CFBuiltinConverters.c

   1 /*
   2  * Copyright (c) 2015 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 /*      CFBuiltinConverters.c
  25         Copyright (c) 1999-2014, Apple Inc. All rights reserved.
  26         Responsibility: Aki Inoue
  27 */
  28
  29 #include "CFStringEncodingConverterExt.h"
  30 #include "CFUniChar.h"
  31 #include "CFUnicodeDecomposition.h"
  32 #include "CFUnicodePrecomposition.h"
  33 #include "CFStringEncodingConverterPriv.h"
  34 #include "CFInternal.h"
  35
  36 #define ParagraphSeparator 0x2029
  37 #define ASCIINewLine 0x0a
  38 static int8_t __CFMapsParagraphSeparator = -1;
  39
  40 CF_INLINE bool __CFIsParagraphSeparator(UTF16Char character) {
  41     if (-1 == __CFMapsParagraphSeparator) __CFMapsParagraphSeparator = (1 ? false : true);
  42
  43     return ((__CFMapsParagraphSeparator && (ParagraphSeparator == character)) ? true : false);
  44 }
  45
  46 /* Precomposition */
  47 static const uint32_t __CFLatin1CombiningCharBitmap[] = { // 0x300 ~ 0x35FF
  48     0xFBB94010, 0x01800000, 0x0000000,
  49 };
  50
  51 bool CFStringEncodingIsValidCombiningCharacterForLatin1(UniChar character) {
  52     return ((character >= 0x300) && (character < 0x360) && (__CFLatin1CombiningCharBitmap[(character - 0x300) / 32] & (1 << (31 - ((character - 0x300) % 32)))) ? true : false);
  53 }
  54
  55 UniChar CFStringEncodingPrecomposeLatinCharacter(const UniChar *character, CFIndex numChars, CFIndex *usedChars) {
  56     if (numChars > 0) {
  57         UTF32Char ch = *(character++), nextCh, composedChar;
  58         CFIndex usedCharLen = 1;
  59
  60         if (CFUniCharIsSurrogateHighCharacter(ch) || CFUniCharIsSurrogateLowCharacter(ch)) {
  61             if (usedChars) (*usedChars) = usedCharLen;
  62             return ch;
  63         }
  64
  65         while (usedCharLen < numChars) {
  66             nextCh = *(character++);
  67
  68             if (CFUniCharIsSurrogateHighCharacter(nextCh) || CFUniCharIsSurrogateLowCharacter(nextCh)) break;
  69
  70             if (CFUniCharIsMemberOf(nextCh, kCFUniCharNonBaseCharacterSet) && ((composedChar = CFUniCharPrecomposeCharacter(ch, nextCh)) != 0xFFFD)) {
  71                 if (composedChar > 0xFFFF) { // Non-base
  72                     break;
  73                 } else {
  74                     ch = composedChar;
  75                 }
  76             } else {
  77                 break;
  78             }
  79             ++usedCharLen;
  80         }
  81         if (usedChars) (*usedChars) = usedCharLen;
  82         if (usedCharLen > 1) return ch;
  83     }
  84     return 0xFFFD;
  85 }
  86
  87 /* ASCII */
  88 static bool __CFToASCII(uint32_t flags, UniChar character, uint8_t *byte) {
  89     if (character < 0x80) {
  90         *byte = (uint8_t)character;
  91     } else if (__CFIsParagraphSeparator(character)) {
  92         *byte = ASCIINewLine;
  93     } else {
  94         return false;
  95     }
  96     return true;
  97 }
  98
  99 static bool __CFFromASCII(uint32_t flags, uint8_t byte, UniChar *character) {
 100     if (byte < 0x80) {
 101         *character = (UniChar)byte;
 102         return true;
 103     } else {
 104         return false;
 105     }
 106 }
 107
 108
 109 CF_PRIVATE const CFStringEncodingConverter __CFConverterASCII = {
 110     __CFToASCII, __CFFromASCII, 1, 1, kCFStringEncodingConverterCheapEightBit,
 111     NULL, NULL, NULL, NULL, NULL, NULL,
 112 };
 113
 114 /* ISO Latin 1 (8859-1) */
 115 static bool __CFToISOLatin1(uint32_t flags, UniChar character, uint8_t *byte) {
 116     if (character <= 0xFF) {
 117         *byte = (uint8_t)character;
 118     } else if (__CFIsParagraphSeparator(character)) {
 119         *byte = ASCIINewLine;
 120     } else {
 121         return false;
 122     }
 123
 124     return true;
 125 }
 126
 127 static bool __CFFromISOLatin1(uint32_t flags, uint8_t byte, UniChar *character) {
 128     *character = (UniChar)byte;
 129     return true;
 130 }
 131
 132 static CFIndex __CFToISOLatin1Precompose(uint32_t flags, const UniChar *character, CFIndex numChars, uint8_t *bytes, CFIndex maxByteLen, CFIndex *usedByteLen) {
 133     uint8_t byte;
 134     CFIndex usedCharLen;
 135
 136     if (__CFToISOLatin1(flags, CFStringEncodingPrecomposeLatinCharacter(character, numChars, &usedCharLen), &byte) && byte && (usedCharLen > 1)) {
 137         if (maxByteLen) *bytes = byte;
 138         *usedByteLen = 1;
 139         return usedCharLen;
 140     } else {
 141         return 0;
 142     }
 143 }
 144
 145 CF_PRIVATE const CFStringEncodingConverter __CFConverterISOLatin1 = {
 146     __CFToISOLatin1, __CFFromISOLatin1, 1, 1, kCFStringEncodingConverterCheapEightBit,
 147     NULL, NULL, NULL, NULL, __CFToISOLatin1Precompose, CFStringEncodingIsValidCombiningCharacterForLatin1,
 148 };
 149
 150 /* Mac Roman */
 151 #define NUM_MACROMAN_FROM_UNI 129
 152 static const CFStringEncodingUnicodeTo8BitCharMap macRoman_from_uni[NUM_MACROMAN_FROM_UNI] = {
 153     { 0x00A0, 0xCA }, /* NO-BREAK SPACE */
 154     { 0x00A1, 0xC1 }, /* INVERTED EXCLAMATION MARK */
 155     { 0x00A2, 0xA2 }, /* CENT SIGN */
 156     { 0x00A3, 0xA3 }, /* POUND SIGN */
 157     { 0x00A5, 0xB4 }, /* YEN SIGN */
 158     { 0x00A7, 0xA4 }, /* SECTION SIGN */
 159     { 0x00A8, 0xAC }, /* DIAERESIS */
 160     { 0x00A9, 0xA9 }, /* COPYRIGHT SIGN */
 161     { 0x00AA, 0xBB }, /* FEMININE ORDINAL INDICATOR */
 162     { 0x00AB, 0xC7 }, /* LEFT-POINTING DOUBLE ANGLE QUOTATION MARK */
 163     { 0x00AC, 0xC2 }, /* NOT SIGN */
 164     { 0x00AE, 0xA8 }, /* REGISTERED SIGN */
 165     { 0x00AF, 0xF8 }, /* MACRON */
 166     { 0x00B0, 0xA1 }, /* DEGREE SIGN */
 167     { 0x00B1, 0xB1 }, /* PLUS-MINUS SIGN */
 168     { 0x00B4, 0xAB }, /* ACUTE ACCENT */
 169     { 0x00B5, 0xB5 }, /* MICRO SIGN */
 170     { 0x00B6, 0xA6 }, /* PILCROW SIGN */
 171     { 0x00B7, 0xE1 }, /* MIDDLE DOT */
 172     { 0x00B8, 0xFC }, /* CEDILLA */
 173     { 0x00BA, 0xBC }, /* MASCULINE ORDINAL INDICATOR */
 174     { 0x00BB, 0xC8 }, /* RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK */
 175     { 0x00BF, 0xC0 }, /* INVERTED QUESTION MARK */
 176     { 0x00C0, 0xCB }, /* LATIN CAPITAL LETTER A WITH GRAVE */
 177     { 0x00C1, 0xE7 }, /* LATIN CAPITAL LETTER A WITH ACUTE */
 178     { 0x00C2, 0xE5 }, /* LATIN CAPITAL LETTER A WITH CIRCUMFLEX */
 179     { 0x00C3, 0xCC }, /* LATIN CAPITAL LETTER A WITH TILDE */
 180     { 0x00C4, 0x80 }, /* LATIN CAPITAL LETTER A WITH DIAERESIS */
 181     { 0x00C5, 0x81 }, /* LATIN CAPITAL LETTER A WITH RING ABOVE */
 182     { 0x00C6, 0xAE }, /* LATIN CAPITAL LIGATURE AE */
 183     { 0x00C7, 0x82 }, /* LATIN CAPITAL LETTER C WITH CEDILLA */
 184     { 0x00C8, 0xE9 }, /* LATIN CAPITAL LETTER E WITH GRAVE */
 185     { 0x00C9, 0x83 }, /* LATIN CAPITAL LETTER E WITH ACUTE */
 186     { 0x00CA, 0xE6 }, /* LATIN CAPITAL LETTER E WITH CIRCUMFLEX */
 187     { 0x00CB, 0xE8 }, /* LATIN CAPITAL LETTER E WITH DIAERESIS */
 188     { 0x00CC, 0xED }, /* LATIN CAPITAL LETTER I WITH GRAVE */
 189     { 0x00CD, 0xEA }, /* LATIN CAPITAL LETTER I WITH ACUTE */
 190     { 0x00CE, 0xEB }, /* LATIN CAPITAL LETTER I WITH CIRCUMFLEX */
 191     { 0x00CF, 0xEC }, /* LATIN CAPITAL LETTER I WITH DIAERESIS */
 192     { 0x00D1, 0x84 }, /* LATIN CAPITAL LETTER N WITH TILDE */
 193     { 0x00D2, 0xF1 }, /* LATIN CAPITAL LETTER O WITH GRAVE */
 194     { 0x00D3, 0xEE }, /* LATIN CAPITAL LETTER O WITH ACUTE */
 195     { 0x00D4, 0xEF }, /* LATIN CAPITAL LETTER O WITH CIRCUMFLEX */
 196     { 0x00D5, 0xCD }, /* LATIN CAPITAL LETTER O WITH TILDE */
 197     { 0x00D6, 0x85 }, /* LATIN CAPITAL LETTER O WITH DIAERESIS */
 198     { 0x00D8, 0xAF }, /* LATIN CAPITAL LETTER O WITH STROKE */
 199     { 0x00D9, 0xF4 }, /* LATIN CAPITAL LETTER U WITH GRAVE */
 200     { 0x00DA, 0xF2 }, /* LATIN CAPITAL LETTER U WITH ACUTE */
 201     { 0x00DB, 0xF3 }, /* LATIN CAPITAL LETTER U WITH CIRCUMFLEX */
 202     { 0x00DC, 0x86 }, /* LATIN CAPITAL LETTER U WITH DIAERESIS */
 203     { 0x00DF, 0xA7 }, /* LATIN SMALL LETTER SHARP S */
 204     { 0x00E0, 0x88 }, /* LATIN SMALL LETTER A WITH GRAVE */
 205     { 0x00E1, 0x87 }, /* LATIN SMALL LETTER A WITH ACUTE */
 206     { 0x00E2, 0x89 }, /* LATIN SMALL LETTER A WITH CIRCUMFLEX */
 207     { 0x00E3, 0x8B }, /* LATIN SMALL LETTER A WITH TILDE */
 208     { 0x00E4, 0x8A }, /* LATIN SMALL LETTER A WITH DIAERESIS */
 209     { 0x00E5, 0x8C }, /* LATIN SMALL LETTER A WITH RING ABOVE */
 210     { 0x00E6, 0xBE }, /* LATIN SMALL LIGATURE AE */
 211     { 0x00E7, 0x8D }, /* LATIN SMALL LETTER C WITH CEDILLA */
 212     { 0x00E8, 0x8F }, /* LATIN SMALL LETTER E WITH GRAVE */
 213     { 0x00E9, 0x8E }, /* LATIN SMALL LETTER E WITH ACUTE */
 214     { 0x00EA, 0x90 }, /* LATIN SMALL LETTER E WITH CIRCUMFLEX */
 215     { 0x00EB, 0x91 }, /* LATIN SMALL LETTER E WITH DIAERESIS */
 216     { 0x00EC, 0x93 }, /* LATIN SMALL LETTER I WITH GRAVE */
 217     { 0x00ED, 0x92 }, /* LATIN SMALL LETTER I WITH ACUTE */
 218     { 0x00EE, 0x94 }, /* LATIN SMALL LETTER I WITH CIRCUMFLEX */
 219     { 0x00EF, 0x95 }, /* LATIN SMALL LETTER I WITH DIAERESIS */
 220     { 0x00F1, 0x96 }, /* LATIN SMALL LETTER N WITH TILDE */
 221     { 0x00F2, 0x98 }, /* LATIN SMALL LETTER O WITH GRAVE */
 222     { 0x00F3, 0x97 }, /* LATIN SMALL LETTER O WITH ACUTE */
 223     { 0x00F4, 0x99 }, /* LATIN SMALL LETTER O WITH CIRCUMFLEX */
 224     { 0x00F5, 0x9B }, /* LATIN SMALL LETTER O WITH TILDE */
 225     { 0x00F6, 0x9A }, /* LATIN SMALL LETTER O WITH DIAERESIS */
 226     { 0x00F7, 0xD6 }, /* DIVISION SIGN */
 227     { 0x00F8, 0xBF }, /* LATIN SMALL LETTER O WITH STROKE */
 228     { 0x00F9, 0x9D }, /* LATIN SMALL LETTER U WITH GRAVE */
 229     { 0x00FA, 0x9C }, /* LATIN SMALL LETTER U WITH ACUTE */
 230     { 0x00FB, 0x9E }, /* LATIN SMALL LETTER U WITH CIRCUMFLEX */
 231     { 0x00FC, 0x9F }, /* LATIN SMALL LETTER U WITH DIAERESIS */
 232     { 0x00FF, 0xD8 }, /* LATIN SMALL LETTER Y WITH DIAERESIS */
 233     { 0x0131, 0xF5 }, /* LATIN SMALL LETTER DOTLESS I */
 234     { 0x0152, 0xCE }, /* LATIN CAPITAL LIGATURE OE */
 235     { 0x0153, 0xCF }, /* LATIN SMALL LIGATURE OE */
 236     { 0x0178, 0xD9 }, /* LATIN CAPITAL LETTER Y WITH DIAERESIS */
 237     { 0x0192, 0xC4 }, /* LATIN SMALL LETTER F WITH HOOK */
 238     { 0x02C6, 0xF6 }, /* MODIFIER LETTER CIRCUMFLEX ACCENT */
 239     { 0x02C7, 0xFF }, /* CARON */
 240     { 0x02D8, 0xF9 }, /* BREVE */
 241     { 0x02D9, 0xFA }, /* DOT ABOVE */
 242     { 0x02DA, 0xFB }, /* RING ABOVE */
 243     { 0x02DB, 0xFE }, /* OGONEK */
 244     { 0x02DC, 0xF7 }, /* SMALL TILDE */
 245     { 0x02DD, 0xFD }, /* DOUBLE ACUTE ACCENT */
 246     { 0x03A9, 0xBD }, /* OHM SIGN (Canonical ?) */
 247     { 0x03C0, 0xB9 }, /* GREEK SMALL LETTER PI */
 248     { 0x2013, 0xD0 }, /* EN DASH */
 249     { 0x2014, 0xD1 }, /* EM DASH */
 250     { 0x2018, 0xD4 }, /* LEFT SINGLE QUOTATION MARK */
 251     { 0x2019, 0xD5 }, /* RIGHT SINGLE QUOTATION MARK */
 252     { 0x201A, 0xE2 }, /* SINGLE LOW-9 QUOTATION MARK */
 253     { 0x201C, 0xD2 }, /* LEFT DOUBLE QUOTATION MARK */
 254     { 0x201D, 0xD3 }, /* RIGHT DOUBLE QUOTATION MARK */
 255     { 0x201E, 0xE3 }, /* DOUBLE LOW-9 QUOTATION MARK */
 256     { 0x2020, 0xA0 }, /* DAGGER */
 257     { 0x2021, 0xE0 }, /* DOUBLE DAGGER */
 258     { 0x2022, 0xA5 }, /* BULLET */
 259     { 0x2026, 0xC9 }, /* HORIZONTAL ELLIPSIS */
 260     { 0x2030, 0xE4 }, /* PER MILLE SIGN */
 261     { 0x2039, 0xDC }, /* SINGLE LEFT-POINTING ANGLE QUOTATION MARK */
 262     { 0x203A, 0xDD }, /* SINGLE RIGHT-POINTING ANGLE QUOTATION MARK */
 263     { 0x2044, 0xDA }, /* FRACTION SLASH */
 264     { 0x20AC, 0xDB }, /* EURO SIGN */
 265     { 0x2122, 0xAA }, /* TRADE MARK SIGN */
 266     { 0x2126, 0xBD }, /* OHM SIGN */
 267     { 0x2202, 0xB6 }, /* PARTIAL DIFFERENTIAL */
 268     { 0x2206, 0xC6 }, /* INCREMENT */
 269     { 0x220F, 0xB8 }, /* N-ARY PRODUCT */
 270     { 0x2211, 0xB7 }, /* N-ARY SUMMATION */
 271     { 0x221A, 0xC3 }, /* SQUARE ROOT */
 272     { 0x221E, 0xB0 }, /* INFINITY */
 273     { 0x222B, 0xBA }, /* INTEGRAL */
