| /* |
| ******************************************************************************* |
| * |
| * Copyright (C) 1998-1999, International Business Machines |
| * Corporation and others. All Rights Reserved. |
| * |
| ******************************************************************************* |
| * |
| * ucnv_err.c |
| * Implements error behaviour functions called by T_UConverter_{from,to}Unicode |
| * |
| * |
| * Change history: |
| * |
| * 06/29/2000 helena Major rewrite of the callback APIs. |
| */ |
| |
| #include "ucmp8.h" |
| #include "ucmp16.h" |
| #include "unicode/ucnv_err.h" |
| #include "ucnv_cnv.h" |
| #include "cmemory.h" |
| #include "unicode/ucnv.h" |
| |
| #define VALUE_STRING_LENGTH 32 |
| /*Magic # 32 = 4(number of char in value string) * 8(max number of bytes per char for any converter) */ |
| #define CODEPOINT_STRING_LENGTH 7 |
| #define UNICODE_PERCENT_SIGN_CODEPOINT 0x0025 |
| #define UNICODE_U_CODEPOINT 0x0055 |
| #define UNICODE_X_CODEPOINT 0x0058 |
| |
| |
| #define ToOffset(a) a<=9?(0x0030+a):(0x0030+a+7) |
| |
| UBool |
| CONVERSION_U_SUCCESS (UErrorCode err) |
| { |
| if ((err == U_INVALID_CHAR_FOUND) || (err == U_ILLEGAL_CHAR_FOUND)) return FALSE; |
| else return TRUE; |
| } |
| |
| /*Takes a int32_t and fills in a UChar* string with that number "radix"-based |
| * and padded with "pad" zeroes |
| */ |
| static void itou (UChar * buffer, int32_t i, int32_t radix, int32_t pad) |
| { |
| int32_t length = 0; |
| int32_t num = 0; |
| int8_t digit; |
| int32_t j; |
| UChar temp; |
| |
| while (i >= radix) |
| { |
| num = i / radix; |
| digit = (int8_t) (i - num * radix); |
| buffer[length++] = (UChar) (ToOffset (digit)); |
| i = num; |
| } |
| |
| buffer[length] = (UChar) (ToOffset (i)); |
| |
| while (length < pad) buffer[++length] = (UChar) 0x0030; /*zero padding */ |
| buffer[length--] = (UChar) 0x0000; |
| |
| /*Reverses the string */ |
| for (j = 0; j < (pad / 2); j++) |
| { |
| temp = buffer[length - j]; |
| buffer[length - j] = buffer[j]; |
| buffer[j] = temp; |
| } |
| |
| return; |
| } |
| |
| /*Function Pointer STOPS at the ILLEGAL_SEQUENCE */ |
| void UCNV_FROM_U_CALLBACK_STOP ( |
| void *context, |
| UConverterFromUnicodeArgs *fromUArgs, |
| const UChar* codeUnits, |
| int32_t length, |
| UChar32 codePoint, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| reason = UCNV_ILLEGAL; |
| return; |
| } |
| |
| |
| /*Function Pointer STOPS at the ILLEGAL_SEQUENCE */ |
| void UCNV_TO_U_CALLBACK_STOP ( |
| void *context, |
| UConverterToUnicodeArgs *toUArgs, |
| const char* codePoints, |
| int32_t length, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| reason = UCNV_ILLEGAL; |
| return; |
| } |
| |
| void UCNV_FROM_U_CALLBACK_SKIP ( |
| void *context, |
| UConverterFromUnicodeArgs *fromUArgs, |
| const UChar* codeUnits, |
| int32_t length, |
| UChar32 codePoint, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| *err = U_ZERO_ERROR; |
| } |
| |
| void UCNV_FROM_U_CALLBACK_SUBSTITUTE ( |
| void *context, |
| UConverterFromUnicodeArgs *fromArgs, |
| const UChar* codeUnits, |
| int32_t length, |
| UChar32 codePoint, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| char togo[5]; |
