src/third_party/icu/source/common/ucnv_ct.c

   1 /*
   2 **********************************************************************
   3 *   Copyright (C) 2010-2012, International Business Machines
   4 *   Corporation and others.  All Rights Reserved.
   5 **********************************************************************
   6 *   file name:  ucnv_ct.c
   7 *   encoding:   US-ASCII
   8 *   tab size:   8 (not used)
   9 *   indentation:4
  10 *
  11 *   created on: 2010Dec09
  12 *   created by: Michael Ow
  13 */
  14
  15 #include "unicode/utypes.h"
  16
  17 #if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION && !UCONFIG_NO_NON_HTML5_CONVERSION
  18
  19 #include "unicode/ucnv.h"
  20 #include "unicode/uset.h"
  21 #include "unicode/ucnv_err.h"
  22 #include "unicode/ucnv_cb.h"
  23 #include "unicode/utf16.h"
  24 #include "ucnv_imp.h"
  25 #include "ucnv_bld.h"
  26 #include "ucnv_cnv.h"
  27 #include "ucnvmbcs.h"
  28 #include "cstring.h"
  29 #include "cmemory.h"
  30
  31 #define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
  32
  33 typedef enum {
  34     INVALID = -2,
  35     DO_SEARCH = -1,
  36
  37     COMPOUND_TEXT_SINGLE_0 = 0,
  38     COMPOUND_TEXT_SINGLE_1 = 1,
  39     COMPOUND_TEXT_SINGLE_2 = 2,
  40     COMPOUND_TEXT_SINGLE_3 = 3,
  41
  42     COMPOUND_TEXT_DOUBLE_1 = 4,
  43     COMPOUND_TEXT_DOUBLE_2 = 5,
  44     COMPOUND_TEXT_DOUBLE_3 = 6,
  45     COMPOUND_TEXT_DOUBLE_4 = 7,
  46     COMPOUND_TEXT_DOUBLE_5 = 8,
  47     COMPOUND_TEXT_DOUBLE_6 = 9,
  48     COMPOUND_TEXT_DOUBLE_7 = 10,
  49
  50     COMPOUND_TEXT_TRIPLE_DOUBLE = 11,
  51
  52     IBM_915 = 12,
  53     IBM_916 = 13,
  54     IBM_914 = 14,
  55     IBM_874 = 15,
  56     IBM_912 = 16,
  57     IBM_913 = 17,
  58     ISO_8859_14 = 18,
  59     IBM_923 = 19,
  60     NUM_OF_CONVERTERS = 20
  61 } COMPOUND_TEXT_CONVERTERS;
  62
  63 #define SEARCH_LENGTH 12
  64
  65 static const uint8_t escSeqCompoundText[NUM_OF_CONVERTERS][5] = {
  66     /* Single */
  67     { 0x1B, 0x2D, 0x41, 0, 0 },
  68     { 0x1B, 0x2D, 0x4D, 0, 0 },
  69     { 0x1B, 0x2D, 0x46, 0, 0 },
  70     { 0x1B, 0x2D, 0x47, 0, 0 },
  71
  72     /* Double */
  73     { 0x1B, 0x24, 0x29, 0x41, 0 },
  74     { 0x1B, 0x24, 0x29, 0x42, 0 },
  75     { 0x1B, 0x24, 0x29, 0x43, 0 },
  76     { 0x1B, 0x24, 0x29, 0x44, 0 },
  77     { 0x1B, 0x24, 0x29, 0x47, 0 },
  78     { 0x1B, 0x24, 0x29, 0x48, 0 },
  79     { 0x1B, 0x24, 0x29, 0x49, 0 },
  80
  81     /* Triple/Double */
  82     { 0x1B, 0x25, 0x47, 0, 0 },
  83
  84     /*IBM-915*/
  85     { 0x1B, 0x2D, 0x4C, 0, 0 },
  86     /*IBM-916*/
  87     { 0x1B, 0x2D, 0x48, 0, 0 },
  88     /*IBM-914*/
  89     { 0x1B, 0x2D, 0x44, 0, 0 },
  90     /*IBM-874*/
  91     { 0x1B, 0x2D, 0x54, 0, 0 },
  92     /*IBM-912*/
  93     { 0x1B, 0x2D, 0x42, 0, 0 },
  94     /* IBM-913 */
  95     { 0x1B, 0x2D, 0x43, 0, 0 },
  96     /* ISO-8859_14 */
  97     { 0x1B, 0x2D, 0x5F, 0, 0 },
  98     /* IBM-923 */
  99     { 0x1B, 0x2D, 0x62, 0, 0 },
 100 };
 101
 102 #define ESC_START 0x1B
 103
 104 #define isASCIIRange(codepoint) \
 105         ((codepoint == 0x0000) || (codepoint == 0x0009) || (codepoint == 0x000A) || \
 106          (codepoint >= 0x0020 && codepoint <= 0x007f) || (codepoint >= 0x00A0 && codepoint <= 0x00FF))
 107
 108 #define isIBM915(codepoint) \
 109         ((codepoint >= 0x0401 && codepoint <= 0x045F) || (codepoint == 0x2116))
 110
 111 #define isIBM916(codepoint) \
 112         ((codepoint >= 0x05D0 && codepoint <= 0x05EA) || (codepoint == 0x2017) || (codepoint == 0x203E))
 113