 274     { 0x2248, 0xC5 }, /* ALMOST EQUAL TO */
 275     { 0x2260, 0xAD }, /* NOT EQUAL TO */
 276     { 0x2264, 0xB2 }, /* LESS-THAN OR EQUAL TO */
 277     { 0x2265, 0xB3 }, /* GREATER-THAN OR EQUAL TO */
 278     { 0x25CA, 0xD7 }, /* LOZENGE */
 279     { 0xF8FF, 0xF0 }, /* Apple logo */
 280     { 0xFB01, 0xDE }, /* LATIN SMALL LIGATURE FI */
 281     { 0xFB02, 0xDF }, /* LATIN SMALL LIGATURE FL */
 282 };
 283
 284 static bool __CFToMacRoman(uint32_t flags, UniChar character, uint8_t *byte) {
 285     if (character < 0x80) {
 286         *byte = (uint8_t)character;
 287         return true;
 288     } else {
 289         return CFStringEncodingUnicodeTo8BitEncoding(macRoman_from_uni, NUM_MACROMAN_FROM_UNI, character, byte);
 290     }
 291 }
 292
 293 static const UniChar macRoman_to_uni[128] = {
 294     0x00C4, /* LATIN CAPITAL LETTER A WITH DIAERESIS */
 295     0x00C5, /* LATIN CAPITAL LETTER A WITH RING ABOVE */
 296     0x00C7, /* LATIN CAPITAL LETTER C WITH CEDILLA */
 297     0x00C9, /* LATIN CAPITAL LETTER E WITH ACUTE */
 298     0x00D1, /* LATIN CAPITAL LETTER N WITH TILDE */
 299     0x00D6, /* LATIN CAPITAL LETTER O WITH DIAERESIS */
 300     0x00DC, /* LATIN CAPITAL LETTER U WITH DIAERESIS */
 301     0x00E1, /* LATIN SMALL LETTER A WITH ACUTE */
 302     0x00E0, /* LATIN SMALL LETTER A WITH GRAVE */
 303     0x00E2, /* LATIN SMALL LETTER A WITH CIRCUMFLEX */
 304     0x00E4, /* LATIN SMALL LETTER A WITH DIAERESIS */
 305     0x00E3, /* LATIN SMALL LETTER A WITH TILDE */
 306     0x00E5, /* LATIN SMALL LETTER A WITH RING ABOVE */
 307     0x00E7, /* LATIN SMALL LETTER C WITH CEDILLA */
 308     0x00E9, /* LATIN SMALL LETTER E WITH ACUTE */
 309     0x00E8, /* LATIN SMALL LETTER E WITH GRAVE */
 310     0x00EA, /* LATIN SMALL LETTER E WITH CIRCUMFLEX */
 311     0x00EB, /* LATIN SMALL LETTER E WITH DIAERESIS */
 312     0x00ED, /* LATIN SMALL LETTER I WITH ACUTE */
 313     0x00EC, /* LATIN SMALL LETTER I WITH GRAVE */
 314     0x00EE, /* LATIN SMALL LETTER I WITH CIRCUMFLEX */
 315     0x00EF, /* LATIN SMALL LETTER I WITH DIAERESIS */
 316     0x00F1, /* LATIN SMALL LETTER N WITH TILDE */
 317     0x00F3, /* LATIN SMALL LETTER O WITH ACUTE */
 318     0x00F2, /* LATIN SMALL LETTER O WITH GRAVE */
 319     0x00F4, /* LATIN SMALL LETTER O WITH CIRCUMFLEX */
 320     0x00F6, /* LATIN SMALL LETTER O WITH DIAERESIS */
 321     0x00F5, /* LATIN SMALL LETTER O WITH TILDE */
 322     0x00FA, /* LATIN SMALL LETTER U WITH ACUTE */
 323     0x00F9, /* LATIN SMALL LETTER U WITH GRAVE */
 324     0x00FB, /* LATIN SMALL LETTER U WITH CIRCUMFLEX */
 325     0x00FC, /* LATIN SMALL LETTER U WITH DIAERESIS */
 326     0x2020, /* DAGGER */
 327     0x00B0, /* DEGREE SIGN */
 328     0x00A2, /* CENT SIGN */
 329     0x00A3, /* POUND SIGN */
 330     0x00A7, /* SECTION SIGN */
 331     0x2022, /* BULLET */
 332     0x00B6, /* PILCROW SIGN */
 333     0x00DF, /* LATIN SMALL LETTER SHARP S */
 334     0x00AE, /* REGISTERED SIGN */
 335     0x00A9, /* COPYRIGHT SIGN */
 336     0x2122, /* TRADE MARK SIGN */
 337     0x00B4, /* ACUTE ACCENT */
 338     0x00A8, /* DIAERESIS */
 339     0x2260, /* NOT EQUAL TO */
 340     0x00C6, /* LATIN CAPITAL LIGATURE AE */
 341     0x00D8, /* LATIN CAPITAL LETTER O WITH STROKE */
 342     0x221E, /* INFINITY */
 343     0x00B1, /* PLUS-MINUS SIGN */
 344     0x2264, /* LESS-THAN OR EQUAL TO */
 345     0x2265, /* GREATER-THAN OR EQUAL TO */
 346     0x00A5, /* YEN SIGN */
 347     0x00B5, /* MICRO SIGN */
 348     0x2202, /* PARTIAL DIFFERENTIAL */
 349     0x2211, /* N-ARY SUMMATION */
 350     0x220F, /* N-ARY PRODUCT */
 351     0x03C0, /* GREEK SMALL LETTER PI */
 352     0x222B, /* INTEGRAL */
 353     0x00AA, /* FEMININE ORDINAL INDICATOR */
 354     0x00BA, /* MASCULINE ORDINAL INDICATOR */
 355     0x03A9, /* OHM SIGN (Canonical mapping) */
 356     0x00E6, /* LATIN SMALL LIGATURE AE */
 357     0x00F8, /* LATIN SMALL LETTER O WITH STROKE */
 358     0x00BF, /* INVERTED QUESTION MARK */
 359     0x00A1, /* INVERTED EXCLAMATION MARK */
 360     0x00AC, /* NOT SIGN */
 361     0x221A, /* SQUARE ROOT */
 362     0x0192, /* LATIN SMALL LETTER F WITH HOOK */
 363     0x2248, /* ALMOST EQUAL TO */
 364     0x2206, /* INCREMENT */
 365     0x00AB, /* LEFT-POINTING DOUBLE ANGLE QUOTATION MARK */
 366     0x00BB, /* RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK */
 367     0x2026, /* HORIZONTAL ELLIPSIS */
 368     0x00A0, /* NO-BREAK SPACE */
 369     0x00C0, /* LATIN CAPITAL LETTER A WITH GRAVE */
 370     0x00C3, /* LATIN CAPITAL LETTER A WITH TILDE */
 371     0x00D5, /* LATIN CAPITAL LETTER O WITH TILDE */
 372     0x0152, /* LATIN CAPITAL LIGATURE OE */
 373     0x0153, /* LATIN SMALL LIGATURE OE */
 374     0x2013, /* EN DASH */
 375     0x2014, /* EM DASH */
 376     0x201C, /* LEFT DOUBLE QUOTATION MARK */
 377     0x201D, /* RIGHT DOUBLE QUOTATION MARK */
 378     0x2018, /* LEFT SINGLE QUOTATION MARK */
 379     0x2019, /* RIGHT SINGLE QUOTATION MARK */
 380     0x00F7, /* DIVISION SIGN */
 381     0x25CA, /* LOZENGE */
 382     0x00FF, /* LATIN SMALL LETTER Y WITH DIAERESIS */
 383     0x0178, /* LATIN CAPITAL LETTER Y WITH DIAERESIS */
 384     0x2044, /* FRACTION SLASH */
 385     0x20AC, /* EURO SIGN */
 386     0x2039, /* SINGLE LEFT-POINTING ANGLE QUOTATION MARK */
 387     0x203A, /* SINGLE RIGHT-POINTING ANGLE QUOTATION MARK */
 388     0xFB01, /* LATIN SMALL LIGATURE FI */
 389     0xFB02, /* LATIN SMALL LIGATURE FL */
 390     0x2021, /* DOUBLE DAGGER */
 391     0x00B7, /* MIDDLE DOT */
 392     0x201A, /* SINGLE LOW-9 QUOTATION MARK */
 393     0x201E, /* DOUBLE LOW-9 QUOTATION MARK */
 394     0x2030, /* PER MILLE SIGN */
 395     0x00C2, /* LATIN CAPITAL LETTER A WITH CIRCUMFLEX */
 396     0x00CA, /* LATIN CAPITAL LETTER E WITH CIRCUMFLEX */
 397     0x00C1, /* LATIN CAPITAL LETTER A WITH ACUTE */
 398     0x00CB, /* LATIN CAPITAL LETTER E WITH DIAERESIS */
 399     0x00C8, /* LATIN CAPITAL LETTER E WITH GRAVE */
 400     0x00CD, /* LATIN CAPITAL LETTER I WITH ACUTE */
 401     0x00CE, /* LATIN CAPITAL LETTER I WITH CIRCUMFLEX */
