michael@0: /*
michael@0:  *****************************************************************************
michael@0:  *
michael@0:  *   Copyright (C) 1998-2007, International Business Machines
michael@0:  *   Corporation and others.  All Rights Reserved.
michael@0:  *
michael@0:  *****************************************************************************
michael@0:  *
michael@0:  *  ucnv_err.c
michael@0:  *  Implements error behaviour functions called by T_UConverter_{from,to}Unicode
michael@0:  *
michael@0:  *
michael@0: *   Change history:
michael@0: *
michael@0: *   06/29/2000  helena      Major rewrite of the callback APIs.
michael@0: */
michael@0: 
michael@0: #include "unicode/utypes.h"
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0: #include "unicode/ucnv_err.h"
michael@0: #include "unicode/ucnv_cb.h"
michael@0: #include "ucnv_cnv.h"
michael@0: #include "cmemory.h"
michael@0: #include "unicode/ucnv.h"
michael@0: #include "ustrfmt.h"
michael@0: 
michael@0: #define VALUE_STRING_LENGTH 32
michael@0: /*Magic # 32 = 4(number of char in value string) * 8(max number of bytes per char for any converter) */
michael@0: #define UNICODE_PERCENT_SIGN_CODEPOINT  0x0025
michael@0: #define UNICODE_U_CODEPOINT             0x0055
michael@0: #define UNICODE_X_CODEPOINT             0x0058
michael@0: #define UNICODE_RS_CODEPOINT            0x005C
michael@0: #define UNICODE_U_LOW_CODEPOINT         0x0075
michael@0: #define UNICODE_X_LOW_CODEPOINT         0x0078
michael@0: #define UNICODE_AMP_CODEPOINT           0x0026
michael@0: #define UNICODE_HASH_CODEPOINT          0x0023
michael@0: #define UNICODE_SEMICOLON_CODEPOINT     0x003B
michael@0: #define UNICODE_PLUS_CODEPOINT          0x002B
michael@0: #define UNICODE_LEFT_CURLY_CODEPOINT    0x007B
michael@0: #define UNICODE_RIGHT_CURLY_CODEPOINT   0x007D
michael@0: #define UNICODE_SPACE_CODEPOINT         0x0020
michael@0: #define UCNV_PRV_ESCAPE_ICU         0
michael@0: #define UCNV_PRV_ESCAPE_C           'C'
michael@0: #define UCNV_PRV_ESCAPE_XML_DEC     'D'
michael@0: #define UCNV_PRV_ESCAPE_XML_HEX     'X'
michael@0: #define UCNV_PRV_ESCAPE_JAVA        'J'
michael@0: #define UCNV_PRV_ESCAPE_UNICODE     'U'
michael@0: #define UCNV_PRV_ESCAPE_CSS2        'S'
michael@0: #define UCNV_PRV_STOP_ON_ILLEGAL    'i'
michael@0: 
michael@0: /*Function Pointer STOPS at the ILLEGAL_SEQUENCE */
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_FROM_U_CALLBACK_STOP (
michael@0:                   const void *context,
michael@0:                   UConverterFromUnicodeArgs *fromUArgs,
michael@0:                   const UChar* codeUnits,
michael@0:                   int32_t length,
michael@0:                   UChar32 codePoint,
michael@0:                   UConverterCallbackReason reason,
michael@0:                   UErrorCode * err)
michael@0: {
michael@0:     /* the caller must have set the error code accordingly */
michael@0:     return;
michael@0: }
michael@0: 
michael@0: 
michael@0: /*Function Pointer STOPS at the ILLEGAL_SEQUENCE */
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_TO_U_CALLBACK_STOP (
michael@0:                    const void *context,
michael@0:                    UConverterToUnicodeArgs *toUArgs,
michael@0:                    const char* codePoints,
michael@0:                    int32_t length,
michael@0:                    UConverterCallbackReason reason,
michael@0:                    UErrorCode * err)
michael@0: {
michael@0:     /* the caller must have set the error code accordingly */
michael@0:     return;
michael@0: }
michael@0: 
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_FROM_U_CALLBACK_SKIP (                  
michael@0:                   const void *context,
michael@0:                   UConverterFromUnicodeArgs *fromUArgs,
michael@0:                   const UChar* codeUnits,
michael@0:                   int32_t length,
michael@0:                   UChar32 codePoint,
michael@0:                   UConverterCallbackReason reason,
michael@0:                   UErrorCode * err)
michael@0: {
michael@0:     if (reason <= UCNV_IRREGULAR)
michael@0:     {
michael@0:         if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
michael@0:         {
michael@0:             *err = U_ZERO_ERROR;
michael@0:         }
michael@0:         /* else the caller must have set the error code accordingly. */