 114 #define isCompoundS3(codepoint) \
 115         ((codepoint == 0x060C) || (codepoint == 0x061B) || (codepoint == 0x061F) || (codepoint >= 0x0621 && codepoint <= 0x063A) || \
 116          (codepoint >= 0x0640 && codepoint <= 0x0652) || (codepoint >= 0x0660 && codepoint <= 0x066D) || (codepoint == 0x200B) || \
 117          (codepoint >= 0x0FE70 && codepoint <= 0x0FE72) || (codepoint == 0x0FE74) || (codepoint >= 0x0FE76 && codepoint <= 0x0FEBE))
 118
 119 #define isCompoundS2(codepoint) \
 120         ((codepoint == 0x02BC) || (codepoint == 0x02BD) || (codepoint >= 0x0384 && codepoint <= 0x03CE) || (codepoint == 0x2015))
 121
 122 #define isIBM914(codepoint) \
 123         ((codepoint == 0x0100) || (codepoint == 0x0101) || (codepoint == 0x0112) || (codepoint == 0x0113) || (codepoint == 0x0116) || (codepoint == 0x0117) || \
 124          (codepoint == 0x0122) || (codepoint == 0x0123) || (codepoint >= 0x0128 && codepoint <= 0x012B) || (codepoint == 0x012E) || (codepoint == 0x012F) || \
 125          (codepoint >= 0x0136 && codepoint <= 0x0138) || (codepoint == 0x013B) || (codepoint == 0x013C) || (codepoint == 0x0145) || (codepoint ==  0x0146) || \
 126          (codepoint >= 0x014A && codepoint <= 0x014D) || (codepoint == 0x0156) || (codepoint == 0x0157) || (codepoint >= 0x0166 && codepoint <= 0x016B) || \
 127          (codepoint == 0x0172) || (codepoint == 0x0173))
 128
 129 #define isIBM874(codepoint) \
 130         ((codepoint >= 0x0E01 && codepoint <= 0x0E3A) || (codepoint >= 0x0E3F && codepoint <= 0x0E5B))
 131
 132 #define isIBM912(codepoint) \
 133         ((codepoint >= 0x0102 && codepoint <= 0x0107) || (codepoint >= 0x010C && codepoint <= 0x0111) || (codepoint >= 0x0118 && codepoint <= 0x011B) || \
 134          (codepoint == 0x0139) || (codepoint == 0x013A) || (codepoint == 0x013D) || (codepoint == 0x013E) || (codepoint >= 0x0141 && codepoint <= 0x0144) || \
 135          (codepoint == 0x0147) || (codepoint == 0x0147) || (codepoint == 0x0150) || (codepoint == 0x0151) || (codepoint == 0x0154) || (codepoint == 0x0155) || \
 136          (codepoint >= 0x0158 && codepoint <= 0x015B) || (codepoint == 0x015E) || (codepoint == 0x015F) || (codepoint >= 0x0160 && codepoint <= 0x0165) || \
 137          (codepoint == 0x016E) || (codepoint == 0x016F) || (codepoint == 0x0170) || (codepoint ==  0x0171) || (codepoint >= 0x0179 && codepoint <= 0x017E) || \
 138          (codepoint == 0x02C7) || (codepoint == 0x02D8) || (codepoint == 0x02D9) || (codepoint == 0x02DB) || (codepoint == 0x02DD))
 139
 140 #define isIBM913(codepoint) \
 141         ((codepoint >= 0x0108 && codepoint <= 0x010B) || (codepoint == 0x011C) || \
 142          (codepoint == 0x011D) || (codepoint == 0x0120) || (codepoint == 0x0121) || \
 143          (codepoint >= 0x0124 && codepoint <= 0x0127) || (codepoint == 0x0134) || (codepoint == 0x0135) || \
 144          (codepoint == 0x015C) || (codepoint == 0x015D) || (codepoint == 0x016C) || (codepoint ==  0x016D))
 145
 146 #define isCompoundS1(codepoint) \
 147         ((codepoint == 0x011E) || (codepoint == 0x011F) || (codepoint == 0x0130) || \
 148          (codepoint == 0x0131) || (codepoint >= 0x0218 && codepoint <= 0x021B))
 149
 150 #define isISO8859_14(codepoint) \
 151         ((codepoint >= 0x0174 && codepoint <= 0x0177) || (codepoint == 0x1E0A) || \
 152          (codepoint == 0x1E0B) || (codepoint == 0x1E1E) || (codepoint == 0x1E1F) || \
 153          (codepoint == 0x1E40) || (codepoint == 0x1E41) || (codepoint == 0x1E56) || \
 154          (codepoint == 0x1E57) || (codepoint == 0x1E60) || (codepoint == 0x1E61) || \