 402     0x00CF, /* LATIN CAPITAL LETTER I WITH DIAERESIS */
 403     0x00CC, /* LATIN CAPITAL LETTER I WITH GRAVE */
 404     0x00D3, /* LATIN CAPITAL LETTER O WITH ACUTE */
 405     0x00D4, /* LATIN CAPITAL LETTER O WITH CIRCUMFLEX */
 406     0xF8FF, /* Apple logo */
 407     0x00D2, /* LATIN CAPITAL LETTER O WITH GRAVE */
 408     0x00DA, /* LATIN CAPITAL LETTER U WITH ACUTE */
 409     0x00DB, /* LATIN CAPITAL LETTER U WITH CIRCUMFLEX */
 410     0x00D9, /* LATIN CAPITAL LETTER U WITH GRAVE */
 411     0x0131, /* LATIN SMALL LETTER DOTLESS I */
 412     0x02C6, /* MODIFIER LETTER CIRCUMFLEX ACCENT */
 413     0x02DC, /* SMALL TILDE */
 414     0x00AF, /* MACRON */
 415     0x02D8, /* BREVE */
 416     0x02D9, /* DOT ABOVE */
 417     0x02DA, /* RING ABOVE */
 418     0x00B8, /* CEDILLA */
 419     0x02DD, /* DOUBLE ACUTE ACCENT */
 420     0x02DB, /* OGONEK */
 421     0x02C7, /* CARON */
 422 };
 423
 424 static bool __CFFromMacRoman(uint32_t flags, uint8_t byte, UniChar *character) {
 425     *character = (byte < 0x80 ? (UniChar)byte : macRoman_to_uni[byte - 0x80]);
 426     return true;
 427 }
 428
 429 static CFIndex __CFToMacRomanPrecompose(uint32_t flags, const UniChar *character, CFIndex numChars, uint8_t *bytes, CFIndex maxByteLen, CFIndex *usedByteLen) {
 430     uint8_t byte;
 431     CFIndex usedCharLen;
 432
 433     if (__CFToMacRoman(flags, CFStringEncodingPrecomposeLatinCharacter(character, numChars, &usedCharLen), &byte) && byte && (usedCharLen > 1)) {
 434         if (maxByteLen) *bytes = byte;
 435         *usedByteLen = 1;
 436         return usedCharLen;
 437     } else {
 438         return 0;
 439     }
 440 }
 441
 442 CF_PRIVATE const CFStringEncodingConverter __CFConverterMacRoman = {
 443     __CFToMacRoman, __CFFromMacRoman, 1, 1, kCFStringEncodingConverterCheapEightBit,
 444     NULL, NULL, NULL, NULL, __CFToMacRomanPrecompose, CFStringEncodingIsValidCombiningCharacterForLatin1,
 445 };
 446
 447 /* Win Latin1 (ANSI CodePage 1252) */
 448 #define NUM_1252_FROM_UNI 27
 449 static const CFStringEncodingUnicodeTo8BitCharMap cp1252_from_uni[NUM_1252_FROM_UNI] = {
 450     {0x0152, 0x8C}, // LATIN CAPITAL LIGATURE OE
 451     {0x0153, 0x9C}, // LATIN SMALL LIGATURE OE
 452     {0x0160, 0x8A}, // LATIN CAPITAL LETTER S WITH CARON
 453     {0x0161, 0x9A}, // LATIN SMALL LETTER S WITH CARON
 454     {0x0178, 0x9F}, // LATIN CAPITAL LETTER Y WITH DIAERESIS
 455     {0x017D, 0x8E}, // LATIN CAPITAL LETTER Z WITH CARON
 456     {0x017E, 0x9E}, // LATIN SMALL LETTER Z WITH CARON
 457     {0x0192, 0x83}, // LATIN SMALL LETTER F WITH HOOK
 458     {0x02C6, 0x88}, // MODIFIER LETTER CIRCUMFLEX ACCENT
 459     {0x02DC, 0x98}, // SMALL TILDE
 460     {0x2013, 0x96}, // EN DASH
 461     {0x2014, 0x97}, // EM DASH
 462     {0x2018, 0x91}, // LEFT SINGLE QUOTATION MARK
 463     {0x2019, 0x92}, // RIGHT SINGLE QUOTATION MARK
 464     {0x201A, 0x82}, // SINGLE LOW-9 QUOTATION MARK
 465     {0x201C, 0x93}, // LEFT DOUBLE QUOTATION MARK
 466     {0x201D, 0x94}, // RIGHT DOUBLE QUOTATION MARK
 467     {0x201E, 0x84}, // DOUBLE LOW-9 QUOTATION MARK
 468     {0x2020, 0x86}, // DAGGER
 469     {0x2021, 0x87}, // DOUBLE DAGGER
 470     {0x2022, 0x95}, // BULLET
 471     {0x2026, 0x85}, // HORIZONTAL ELLIPSIS
 472     {0x2030, 0x89}, // PER MILLE SIGN
 473     {0x2039, 0x8B}, // SINGLE LEFT-POINTING ANGLE QUOTATION MARK
 474     {0x203A, 0x9B}, // SINGLE RIGHT-POINTING ANGLE QUOTATION MARK
 475     {0x20AC, 0x80}, // EURO SIGN
 476     {0x2122, 0x99}, // TRADE MARK SIGN
 477 };
 478
 479 static bool __CFToWinLatin1(uint32_t flags, UniChar character, uint8_t *byte) {
 480     if ((character < 0x80) || ((character > 0x9F) && (character <= 0x00FF))) {
 481         *byte = (uint8_t)character;
 482         return true;
 483     }
 484     return CFStringEncodingUnicodeTo8BitEncoding(cp1252_from_uni, NUM_1252_FROM_UNI, character, byte);
 485 }
 486
 487 static const uint16_t cp1252_to_uni[32] = {
 488     0x20AC, //  EURO SIGN
 489     0xFFFD, //  NOT USED
 490     0x201A, //  SINGLE LOW-9 QUOTATION MARK
 491     0x0192, //  LATIN SMALL LETTER F WITH HOOK
 492     0x201E, //  DOUBLE LOW-9 QUOTATION MARK
 493     0x2026, //  HORIZONTAL ELLIPSIS
 494     0x2020, //  DAGGER
 495     0x2021, //  DOUBLE DAGGER
 496     0x02C6, //  MODIFIER LETTER CIRCUMFLEX ACCENT
 497     0x2030, //  PER MILLE SIGN
 498     0x0160, //  LATIN CAPITAL LETTER S WITH CARON
 499     0x2039, //  SINGLE LEFT-POINTING ANGLE QUOTATION MARK
 500     0x0152, //  LATIN CAPITAL LIGATURE OE
 501     0xFFFD, //  NOT USED
 502     0x017D, //  LATIN CAPITAL LETTER Z WITH CARON
 503     0xFFFD, //  NOT USED
 504     0xFFFD, //  NOT USED
 505     0x2018, //  LEFT SINGLE QUOTATION MARK
 506     0x2019, //  RIGHT SINGLE QUOTATION MARK
 507     0x201C, //  LEFT DOUBLE QUOTATION MARK
 508     0x201D, //  RIGHT DOUBLE QUOTATION MARK
 509     0x2022, //  BULLET
 510     0x2013, //  EN DASH
 511     0x2014, //  EM DASH
 512     0x02DC, //  SMALL TILDE
 513     0x2122, //  TRADE MARK SIGN
 514     0x0161, //  LATIN SMALL LETTER S WITH CARON
 515     0x203A, //  SINGLE RIGHT-POINTING ANGLE QUOTATION MARK
 516     0x0153, //  LATIN SMALL LIGATURE OE
 517     0xFFFD, //  NOT USED
 518     0x017E, //  LATIN SMALL LETTER Z WITH CARON
 519     0x0178, //  LATIN CAPITAL LETTER Y WITH DIAERESIS
 520 };
 521
 522 static bool __CFFromWinLatin1(uint32_t flags, uint8_t byte, UniChar *character) {
 523     *character = (byte < 0x80 || byte > 0x9F ? (UniChar)byte : cp1252_to_uni[byte - 0x80]);
 524     return (*character != 0xFFFD);
 525 }
 526
 527 static CFIndex __CFToWinLatin1Precompose(uint32_t flags, const UniChar *character, CFIndex numChars, uint8_t *bytes, CFIndex maxByteLen, CFIndex *usedByteLen) {
 528     uint8_t byte;
 529     CFIndex usedCharLen;
 530
 531     if (__CFToWinLatin1(flags, CFStringEncodingPrecomposeLatinCharacter(character, numChars, &usedCharLen), &byte) && byte && (usedCharLen > 1)) {
 532         if (maxByteLen) *bytes = byte;
 533         *usedByteLen = 1;