| int32_t togoLen; |
| |
| |
| |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| |
| /*In case we're dealing with a modal converter a la UCNV_EBCDIC_STATEFUL, |
| we need to make sure that the emitting of the substitution charater in the right mode*/ |
| uprv_memcpy(togo, fromArgs->converter->subChar, togoLen = fromArgs->converter->subCharLen); |
| if (ucnv_getType(fromArgs->converter) == UCNV_EBCDIC_STATEFUL) |
| { |
| if ((fromArgs->converter->fromUnicodeStatus)&&(togoLen != 2)) |
| { |
| togo[0] = UCNV_SI; |
| togo[1] = fromArgs->converter->subChar[0]; |
| togo[2] = UCNV_SO; |
| togoLen = 3; |
| } |
| else if (!(fromArgs->converter->fromUnicodeStatus)&&(togoLen != 1)) |
| { |
| togo[0] = UCNV_SO; |
| togo[1] = fromArgs->converter->subChar[0]; |
| togo[2] = fromArgs->converter->subChar[1]; |
| togo[3] = UCNV_SI; |
| togoLen = 4; |
| } |
| } |
| |
| /*if we have enough space on the output buffer we just copy |
| the subchar there and update the pointer */ |
| if ((fromArgs->targetLimit - *(fromArgs->pTarget)) >= togoLen) |
| { |
| uprv_memcpy (*(fromArgs->pTarget), togo, togoLen); |
| *(fromArgs->pTarget) += togoLen; |
| *err = U_ZERO_ERROR; |
| if (fromArgs->offsets) |
| { |
| int i=0; |
| for (i=0;i<togoLen;i++) fromArgs->offsets[i]=0; |
| fromArgs->offsets += togoLen; |
| } |
| } |
| else |
| { |
| /*if we don't have enough space on the output buffer |
| *we copy as much as we can to it, update that pointer. |
| *copy the rest in the internal buffer, and increase the |
| *length marker |
| */ |
| uprv_memcpy (*(fromArgs->pTarget), togo, (fromArgs->targetLimit - *(fromArgs->pTarget))); |
| if (fromArgs->offsets) |
| { |
| int i=0; |
| for (i=0;i<(fromArgs->targetLimit - *(fromArgs->pTarget));i++) fromArgs->offsets[i]=0; |
| fromArgs->offsets += (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| } |
| uprv_memcpy (fromArgs->converter->charErrorBuffer + fromArgs->converter->charErrorBufferLength, |
| togo + (fromArgs->targetLimit - *(fromArgs->pTarget)), |
| togoLen - (fromArgs->targetLimit - *(fromArgs->pTarget))); |
| fromArgs->converter->charErrorBufferLength += togoLen - (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| *(fromArgs->pTarget) += (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| *err = U_INDEX_OUTOFBOUNDS_ERROR; |
| } |
| |
| return; |
| |
| } |
| |
| /*uses itou to get a unicode escape sequence of the offensive sequence, |
| *uses a clean copy (resetted) of the converter, to convert that unicode |
| *escape sequence to the target codepage (if conversion failure happens then |
| *we revert to substituting with subchar) |
| */ |
| void UCNV_FROM_U_CALLBACK_ESCAPE ( |
| void *context, |
| UConverterFromUnicodeArgs *fromArgs, |
| const UChar *codeUnits, |
| int32_t length, |
| UChar32 codePoint, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| |
| UChar valueString[VALUE_STRING_LENGTH]; |
| int32_t valueStringLength = 0; |
| UChar codepoint[CODEPOINT_STRING_LENGTH]; |
| int32_t i = 0; |
| /*Makes a bitwise copy of the converter passwd in */ |
| UConverter myConverter = *(fromArgs->converter); |
| char myTarget[VALUE_STRING_LENGTH]; |
| char *myTargetAlias = myTarget; |