michael@0:     }
michael@0:     /* else ignore the reset, close and clone calls. */
michael@0: }
michael@0: 
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_FROM_U_CALLBACK_SUBSTITUTE (
michael@0:                   const void *context,
michael@0:                   UConverterFromUnicodeArgs *fromArgs,
michael@0:                   const UChar* codeUnits,
michael@0:                   int32_t length,
michael@0:                   UChar32 codePoint,
michael@0:                   UConverterCallbackReason reason,
michael@0:                   UErrorCode * err)
michael@0: {
michael@0:     if (reason <= UCNV_IRREGULAR)
michael@0:     {
michael@0:         if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
michael@0:         {
michael@0:             *err = U_ZERO_ERROR;
michael@0:             ucnv_cbFromUWriteSub(fromArgs, 0, err);
michael@0:         }
michael@0:         /* else the caller must have set the error code accordingly. */
michael@0:     }
michael@0:     /* else ignore the reset, close and clone calls. */
michael@0: }
michael@0: 
michael@0: /*uses uprv_itou to get a unicode escape sequence of the offensive sequence,
michael@0:  *uses a clean copy (resetted) of the converter, to convert that unicode
michael@0:  *escape sequence to the target codepage (if conversion failure happens then
michael@0:  *we revert to substituting with subchar)
michael@0:  */
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_FROM_U_CALLBACK_ESCAPE (
michael@0:                          const void *context,
michael@0:                          UConverterFromUnicodeArgs *fromArgs,
michael@0:                          const UChar *codeUnits,
michael@0:                          int32_t length,
michael@0:                          UChar32 codePoint,
michael@0:                          UConverterCallbackReason reason,
michael@0:                          UErrorCode * err)
michael@0: {
michael@0: 
michael@0:   UChar valueString[VALUE_STRING_LENGTH];
michael@0:   int32_t valueStringLength = 0;
michael@0:   int32_t i = 0;
michael@0: 
michael@0:   const UChar *myValueSource = NULL;
michael@0:   UErrorCode err2 = U_ZERO_ERROR;
michael@0:   UConverterFromUCallback original = NULL;
michael@0:   const void *originalContext;
michael@0: 
michael@0:   UConverterFromUCallback ignoredCallback = NULL;
michael@0:   const void *ignoredContext;
michael@0:   
michael@0:   if (reason > UCNV_IRREGULAR)
michael@0:   {
michael@0:       return;
michael@0:   }
michael@0: 
michael@0:   ucnv_setFromUCallBack (fromArgs->converter,
michael@0:                      (UConverterFromUCallback) UCNV_FROM_U_CALLBACK_SUBSTITUTE,
michael@0:                      NULL,
michael@0:                      &original,
michael@0:                      &originalContext,
michael@0:                      &err2);
michael@0:   
michael@0:   if (U_FAILURE (err2))
michael@0:   {
michael@0:     *err = err2;
michael@0:     return;
michael@0:   } 
michael@0:   if(context==NULL)
michael@0:   { 
michael@0:       while (i < length)
michael@0:       {
michael@0:         valueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT;  /* adding % */
michael@0:         valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT; /* adding U */
michael@0:         valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
michael@0:       }
michael@0:   }
michael@0:   else
michael@0:   {
michael@0:       switch(*((char*)context))
michael@0:       {
michael@0:       case UCNV_PRV_ESCAPE_JAVA:
michael@0:           while (i < length)
michael@0:           {
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_U_LOW_CODEPOINT; /* adding u */
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
michael@0:           }
michael@0:           break;
michael@0: 
michael@0:       case UCNV_PRV_ESCAPE_C:
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
michael@0: 
michael@0:           if(length==2){
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT; /* adding U */
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 8);
michael@0: 
michael@0:           }
michael@0:           else{
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_U_LOW_CODEPOINT; /* adding u */
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 4);
michael@0:           }
michael@0:           break;
michael@0: 