 534         return usedCharLen;
 535     } else {
 536         return 0;
 537     }
 538 }
 539
 540 CF_PRIVATE const CFStringEncodingConverter __CFConverterWinLatin1 = {
 541     __CFToWinLatin1, __CFFromWinLatin1, 1, 1, kCFStringEncodingConverterCheapEightBit,
 542     NULL, NULL, NULL, NULL, __CFToWinLatin1Precompose, CFStringEncodingIsValidCombiningCharacterForLatin1,
 543 };
 544
 545 /* NEXTSTEP Encoding */
 546 #define NUM_NEXTSTEP_FROM_UNI   127
 547
 548 static const CFStringEncodingUnicodeTo8BitCharMap nextstep_from_tab[NUM_NEXTSTEP_FROM_UNI] = {
 549         { 0x00a0, 0x80 },
 550         { 0x00a1, 0xa1 },
 551         { 0x00a2, 0xa2 },
 552         { 0x00a3, 0xa3 },
 553         { 0x00a4, 0xa8 },
 554         { 0x00a5, 0xa5 },
 555         { 0x00a6, 0xb5 },
 556         { 0x00a7, 0xa7 },
 557         { 0x00a8, 0xc8 },
 558         { 0x00a9, 0xa0 },
 559         { 0x00aa, 0xe3 },
 560         { 0x00ab, 0xab },
 561         { 0x00ac, 0xbe },
 562 /*      { 0x00ad, 0x2d }, <= 96/10/25 rick removed; converts soft-hyphen to hyphen! */
 563         { 0x00ae, 0xb0 },
 564         { 0x00af, 0xc5 },
 565         { 0x00b1, 0xd1 },
 566         { 0x00b2, 0xc9 },
 567         { 0x00b3, 0xcc },
 568         { 0x00b4, 0xc2 },
 569         { 0x00b5, 0x9d },
 570         { 0x00b6, 0xb6 },
 571         { 0x00b7, 0xb4 },
 572         { 0x00b8, 0xcb },
 573         { 0x00b9, 0xc0 },
 574         { 0x00ba, 0xeb },
 575         { 0x00bb, 0xbb },
 576         { 0x00bc, 0xd2 },
 577         { 0x00bd, 0xd3 },
 578         { 0x00be, 0xd4 },
 579         { 0x00bf, 0xbf },
 580         { 0x00c0, 0x81 },
 581         { 0x00c1, 0x82 },
 582         { 0x00c2, 0x83 },
 583         { 0x00c3, 0x84 },
 584         { 0x00c4, 0x85 },
 585         { 0x00c5, 0x86 },
 586         { 0x00c6, 0xe1 },
 587         { 0x00c7, 0x87 },
 588         { 0x00c8, 0x88 },
 589         { 0x00c9, 0x89 },
 590         { 0x00ca, 0x8a },
 591         { 0x00cb, 0x8b },
 592         { 0x00cc, 0x8c },
 593         { 0x00cd, 0x8d },
 594         { 0x00ce, 0x8e },
 595         { 0x00cf, 0x8f },
 596         { 0x00d0, 0x90 },
 597         { 0x00d1, 0x91 },
 598         { 0x00d2, 0x92 },
 599         { 0x00d3, 0x93 },
 600         { 0x00d4, 0x94 },
 601         { 0x00d5, 0x95 },
 602         { 0x00d6, 0x96 },
 603         { 0x00d7, 0x9e },
 604         { 0x00d8, 0xe9 },
 605         { 0x00d9, 0x97 },
 606         { 0x00da, 0x98 },
 607         { 0x00db, 0x99 },
 608         { 0x00dc, 0x9a },
 609         { 0x00dd, 0x9b },
 610         { 0x00de, 0x9c },
 611         { 0x00df, 0xfb },
 612         { 0x00e0, 0xd5 },
 613         { 0x00e1, 0xd6 },
 614         { 0x00e2, 0xd7 },
 615         { 0x00e3, 0xd8 },
 616         { 0x00e4, 0xd9 },
 617         { 0x00e5, 0xda },
 618         { 0x00e6, 0xf1 },
 619         { 0x00e7, 0xdb },
 620         { 0x00e8, 0xdc },
 621         { 0x00e9, 0xdd },
 622         { 0x00ea, 0xde },
 623         { 0x00eb, 0xdf },
 624         { 0x00ec, 0xe0 },
 625         { 0x00ed, 0xe2 },
 626         { 0x00ee, 0xe4 },
 627         { 0x00ef, 0xe5 },
 628         { 0x00f0, 0xe6 },
 629         { 0x00f1, 0xe7 },
 630         { 0x00f2, 0xec },
 631         { 0x00f3, 0xed },
 632         { 0x00f4, 0xee },
 633         { 0x00f5, 0xef },
 634         { 0x00f6, 0xf0 },
 635         { 0x00f7, 0x9f },
 636         { 0x00f8, 0xf9 },
 637         { 0x00f9, 0xf2 },
 638         { 0x00fa, 0xf3 },
 639         { 0x00fb, 0xf4 },
 640         { 0x00fc, 0xf6 },
 641         { 0x00fd, 0xf7 },
 642         { 0x00fe, 0xfc },
 643         { 0x00ff, 0xfd },
 644         { 0x0131, 0xf5 },
 645         { 0x0141, 0xe8 },
 646         { 0x0142, 0xf8 },
 647         { 0x0152, 0xea },
 648         { 0x0153, 0xfa },
 649         { 0x0192, 0xa6 },
 650         { 0x02c6, 0xc3 },
 651         { 0x02c7, 0xcf },
 652         { 0x02cb, 0xc1 },
 653         { 0x02d8, 0xc6 },
 654         { 0x02d9, 0xc7 },
 655         { 0x02da, 0xca },
 656         { 0x02db, 0xce },
 657         { 0x02dc, 0xc4 },
 658         { 0x02dd, 0xcd },
 659         { 0x2013, 0xb1 },
 660         { 0x2014, 0xd0 },
 661         { 0x2019, 0xa9 },
 662         { 0x201a, 0xb8 },
 663         { 0x201c, 0xaa },
 664         { 0x201d, 0xba },
 665         { 0x201e, 0xb9 },
 666         { 0x2020, 0xb2 },
 667         { 0x2021, 0xb3 },
 668         { 0x2022, 0xb7 },
 669         { 0x2026, 0xbc },
 670         { 0x2030, 0xbd },
 671         { 0x2039, 0xac },
 672         { 0x203a, 0xad },
 673         { 0x2044, 0xa4 },
 674         { 0xfb01, 0xae },
 675         { 0xfb02, 0xaf },
 676         { 0xfffd, 0xff },
 677 };
 678
 679 static bool __CFToNextStepLatin(uint32_t flags, UniChar character, uint8_t *byte) {
 680     if (character < 0x80) {
 681         *byte = (uint8_t)character;
 682         return true;
 683     } else if (__CFIsParagraphSeparator(character)) {
 684         *byte = ASCIINewLine;
 685         return true;
 686     } else {
 687         return CFStringEncodingUnicodeTo8BitEncoding(nextstep_from_tab, NUM_NEXTSTEP_FROM_UNI, character, byte);
 688     }
 689 };
 690
 691 static const UniChar NSToPrecompUnicodeTable[128] = {
 692         /* NextStep Encoding    Unicode */
 693         /*  128 figspace */     0x00a0,         /* 0x2007 is fig space */
 694         /*  129 Agrave */       0x00c0,
 695         /*  130 Aacute */       0x00c1,
 696         /*  131 Acircumflex */  0x00c2,
 697         /*  132 Atilde */       0x00c3,
 698         /*  133 Adieresis */    0x00c4,
 699         /*  134 Aring */        0x00c5,
 700         /*  135 Ccedilla */     0x00c7,
 701         /*  136 Egrave */       0x00c8,
 702         /*  137 Eacute */       0x00c9,
 703         /*  138 Ecircumflex */  0x00ca,
 704         /*  139 Edieresis */    0x00cb,
 705         /*  140 Igrave */       0x00cc,
 706         /*  141 Iacute */       0x00cd,
 707         /*  142 Icircumflex */  0x00ce,
 708         /*  143 Idieresis */    0x00cf,
 709         /*  144 Eth */          0x00d0,
 710         /*  145 Ntilde */       0x00d1,
 711         /*  146 Ograve */       0x00d2,