 155          (codepoint == 0x1E6A) || (codepoint == 0x1E6B) || (codepoint == 0x1EF2) || \
 156          (codepoint == 0x1EF3) || (codepoint >= 0x1E80 && codepoint <= 0x1E85))
 157
 158 #define isIBM923(codepoint) \
 159         ((codepoint == 0x0152) || (codepoint == 0x0153) || (codepoint == 0x0178) || (codepoint == 0x20AC))
 160
 161
 162 typedef struct{
 163     UConverterSharedData *myConverterArray[NUM_OF_CONVERTERS];
 164     COMPOUND_TEXT_CONVERTERS state;
 165 } UConverterDataCompoundText;
 166
 167 /*********** Compound Text Converter Protos ***********/
 168 static void
 169 _CompoundTextOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode);
 170
 171 static void
 172  _CompoundTextClose(UConverter *converter);
 173
 174 static void
 175 _CompoundTextReset(UConverter *converter, UConverterResetChoice choice);
 176
 177 static const char*
 178 _CompoundTextgetName(const UConverter* cnv);
 179
 180
 181 static int32_t findNextEsc(const char *source, const char *sourceLimit) {
 182     int32_t length = sourceLimit - source;
 183     int32_t i;
 184     for (i = 1; i < length; i++) {
 185         if (*(source + i) == 0x1B) {
 186             return i;
 187         }
 188     }
 189
 190     return length;
 191 }
 192
 193 static COMPOUND_TEXT_CONVERTERS getState(int codepoint) {
 194     COMPOUND_TEXT_CONVERTERS state = DO_SEARCH;
 195
 196     if (isASCIIRange(codepoint)) {
 197         state = COMPOUND_TEXT_SINGLE_0;
 198     } else if (isIBM912(codepoint)) {
 199         state = IBM_912;
 200     }else if (isIBM913(codepoint)) {
 201         state = IBM_913;
 202     } else if (isISO8859_14(codepoint)) {
 203         state = ISO_8859_14;
 204     } else if (isIBM923(codepoint)) {
 205         state = IBM_923;
 206     } else if (isIBM874(codepoint)) {
 207         state = IBM_874;
 208     } else if (isIBM914(codepoint)) {
 209         state = IBM_914;
 210     } else if (isCompoundS2(codepoint)) {
 211         state = COMPOUND_TEXT_SINGLE_2;
 212     } else if (isCompoundS3(codepoint)) {
 213         state = COMPOUND_TEXT_SINGLE_3;
 214     } else if (isIBM916(codepoint)) {
 215         state = IBM_916;
 216     } else if (isIBM915(codepoint)) {
 217         state = IBM_915;
 218     } else if (isCompoundS1(codepoint)) {
 219         state = COMPOUND_TEXT_SINGLE_1;
 220     }
 221
 222     return state;
 223 }
 224
 225 static COMPOUND_TEXT_CONVERTERS findStateFromEscSeq(const char* source, const char* sourceLimit, const uint8_t* toUBytesBuffer, int32_t toUBytesBufferLength, UErrorCode *err) {
 226     COMPOUND_TEXT_CONVERTERS state = INVALID;
 227     UBool matchFound = FALSE;
 228     int32_t i, n, offset = toUBytesBufferLength;
 229
 230     for (i = 0; i < NUM_OF_CONVERTERS; i++) {
 231         matchFound = TRUE;
 232         for (n = 0; escSeqCompoundText[i][n] != 0; n++) {
 233             if (n < toUBytesBufferLength) {
 234                 if (toUBytesBuffer[n] != escSeqCompoundText[i][n]) {
 235                     matchFound = FALSE;
 236                     break;
 237                 }
 238             } else if ((source + (n - offset)) >= sourceLimit) {
 239                 *err = U_TRUNCATED_CHAR_FOUND;
 240                 matchFound = FALSE;
 241                 break;
 242             } else if (*(source + (n - offset)) != escSeqCompoundText[i][n]) {
 243                 matchFound = FALSE;
 244                 break;
 245             }
 246         }
 247
 248         if (matchFound) {
 249             break;
 250         }
 251     }
 252