michael@0:       case UCNV_PRV_ESCAPE_XML_DEC:
michael@0: 
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
michael@0:           if(length==2){
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 10, 0);
michael@0:           }
michael@0:           else{
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 10, 0);
michael@0:           }
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
michael@0:           break;
michael@0: 
michael@0:       case UCNV_PRV_ESCAPE_XML_HEX:
michael@0: 
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
michael@0:           if(length==2){
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 0);
michael@0:           }
michael@0:           else{
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 0);
michael@0:           }
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
michael@0:           break;
michael@0: 
michael@0:       case UCNV_PRV_ESCAPE_UNICODE:
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_LEFT_CURLY_CODEPOINT;    /* adding { */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT;    /* adding U */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_PLUS_CODEPOINT; /* adding + */
michael@0:           if (length == 2) {
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 4);
michael@0:           } else {
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[0], 16, 4);
michael@0:           }
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_RIGHT_CURLY_CODEPOINT;    /* adding } */
michael@0:           break;
michael@0: 
michael@0:       case UCNV_PRV_ESCAPE_CSS2:
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
michael@0:           valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, codePoint, 16, 0);
michael@0:           /* Always add space character, becase the next character might be whitespace,
michael@0:              which would erroneously be considered the termination of the escape sequence. */
michael@0:           valueString[valueStringLength++] = (UChar) UNICODE_SPACE_CODEPOINT;
michael@0:           break;
michael@0: 
michael@0:       default:
michael@0:           while (i < length)
michael@0:           {
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT;  /* adding % */
michael@0:               valueString[valueStringLength++] = (UChar) UNICODE_U_CODEPOINT;             /* adding U */
michael@0:               valueStringLength += uprv_itou (valueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint16_t)codeUnits[i++], 16, 4);
michael@0:           }
michael@0:       }
michael@0:   }  
michael@0:   myValueSource = valueString;
michael@0: 
michael@0:   /* reset the error */
michael@0:   *err = U_ZERO_ERROR;
michael@0: 
michael@0:   ucnv_cbFromUWriteUChars(fromArgs, &myValueSource, myValueSource+valueStringLength, 0, err);
michael@0: 
michael@0:   ucnv_setFromUCallBack (fromArgs->converter,
michael@0:                          original,
michael@0:                          originalContext,
michael@0:                          &ignoredCallback,
michael@0:                          &ignoredContext,
michael@0:                          &err2);
michael@0:   if (U_FAILURE (err2))
michael@0:   {
michael@0:       *err = err2;
michael@0:       return;
michael@0:   }
michael@0: 
michael@0:   return;
michael@0: }
michael@0: 
michael@0: 
michael@0: 
michael@0: U_CAPI void  U_EXPORT2
michael@0: UCNV_TO_U_CALLBACK_SKIP (
michael@0:                  const void *context,
michael@0:                  UConverterToUnicodeArgs *toArgs,
michael@0:                  const char* codeUnits,
michael@0:                  int32_t length,
michael@0:                  UConverterCallbackReason reason,
michael@0:                  UErrorCode * err)
michael@0: {
michael@0:     if (reason <= UCNV_IRREGULAR)
michael@0:     {
michael@0:         if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
michael@0:         {
michael@0:             *err = U_ZERO_ERROR;
michael@0:         }
michael@0:         /* else the caller must have set the error code accordingly. */
michael@0:     }
michael@0:     /* else ignore the reset, close and clone calls. */
michael@0: }
michael@0: 
michael@0: U_CAPI void    U_EXPORT2
michael@0: UCNV_TO_U_CALLBACK_SUBSTITUTE (
michael@0:                  const void *context,
michael@0:                  UConverterToUnicodeArgs *toArgs,