 712         /*  147 Oacute */       0x00d3,
 713         /*  148 Ocircumflex */  0x00d4,
 714         /*  149 Otilde */       0x00d5,
 715         /*  150 Odieresis */    0x00d6,
 716         /*  151 Ugrave */       0x00d9,
 717         /*  152 Uacute */       0x00da,
 718         /*  153 Ucircumflex */  0x00db,
 719         /*  154 Udieresis */    0x00dc,
 720         /*  155 Yacute */       0x00dd,
 721         /*  156 Thorn */        0x00de,
 722         /*  157 mu */           0x00b5,
 723         /*  158 multiply */     0x00d7,
 724         /*  159 divide */       0x00f7,
 725         /*  160 copyright */    0x00a9,
 726         /*  161 exclamdown */   0x00a1,
 727         /*  162 cent */         0x00a2,
 728         /*  163 sterling */     0x00a3,
 729         /*  164 fraction */     0x2044,
 730         /*  165 yen */          0x00a5,
 731         /*  166 florin */       0x0192,
 732         /*  167 section */      0x00a7,
 733         /*  168 currency */     0x00a4,
 734         /*  169 quotesingle */  0x2019,
 735         /*  170 quotedblleft */ 0x201c,
 736         /*  171 guillemotleft */ 0x00ab,
 737         /*  172 guilsinglleft */ 0x2039,
 738         /*  173 guilsinglright */ 0x203a,
 739         /*  174 fi */           0xFB01,
 740         /*  175 fl */           0xFB02,
 741         /*  176 registered */   0x00ae,
 742         /*  177 endash */       0x2013,
 743         /*  178 dagger */       0x2020,
 744         /*  179 daggerdbl */    0x2021,
 745         /*  180 periodcentered */ 0x00b7,
 746         /*  181 brokenbar */    0x00a6,
 747         /*  182 paragraph */    0x00b6,
 748         /*  183 bullet */       0x2022,
 749         /*  184 quotesinglbase */ 0x201a,
 750         /*  185 quotedblbase */ 0x201e,
 751         /*  186 quotedblright */ 0x201d,
 752         /*  187 guillemotright */ 0x00bb,
 753         /*  188 ellipsis */     0x2026,
 754         /*  189 perthousand */  0x2030,
 755         /*  190 logicalnot */   0x00ac,
 756         /*  191 questiondown */ 0x00bf,
 757         /*  192 onesuperior */  0x00b9,
 758         /*  193 grave */        0x02cb,
 759         /*  194 acute */        0x00b4,
 760         /*  195 circumflex */   0x02c6,
 761         /*  196 tilde */        0x02dc,
 762         /*  197 macron */       0x00af,
 763         /*  198 breve */        0x02d8,
 764         /*  199 dotaccent */    0x02d9,
 765         /*  200 dieresis */     0x00a8,
 766         /*  201 twosuperior */  0x00b2,
 767         /*  202 ring */         0x02da,
 768         /*  203 cedilla */      0x00b8,
 769         /*  204 threesuperior */ 0x00b3,
 770         /*  205 hungarumlaut */ 0x02dd,
 771         /*  206 ogonek */       0x02db,
 772         /*  207 caron */        0x02c7,
 773         /*  208 emdash */       0x2014,
 774         /*  209 plusminus */    0x00b1,
 775         /*  210 onequarter */   0x00bc,
 776         /*  211 onehalf */      0x00bd,
 777         /*  212 threequarters */ 0x00be,
 778         /*  213 agrave */       0x00e0,
 779         /*  214 aacute */       0x00e1,
 780         /*  215 acircumflex */  0x00e2,
 781         /*  216 atilde */       0x00e3,
 782         /*  217 adieresis */    0x00e4,
 783         /*  218 aring */        0x00e5,
 784         /*  219 ccedilla */     0x00e7,
 785         /*  220 egrave */       0x00e8,
 786         /*  221 eacute */       0x00e9,
 787         /*  222 ecircumflex */  0x00ea,
 788         /*  223 edieresis */    0x00eb,
 789         /*  224 igrave */       0x00ec,
 790         /*  225 AE */           0x00c6,
 791         /*  226 iacute */       0x00ed,
 792         /*  227 ordfeminine */  0x00aa,
 793         /*  228 icircumflex */  0x00ee,
 794         /*  229 idieresis */    0x00ef,
 795         /*  230 eth */          0x00f0,
 796         /*  231 ntilde */       0x00f1,
 797         /*  232 Lslash */       0x0141,
 798         /*  233 Oslash */       0x00d8,
 799         /*  234 OE */           0x0152,
 800         /*  235 ordmasculine */ 0x00ba,
 801         /*  236 ograve */       0x00f2,
 802         /*  237 oacute */       0x00f3,
 803         /*  238 ocircumflex */  0x00f4,
 804         /*  239 otilde */       0x00f5,
 805         /*  240 odieresis */    0x00f6,
 806         /*  241 ae */           0x00e6,
 807         /*  242 ugrave */       0x00f9,
 808         /*  243 uacute */       0x00fa,
 809         /*  244 ucircumflex */  0x00fb,
 810         /*  245 dotlessi */     0x0131,
 811         /*  246 udieresis */    0x00fc,
 812         /*  247 yacute */       0x00fd,
 813         /*  248 lslash */       0x0142,
 814         /*  249 oslash */       0x00f8,
 815         /*  250 oe */           0x0153,
 816         /*  251 germandbls */   0x00df,
 817         /*  252 thorn */        0x00fe,
 818         /*  253 ydieresis */    0x00ff,
 819         /*  254 .notdef */      0xFFFD,
 820         /*  255 .notdef */      0xFFFD
 821 };
 822
 823 static bool __CFFromNextStepLatin(uint32_t flags, uint8_t byte, UniChar *character) {
 824     return ((*character = (byte < 0x80 ? (UniChar)byte : NSToPrecompUnicodeTable[byte - 0x80])) != 0xFFFD);
 825 }
 826
 827 static CFIndex __CFToNextStepLatinPrecompose(uint32_t flags, const UniChar *character, CFIndex numChars, uint8_t *bytes, CFIndex maxByteLen, CFIndex *usedByteLen) {
 828     uint8_t byte;
 829     CFIndex usedCharLen;
 830
 831     if (__CFToNextStepLatin(flags, CFStringEncodingPrecomposeLatinCharacter(character, numChars, &usedCharLen), &byte) && byte && (usedCharLen > 1)) {
 832         if (maxByteLen) *bytes = byte;
 833         *usedByteLen = 1;
 834         return usedCharLen;
 835     } else {
 836         return 0;
 837     }
 838 }
 839
 840 CF_PRIVATE const CFStringEncodingConverter __CFConverterNextStepLatin = {
 841     __CFToNextStepLatin, __CFFromNextStepLatin, 1, 1, kCFStringEncodingConverterCheapEightBit,
 842     NULL, NULL, NULL, NULL, __CFToNextStepLatinPrecompose, CFStringEncodingIsValidCombiningCharacterForLatin1,
 843 };
 844
 845 /* UTF8 */
 846 /*
 847  * Copyright 2001 Unicode, Inc.