 253     if (matchFound) {
 254         state = (COMPOUND_TEXT_CONVERTERS)i;
 255     }
 256
 257     return state;
 258 }
 259
 260 static void
 261 _CompoundTextOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode){
 262     cnv->extraInfo = uprv_malloc (sizeof (UConverterDataCompoundText));
 263     if (cnv->extraInfo != NULL) {
 264         UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) cnv->extraInfo;
 265
 266         UConverterNamePieces stackPieces;
 267         UConverterLoadArgs stackArgs={ (int32_t)sizeof(UConverterLoadArgs) };
 268
 269         myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_0] = NULL;
 270         myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_1] = ucnv_loadSharedData("icu-internal-compound-s1", &stackPieces, &stackArgs, errorCode);
 271         myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_2] = ucnv_loadSharedData("icu-internal-compound-s2", &stackPieces, &stackArgs, errorCode);
 272         myConverterData->myConverterArray[COMPOUND_TEXT_SINGLE_3] = ucnv_loadSharedData("icu-internal-compound-s3", &stackPieces, &stackArgs, errorCode);
 273         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_1] = ucnv_loadSharedData("icu-internal-compound-d1", &stackPieces, &stackArgs, errorCode);
 274         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_2] = ucnv_loadSharedData("icu-internal-compound-d2", &stackPieces, &stackArgs, errorCode);
 275         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_3] = ucnv_loadSharedData("icu-internal-compound-d3", &stackPieces, &stackArgs, errorCode);
 276         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_4] = ucnv_loadSharedData("icu-internal-compound-d4", &stackPieces, &stackArgs, errorCode);
 277         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_5] = ucnv_loadSharedData("icu-internal-compound-d5", &stackPieces, &stackArgs, errorCode);
 278         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_6] = ucnv_loadSharedData("icu-internal-compound-d6", &stackPieces, &stackArgs, errorCode);
 279         myConverterData->myConverterArray[COMPOUND_TEXT_DOUBLE_7] = ucnv_loadSharedData("icu-internal-compound-d7", &stackPieces, &stackArgs, errorCode);
 280         myConverterData->myConverterArray[COMPOUND_TEXT_TRIPLE_DOUBLE] = ucnv_loadSharedData("icu-internal-compound-t", &stackPieces, &stackArgs, errorCode);
 281
 282         myConverterData->myConverterArray[IBM_915] = ucnv_loadSharedData("ibm-915_P100-1995", &stackPieces, &stackArgs, errorCode);
 283         myConverterData->myConverterArray[IBM_916] = ucnv_loadSharedData("ibm-916_P100-1995", &stackPieces, &stackArgs, errorCode);
 284         myConverterData->myConverterArray[IBM_914] = ucnv_loadSharedData("ibm-914_P100-1995", &stackPieces, &stackArgs, errorCode);
 285         myConverterData->myConverterArray[IBM_874] = ucnv_loadSharedData("ibm-874_P100-1995", &stackPieces, &stackArgs, errorCode);
 286         myConverterData->myConverterArray[IBM_912] = ucnv_loadSharedData("ibm-912_P100-1995", &stackPieces, &stackArgs, errorCode);
 287         myConverterData->myConverterArray[IBM_913] = ucnv_loadSharedData("ibm-913_P100-2000", &stackPieces, &stackArgs, errorCode);
 288         myConverterData->myConverterArray[ISO_8859_14] = ucnv_loadSharedData("iso-8859_14-1998", &stackPieces, &stackArgs, errorCode);
 289         myConverterData->myConverterArray[IBM_923] = ucnv_loadSharedData("ibm-923_P100-1998", &stackPieces, &stackArgs, errorCode);
 290
 291         if (U_FAILURE(*errorCode) || pArgs->onlyTestIsLoadable) {