| const UChar *myValueSource = NULL; |
| UErrorCode err2 = U_ZERO_ERROR; |
| uint32_t myFromUnicodeStatus = fromArgs->converter->fromUnicodeStatus; |
| UConverterFromUCallback original = NULL; |
| |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| |
| ucnv_reset (&myConverter); |
| myConverter.fromUnicodeStatus = myFromUnicodeStatus; |
| |
| ucnv_setFromUCallBack (&myConverter, |
| (UConverterFromUCallback) UCNV_FROM_U_CALLBACK_STOP, |
| NULL, /* To Do for HSYS: context is null? */ |
| original, |
| NULL, |
| &err2); |
| if (U_FAILURE (err2)) |
| { |
| *err = err2; |
| return; |
| } |
| |
| codepoint[0] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */ |
| codepoint[1] = (UChar) UNICODE_U_CODEPOINT; /* adding U */ |
| |
| while (i < fromArgs->converter->invalidUCharLength) |
| { |
| itou (codepoint + 2, fromArgs->converter->invalidUCharBuffer[i++], 16, 4); |
| uprv_memcpy (valueString + valueStringLength, codepoint, sizeof (UChar) * 6); |
| valueStringLength += CODEPOINT_STRING_LENGTH - 1; |
| } |
| |
| myValueSource = valueString; |
| |
| /*converts unicode escape sequence */ |
| ucnv_fromUnicode (&myConverter, |
| &myTargetAlias, |
| myTargetAlias + VALUE_STRING_LENGTH, |
| &myValueSource, |
| myValueSource + CODEPOINT_STRING_LENGTH - 1, |
| NULL, |
| TRUE, |
| &err2); |
| |
| if (U_FAILURE (err2)) |
| { |
| UCNV_FROM_U_CALLBACK_SUBSTITUTE ( |
| NULL, /* TO do for HSYS: context */ |
| fromArgs, |
| codeUnits, |
| length, |
| codePoint, |
| reason, |
| err); |
| return; |
| } |
| |
| |
| |
| valueStringLength = myTargetAlias - myTarget; |
| |
| /*if we have enough space on the output buffer we just copy |
| * the subchar there and update the pointer |
| */ |
| if ((fromArgs->targetLimit - *(fromArgs->pTarget)) >= valueStringLength) |
| { |
| uprv_memcpy (*(fromArgs->pTarget), myTarget, valueStringLength); |
| *(fromArgs->pTarget) += valueStringLength; |
| *err = U_ZERO_ERROR; |
| |
| if (fromArgs->offsets) |
| { |
| int j=0; |
| for (j=0;j<valueStringLength;j++) fromArgs->offsets[j]=0; |
| fromArgs->offsets += valueStringLength; |
| } |
| } |
| else |
| { |
| /*if we don't have enough space on the output buffer |
| *we copy as much as we can to it, update that pointer. |
| *copy the rest in the internal buffer, and increase the |
| *length marker |
| */ |
| |
| if (fromArgs->offsets) |
| { |
| int j=0; |
| for (j=0;j<(fromArgs->targetLimit - *(fromArgs->pTarget));j++) fromArgs->offsets[j]=0; |
| fromArgs->offsets += (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| } |
| uprv_memcpy (*(fromArgs->pTarget), myTarget, (fromArgs->targetLimit - *(fromArgs->pTarget))); |
| uprv_memcpy (fromArgs->converter->charErrorBuffer + fromArgs->converter->charErrorBufferLength, |
| myTarget + (fromArgs->targetLimit - *(fromArgs->pTarget)), |
| valueStringLength - (fromArgs->targetLimit - *(fromArgs->pTarget))); |
| fromArgs->converter->charErrorBufferLength += valueStringLength - (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| *(fromArgs->pTarget) += (fromArgs->targetLimit - *(fromArgs->pTarget)); |
| *err = U_INDEX_OUTOFBOUNDS_ERROR; |
| } |
| |
| return; |
| } |
| |
| |
| |
| void UCNV_TO_U_CALLBACK_SKIP ( |
| void *context, |
| UConverterToUnicodeArgs *toArgs, |
| const char* codeUnits, |
| int32_t length, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| *err = U_ZERO_ERROR; |
| } |
| |
| void UCNV_TO_U_CALLBACK_SUBSTITUTE ( |
| void *context, |
| UConverterToUnicodeArgs *toArgs, |
| const char* codeUnits, |
| int32_t length, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| |
| if ((toArgs->targetLimit - *(toArgs->pTarget)) >= 1) |
| { |
| **(toArgs->pTarget) = 0xFFFD; |
| (*(toArgs->pTarget))++; |
| if (toArgs->offsets) *(toArgs->offsets) = 0; |
| *err = U_ZERO_ERROR; |
| } |
| else |
| { |
| toArgs->converter->UCharErrorBuffer[toArgs->converter->UCharErrorBufferLength] = 0xFFFD; |
| toArgs->converter->UCharErrorBufferLength++; |
| *err = U_INDEX_OUTOFBOUNDS_ERROR; |
| } |
| |
| return; |
| |
| } |
| |
| /*uses itou to get a unicode escape sequence of the offensive sequence, |
| *and uses that as the substitution sequence |
| */ |
| void UCNV_TO_U_CALLBACK_ESCAPE ( |
| void *context, |
| UConverterToUnicodeArgs *toArgs, |
| const char* codeUnits, |
| int32_t length, |
| UConverterCallbackReason reason, |
| UErrorCode * err) |
| { |
| UChar uniValueString[VALUE_STRING_LENGTH]; |
| int32_t valueStringLength = 0; |
| UChar codepoint[CODEPOINT_STRING_LENGTH]; |
| int32_t i = 0; |
| |
| if (CONVERSION_U_SUCCESS (*err)) return; |
| |
| codepoint[0] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */ |
| codepoint[1] = (UChar) UNICODE_X_CODEPOINT; /* adding X */ |
| |
| while (i < toArgs->converter->invalidCharLength) |
| { |
| itou (codepoint + 2, toArgs->converter->invalidCharBuffer[i++], 16, 2); |
| uprv_memcpy (uniValueString + valueStringLength, codepoint, sizeof (UChar) * 4); |
| valueStringLength += 4; |
| } |
| |
| if ((toArgs->targetLimit - *(toArgs->pTarget)) >= valueStringLength) |
| { |
| /*if we have enough space on the output buffer we just copy |
| * the subchar there and update the pointer |
| */ |
| uprv_memcpy (*(toArgs->pTarget), uniValueString, (sizeof (UChar)) * (valueStringLength)); |
| if (toArgs->offsets) |
| { |
| for (i = 0; i < valueStringLength; i++) toArgs->offsets[i] = 0; |
| } |
| *(toArgs->pTarget) += valueStringLength; |
| |
| *err = U_ZERO_ERROR; |
| } |
| else |
| { |
| /*if we don't have enough space on the output buffer |
| *we copy as much as we can to it, update that pointer. |
| *copy the rest in the internal buffer, and increase the |
| *length marker |
| */ |
| uprv_memcpy (*(toArgs->pTarget), uniValueString, (sizeof (UChar)) * (toArgs->targetLimit - *(toArgs->pTarget))); |
| if (toArgs->offsets) |
| { |
| for (i = 0; i < (toArgs->targetLimit - *(toArgs->pTarget)); i++) toArgs->offsets[i] = 0; |
| } |
| |
| |
| uprv_memcpy (toArgs->converter->UCharErrorBuffer, |
| uniValueString + (toArgs->targetLimit - *(toArgs->pTarget)), |
| (sizeof (UChar)) * (valueStringLength - (toArgs->targetLimit - *(toArgs->pTarget)))); |
| toArgs->converter->UCharErrorBufferLength += valueStringLength - (toArgs->targetLimit - *(toArgs->pTarget)); |
| *(toArgs->pTarget) += (toArgs->targetLimit - *(toArgs->pTarget)); |
| *err = U_INDEX_OUTOFBOUNDS_ERROR; |
| } |
| |
| return; |
| } |