 848  *
 849  * Disclaimer
 850  *
 851  * This source code is provided as is by Unicode, Inc. No claims are
 852  * made as to fitness for any particular purpose. No warranties of any
 853  * kind are expressed or implied. The recipient agrees to determine
 854  * applicability of information provided. If this file has been
 855  * purchased on magnetic or optical media from Unicode, Inc., the
 856  * sole remedy for any claim will be exchange of defective media
 857  * within 90 days of receipt.
 858  *
 859  * Limitations on Rights to Redistribute This Code
 860  *
 861  * Unicode, Inc. hereby grants the right to freely use the information
 862  * supplied in this file in the creation of products supporting the
 863  * Unicode Standard, and to make copies of this file in any form
 864  * for internal or external distribution as long as this notice
 865  * remains attached.
 866  */
 867
 868 static const uint32_t kReplacementCharacter =   0x0000FFFDUL;
 869 static const uint32_t kMaximumUCS2 =            0x0000FFFFUL;
 870 static const uint32_t kMaximumUTF16 =           0x0010FFFFUL;
 871 static const uint32_t kMaximumUCS4 =            0x7FFFFFFFUL;
 872
 873 static const int halfShift                      = 10;
 874 static const uint32_t halfBase          = 0x0010000UL;
 875 static const uint32_t halfMask          = 0x3FFUL;
 876 static const uint32_t kSurrogateHighStart       = 0xD800UL;
 877 static const uint32_t kSurrogateHighEnd = 0xDBFFUL;
 878 static const uint32_t kSurrogateLowStart        = 0xDC00UL;
 879 static const uint32_t kSurrogateLowEnd  = 0xDFFFUL;
 880
 881 /*
 882  * Index into the table below with the first byte of a UTF-8 sequence to
 883  * get the number of trailing bytes that are supposed to follow it.
 884  */
 885 static const char trailingBytesForUTF8[256] = {
 886         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 887         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 888         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 889         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 890         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 891         0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
 892         1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
 893         2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5
 894 };
 895
 896 /*
 897  * Magic values subtracted from a buffer value during UTF8 conversion.
 898  * This table contains as many values as there might be trailing bytes
 899  * in a UTF-8 sequence.
 900  */
 901 static const UTF32Char offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL,
 902                                          0x03C82080UL, 0xFA082080UL, 0x82082080UL };
 903
 904 static const uint8_t firstByteMark[7] = { 0x00, 0x00, 0xC0, 0xE0, 0xF0, 0xF8, 0xFC };
 905
 906 /* This code is similar in effect to making successive calls on the mbtowc and wctomb routines in FSS-UTF. However, it is considerably different in code:
 907         * it is adapted to be consistent with UTF16,
 908         * constants have been gathered.
 909         * loops & conditionals have been removed as much as possible for
 910         * efficiency, in favor of drop-through switch statements.
 911 */
 912
 913 CF_INLINE uint16_t __CFUTF8BytesToWriteForCharacter(uint32_t ch) {
 914     if (ch < 0x80) return  1;
 915     else if (ch < 0x800) return 2;
 916     else if (ch < 0x10000) return 3;
 917     else if (ch < 0x200000) return 4;
 918     else if (ch < 0x4000000) return 5;
 919     else if (ch <= kMaximumUCS4) return 6;
 920     else return 0;
 921 }
 922
 923 CF_INLINE uint16_t __CFToUTF8Core(uint32_t ch, uint8_t *bytes, uint32_t maxByteLen) {
 924     uint16_t bytesToWrite = __CFUTF8BytesToWriteForCharacter(ch);
 925     const uint32_t byteMask = 0xBF;
 926     const uint32_t byteMark = 0x80;
 927
 928     if (!bytesToWrite) {
 929         bytesToWrite = 2;
 930         ch = kReplacementCharacter;
 931     }
 932
 933     if (maxByteLen < bytesToWrite) return 0;
 934
 935     switch (bytesToWrite) {     /* note: code falls through cases! */
 936         case 6: bytes[5] = (ch | byteMark) & byteMask; ch >>= 6;
 937         case 5: bytes[4] = (ch | byteMark) & byteMask; ch >>= 6;
 938         case 4: bytes[3] = (ch | byteMark) & byteMask; ch >>= 6;
 939         case 3: bytes[2] = (ch | byteMark) & byteMask; ch >>= 6;
 940         case 2: bytes[1] = (ch | byteMark) & byteMask; ch >>= 6;
 941         case 1: bytes[0] =  ch | firstByteMark[bytesToWrite];
 942     }
 943     return bytesToWrite;
 944 }
 945
 946 static CFIndex __CFToUTF8(uint32_t flags, const UniChar *characters, CFIndex numChars, uint8_t *bytes, CFIndex maxByteLen, CFIndex *usedByteLen) {
 947     uint16_t bytesWritten;
 948     uint32_t ch;
 949     const UniChar *beginCharacter = characters;
 950     const UniChar *endCharacter = characters + numChars;
 951     const uint8_t *beginBytes = bytes;
 952     const uint8_t *endBytes = bytes + maxByteLen;
 953     bool isStrict = (flags & kCFStringEncodingUseHFSPlusCanonical ? false : true);
 954
 955     while ((characters < endCharacter) && (!maxByteLen || (bytes < endBytes))) {
 956         ch = *(characters++);
 957
 958         if (ch < 0x80) { // ASCII
 959             if (maxByteLen) *bytes = ch;
 960             ++bytes;
 961         } else {
 962             if (ch >= kSurrogateHighStart) {
 963                 if (ch <= kSurrogateHighEnd) {
 964                     if ((characters < endCharacter) && ((*characters >= kSurrogateLowStart) && (*characters <= kSurrogateLowEnd))) {
 965                         ch = ((ch - kSurrogateHighStart) << halfShift) + (*(characters++) - kSurrogateLowStart) + halfBase;
 966                     } else if (isStrict) {
 967                         --characters;
 968                         break;
 969                     }
 970                 } else if (isStrict && (ch <= kSurrogateLowEnd)) {
 971                     --characters;
 972                     break;
 973                 }
 974             }
 975
 976             if (!(bytesWritten = (maxByteLen ? __CFToUTF8Core(ch, bytes, endBytes - bytes) : __CFUTF8BytesToWriteForCharacter(ch)))) {
 977                 characters -= (ch < 0x10000 ? 1 : 2);
 978                 break;
 979             }
 980             bytes += bytesWritten;
 981         }
 982     }
 983
 984     if (usedByteLen) *usedByteLen = bytes - beginBytes;
 985     return characters - beginCharacter;
 986 }
 987
 988 /*
 989  * Utility routine to tell whether a sequence of bytes is legal UTF-8.
 990  * This must be called with the length pre-determined by the first byte.
 991  * If not calling this from ConvertUTF8to*, then the length can be set by:
 992  *      length = trailingBytesForUTF8[*source]+1;
 993  * and the sequence is illegal right away if there aren't that many bytes
 994  * available.
 995  * If presented with a length > 4, this returns false.  The Unicode
 996  * definition of UTF-8 goes up to 4-byte sequences.
 997  */
 998
 999 CF_INLINE bool __CFIsLegalUTF8(const uint8_t *source, CFIndex length) {
1000     if (length > 4) return false;
1001
1002     const uint8_t *srcptr = source+length;
1003     uint8_t head = *source;
1004
1005     while (--srcptr > source) if ((*srcptr & 0xC0) != 0x80) return false;
1006
1007     if (((head >= 0x80) && (head < 0xC2)) || (head > 0xF4)) return false;
1008
1009     if (((head == 0xE0) && (*(source + 1) < 0xA0)) || ((head == 0xED) && (*(source + 1) > 0x9F)) || ((head == 0xF0) && (*(source + 1) < 0x90)) || ((head == 0xF4) && (*(source + 1) > 0x8F))) return false;
1010     return true;
1011 }
1012
1013 static CFIndex __CFFromUTF8(uint32_t flags, const uint8_t *bytes, CFIndex numBytes, UniChar *characters, CFIndex maxCharLen, CFIndex *usedCharLen) {
1014     const uint8_t *source = bytes;
1015     uint16_t extraBytesToRead;
1016     CFIndex theUsedCharLen = 0;
1017     uint32_t ch;
1018     bool isHFSPlus = (flags & kCFStringEncodingUseHFSPlusCanonical ? true : false);
1019     bool needsToDecompose = (flags & kCFStringEncodingUseCanonical || isHFSPlus ? true : false);
1020     bool strictUTF8 = (flags & kCFStringEncodingLenientUTF8Conversion ? false : true);
1021     UTF32Char decomposed[MAX_DECOMPOSED_LENGTH];
1022     CFIndex decompLength;
1023     bool isStrict = !isHFSPlus;
1024
1025     while (numBytes && (!maxCharLen || (theUsedCharLen < maxCharLen))) {
1026         extraBytesToRead = trailingBytesForUTF8[*source];
1027
1028         if (extraBytesToRead > --numBytes) break;
1029         numBytes -= extraBytesToRead;
1030
1031         /* Do this check whether lenient or strict */
1032         // We need to allow 0xA9 (copyright in MacRoman and Unicode) not to break existing apps
1033         // Will use a flag passed in from upper layers to switch restriction mode for this case in the next release
1034         if ((extraBytesToRead > 3) || (strictUTF8 && !__CFIsLegalUTF8(source, extraBytesToRead + 1))) {
1035             if ((*source == 0xA9) || (flags & kCFStringEncodingAllowLossyConversion)) {
1036                 numBytes += extraBytesToRead;
1037                 ++source;
1038                 if (maxCharLen) *(characters++) = (UTF16Char)kReplacementCharacter;
1039                 ++theUsedCharLen;
1040                 continue;
1041             } else {
1042                 break;
1043             }
1044         }
1045
1046         ch = 0;
1047         /*
1048          * The cases all fall through. See "Note A" below.