 292             _CompoundTextClose(cnv);
 293             return;
 294         }
 295
 296         myConverterData->state = (COMPOUND_TEXT_CONVERTERS)0;
 297     } else {
 298         *errorCode = U_MEMORY_ALLOCATION_ERROR;
 299     }
 300 }
 301
 302
 303 static void
 304 _CompoundTextClose(UConverter *converter) {
 305     UConverterDataCompoundText* myConverterData = (UConverterDataCompoundText*)(converter->extraInfo);
 306     int32_t i;
 307
 308     if (converter->extraInfo != NULL) {
 309         /*close the array of converter pointers and free the memory*/
 310         for (i = 0; i < NUM_OF_CONVERTERS; i++) {
 311             if (myConverterData->myConverterArray[i] != NULL) {
 312                 ucnv_unloadSharedDataIfReady(myConverterData->myConverterArray[i]);
 313             }
 314         }
 315
 316         uprv_free(converter->extraInfo);
 317     }
 318 }
 319
 320 static void
 321 _CompoundTextReset(UConverter *converter, UConverterResetChoice choice) {
 322 }
 323
 324 static const char*
 325 _CompoundTextgetName(const UConverter* cnv){
 326     return "x11-compound-text";
 327 }
 328
 329 static void
 330 UConverter_fromUnicode_CompoundText_OFFSETS(UConverterFromUnicodeArgs* args, UErrorCode* err){
 331     UConverter *cnv = args->converter;
 332     uint8_t *target = (uint8_t *) args->target;
 333     const uint8_t *targetLimit = (const uint8_t *) args->targetLimit;
 334     const UChar* source = args->source;
 335     const UChar* sourceLimit = args->sourceLimit;
 336     /* int32_t* offsets = args->offsets; */
 337     UChar32 sourceChar;
 338     UBool useFallback = cnv->useFallback;
 339     uint8_t tmpTargetBuffer[7];
 340     int32_t tmpTargetBufferLength = 0;
 341     COMPOUND_TEXT_CONVERTERS currentState, tmpState;
 342     uint32_t pValue;
 343     int32_t pValueLength = 0;
 344     int32_t i, n, j;
 345
 346     UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) cnv->extraInfo;
 347
 348     currentState = myConverterData->state;
 349
 350     /* check if the last codepoint of previous buffer was a lead surrogate*/
 351     if((sourceChar = cnv->fromUChar32)!=0 && target< targetLimit) {
 352         goto getTrail;
 353     }
 354
 355     while( source < sourceLimit){
 356         if(target < targetLimit){
 357
 358             sourceChar  = *(source++);
 359             /*check if the char is a First surrogate*/
 360              if(U16_IS_SURROGATE(sourceChar)) {
 361                 if(U16_IS_SURROGATE_LEAD(sourceChar)) {
 362 getTrail:
 363                     /*look ahead to find the trail surrogate*/
 364                     if(source < sourceLimit) {
 365                         /* test the following code unit */
 366                         UChar trail=(UChar) *source;
 367                         if(U16_IS_TRAIL(trail)) {
 368                             source++;
 369                             sourceChar=U16_GET_SUPPLEMENTARY(sourceChar, trail);
 370                             cnv->fromUChar32=0x00;
 371                             /* convert this supplementary code point */
 372                             /* exit this condition tree */
 373                         } else {
 374                             /* this is an unmatched lead code unit (1st surrogate) */
 375                             /* callback(illegal) */
 376                             *err=U_ILLEGAL_CHAR_FOUND;
 377                             cnv->fromUChar32=sourceChar;
 378                             break;
 379                         }
 380                     } else {