michael@0:                  const char* codeUnits,
michael@0:                  int32_t length,
michael@0:                  UConverterCallbackReason reason,
michael@0:                  UErrorCode * err)
michael@0: {
michael@0:     if (reason <= UCNV_IRREGULAR)
michael@0:     {
michael@0:         if (context == NULL || (*((char*)context) == UCNV_PRV_STOP_ON_ILLEGAL && reason == UCNV_UNASSIGNED))
michael@0:         {
michael@0:             *err = U_ZERO_ERROR;
michael@0:             ucnv_cbToUWriteSub(toArgs,0,err);
michael@0:         }
michael@0:         /* else the caller must have set the error code accordingly. */
michael@0:     }
michael@0:     /* else ignore the reset, close and clone calls. */
michael@0: }
michael@0: 
michael@0: /*uses uprv_itou to get a unicode escape sequence of the offensive sequence,
michael@0:  *and uses that as the substitution sequence
michael@0:  */
michael@0: U_CAPI void   U_EXPORT2
michael@0: UCNV_TO_U_CALLBACK_ESCAPE (
michael@0:                  const void *context,
michael@0:                  UConverterToUnicodeArgs *toArgs,
michael@0:                  const char* codeUnits,
michael@0:                  int32_t length,
michael@0:                  UConverterCallbackReason reason,
michael@0:                  UErrorCode * err)
michael@0: {
michael@0:     UChar uniValueString[VALUE_STRING_LENGTH];
michael@0:     int32_t valueStringLength = 0;
michael@0:     int32_t i = 0;
michael@0: 
michael@0:     if (reason > UCNV_IRREGULAR)
michael@0:     {
michael@0:         return;
michael@0:     }
michael@0: 
michael@0:     if(context==NULL)
michael@0:     {    
michael@0:         while (i < length)
michael@0:         {
michael@0:             uniValueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */
michael@0:             uniValueString[valueStringLength++] = (UChar) UNICODE_X_CODEPOINT;    /* adding X */
michael@0:             valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t) codeUnits[i++], 16, 2);
michael@0:         }
michael@0:     }
michael@0:     else
michael@0:     {
michael@0:         switch(*((char*)context))
michael@0:         {
michael@0:         case UCNV_PRV_ESCAPE_XML_DEC:
michael@0:             while (i < length)
michael@0:             {
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
michael@0:                 valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 10, 0);
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
michael@0:             }
michael@0:             break;
michael@0: 
michael@0:         case UCNV_PRV_ESCAPE_XML_HEX:
michael@0:             while (i < length)
michael@0:             {
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_AMP_CODEPOINT;   /* adding & */
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_HASH_CODEPOINT;  /* adding # */
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
michael@0:                 valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 16, 0);
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_SEMICOLON_CODEPOINT; /* adding ; */
michael@0:             }
michael@0:             break;
michael@0:         case UCNV_PRV_ESCAPE_C:
michael@0:             while (i < length)
michael@0:             {
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_RS_CODEPOINT;    /* adding \ */
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_X_LOW_CODEPOINT; /* adding x */
michael@0:                 valueStringLength += uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t)codeUnits[i++], 16, 2);
michael@0:             }
michael@0:             break;
michael@0:         default:
michael@0:             while (i < length)
michael@0:             {
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_PERCENT_SIGN_CODEPOINT; /* adding % */
michael@0:                 uniValueString[valueStringLength++] = (UChar) UNICODE_X_CODEPOINT;    /* adding X */
michael@0:                 uprv_itou (uniValueString + valueStringLength, VALUE_STRING_LENGTH - valueStringLength, (uint8_t) codeUnits[i++], 16, 2);
michael@0:                 valueStringLength += 2;
michael@0:             }
michael@0:         }
michael@0:     }
michael@0:     /* reset the error */
michael@0:     *err = U_ZERO_ERROR;
michael@0: 
michael@0:     ucnv_cbToUWriteUChars(toArgs, uniValueString, valueStringLength, 0, err);
michael@0: }
michael@0: 
michael@0: #endif