1049          */
1050         switch (extraBytesToRead) {
1051             case 3:     ch += *source++; ch <<= 6;
1052             case 2:     ch += *source++; ch <<= 6;
1053             case 1:     ch += *source++; ch <<= 6;
1054             case 0:     ch += *source++;
1055         }
1056         ch -= offsetsFromUTF8[extraBytesToRead];
1057
1058         if (ch <= kMaximumUCS2) {
1059             if (isStrict && (ch >= kSurrogateHighStart && ch <= kSurrogateLowEnd)) {
1060                 source -= (extraBytesToRead + 1);
1061                 break;
1062             }
1063             if (needsToDecompose && CFUniCharIsDecomposableCharacter(ch, isHFSPlus)) {
1064                 decompLength = CFUniCharDecomposeCharacter(ch, decomposed, MAX_DECOMPOSED_LENGTH);
1065
1066                 if (maxCharLen) {
1067                     if (!CFUniCharFillDestinationBuffer(decomposed, decompLength, (void **)&characters, maxCharLen, &theUsedCharLen, kCFUniCharUTF16Format)) break;
1068                 } else {
1069                     theUsedCharLen += decompLength;
1070                 }
1071             } else {
1072                 if (maxCharLen) *(characters++) = (UTF16Char)ch;
1073                 ++theUsedCharLen;
1074             }
1075         } else if (ch > kMaximumUTF16) {
1076             if (isStrict) {
1077                 source -= (extraBytesToRead + 1);
1078                 break;
1079             }
1080             if (maxCharLen) *(characters++) = (UTF16Char)kReplacementCharacter;
1081             ++theUsedCharLen;
1082         } else {
1083             if (needsToDecompose && CFUniCharIsDecomposableCharacter(ch, isHFSPlus)) {
1084                 decompLength = CFUniCharDecomposeCharacter(ch, decomposed, MAX_DECOMPOSED_LENGTH);
1085
1086                 if (maxCharLen) {
1087                     if (!CFUniCharFillDestinationBuffer(decomposed, decompLength, (void **)&characters, maxCharLen, &theUsedCharLen, kCFUniCharUTF16Format)) break;
1088                 } else {
1089                     while (--decompLength >= 0) theUsedCharLen += (decomposed[decompLength] < 0x10000 ? 1 : 2);
1090                 }
1091             } else {
1092                 if (maxCharLen) {
1093                     if ((theUsedCharLen + 2) > maxCharLen) break;
1094                     ch -= halfBase;
1095                     *(characters++) = (ch >> halfShift) + kSurrogateHighStart;
1096                     *(characters++) = (ch & halfMask) + kSurrogateLowStart;
1097                 }
1098                 theUsedCharLen += 2;
1099             }
1100         }
1101     }
1102
1103     if (usedCharLen) *usedCharLen = theUsedCharLen;
1104
1105     return source - bytes;
1106 }
1107
1108 static CFIndex __CFToUTF8Len(uint32_t flags, const UniChar *characters, CFIndex numChars) {
1109     uint32_t bytesToWrite = 0;
1110     uint32_t ch;
1111
1112     while (numChars) {
1113         ch = *characters++;
1114         numChars--;
1115         if ((ch >= kSurrogateHighStart && ch <= kSurrogateHighEnd) && numChars && (*characters >= kSurrogateLowStart && *characters <= kSurrogateLowEnd)) {
1116             ch = ((ch - kSurrogateHighStart) << halfShift) + (*characters++ - kSurrogateLowStart) + halfBase;
1117             numChars--;
1118         }
1119         bytesToWrite += __CFUTF8BytesToWriteForCharacter(ch);
1120     }
1121
1122     return bytesToWrite;
1123 }
1124
1125 static CFIndex __CFFromUTF8Len(uint32_t flags, const uint8_t *source, CFIndex numBytes) {
1126     uint16_t extraBytesToRead;
1127     CFIndex theUsedCharLen = 0;
1128     uint32_t ch;
1129     bool isHFSPlus = (flags & kCFStringEncodingUseHFSPlusCanonical ? true : false);
1130     bool needsToDecompose = (flags & kCFStringEncodingUseCanonical || isHFSPlus ? true : false);
1131     bool strictUTF8 = (flags & kCFStringEncodingLenientUTF8Conversion ? false : true);
1132     UTF32Char decomposed[MAX_DECOMPOSED_LENGTH];
1133     CFIndex decompLength;
1134     bool isStrict = !isHFSPlus;
1135
1136     while (numBytes) {
1137         extraBytesToRead = trailingBytesForUTF8[*source];
1138
1139         if (extraBytesToRead > --numBytes) break;
1140         numBytes -= extraBytesToRead;
1141
1142         /* Do this check whether lenient or strict */
1143         // We need to allow 0xA9 (copyright in MacRoman and Unicode) not to break existing apps
1144         // Will use a flag passed in from upper layers to switch restriction mode for this case in the next release
1145         if ((extraBytesToRead > 3) || (strictUTF8 && !__CFIsLegalUTF8(source, extraBytesToRead + 1))) {
1146             if ((*source == 0xA9) || (flags & kCFStringEncodingAllowLossyConversion)) {
1147                 numBytes += extraBytesToRead;
1148                 ++source;
1149                 ++theUsedCharLen;
1150                 continue;
1151             } else {
1152                 break;
1153             }
1154         }
1155
1156
1157         ch = 0;
1158         /*
1159          * The cases all fall through. See "Note A" below.
1160          */
1161         switch (extraBytesToRead) {
1162             case 3:     ch += *source++; ch <<= 6;
1163             case 2:     ch += *source++; ch <<= 6;
1164             case 1:     ch += *source++; ch <<= 6;
1165             case 0:     ch += *source++;
1166         }
1167         ch -= offsetsFromUTF8[extraBytesToRead];
1168
1169         if (ch <= kMaximumUCS2) {
1170             if (isStrict && (ch >= kSurrogateHighStart && ch <= kSurrogateLowEnd)) {
1171                 break;
1172             }
1173             if (needsToDecompose && CFUniCharIsDecomposableCharacter(ch, isHFSPlus)) {
1174                 decompLength = CFUniCharDecomposeCharacter(ch, decomposed, MAX_DECOMPOSED_LENGTH);
1175                 theUsedCharLen += decompLength;
1176             } else {
1177                 ++theUsedCharLen;
1178             }
1179         } else if (ch > kMaximumUTF16) {
1180             ++theUsedCharLen;
1181         } else {
1182             if (needsToDecompose && CFUniCharIsDecomposableCharacter(ch, isHFSPlus)) {
1183                 decompLength = CFUniCharDecomposeCharacter(ch, decomposed, MAX_DECOMPOSED_LENGTH);
1184                 while (--decompLength >= 0) theUsedCharLen += (decomposed[decompLength] < 0x10000 ? 1 : 2);
1185             } else {
1186                 theUsedCharLen += 2;
1187             }
1188         }
1189     }
1190
1191     return theUsedCharLen;
1192 }
1193
1194 CF_PRIVATE const CFStringEncodingConverter __CFConverterUTF8 = {
1195     __CFToUTF8, __CFFromUTF8, 3, 2, kCFStringEncodingConverterStandard,
1196     __CFToUTF8Len, __CFFromUTF8Len, NULL, NULL, NULL, NULL,
1197 };