 257                 currentChar -= 0x10000;
 258                 *(precomposed++) = (UTF16Char)((currentChar >> 10) + 0xD800UL);
 259                 *(precomposed++) = (UTF16Char)((currentChar & 0x3FF) + 0xDC00UL);
 260                 currentBaseIsBMP = false;
 261             } else {
 262                 ++usedLength;
 263                 if (usedLength > maxLength) break;
 264                 *(precomposed++) = (UTF16Char)currentChar;
 265                 currentBaseIsBMP = true;
 266             }
 267         }
 268     }
 269
 270     if (precomposedChar != 0xFFFD) {
 271         if (currentChar > 0xFFFF) { // Non-BMP
 272             if (lastChar < 0x10000) { // Last char was BMP
 273                 ++usedLength;
 274                 if (usedLength > maxLength) {
 275                     if (consumedLength) *consumedLength = originalLength - length;
 276                     if (filledLength) *filledLength = usedLength;
 277                     return false;
 278                 }
 279                 __CFUniCharMoveBufferFromEnd0(currentBase + 1, precomposed - (currentBase + 1), 1);
 280             }
 281             precomposedChar -= 0x10000;
 282             *currentBase = (UTF16Char)((precomposedChar >> 10) + 0xD800UL);
 283             *(currentBase + 1) = (UTF16Char)((precomposedChar & 0x3FF) + 0xDC00UL);
 284         } else {
 285             if (lastChar > 0xFFFF) { // Last char was Non-BMP
 286                 --usedLength;
 287                 memmove(currentBase + 1, currentBase + 2, (precomposed - (currentBase + 2)) * sizeof(UTF16Char));
 288             }
 289             *(currentBase) = (UTF16Char)precomposedChar;
 290         }
 291     }
 292
 293     if (consumedLength) *consumedLength = originalLength - length;
 294     if (filledLength) *filledLength = usedLength;
 295
 296     return true;
 297 }
 298
 299 #undef __CFUniCharIsNonBaseCharacter
 300 #undef HANGUL_SBASE
 301 #undef HANGUL_LBASE
 302 #undef HANGUL_VBASE
 303 #undef HANGUL_TBASE
 304 #undef HANGUL_SCOUNT
 305 #undef HANGUL_LCOUNT
 306 #undef HANGUL_VCOUNT
 307 #undef HANGUL_TCOUNT
 308 #undef HANGUL_NCOUNT
 309