 381                         /* no more input */
 382                         cnv->fromUChar32=sourceChar;
 383                         break;
 384                     }
 385                 } else {
 386                     /* this is an unmatched trail code unit (2nd surrogate) */
 387                     /* callback(illegal) */
 388                     *err=U_ILLEGAL_CHAR_FOUND;
 389                     cnv->fromUChar32=sourceChar;
 390                     break;
 391                 }
 392             }
 393
 394              tmpTargetBufferLength = 0;
 395              tmpState = getState(sourceChar);
 396
 397              if (tmpState != DO_SEARCH && currentState != tmpState) {
 398                  /* Get escape sequence if necessary */
 399                  currentState = tmpState;
 400                  for (i = 0; escSeqCompoundText[currentState][i] != 0; i++) {
 401                      tmpTargetBuffer[tmpTargetBufferLength++] = escSeqCompoundText[currentState][i];
 402                  }
 403              }
 404
 405              if (tmpState == DO_SEARCH) {
 406                  /* Test all available converters */
 407                  for (i = 1; i < SEARCH_LENGTH; i++) {
 408                      pValueLength = ucnv_MBCSFromUChar32(myConverterData->myConverterArray[i], sourceChar, &pValue, useFallback);
 409                      if (pValueLength > 0) {
 410                          tmpState = (COMPOUND_TEXT_CONVERTERS)i;
 411                          if (currentState != tmpState) {
 412                              currentState = tmpState;
 413                              for (j = 0; escSeqCompoundText[currentState][j] != 0; j++) {
 414                                  tmpTargetBuffer[tmpTargetBufferLength++] = escSeqCompoundText[currentState][j];
 415                              }
 416                          }
 417                          for (n = (pValueLength - 1); n >= 0; n--) {
 418                              tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)(pValue >> (n * 8));
 419                          }
 420                          break;
 421                      }
 422                  }
 423              } else if (tmpState == COMPOUND_TEXT_SINGLE_0) {
 424                  tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)sourceChar;
 425              } else {
 426                  pValueLength = ucnv_MBCSFromUChar32(myConverterData->myConverterArray[currentState], sourceChar, &pValue, useFallback);
 427                  if (pValueLength > 0) {
 428                      for (n = (pValueLength - 1); n >= 0; n--) {
 429                          tmpTargetBuffer[tmpTargetBufferLength++] = (uint8_t)(pValue >> (n * 8));
 430                      }
 431                  }
 432              }
 433
 434              for (i = 0; i < tmpTargetBufferLength; i++) {
 435                  if (target < targetLimit) {
 436                      *target++ = tmpTargetBuffer[i];
 437                  } else {
 438                      *err = U_BUFFER_OVERFLOW_ERROR;
 439                      break;
 440                  }
 441              }
 442
 443              if (*err == U_BUFFER_OVERFLOW_ERROR) {
 444                  for (; i < tmpTargetBufferLength; i++) {
 445                      args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = tmpTargetBuffer[i];
 446                  }
 447              }
 448         } else {
 449             *err = U_BUFFER_OVERFLOW_ERROR;
 450             break;
 451         }
 452     }
 453
 454     /*save the state and return */
 455     myConverterData->state = currentState;
 456     args->source = source;
 457     args->target = (char*)target;
 458 }
 459
 460
 461 static void
 462 UConverter_toUnicode_CompoundText_OFFSETS(UConverterToUnicodeArgs *args,
 463                                                UErrorCode* err){
 464     const char *mySource = (char *) args->source;
 465     UChar *myTarget = args->target;
 466     const char *mySourceLimit = args->sourceLimit;
 467     const char *tmpSourceLimit = mySourceLimit;
 468     uint32_t mySourceChar = 0x0000;
 469     COMPOUND_TEXT_CONVERTERS currentState, tmpState;
 470     int32_t sourceOffset = 0;
 471     UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *) args->converter->extraInfo;
 472     UConverterSharedData* savedSharedData = NULL;
 473
 474     UConverterToUnicodeArgs subArgs;
 475     int32_t minArgsSize;
 476
 477     /* set up the subconverter arguments */
 478     if(args->size<sizeof(UConverterToUnicodeArgs)) {
 479         minArgsSize = args->size;
 480     } else {
 481         minArgsSize = (int32_t)sizeof(UConverterToUnicodeArgs);
 482     }
 483
 484     uprv_memcpy(&subArgs, args, minArgsSize);
 485     subArgs.size = (uint16_t)minArgsSize;
 486
 487     currentState = tmpState =  myConverterData->state;
 488
 489     while(mySource < mySourceLimit){
 490         if(myTarget < args->targetLimit){
 491             if (args->converter->toULength > 0) {
 492                 mySourceChar = args->converter->toUBytes[0];
 493             } else {
 494                 mySourceChar = (uint8_t)*mySource;
 495             }
 496
 497             if (mySourceChar == ESC_START) {
 498                 tmpState = findStateFromEscSeq(mySource, mySourceLimit, args->converter->toUBytes, args->converter->toULength, err);
 499
 500                 if (*err == U_TRUNCATED_CHAR_FOUND) {
 501                     for (; mySource < mySourceLimit;) {
 502                         args->converter->toUBytes[args->converter->toULength++] = *mySource++;
 503                     }
 504                     *err = U_ZERO_ERROR;
 505                     break;
 506                 } else if (tmpState == INVALID) {
 507                     if (args->converter->toULength == 0) {
 508                         mySource++; /* skip over the 0x1b byte */
 509                     }
 510                     *err = U_ILLEGAL_CHAR_FOUND;
 511                     break;
 512                 }
 513
 514                 if (tmpState != currentState) {
 515                     currentState = tmpState;
 516                 }
 517
 518                 sourceOffset = uprv_strlen((char*)escSeqCompoundText[currentState]) - args->converter->toULength;
 519
 520                 mySource += sourceOffset;
 521
 522                 args->converter->toULength = 0;
 523             }
 524
 525             if (currentState == COMPOUND_TEXT_SINGLE_0) {
 526                 while (mySource < mySourceLimit) {
 527                     if (*mySource == ESC_START) {
 528                         break;
 529                     }
 530                     if (myTarget < args->targetLimit) {
 531                         *myTarget++ = 0x00ff&(*mySource++);
 532                     } else {
 533                         *err = U_BUFFER_OVERFLOW_ERROR;
 534                         break;
 535                     }
 536                 }
 537             } else if (mySource < mySourceLimit){
 538                 sourceOffset = findNextEsc(mySource, mySourceLimit);
 539
 540                 tmpSourceLimit = mySource + sourceOffset;
 541
 542                 subArgs.source = mySource;
 543                 subArgs.sourceLimit = tmpSourceLimit;
 544                 subArgs.target = myTarget;
 545                 savedSharedData = subArgs.converter->sharedData;
 546                 subArgs.converter->sharedData = myConverterData->myConverterArray[currentState];
 547
 548                 ucnv_MBCSToUnicodeWithOffsets(&subArgs, err);
 549
 550                 subArgs.converter->sharedData = savedSharedData;
 551
 552                 mySource = subArgs.source;
 553                 myTarget = subArgs.target;
 554
 555                 if (U_FAILURE(*err)) {
 556                     if(*err == U_BUFFER_OVERFLOW_ERROR) {
 557                         if(subArgs.converter->UCharErrorBufferLength > 0) {
 558                             uprv_memcpy(args->converter->UCharErrorBuffer, subArgs.converter->UCharErrorBuffer,
 559                                         subArgs.converter->UCharErrorBufferLength);
 560                         }
 561                         args->converter->UCharErrorBufferLength=subArgs.converter->UCharErrorBufferLength;
 562                         subArgs.converter->UCharErrorBufferLength = 0;
 563                     }
 564                     break;
 565                 }
 566             }
 567         } else {
 568             *err = U_BUFFER_OVERFLOW_ERROR;
 569             break;
 570         }
 571     }
 572     myConverterData->state = currentState;
 573     args->target = myTarget;
 574     args->source = mySource;
 575 }
 576
 577 static void
 578 _CompoundText_GetUnicodeSet(const UConverter *cnv,
 579                     const USetAdder *sa,
 580                     UConverterUnicodeSet which,
 581                     UErrorCode *pErrorCode) {
 582     UConverterDataCompoundText *myConverterData = (UConverterDataCompoundText *)cnv->extraInfo;
 583     int32_t i;
 584
 585     for (i = 1; i < NUM_OF_CONVERTERS; i++) {
 586         ucnv_MBCSGetUnicodeSetForUnicode(myConverterData->myConverterArray[i], sa, which, pErrorCode);
 587     }
 588     sa->add(sa->set, 0x0000);
 589     sa->add(sa->set, 0x0009);
 590     sa->add(sa->set, 0x000A);
 591     sa->addRange(sa->set, 0x0020, 0x007F);
 592     sa->addRange(sa->set, 0x00A0, 0x00FF);
 593 }
 594
 595 static const UConverterImpl _CompoundTextImpl = {
 596
 597     UCNV_COMPOUND_TEXT,
 598
 599     NULL,
 600     NULL,
 601
 602     _CompoundTextOpen,
 603     _CompoundTextClose,
 604     _CompoundTextReset,
 605
 606     UConverter_toUnicode_CompoundText_OFFSETS,
 607     UConverter_toUnicode_CompoundText_OFFSETS,
 608     UConverter_fromUnicode_CompoundText_OFFSETS,
 609     UConverter_fromUnicode_CompoundText_OFFSETS,
 610     NULL,
 611
 612     NULL,
 613     _CompoundTextgetName,
 614     NULL,
 615     NULL,
 616     _CompoundText_GetUnicodeSet
 617 };
 618 static const UConverterStaticData _CompoundTextStaticData = {
 619     sizeof(UConverterStaticData),
 620     "COMPOUND_TEXT",
 621     0,
 622     UCNV_IBM,
 623     UCNV_COMPOUND_TEXT,
 624     1,
 625     6,
 626     { 0xef, 0, 0, 0 },
 627     1,
 628     FALSE,
 629     FALSE,
 630     0,
 631     0,
 632     { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 } /* reserved */
 633 };
 634 const UConverterSharedData _CompoundTextData = {
 635     sizeof(UConverterSharedData),
 636     ~((uint32_t) 0),
 637     NULL,
 638     NULL,
 639     &_CompoundTextStaticData,
 640     FALSE,
 641     &_CompoundTextImpl,
 642     0
 643 };
 644
 645 #endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */