1999-08-16 21:50:52 +00:00
|
|
|
/*
|
2001-03-21 20:44:20 +00:00
|
|
|
******************************************************************************
|
2000-01-13 23:54:23 +00:00
|
|
|
*
|
2007-01-24 23:27:45 +00:00
|
|
|
* Copyright (C) 1998-2007, International Business Machines
|
2000-01-13 23:54:23 +00:00
|
|
|
* Corporation and others. All Rights Reserved.
|
|
|
|
*
|
2001-03-21 20:44:20 +00:00
|
|
|
******************************************************************************
|
2000-01-13 23:54:23 +00:00
|
|
|
*
|
|
|
|
* ucnv.c:
|
2001-03-17 22:13:27 +00:00
|
|
|
* Implements APIs for the ICU's codeset conversion library;
|
|
|
|
* mostly calls through internal functions;
|
|
|
|
* created by Bertrand A. Damiba
|
2000-01-13 23:54:23 +00:00
|
|
|
*
|
|
|
|
* Modification History:
|
|
|
|
*
|
|
|
|
* Date Name Description
|
|
|
|
* 04/04/99 helena Fixed internal header inclusion.
|
2000-05-12 19:59:03 +00:00
|
|
|
* 05/09/00 helena Added implementation to handle fallback mappings.
|
2000-06-20 23:42:20 +00:00
|
|
|
* 06/20/2000 helena OS/400 port changes; mostly typecast.
|
2000-01-13 23:54:23 +00:00
|
|
|
*/
|
2001-09-28 22:22:59 +00:00
|
|
|
|
|
|
|
#include "unicode/utypes.h"
|
2004-08-26 22:51:40 +00:00
|
|
|
|
|
|
|
#if !UCONFIG_NO_CONVERSION
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
#include "unicode/ustring.h"
|
1999-12-28 23:39:02 +00:00
|
|
|
#include "unicode/ucnv.h"
|
2001-09-28 22:22:59 +00:00
|
|
|
#include "unicode/ucnv_err.h"
|
2003-04-05 01:33:02 +00:00
|
|
|
#include "unicode/uset.h"
|
2004-10-18 02:43:33 +00:00
|
|
|
#include "putilimp.h"
|
1999-08-16 21:50:52 +00:00
|
|
|
#include "cmemory.h"
|
|
|
|
#include "cstring.h"
|
2003-07-22 04:22:57 +00:00
|
|
|
#include "uassert.h"
|
2003-11-21 23:24:11 +00:00
|
|
|
#include "utracimp.h"
|
2001-09-28 22:22:59 +00:00
|
|
|
#include "ustr_imp.h"
|
|
|
|
#include "ucnv_imp.h"
|
|
|
|
#include "ucnv_cnv.h"
|
2000-06-27 20:47:56 +00:00
|
|
|
#include "ucnv_bld.h"
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-05-19 00:55:13 +00:00
|
|
|
/* size of intermediate and preflighting buffers in ucnv_convert() */
|
2003-04-18 23:11:39 +00:00
|
|
|
#define CHUNK_SIZE 1024
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-03-16 20:49:41 +00:00
|
|
|
typedef struct UAmbiguousConverter {
|
|
|
|
const char *name;
|
2001-08-24 00:22:07 +00:00
|
|
|
const UChar variant5c;
|
2001-03-16 20:49:41 +00:00
|
|
|
} UAmbiguousConverter;
|
|
|
|
|
|
|
|
static const UAmbiguousConverter ambiguousConverters[]={
|
2007-08-24 09:42:49 +00:00
|
|
|
{ "ibm-897_P100-1995", 0xa5 },
|
2007-05-10 06:40:31 +00:00
|
|
|
{ "ibm-942_P120-1999", 0xa5 },
|
|
|
|
{ "ibm-943_P130-1999", 0xa5 },
|
2007-08-24 09:42:49 +00:00
|
|
|
{ "ibm-946_P100-1995", 0xa5 },
|
2007-05-10 06:40:31 +00:00
|
|
|
{ "ibm-33722_P120-1999", 0xa5 },
|
2007-08-24 09:42:49 +00:00
|
|
|
/*{ "ibm-54191_P100-2006", 0xa5 },*/
|
|
|
|
/*{ "ibm-62383_P100-2007", 0xa5 },*/
|
|
|
|
/*{ "ibm-891_P100-1995", 0x20a9 },*/
|
|
|
|
{ "ibm-944_P100-1995", 0x20a9 },
|
2007-05-10 06:40:31 +00:00
|
|
|
{ "ibm-949_P110-1999", 0x20a9 },
|
2003-04-05 09:50:53 +00:00
|
|
|
{ "ibm-1363_P110-1997", 0x20a9 },
|
2001-03-16 20:49:41 +00:00
|
|
|
{ "ISO_2022,locale=ko,version=0", 0x20a9 }
|
2001-01-23 18:10:46 +00:00
|
|
|
};
|
|
|
|
|
1999-08-16 21:50:52 +00:00
|
|
|
/*Calls through createConverter */
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverter* U_EXPORT2
|
|
|
|
ucnv_open (const char *name,
|
2000-08-11 01:27:17 +00:00
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2002-08-05 22:55:17 +00:00
|
|
|
UConverter *r;
|
|
|
|
|
2001-09-05 23:23:11 +00:00
|
|
|
if (err == NULL || U_FAILURE (*err)) {
|
|
|
|
return NULL;
|
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
r = ucnv_createConverter(NULL, name, err);
|
2002-08-05 22:55:17 +00:00
|
|
|
return r;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2002-07-17 02:34:50 +00:00
|
|
|
U_CAPI UConverter* U_EXPORT2
|
|
|
|
ucnv_openPackage (const char *packageName, const char *converterName, UErrorCode * err)
|
|
|
|
{
|
|
|
|
return ucnv_createConverterFromPackage(packageName, converterName, err);
|
|
|
|
}
|
|
|
|
|
1999-08-16 21:50:52 +00:00
|
|
|
/*Extracts the UChar* to a char* and calls through createConverter */
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverter* U_EXPORT2
|
|
|
|
ucnv_openU (const UChar * name,
|
2000-08-11 01:27:17 +00:00
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-09-05 23:23:11 +00:00
|
|
|
char asciiName[UCNV_MAX_CONVERTER_NAME_LENGTH];
|
|
|
|
|
2002-07-03 19:04:43 +00:00
|
|
|
if (err == NULL || U_FAILURE(*err))
|
2001-09-05 23:23:11 +00:00
|
|
|
return NULL;
|
|
|
|
if (name == NULL)
|
|
|
|
return ucnv_open (NULL, err);
|
2002-07-03 19:04:43 +00:00
|
|
|
if (u_strlen(name) >= UCNV_MAX_CONVERTER_NAME_LENGTH)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-09-05 23:23:11 +00:00
|
|
|
*err = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return NULL;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2002-07-03 19:04:43 +00:00
|
|
|
return ucnv_open(u_austrcpy(asciiName, name), err);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2005-05-08 07:56:44 +00:00
|
|
|
/* Copy the string that is represented by the UConverterPlatform enum
|
|
|
|
* @param platformString An output buffer
|
|
|
|
* @param platform An enum representing a platform
|
|
|
|
* @return the length of the copied string.
|
|
|
|
*/
|
|
|
|
static int32_t
|
|
|
|
ucnv_copyPlatformString(char *platformString, UConverterPlatform pltfrm)
|
|
|
|
{
|
|
|
|
switch (pltfrm)
|
|
|
|
{
|
|
|
|
case UCNV_IBM:
|
|
|
|
uprv_strcpy(platformString, "ibm-");
|
|
|
|
return 4;
|
|
|
|
case UCNV_UNKNOWN:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* default to empty string */
|
|
|
|
*platformString = 0;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
1999-08-16 21:50:52 +00:00
|
|
|
/*Assumes a $platform-#codepage.$CONVERTER_FILE_EXTENSION scheme and calls
|
|
|
|
*through createConverter*/
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverter* U_EXPORT2
|
|
|
|
ucnv_openCCSID (int32_t codepage,
|
|
|
|
UConverterPlatform platform,
|
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-09-05 23:23:11 +00:00
|
|
|
char myName[UCNV_MAX_CONVERTER_NAME_LENGTH];
|
|
|
|
int32_t myNameLen;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2002-07-03 19:04:43 +00:00
|
|
|
if (err == NULL || U_FAILURE (*err))
|
2001-09-05 23:23:11 +00:00
|
|
|
return NULL;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2002-07-03 16:54:59 +00:00
|
|
|
/* ucnv_copyPlatformString could return "ibm-" or "cp" */
|
|
|
|
myNameLen = ucnv_copyPlatformString(myName, platform);
|
|
|
|
T_CString_integerToString(myName + myNameLen, codepage, 10);
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
return ucnv_createConverter(NULL, myName, err);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-02-16 22:22:40 +00:00
|
|
|
/* Creating a temporary stack-based object that can be used in one thread,
|
|
|
|
and created from a converter that is shared across threads.
|
|
|
|
*/
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverter* U_EXPORT2
|
|
|
|
ucnv_safeClone(const UConverter* cnv, void *stackBuffer, int32_t *pBufferSize, UErrorCode *status)
|
2001-02-16 22:22:40 +00:00
|
|
|
{
|
2003-05-13 21:05:05 +00:00
|
|
|
UConverter *localConverter, *allocatedConverter;
|
2001-02-16 22:22:40 +00:00
|
|
|
int32_t bufferSizeNeeded;
|
2001-09-26 21:09:18 +00:00
|
|
|
char *stackBufferChars = (char *)stackBuffer;
|
2002-07-31 20:53:03 +00:00
|
|
|
UErrorCode cbErr;
|
|
|
|
UConverterToUnicodeArgs toUArgs = {
|
|
|
|
sizeof(UConverterToUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
UConverterFromUnicodeArgs fromUArgs = {
|
|
|
|
sizeof(UConverterFromUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
2001-02-16 22:22:40 +00:00
|
|
|
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_ENTRY_OC(UTRACE_UCNV_CLONE);
|
|
|
|
|
2001-02-16 22:22:40 +00:00
|
|
|
if (status == NULL || U_FAILURE(*status)){
|
2003-11-26 03:09:51 +00:00
|
|
|
UTRACE_EXIT_STATUS(status? *status: U_ILLEGAL_ARGUMENT_ERROR);
|
2001-02-16 22:22:40 +00:00
|
|
|
return 0;
|
|
|
|
}
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2001-02-16 22:22:40 +00:00
|
|
|
if (!pBufferSize || !cnv){
|
2003-11-21 23:24:11 +00:00
|
|
|
*status = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
UTRACE_EXIT_STATUS(*status);
|
2001-02-16 22:22:40 +00:00
|
|
|
return 0;
|
|
|
|
}
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_DATA3(UTRACE_OPEN_CLOSE, "clone converter %s at %p into stackBuffer %p",
|
|
|
|
ucnv_getName(cnv, status), cnv, stackBuffer);
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2001-03-21 20:44:20 +00:00
|
|
|
if (cnv->sharedData->impl->safeClone != NULL) {
|
|
|
|
/* call the custom safeClone function for sizing */
|
|
|
|
bufferSizeNeeded = 0;
|
2003-12-16 05:16:55 +00:00
|
|
|
cnv->sharedData->impl->safeClone(cnv, NULL, &bufferSizeNeeded, status);
|
2001-03-21 20:44:20 +00:00
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
2002-07-31 20:53:03 +00:00
|
|
|
/* inherent sizing */
|
2001-03-21 20:44:20 +00:00
|
|
|
bufferSizeNeeded = sizeof(UConverter);
|
|
|
|
}
|
|
|
|
|
2001-09-26 21:09:18 +00:00
|
|
|
if (*pBufferSize <= 0){ /* 'preflighting' request - set needed size into *pBufferSize */
|
2001-03-21 20:44:20 +00:00
|
|
|
*pBufferSize = bufferSizeNeeded;
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_EXIT_VALUE(bufferSizeNeeded);
|
2001-03-21 20:44:20 +00:00
|
|
|
return 0;
|
2001-02-16 22:22:40 +00:00
|
|
|
}
|
|
|
|
|
2002-07-31 20:53:03 +00:00
|
|
|
|
2003-12-16 05:16:55 +00:00
|
|
|
/* Pointers on 64-bit platforms need to be aligned
|
|
|
|
* on a 64-bit boundary in memory.
|
|
|
|
*/
|
|
|
|
if (U_ALIGNMENT_OFFSET(stackBuffer) != 0) {
|
|
|
|
int32_t offsetUp = (int32_t)U_ALIGNMENT_OFFSET_UP(stackBufferChars);
|
|
|
|
if(*pBufferSize > offsetUp) {
|
|
|
|
*pBufferSize -= offsetUp;
|
|
|
|
stackBufferChars += offsetUp;
|
|
|
|
} else {
|
|
|
|
/* prevent using the stack buffer but keep the size > 0 so that we do not just preflight */
|
|
|
|
*pBufferSize = 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
stackBuffer = (void *)stackBufferChars;
|
|
|
|
|
2002-07-31 20:53:03 +00:00
|
|
|
/* Now, see if we must allocate any memory */
|
2001-02-16 22:22:40 +00:00
|
|
|
if (*pBufferSize < bufferSizeNeeded || stackBuffer == NULL)
|
|
|
|
{
|
2001-03-21 20:44:20 +00:00
|
|
|
/* allocate one here...*/
|
2003-05-13 21:05:05 +00:00
|
|
|
localConverter = allocatedConverter = (UConverter *) uprv_malloc (bufferSizeNeeded);
|
2002-07-31 20:53:03 +00:00
|
|
|
|
|
|
|
if(localConverter == NULL) {
|
|
|
|
*status = U_MEMORY_ALLOCATION_ERROR;
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_EXIT_STATUS(*status);
|
2002-07-31 20:53:03 +00:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (U_SUCCESS(*status)) {
|
2002-08-21 19:12:24 +00:00
|
|
|
*status = U_SAFECLONE_ALLOCATED_WARNING;
|
2001-03-21 20:44:20 +00:00
|
|
|
}
|
2002-07-31 20:53:03 +00:00
|
|
|
|
|
|
|
/* record the fact that memory was allocated */
|
|
|
|
*pBufferSize = bufferSizeNeeded;
|
2001-02-16 22:22:40 +00:00
|
|
|
} else {
|
2002-07-31 20:53:03 +00:00
|
|
|
/* just use the stack buffer */
|
|
|
|
localConverter = (UConverter*) stackBuffer;
|
2003-05-13 21:05:05 +00:00
|
|
|
allocatedConverter = NULL;
|
2002-07-31 20:53:03 +00:00
|
|
|
}
|
|
|
|
|
2003-12-16 05:16:55 +00:00
|
|
|
uprv_memset(localConverter, 0, bufferSizeNeeded);
|
|
|
|
|
2002-08-05 22:55:17 +00:00
|
|
|
/* Copy initial state */
|
|
|
|
uprv_memcpy(localConverter, cnv, sizeof(UConverter));
|
2003-05-13 21:05:05 +00:00
|
|
|
localConverter->isCopyLocal = localConverter->isExtraLocal = FALSE;
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2006-07-05 23:08:50 +00:00
|
|
|
/* copy the substitution string */
|
|
|
|
if (cnv->subChars == (uint8_t *)cnv->subUChars) {
|
|
|
|
localConverter->subChars = (uint8_t *)localConverter->subUChars;
|
|
|
|
} else {
|
|
|
|
localConverter->subChars = (uint8_t *)uprv_malloc(UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
|
|
|
|
if (localConverter->subChars == NULL) {
|
|
|
|
uprv_free(allocatedConverter);
|
|
|
|
UTRACE_EXIT_STATUS(*status);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
uprv_memcpy(localConverter->subChars, cnv->subChars, UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
|
|
|
|
}
|
|
|
|
|
2002-08-05 22:55:17 +00:00
|
|
|
/* now either call the safeclone fcn or not */
|
|
|
|
if (cnv->sharedData->impl->safeClone != NULL) {
|
|
|
|
/* call the custom safeClone function */
|
|
|
|
localConverter = cnv->sharedData->impl->safeClone(cnv, localConverter, pBufferSize, status);
|
|
|
|
}
|
|
|
|
|
2003-05-13 21:05:05 +00:00
|
|
|
if(localConverter==NULL || U_FAILURE(*status)) {
|
2006-07-05 23:08:50 +00:00
|
|
|
if (allocatedConverter != NULL && allocatedConverter->subChars != (uint8_t *)allocatedConverter->subUChars) {
|
|
|
|
uprv_free(allocatedConverter->subChars);
|
|
|
|
}
|
2003-05-13 21:05:05 +00:00
|
|
|
uprv_free(allocatedConverter);
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_EXIT_STATUS(*status);
|
2003-05-13 21:05:05 +00:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2002-07-31 20:53:03 +00:00
|
|
|
/* increment refcount of shared data if needed */
|
2003-05-09 21:33:17 +00:00
|
|
|
/*
|
|
|
|
Checking whether it's an algorithic converter is okay
|
|
|
|
in multithreaded applications because the value never changes.
|
|
|
|
Don't check referenceCounter for any other value.
|
|
|
|
*/
|
|
|
|
if (cnv->sharedData->referenceCounter != ~0) {
|
|
|
|
ucnv_incrementRefCount(cnv->sharedData);
|
|
|
|
}
|
2002-07-31 20:53:03 +00:00
|
|
|
|
|
|
|
if(localConverter == (UConverter*)stackBuffer) {
|
|
|
|
/* we're using user provided data - set to not destroy */
|
|
|
|
localConverter->isCopyLocal = TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* allow callback functions to handle any memory allocation */
|
|
|
|
toUArgs.converter = fromUArgs.converter = localConverter;
|
|
|
|
cbErr = U_ZERO_ERROR;
|
|
|
|
cnv->fromCharErrorBehaviour(cnv->toUContext, &toUArgs, NULL, 0, UCNV_CLONE, &cbErr);
|
|
|
|
cbErr = U_ZERO_ERROR;
|
|
|
|
cnv->fromUCharErrorBehaviour(cnv->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_CLONE, &cbErr);
|
|
|
|
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_EXIT_PTR_STATUS(localConverter, *status);
|
2001-03-21 20:44:20 +00:00
|
|
|
return localConverter;
|
2001-02-16 22:22:40 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
1999-08-16 21:50:52 +00:00
|
|
|
/*Decreases the reference counter in the shared immutable section of the object
|
|
|
|
*and frees the mutable part*/
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_close (UConverter * converter)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2002-08-05 22:55:17 +00:00
|
|
|
UErrorCode errorCode = U_ZERO_ERROR;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_ENTRY_OC(UTRACE_UCNV_CLOSE);
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
if (converter == NULL)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_EXIT();
|
2001-10-04 01:04:28 +00:00
|
|
|
return;
|
|
|
|
}
|
1999-09-30 23:47:52 +00:00
|
|
|
|
2003-11-21 23:24:11 +00:00
|
|
|
UTRACE_DATA3(UTRACE_OPEN_CLOSE, "close converter %s at %p, isCopyLocal=%b",
|
|
|
|
ucnv_getName(converter, &errorCode), converter, converter->isCopyLocal);
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2006-06-12 06:28:31 +00:00
|
|
|
/* In order to speed up the close, only call the callbacks when they have been changed.
|
|
|
|
This performance check will only work when the callbacks are set within a shared library
|
|
|
|
or from user code that statically links this code. */
|
|
|
|
/* first, notify the callback functions that the converter is closed */
|
|
|
|
if (converter->fromCharErrorBehaviour != UCNV_TO_U_DEFAULT_CALLBACK) {
|
|
|
|
UConverterToUnicodeArgs toUArgs = {
|
|
|
|
sizeof(UConverterToUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
|
|
|
|
toUArgs.converter = converter;
|
|
|
|
errorCode = U_ZERO_ERROR;
|
|
|
|
converter->fromCharErrorBehaviour(converter->toUContext, &toUArgs, NULL, 0, UCNV_CLOSE, &errorCode);
|
|
|
|
}
|
|
|
|
if (converter->fromUCharErrorBehaviour != UCNV_FROM_U_DEFAULT_CALLBACK) {
|
|
|
|
UConverterFromUnicodeArgs fromUArgs = {
|
|
|
|
sizeof(UConverterFromUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
fromUArgs.converter = converter;
|
|
|
|
errorCode = U_ZERO_ERROR;
|
|
|
|
converter->fromUCharErrorBehaviour(converter->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_CLOSE, &errorCode);
|
|
|
|
}
|
2000-08-14 23:11:00 +00:00
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
if (converter->sharedData->impl->close != NULL) {
|
|
|
|
converter->sharedData->impl->close(converter);
|
|
|
|
}
|
2002-07-17 02:34:50 +00:00
|
|
|
|
2006-07-05 23:08:50 +00:00
|
|
|
if (converter->subChars != (uint8_t *)converter->subUChars) {
|
|
|
|
uprv_free(converter->subChars);
|
|
|
|
}
|
|
|
|
|
2003-05-09 21:33:17 +00:00
|
|
|
/*
|
|
|
|
Checking whether it's an algorithic converter is okay
|
|
|
|
in multithreaded applications because the value never changes.
|
|
|
|
Don't check referenceCounter for any other value.
|
|
|
|
*/
|
|
|
|
if (converter->sharedData->referenceCounter != ~0) {
|
|
|
|
ucnv_unloadSharedDataIfReady(converter->sharedData);
|
|
|
|
}
|
2002-08-05 22:55:17 +00:00
|
|
|
|
2002-07-17 02:34:50 +00:00
|
|
|
if(!converter->isCopyLocal){
|
2006-06-12 06:28:31 +00:00
|
|
|
uprv_free(converter);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2003-11-21 23:24:11 +00:00
|
|
|
|
|
|
|
UTRACE_EXIT();
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
1999-11-23 23:02:55 +00:00
|
|
|
/*returns a single Name from the list, will return NULL if out of bounds
|
1999-08-16 21:50:52 +00:00
|
|
|
*/
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI const char* U_EXPORT2
|
|
|
|
ucnv_getAvailableName (int32_t n)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2005-05-08 07:56:44 +00:00
|
|
|
if (0 <= n && n <= 0xffff) {
|
|
|
|
UErrorCode err = U_ZERO_ERROR;
|
|
|
|
const char *name = ucnv_bld_getAvailableConverter((uint16_t)n, &err);
|
|
|
|
if (U_SUCCESS(err)) {
|
|
|
|
return name;
|
|
|
|
}
|
1999-11-23 23:02:55 +00:00
|
|
|
}
|
2005-05-08 07:56:44 +00:00
|
|
|
return NULL;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_countAvailable ()
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
UErrorCode err = U_ZERO_ERROR;
|
2005-05-08 07:56:44 +00:00
|
|
|
return ucnv_bld_countAvailableConverters(&err);
|
2000-11-17 19:28:31 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_getSubstChars (const UConverter * converter,
|
|
|
|
char *mySubChar,
|
|
|
|
int8_t * len,
|
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2006-07-05 23:08:50 +00:00
|
|
|
if (converter->subCharLen <= 0) {
|
|
|
|
/* Unicode string or empty string from ucnv_setSubstString(). */
|
|
|
|
*len = 0;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
if (*len < converter->subCharLen) /*not enough space in subChars */
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
*err = U_INDEX_OUTOFBOUNDS_ERROR;
|
|
|
|
return;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2006-07-05 23:08:50 +00:00
|
|
|
uprv_memcpy (mySubChar, converter->subChars, converter->subCharLen); /*fills in the subchars */
|
2001-10-04 01:04:28 +00:00
|
|
|
*len = converter->subCharLen; /*store # of bytes copied to buffer */
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_setSubstChars (UConverter * converter,
|
|
|
|
const char *mySubChar,
|
|
|
|
int8_t len,
|
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return;
|
|
|
|
|
|
|
|
/*Makes sure that the subChar is within the codepages char length boundaries */
|
|
|
|
if ((len > converter->sharedData->staticData->maxBytesPerChar)
|
|
|
|
|| (len < converter->sharedData->staticData->minBytesPerChar))
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
*err = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2001-10-04 01:04:28 +00:00
|
|
|
|
2006-07-05 23:08:50 +00:00
|
|
|
uprv_memcpy (converter->subChars, mySubChar, len); /*copies the subchars */
|
2001-10-04 01:04:28 +00:00
|
|
|
converter->subCharLen = len; /*sets the new len */
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
/*
|
|
|
|
* There is currently (2001Feb) no separate API to set/get subChar1.
|
|
|
|
* In order to always have subChar written after it is explicitly set,
|
|
|
|
* we set subChar1 to 0.
|
|
|
|
*/
|
|
|
|
converter->subChar1 = 0;
|
|
|
|
|
|
|
|
return;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2007-08-19 08:59:28 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
2006-07-05 23:08:50 +00:00
|
|
|
ucnv_setSubstString(UConverter *cnv,
|
|
|
|
const UChar *s,
|
|
|
|
int32_t length,
|
|
|
|
UErrorCode *err) {
|
|
|
|
UAlignedMemory cloneBuffer[U_CNV_SAFECLONE_BUFFERSIZE / sizeof(UAlignedMemory) + 1];
|
|
|
|
char chars[UCNV_ERROR_BUFFER_LENGTH];
|
|
|
|
|
|
|
|
UConverter *clone;
|
|
|
|
uint8_t *subChars;
|
|
|
|
int32_t cloneSize, length8;
|
|
|
|
|
|
|
|
/* Let the following functions check all arguments. */
|
|
|
|
cloneSize = sizeof(cloneBuffer);
|
|
|
|
clone = ucnv_safeClone(cnv, cloneBuffer, &cloneSize, err);
|
|
|
|
ucnv_setFromUCallBack(clone, UCNV_FROM_U_CALLBACK_STOP, NULL, NULL, NULL, err);
|
|
|
|
length8 = ucnv_fromUChars(clone, chars, (int32_t)sizeof(chars), s, length, err);
|
|
|
|
ucnv_close(clone);
|
|
|
|
if (U_FAILURE(*err)) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2006-08-02 22:30:16 +00:00
|
|
|
if (cnv->sharedData->impl->writeSub == NULL
|
2006-07-28 22:58:29 +00:00
|
|
|
#if !UCONFIG_NO_LEGACY_CONVERSION
|
2006-08-02 22:30:16 +00:00
|
|
|
|| (cnv->sharedData->staticData->conversionType == UCNV_MBCS &&
|
2006-07-05 23:08:50 +00:00
|
|
|
ucnv_MBCSGetType(cnv) != UCNV_EBCDIC_STATEFUL)
|
2006-07-28 22:58:29 +00:00
|
|
|
#endif
|
2006-07-05 23:08:50 +00:00
|
|
|
) {
|
|
|
|
/* The converter is not stateful. Store the charset bytes as a fixed string. */
|
|
|
|
subChars = (uint8_t *)chars;
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* The converter has a non-default writeSub() function, indicating
|
|
|
|
* that it is stateful.
|
|
|
|
* Store the Unicode string for on-the-fly conversion for correct
|
|
|
|
* state handling.
|
|
|
|
*/
|
|
|
|
if (length > UCNV_ERROR_BUFFER_LENGTH) {
|
|
|
|
/*
|
|
|
|
* Should not occur. The converter should output at least one byte
|
|
|
|
* per UChar, which means that ucnv_fromUChars() should catch all
|
|
|
|
* overflows.
|
|
|
|
*/
|
|
|
|
*err = U_BUFFER_OVERFLOW_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
subChars = (uint8_t *)s;
|
|
|
|
if (length < 0) {
|
|
|
|
length = u_strlen(s);
|
|
|
|
}
|
|
|
|
length8 = length * U_SIZEOF_UCHAR;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* For storing the substitution string, select either the small buffer inside
|
|
|
|
* UConverter or allocate a subChars buffer.
|
|
|
|
*/
|
|
|
|
if (length8 > UCNV_MAX_SUBCHAR_LEN) {
|
|
|
|
/* Use a separate buffer for the string. Outside UConverter to not make it too large. */
|
|
|
|
if (cnv->subChars == (uint8_t *)cnv->subUChars) {
|
|
|
|
/* Allocate a new buffer for the string. */
|
|
|
|
cnv->subChars = (uint8_t *)uprv_malloc(UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
|
|
|
|
if (cnv->subChars == NULL) {
|
|
|
|
cnv->subChars = (uint8_t *)cnv->subUChars;
|
|
|
|
*err = U_MEMORY_ALLOCATION_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
uprv_memset(cnv->subChars, 0, UCNV_ERROR_BUFFER_LENGTH * U_SIZEOF_UCHAR);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Copy the substitution string into the UConverter or its subChars buffer. */
|
|
|
|
if (length8 == 0) {
|
|
|
|
cnv->subCharLen = 0;
|
|
|
|
} else {
|
|
|
|
uprv_memcpy(cnv->subChars, subChars, length8);
|
|
|
|
if (subChars == (uint8_t *)chars) {
|
|
|
|
cnv->subCharLen = (int8_t)length8;
|
|
|
|
} else /* subChars == s */ {
|
|
|
|
cnv->subCharLen = (int8_t)-length;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* See comment in ucnv_setSubstChars(). */
|
|
|
|
cnv->subChar1 = 0;
|
|
|
|
}
|
|
|
|
|
1999-08-16 21:50:52 +00:00
|
|
|
/*resets the internal states of a converter
|
|
|
|
*goal : have the same behaviour than a freshly created converter
|
|
|
|
*/
|
2003-07-22 04:22:57 +00:00
|
|
|
static void _reset(UConverter *converter, UConverterResetChoice choice,
|
|
|
|
UBool callCallback) {
|
2001-10-04 01:04:28 +00:00
|
|
|
if(converter == NULL) {
|
|
|
|
return;
|
|
|
|
}
|
2000-07-15 20:47:31 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
if(callCallback) {
|
|
|
|
/* first, notify the callback functions that the converter is reset */
|
|
|
|
UConverterToUnicodeArgs toUArgs = {
|
|
|
|
sizeof(UConverterToUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
UConverterFromUnicodeArgs fromUArgs = {
|
|
|
|
sizeof(UConverterFromUnicodeArgs),
|
|
|
|
TRUE,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL,
|
|
|
|
NULL
|
|
|
|
};
|
|
|
|
UErrorCode errorCode;
|
|
|
|
|
|
|
|
toUArgs.converter = fromUArgs.converter = converter;
|
|
|
|
if(choice<=UCNV_RESET_TO_UNICODE) {
|
|
|
|
errorCode = U_ZERO_ERROR;
|
|
|
|
converter->fromCharErrorBehaviour(converter->toUContext, &toUArgs, NULL, 0, UCNV_RESET, &errorCode);
|
|
|
|
}
|
|
|
|
if(choice!=UCNV_RESET_TO_UNICODE) {
|
|
|
|
errorCode = U_ZERO_ERROR;
|
|
|
|
converter->fromUCharErrorBehaviour(converter->fromUContext, &fromUArgs, NULL, 0, 0, UCNV_RESET, &errorCode);
|
|
|
|
}
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2000-12-20 21:18:46 +00:00
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
/* now reset the converter itself */
|
|
|
|
if(choice<=UCNV_RESET_TO_UNICODE) {
|
|
|
|
converter->toUnicodeStatus = converter->sharedData->toUnicodeStatus;
|
2003-07-22 04:22:57 +00:00
|
|
|
converter->mode = 0;
|
2002-05-25 00:32:25 +00:00
|
|
|
converter->toULength = 0;
|
|
|
|
converter->invalidCharLength = converter->UCharErrorBufferLength = 0;
|
2003-10-25 00:29:13 +00:00
|
|
|
converter->preToULength = 0;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
|
|
|
if(choice!=UCNV_RESET_TO_UNICODE) {
|
|
|
|
converter->fromUnicodeStatus = 0;
|
2003-07-24 00:28:47 +00:00
|
|
|
converter->fromUChar32 = 0;
|
2002-05-25 00:32:25 +00:00
|
|
|
converter->invalidUCharLength = converter->charErrorBufferLength = 0;
|
2003-10-25 00:29:13 +00:00
|
|
|
converter->preFromUFirstCP = U_SENTINEL;
|
|
|
|
converter->preFromULength = 0;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (converter->sharedData->impl->reset != NULL) {
|
|
|
|
/* call the custom reset function */
|
|
|
|
converter->sharedData->impl->reset(converter, choice);
|
|
|
|
}
|
2000-11-17 02:40:07 +00:00
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_reset(UConverter *converter)
|
2001-10-04 01:04:28 +00:00
|
|
|
{
|
2003-07-22 04:22:57 +00:00
|
|
|
_reset(converter, UCNV_RESET_BOTH, TRUE);
|
2000-11-17 02:40:07 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_resetToUnicode(UConverter *converter)
|
2001-10-04 01:04:28 +00:00
|
|
|
{
|
2003-07-22 04:22:57 +00:00
|
|
|
_reset(converter, UCNV_RESET_TO_UNICODE, TRUE);
|
2000-12-20 21:18:46 +00:00
|
|
|
}
|
2000-11-17 02:40:07 +00:00
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_resetFromUnicode(UConverter *converter)
|
2001-10-04 01:04:28 +00:00
|
|
|
{
|
2003-07-22 04:22:57 +00:00
|
|
|
_reset(converter, UCNV_RESET_FROM_UNICODE, TRUE);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI int8_t U_EXPORT2
|
|
|
|
ucnv_getMaxCharSize (const UConverter * converter)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2003-10-30 22:54:18 +00:00
|
|
|
return converter->maxBytesPerUChar;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI int8_t U_EXPORT2
|
|
|
|
ucnv_getMinCharSize (const UConverter * converter)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
return converter->sharedData->staticData->minBytesPerChar;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI const char* U_EXPORT2
|
|
|
|
ucnv_getName (const UConverter * converter, UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
|
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return NULL;
|
|
|
|
if(converter->sharedData->impl->getName){
|
|
|
|
const char* temp= converter->sharedData->impl->getName(converter);
|
|
|
|
if(temp)
|
|
|
|
return temp;
|
|
|
|
}
|
|
|
|
return converter->sharedData->staticData->name;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2004-07-18 09:57:18 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_getCCSID(const UConverter * converter,
|
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2004-07-18 09:57:18 +00:00
|
|
|
int32_t ccsid;
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return -1;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2004-07-18 09:57:18 +00:00
|
|
|
ccsid = converter->sharedData->staticData->codepage;
|
|
|
|
if (ccsid == 0) {
|
2004-11-01 17:41:49 +00:00
|
|
|
/* Rare case. This is for cases like gb18030,
|
|
|
|
which doesn't have an IBM cannonical name, but does have an IBM alias. */
|
2004-07-18 09:57:18 +00:00
|
|
|
const char *standardName = ucnv_getStandardName(ucnv_getName(converter, err), "IBM", err);
|
|
|
|
if (U_SUCCESS(*err) && standardName) {
|
2004-11-01 19:13:35 +00:00
|
|
|
const char *ccsidStr = uprv_strchr(standardName, '-');
|
|
|
|
if (ccsidStr) {
|
|
|
|
ccsid = (int32_t)atol(ccsidStr+1); /* +1 to skip '-' */
|
|
|
|
}
|
2004-07-18 09:57:18 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return ccsid;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverterPlatform U_EXPORT2
|
|
|
|
ucnv_getPlatform (const UConverter * converter,
|
2000-08-11 01:27:17 +00:00
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return UCNV_UNKNOWN;
|
|
|
|
|
|
|
|
return (UConverterPlatform)converter->sharedData->staticData->platform;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2000-07-14 01:01:14 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_getToUCallBack (const UConverter * converter,
|
|
|
|
UConverterToUCallback *action,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void **context)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
*action = converter->fromCharErrorBehaviour;
|
|
|
|
*context = converter->toUContext;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2000-07-14 01:01:14 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_getFromUCallBack (const UConverter * converter,
|
|
|
|
UConverterFromUCallback *action,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void **context)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
*action = converter->fromUCharErrorBehaviour;
|
|
|
|
*context = converter->fromUContext;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_setToUCallBack (UConverter * converter,
|
2000-08-11 01:27:17 +00:00
|
|
|
UConverterToUCallback newAction,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void* newContext,
|
2000-08-11 01:27:17 +00:00
|
|
|
UConverterToUCallback *oldAction,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void** oldContext,
|
2000-08-11 01:27:17 +00:00
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return;
|
2002-01-09 20:39:35 +00:00
|
|
|
if (oldAction) *oldAction = converter->fromCharErrorBehaviour;
|
2001-10-04 01:04:28 +00:00
|
|
|
converter->fromCharErrorBehaviour = newAction;
|
2002-01-09 20:39:35 +00:00
|
|
|
if (oldContext) *oldContext = converter->toUContext;
|
2001-10-04 01:04:28 +00:00
|
|
|
converter->toUContext = newContext;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_setFromUCallBack (UConverter * converter,
|
2000-08-11 01:27:17 +00:00
|
|
|
UConverterFromUCallback newAction,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void* newContext,
|
2000-08-11 01:27:17 +00:00
|
|
|
UConverterFromUCallback *oldAction,
|
2001-06-26 22:51:14 +00:00
|
|
|
const void** oldContext,
|
2000-08-11 01:27:17 +00:00
|
|
|
UErrorCode * err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2001-10-04 01:04:28 +00:00
|
|
|
if (U_FAILURE (*err))
|
|
|
|
return;
|
2002-01-09 20:39:35 +00:00
|
|
|
if (oldAction) *oldAction = converter->fromUCharErrorBehaviour;
|
2001-10-04 01:04:28 +00:00
|
|
|
converter->fromUCharErrorBehaviour = newAction;
|
2002-01-09 20:39:35 +00:00
|
|
|
if (oldContext) *oldContext = converter->fromUContext;
|
2001-10-04 01:04:28 +00:00
|
|
|
converter->fromUContext = newContext;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2000-08-11 01:27:17 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
static void
|
|
|
|
_updateOffsets(int32_t *offsets, int32_t length,
|
|
|
|
int32_t sourceIndex, int32_t errorInputLength) {
|
|
|
|
int32_t *limit;
|
|
|
|
int32_t delta, offset;
|
|
|
|
|
|
|
|
if(sourceIndex>=0) {
|
|
|
|
/*
|
|
|
|
* adjust each offset by adding the previous sourceIndex
|
|
|
|
* minus the length of the input sequence that caused an
|
|
|
|
* error, if any
|
|
|
|
*/
|
|
|
|
delta=sourceIndex-errorInputLength;
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* set each offset to -1 because this conversion function
|
|
|
|
* does not handle offsets
|
|
|
|
*/
|
|
|
|
delta=-1;
|
|
|
|
}
|
|
|
|
|
|
|
|
limit=offsets+length;
|
|
|
|
if(delta==0) {
|
|
|
|
/* most common case, nothing to do */
|
|
|
|
} else if(delta>0) {
|
|
|
|
/* add the delta to each offset (but not if the offset is <0) */
|
|
|
|
while(offsets<limit) {
|
|
|
|
offset=*offsets;
|
|
|
|
if(offset>=0) {
|
|
|
|
*offsets=offset+delta;
|
|
|
|
}
|
|
|
|
++offsets;
|
|
|
|
}
|
|
|
|
} else /* delta<0 */ {
|
|
|
|
/*
|
|
|
|
* set each offset to -1 because this conversion function
|
|
|
|
* does not handle offsets
|
|
|
|
* or the error input sequence started in a previous buffer
|
|
|
|
*/
|
|
|
|
while(offsets<limit) {
|
|
|
|
*offsets++=-1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ucnv_fromUnicode --------------------------------------------------------- */
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
/*
|
|
|
|
* Implementation note for m:n conversions
|
|
|
|
*
|
|
|
|
* While collecting source units to find the longest match for m:n conversion,
|
|
|
|
* some source units may need to be stored for a partial match.
|
|
|
|
* When a second buffer does not yield a match on all of the previously stored
|
|
|
|
* source units, then they must be "replayed", i.e., fed back into the converter.
|
|
|
|
*
|
|
|
|
* The code relies on the fact that replaying will not nest -
|
|
|
|
* converting a replay buffer will not result in a replay.
|
|
|
|
* This is because a replay is necessary only after the _continuation_ of a
|
|
|
|
* partial match failed, but a replay buffer is converted as a whole.
|
|
|
|
* It may result in some of its units being stored again for a partial match,
|
|
|
|
* but there will not be a continuation _during_ the replay which could fail.
|
|
|
|
*
|
|
|
|
* It is conceivable that a callback function could call the converter
|
|
|
|
* recursively in a way that causes another replay to be stored, but that
|
|
|
|
* would be an error in the callback function.
|
|
|
|
* Such violations will cause assertion failures in a debug build,
|
|
|
|
* and wrong output, but they will not cause a crash.
|
|
|
|
*/
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
static void
|
|
|
|
_fromUnicodeWithCallback(UConverterFromUnicodeArgs *pArgs, UErrorCode *err) {
|
|
|
|
UConverterFromUnicode fromUnicode;
|
|
|
|
UConverter *cnv;
|
|
|
|
const UChar *s;
|
|
|
|
char *t;
|
|
|
|
int32_t *offsets;
|
|
|
|
int32_t sourceIndex;
|
|
|
|
int32_t errorInputLength;
|
|
|
|
UBool converterSawEndOfInput, calledCallback;
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
/* variables for m:n conversion */
|
|
|
|
UChar replay[UCNV_EXT_MAX_UCHARS];
|
|
|
|
const UChar *realSource, *realSourceLimit;
|
|
|
|
int32_t realSourceIndex;
|
|
|
|
UBool realFlush;
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
cnv=pArgs->converter;
|
|
|
|
s=pArgs->source;
|
|
|
|
t=pArgs->target;
|
|
|
|
offsets=pArgs->offsets;
|
|
|
|
|
|
|
|
/* get the converter implementation function */
|
|
|
|
sourceIndex=0;
|
|
|
|
if(offsets==NULL) {
|
|
|
|
fromUnicode=cnv->sharedData->impl->fromUnicode;
|
|
|
|
} else {
|
|
|
|
fromUnicode=cnv->sharedData->impl->fromUnicodeWithOffsets;
|
|
|
|
if(fromUnicode==NULL) {
|
|
|
|
/* there is no WithOffsets implementation */
|
|
|
|
fromUnicode=cnv->sharedData->impl->fromUnicode;
|
|
|
|
/* we will write -1 for each offset */
|
|
|
|
sourceIndex=-1;
|
|
|
|
}
|
|
|
|
}
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(cnv->preFromULength>=0) {
|
|
|
|
/* normal mode */
|
|
|
|
realSource=NULL;
|
2003-12-04 00:34:51 +00:00
|
|
|
|
|
|
|
/* avoid compiler warnings - not otherwise necessary, and the values do not matter */
|
|
|
|
realSourceLimit=NULL;
|
|
|
|
realFlush=FALSE;
|
|
|
|
realSourceIndex=0;
|
2003-10-25 00:29:13 +00:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Previous m:n conversion stored source units from a partial match
|
|
|
|
* and failed to consume all of them.
|
|
|
|
* We need to "replay" them from a temporary buffer and convert them first.
|
|
|
|
*/
|
|
|
|
realSource=pArgs->source;
|
|
|
|
realSourceLimit=pArgs->sourceLimit;
|
|
|
|
realFlush=pArgs->flush;
|
|
|
|
realSourceIndex=sourceIndex;
|
|
|
|
|
|
|
|
uprv_memcpy(replay, cnv->preFromU, -cnv->preFromULength*U_SIZEOF_UCHAR);
|
|
|
|
pArgs->source=replay;
|
|
|
|
pArgs->sourceLimit=replay-cnv->preFromULength;
|
|
|
|
pArgs->flush=FALSE;
|
|
|
|
sourceIndex=-1;
|
|
|
|
|
|
|
|
cnv->preFromULength=0;
|
|
|
|
}
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
/*
|
2003-07-22 04:22:57 +00:00
|
|
|
* loop for conversion and error handling
|
|
|
|
*
|
|
|
|
* loop {
|
|
|
|
* convert
|
|
|
|
* loop {
|
|
|
|
* update offsets
|
|
|
|
* handle end of input
|
|
|
|
* handle errors/call callback
|
|
|
|
* }
|
|
|
|
* }
|
|
|
|
*/
|
|
|
|
for(;;) {
|
2007-02-06 05:24:14 +00:00
|
|
|
if(U_SUCCESS(*err)) {
|
|
|
|
/* convert */
|
|
|
|
fromUnicode(pArgs, err);
|
2003-07-22 04:22:57 +00:00
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* set a flag for whether the converter
|
|
|
|
* successfully processed the end of the input
|
|
|
|
*
|
|
|
|
* need not check cnv->preFromULength==0 because a replay (<0) will cause
|
|
|
|
* s<sourceLimit before converterSawEndOfInput is checked
|
|
|
|
*/
|
|
|
|
converterSawEndOfInput=
|
|
|
|
(UBool)(U_SUCCESS(*err) &&
|
|
|
|
pArgs->flush && pArgs->source==pArgs->sourceLimit &&
|
|
|
|
cnv->fromUChar32==0);
|
|
|
|
} else {
|
|
|
|
/* handle error from ucnv_convertEx() */
|
|
|
|
converterSawEndOfInput=FALSE;
|
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* no callback called yet for this iteration */
|
|
|
|
calledCallback=FALSE;
|
|
|
|
|
|
|
|
/* no sourceIndex adjustment for conversion, only for callback output */
|
|
|
|
errorInputLength=0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* loop for offsets and error handling
|
|
|
|
*
|
|
|
|
* iterates at most 3 times:
|
|
|
|
* 1. to clean up after the conversion function
|
|
|
|
* 2. after the callback
|
|
|
|
* 3. after the callback again if there was truncated input
|
|
|
|
*/
|
|
|
|
for(;;) {
|
|
|
|
/* update offsets if we write any */
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
int32_t length=(int32_t)(pArgs->target-t);
|
|
|
|
if(length>0) {
|
|
|
|
_updateOffsets(offsets, length, sourceIndex, errorInputLength);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* if a converter handles offsets and updates the offsets
|
|
|
|
* pointer at the end, then pArgs->offset should not change
|
|
|
|
* here;
|
|
|
|
* however, some converters do not handle offsets at all
|
|
|
|
* (sourceIndex<0) or may not update the offsets pointer
|
|
|
|
*/
|
|
|
|
pArgs->offsets=offsets+=length;
|
|
|
|
}
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(sourceIndex>=0) {
|
|
|
|
sourceIndex+=(int32_t)(pArgs->source-s);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if(cnv->preFromULength<0) {
|
|
|
|
/*
|
|
|
|
* switch the source to new replay units (cannot occur while replaying)
|
|
|
|
* after offset handling and before end-of-input and callback handling
|
|
|
|
*/
|
|
|
|
if(realSource==NULL) {
|
|
|
|
realSource=pArgs->source;
|
|
|
|
realSourceLimit=pArgs->sourceLimit;
|
|
|
|
realFlush=pArgs->flush;
|
|
|
|
realSourceIndex=sourceIndex;
|
|
|
|
|
|
|
|
uprv_memcpy(replay, cnv->preFromU, -cnv->preFromULength*U_SIZEOF_UCHAR);
|
|
|
|
pArgs->source=replay;
|
|
|
|
pArgs->sourceLimit=replay-cnv->preFromULength;
|
|
|
|
pArgs->flush=FALSE;
|
|
|
|
if((sourceIndex+=cnv->preFromULength)<0) {
|
|
|
|
sourceIndex=-1;
|
|
|
|
}
|
|
|
|
|
|
|
|
cnv->preFromULength=0;
|
|
|
|
} else {
|
|
|
|
/* see implementation note before _fromUnicodeWithCallback() */
|
|
|
|
U_ASSERT(realSource==NULL);
|
|
|
|
*err=U_INTERNAL_PROGRAM_ERROR;
|
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* update pointers */
|
|
|
|
s=pArgs->source;
|
|
|
|
t=pArgs->target;
|
|
|
|
|
|
|
|
if(U_SUCCESS(*err)) {
|
|
|
|
if(s<pArgs->sourceLimit) {
|
|
|
|
/*
|
|
|
|
* continue with the conversion loop while there is still input left
|
|
|
|
* (continue converting by breaking out of only the inner loop)
|
|
|
|
*/
|
|
|
|
break;
|
2003-10-25 00:29:13 +00:00
|
|
|
} else if(realSource!=NULL) {
|
|
|
|
/* switch back from replaying to the real source and continue */
|
|
|
|
pArgs->source=realSource;
|
|
|
|
pArgs->sourceLimit=realSourceLimit;
|
|
|
|
pArgs->flush=realFlush;
|
|
|
|
sourceIndex=realSourceIndex;
|
|
|
|
|
|
|
|
realSource=NULL;
|
|
|
|
break;
|
2003-07-24 00:28:47 +00:00
|
|
|
} else if(pArgs->flush && cnv->fromUChar32!=0) {
|
2003-07-22 04:22:57 +00:00
|
|
|
/*
|
|
|
|
* the entire input stream is consumed
|
|
|
|
* and there is a partial, truncated input sequence left
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* inject an error and continue with callback handling */
|
|
|
|
*err=U_TRUNCATED_CHAR_FOUND;
|
|
|
|
calledCallback=FALSE; /* new error condition */
|
|
|
|
} else {
|
|
|
|
/* input consumed */
|
|
|
|
if(pArgs->flush) {
|
|
|
|
/*
|
|
|
|
* return to the conversion loop once more if the flush
|
|
|
|
* flag is set and the conversion function has not
|
|
|
|
* successfully processed the end of the input yet
|
|
|
|
*
|
|
|
|
* (continue converting by breaking out of only the inner loop)
|
|
|
|
*/
|
|
|
|
if(!converterSawEndOfInput) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* reset the converter without calling the callback function */
|
|
|
|
_reset(cnv, UCNV_RESET_FROM_UNICODE, FALSE);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* done successfully */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* U_FAILURE(*err) */
|
|
|
|
{
|
|
|
|
UErrorCode e;
|
|
|
|
|
|
|
|
if( calledCallback ||
|
|
|
|
(e=*err)==U_BUFFER_OVERFLOW_ERROR ||
|
|
|
|
(e!=U_INVALID_CHAR_FOUND &&
|
|
|
|
e!=U_ILLEGAL_CHAR_FOUND &&
|
|
|
|
e!=U_TRUNCATED_CHAR_FOUND)
|
|
|
|
) {
|
|
|
|
/*
|
|
|
|
* the callback did not or cannot resolve the error:
|
|
|
|
* set output pointers and return
|
|
|
|
*
|
|
|
|
* the check for buffer overflow is redundant but it is
|
|
|
|
* a high-runner case and hopefully documents the intent
|
|
|
|
* well
|
2003-10-25 00:29:13 +00:00
|
|
|
*
|
|
|
|
* if we were replaying, then the replay buffer must be
|
|
|
|
* copied back into the UConverter
|
|
|
|
* and the real arguments must be restored
|
2003-07-22 04:22:57 +00:00
|
|
|
*/
|
2003-10-25 00:29:13 +00:00
|
|
|
if(realSource!=NULL) {
|
|
|
|
int32_t length;
|
|
|
|
|
|
|
|
U_ASSERT(cnv->preFromULength==0);
|
|
|
|
|
|
|
|
length=(int32_t)(pArgs->sourceLimit-pArgs->source);
|
|
|
|
if(length>0) {
|
|
|
|
uprv_memcpy(cnv->preFromU, pArgs->source, length*U_SIZEOF_UCHAR);
|
|
|
|
cnv->preFromULength=(int8_t)-length;
|
|
|
|
}
|
|
|
|
|
|
|
|
pArgs->source=realSource;
|
|
|
|
pArgs->sourceLimit=realSourceLimit;
|
|
|
|
pArgs->flush=realFlush;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* callback handling */
|
|
|
|
{
|
|
|
|
UChar32 codePoint;
|
|
|
|
|
2003-07-24 00:28:47 +00:00
|
|
|
/* get and write the code point */
|
|
|
|
codePoint=cnv->fromUChar32;
|
|
|
|
errorInputLength=0;
|
|
|
|
U16_APPEND_UNSAFE(cnv->invalidUCharBuffer, errorInputLength, codePoint);
|
|
|
|
cnv->invalidUCharLength=(int8_t)errorInputLength;
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* set the converter state to deal with the next character */
|
2003-07-24 00:28:47 +00:00
|
|
|
cnv->fromUChar32=0;
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* call the callback function */
|
|
|
|
cnv->fromUCharErrorBehaviour(cnv->fromUContext, pArgs,
|
|
|
|
cnv->invalidUCharBuffer, errorInputLength, codePoint,
|
|
|
|
*err==U_INVALID_CHAR_FOUND ? UCNV_UNASSIGNED : UCNV_ILLEGAL,
|
|
|
|
err);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* loop back to the offset handling
|
|
|
|
*
|
|
|
|
* this flag will indicate after offset handling
|
|
|
|
* that a callback was called;
|
|
|
|
* if the callback did not resolve the error, then we return
|
|
|
|
*/
|
|
|
|
calledCallback=TRUE;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* Output the fromUnicode overflow buffer.
|
|
|
|
* Call this function if(cnv->charErrorBufferLength>0).
|
|
|
|
* @return TRUE if overflow
|
|
|
|
*/
|
|
|
|
static UBool
|
|
|
|
ucnv_outputOverflowFromUnicode(UConverter *cnv,
|
|
|
|
char **target, const char *targetLimit,
|
|
|
|
int32_t **pOffsets,
|
|
|
|
UErrorCode *err) {
|
|
|
|
int32_t *offsets;
|
|
|
|
char *overflow, *t;
|
|
|
|
int32_t i, length;
|
|
|
|
|
|
|
|
t=*target;
|
|
|
|
if(pOffsets!=NULL) {
|
|
|
|
offsets=*pOffsets;
|
|
|
|
} else {
|
|
|
|
offsets=NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
overflow=(char *)cnv->charErrorBuffer;
|
|
|
|
length=cnv->charErrorBufferLength;
|
|
|
|
i=0;
|
|
|
|
while(i<length) {
|
|
|
|
if(t==targetLimit) {
|
|
|
|
/* the overflow buffer contains too much, keep the rest */
|
|
|
|
int32_t j=0;
|
|
|
|
|
|
|
|
do {
|
|
|
|
overflow[j++]=overflow[i++];
|
|
|
|
} while(i<length);
|
|
|
|
|
|
|
|
cnv->charErrorBufferLength=(int8_t)j;
|
|
|
|
*target=t;
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*pOffsets=offsets;
|
|
|
|
}
|
|
|
|
*err=U_BUFFER_OVERFLOW_ERROR;
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* copy the overflow contents to the target */
|
|
|
|
*t++=overflow[i++];
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*offsets++=-1; /* no source index available for old output */
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* the overflow buffer is completely copied to the target */
|
|
|
|
cnv->charErrorBufferLength=0;
|
|
|
|
*target=t;
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*pOffsets=offsets;
|
|
|
|
}
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_fromUnicode(UConverter *cnv,
|
|
|
|
char **target, const char *targetLimit,
|
|
|
|
const UChar **source, const UChar *sourceLimit,
|
|
|
|
int32_t *offsets,
|
|
|
|
UBool flush,
|
|
|
|
UErrorCode *err) {
|
|
|
|
UConverterFromUnicodeArgs args;
|
|
|
|
const UChar *s;
|
|
|
|
char *t;
|
|
|
|
|
|
|
|
/* check parameters */
|
|
|
|
if(err==NULL || U_FAILURE(*err)) {
|
2001-10-04 01:04:28 +00:00
|
|
|
return;
|
|
|
|
}
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
if(cnv==NULL || target==NULL || source==NULL) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
2001-10-04 01:04:28 +00:00
|
|
|
return;
|
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
s=*source;
|
|
|
|
t=*target;
|
2000-02-08 23:41:16 +00:00
|
|
|
|
2007-02-27 08:48:19 +00:00
|
|
|
if ((const void *)U_MAX_PTR(sourceLimit) == (const void *)sourceLimit) {
|
|
|
|
/*
|
|
|
|
Prevent code from going into an infinite loop in case we do hit this
|
|
|
|
limit. The limit pointer is expected to be on a UChar * boundary.
|
|
|
|
This also prevents the next argument check from failing.
|
|
|
|
*/
|
|
|
|
sourceLimit = (const UChar *)(((const char *)sourceLimit) - 1);
|
|
|
|
}
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
/*
|
2006-11-14 22:59:46 +00:00
|
|
|
* All these conditions should never happen.
|
|
|
|
*
|
|
|
|
* 1) Make sure that the limits are >= to the address source or target
|
|
|
|
*
|
|
|
|
* 2) Make sure that the buffer sizes do not exceed the number range for
|
2003-07-22 04:22:57 +00:00
|
|
|
* int32_t because some functions use the size (in units or bytes)
|
|
|
|
* rather than comparing pointers, and because offsets are int32_t values.
|
|
|
|
*
|
|
|
|
* size_t is guaranteed to be unsigned and large enough for the job.
|
|
|
|
*
|
|
|
|
* Return with an error instead of adjusting the limits because we would
|
|
|
|
* not be able to maintain the semantics that either the source must be
|
|
|
|
* consumed or the target filled (unless an error occurs).
|
|
|
|
* An adjustment would be targetLimit=t+0x7fffffff; for example.
|
2006-11-14 22:59:46 +00:00
|
|
|
*
|
|
|
|
* 3) Make sure that the user didn't incorrectly cast a UChar * pointer
|
|
|
|
* to a char * pointer and provide an incomplete UChar code unit.
|
2003-07-22 04:22:57 +00:00
|
|
|
*/
|
2006-11-14 22:59:46 +00:00
|
|
|
if (sourceLimit<s || targetLimit<t ||
|
2003-07-22 04:22:57 +00:00
|
|
|
((size_t)(sourceLimit-s)>(size_t)0x3fffffff && sourceLimit>s) ||
|
2006-11-14 22:59:46 +00:00
|
|
|
((size_t)(targetLimit-t)>(size_t)0x7fffffff && targetLimit>t) ||
|
|
|
|
(((const char *)sourceLimit-(const char *)s) & 1) != 0)
|
|
|
|
{
|
2003-07-22 04:22:57 +00:00
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/* output the target overflow buffer */
|
|
|
|
if( cnv->charErrorBufferLength>0 &&
|
|
|
|
ucnv_outputOverflowFromUnicode(cnv, target, targetLimit, &offsets, err)
|
|
|
|
) {
|
|
|
|
/* U_BUFFER_OVERFLOW_ERROR */
|
|
|
|
return;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2007-02-06 05:24:14 +00:00
|
|
|
/* *target may have moved, therefore stop using t */
|
2003-04-18 23:11:39 +00:00
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(!flush && s==sourceLimit && cnv->preFromULength>=0) {
|
2003-04-18 23:11:39 +00:00
|
|
|
/* the overflow buffer is emptied and there is no new input: we are done */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/*
|
|
|
|
* Do not simply return with a buffer overflow error if
|
|
|
|
* !flush && t==targetLimit
|
|
|
|
* because it is possible that the source will not generate any output.
|
|
|
|
* For example, the skip callback may be called;
|
|
|
|
* it does not output anything.
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* prepare the converter arguments */
|
|
|
|
args.converter=cnv;
|
|
|
|
args.flush=flush;
|
|
|
|
args.offsets=offsets;
|
|
|
|
args.source=s;
|
|
|
|
args.sourceLimit=sourceLimit;
|
2007-02-06 05:24:14 +00:00
|
|
|
args.target=*target;
|
2003-07-22 04:22:57 +00:00
|
|
|
args.targetLimit=targetLimit;
|
|
|
|
args.size=sizeof(args);
|
|
|
|
|
|
|
|
_fromUnicodeWithCallback(&args, err);
|
|
|
|
|
|
|
|
*source=args.source;
|
|
|
|
*target=args.target;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* ucnv_toUnicode() --------------------------------------------------------- */
|
|
|
|
|
|
|
|
static void
|
|
|
|
_toUnicodeWithCallback(UConverterToUnicodeArgs *pArgs, UErrorCode *err) {
|
|
|
|
UConverterToUnicode toUnicode;
|
|
|
|
UConverter *cnv;
|
|
|
|
const char *s;
|
|
|
|
UChar *t;
|
|
|
|
int32_t *offsets;
|
|
|
|
int32_t sourceIndex;
|
|
|
|
int32_t errorInputLength;
|
|
|
|
UBool converterSawEndOfInput, calledCallback;
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
/* variables for m:n conversion */
|
|
|
|
char replay[UCNV_EXT_MAX_BYTES];
|
|
|
|
const char *realSource, *realSourceLimit;
|
|
|
|
int32_t realSourceIndex;
|
|
|
|
UBool realFlush;
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
cnv=pArgs->converter;
|
|
|
|
s=pArgs->source;
|
|
|
|
t=pArgs->target;
|
|
|
|
offsets=pArgs->offsets;
|
|
|
|
|
|
|
|
/* get the converter implementation function */
|
|
|
|
sourceIndex=0;
|
|
|
|
if(offsets==NULL) {
|
|
|
|
toUnicode=cnv->sharedData->impl->toUnicode;
|
|
|
|
} else {
|
|
|
|
toUnicode=cnv->sharedData->impl->toUnicodeWithOffsets;
|
|
|
|
if(toUnicode==NULL) {
|
|
|
|
/* there is no WithOffsets implementation */
|
|
|
|
toUnicode=cnv->sharedData->impl->toUnicode;
|
|
|
|
/* we will write -1 for each offset */
|
|
|
|
sourceIndex=-1;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
}
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(cnv->preToULength>=0) {
|
|
|
|
/* normal mode */
|
|
|
|
realSource=NULL;
|
2003-12-04 00:34:51 +00:00
|
|
|
|
|
|
|
/* avoid compiler warnings - not otherwise necessary, and the values do not matter */
|
|
|
|
realSourceLimit=NULL;
|
|
|
|
realFlush=FALSE;
|
|
|
|
realSourceIndex=0;
|
2003-10-25 00:29:13 +00:00
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Previous m:n conversion stored source units from a partial match
|
|
|
|
* and failed to consume all of them.
|
|
|
|
* We need to "replay" them from a temporary buffer and convert them first.
|
|
|
|
*/
|
|
|
|
realSource=pArgs->source;
|
|
|
|
realSourceLimit=pArgs->sourceLimit;
|
|
|
|
realFlush=pArgs->flush;
|
|
|
|
realSourceIndex=sourceIndex;
|
|
|
|
|
|
|
|
uprv_memcpy(replay, cnv->preToU, -cnv->preToULength);
|
|
|
|
pArgs->source=replay;
|
|
|
|
pArgs->sourceLimit=replay-cnv->preToULength;
|
|
|
|
pArgs->flush=FALSE;
|
|
|
|
sourceIndex=-1;
|
|
|
|
|
|
|
|
cnv->preToULength=0;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/*
|
|
|
|
* loop for conversion and error handling
|
|
|
|
*
|
|
|
|
* loop {
|
|
|
|
* convert
|
|
|
|
* loop {
|
|
|
|
* update offsets
|
|
|
|
* handle end of input
|
|
|
|
* handle errors/call callback
|
|
|
|
* }
|
|
|
|
* }
|
|
|
|
*/
|
|
|
|
for(;;) {
|
2003-08-01 14:39:14 +00:00
|
|
|
if(U_SUCCESS(*err)) {
|
|
|
|
/* convert */
|
|
|
|
toUnicode(pArgs, err);
|
2003-07-22 04:22:57 +00:00
|
|
|
|
2003-08-01 14:39:14 +00:00
|
|
|
/*
|
|
|
|
* set a flag for whether the converter
|
|
|
|
* successfully processed the end of the input
|
2003-10-31 04:59:36 +00:00
|
|
|
*
|
2003-10-31 16:36:32 +00:00
|
|
|
* need not check cnv->preToULength==0 because a replay (<0) will cause
|
2003-10-31 04:59:36 +00:00
|
|
|
* s<sourceLimit before converterSawEndOfInput is checked
|
2003-08-01 14:39:14 +00:00
|
|
|
*/
|
|
|
|
converterSawEndOfInput=
|
|
|
|
(UBool)(U_SUCCESS(*err) &&
|
|
|
|
pArgs->flush && pArgs->source==pArgs->sourceLimit &&
|
|
|
|
cnv->toULength==0);
|
|
|
|
} else {
|
2007-02-06 05:24:14 +00:00
|
|
|
/* handle error from getNextUChar() or ucnv_convertEx() */
|
2003-08-01 14:39:14 +00:00
|
|
|
converterSawEndOfInput=FALSE;
|
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* no callback called yet for this iteration */
|
|
|
|
calledCallback=FALSE;
|
|
|
|
|
|
|
|
/* no sourceIndex adjustment for conversion, only for callback output */
|
|
|
|
errorInputLength=0;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* loop for offsets and error handling
|
|
|
|
*
|
|
|
|
* iterates at most 3 times:
|
|
|
|
* 1. to clean up after the conversion function
|
|
|
|
* 2. after the callback
|
|
|
|
* 3. after the callback again if there was truncated input
|
|
|
|
*/
|
|
|
|
for(;;) {
|
|
|
|
/* update offsets if we write any */
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
int32_t length=(int32_t)(pArgs->target-t);
|
|
|
|
if(length>0) {
|
|
|
|
_updateOffsets(offsets, length, sourceIndex, errorInputLength);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* if a converter handles offsets and updates the offsets
|
|
|
|
* pointer at the end, then pArgs->offset should not change
|
|
|
|
* here;
|
|
|
|
* however, some converters do not handle offsets at all
|
|
|
|
* (sourceIndex<0) or may not update the offsets pointer
|
|
|
|
*/
|
|
|
|
pArgs->offsets=offsets+=length;
|
|
|
|
}
|
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(sourceIndex>=0) {
|
|
|
|
sourceIndex+=(int32_t)(pArgs->source-s);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if(cnv->preToULength<0) {
|
|
|
|
/*
|
|
|
|
* switch the source to new replay units (cannot occur while replaying)
|
|
|
|
* after offset handling and before end-of-input and callback handling
|
|
|
|
*/
|
|
|
|
if(realSource==NULL) {
|
|
|
|
realSource=pArgs->source;
|
|
|
|
realSourceLimit=pArgs->sourceLimit;
|
|
|
|
realFlush=pArgs->flush;
|
|
|
|
realSourceIndex=sourceIndex;
|
|
|
|
|
|
|
|
uprv_memcpy(replay, cnv->preToU, -cnv->preToULength);
|
|
|
|
pArgs->source=replay;
|
|
|
|
pArgs->sourceLimit=replay-cnv->preToULength;
|
|
|
|
pArgs->flush=FALSE;
|
|
|
|
if((sourceIndex+=cnv->preToULength)<0) {
|
|
|
|
sourceIndex=-1;
|
|
|
|
}
|
|
|
|
|
|
|
|
cnv->preToULength=0;
|
|
|
|
} else {
|
|
|
|
/* see implementation note before _fromUnicodeWithCallback() */
|
|
|
|
U_ASSERT(realSource==NULL);
|
|
|
|
*err=U_INTERNAL_PROGRAM_ERROR;
|
|
|
|
}
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* update pointers */
|
|
|
|
s=pArgs->source;
|
|
|
|
t=pArgs->target;
|
|
|
|
|
|
|
|
if(U_SUCCESS(*err)) {
|
|
|
|
if(s<pArgs->sourceLimit) {
|
|
|
|
/*
|
|
|
|
* continue with the conversion loop while there is still input left
|
|
|
|
* (continue converting by breaking out of only the inner loop)
|
|
|
|
*/
|
|
|
|
break;
|
2003-10-25 00:29:13 +00:00
|
|
|
} else if(realSource!=NULL) {
|
|
|
|
/* switch back from replaying to the real source and continue */
|
|
|
|
pArgs->source=realSource;
|
|
|
|
pArgs->sourceLimit=realSourceLimit;
|
|
|
|
pArgs->flush=realFlush;
|
|
|
|
sourceIndex=realSourceIndex;
|
|
|
|
|
|
|
|
realSource=NULL;
|
|
|
|
break;
|
2003-07-22 04:22:57 +00:00
|
|
|
} else if(pArgs->flush && cnv->toULength>0) {
|
|
|
|
/*
|
|
|
|
* the entire input stream is consumed
|
|
|
|
* and there is a partial, truncated input sequence left
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* inject an error and continue with callback handling */
|
|
|
|
*err=U_TRUNCATED_CHAR_FOUND;
|
|
|
|
calledCallback=FALSE; /* new error condition */
|
|
|
|
} else {
|
|
|
|
/* input consumed */
|
|
|
|
if(pArgs->flush) {
|
|
|
|
/*
|
|
|
|
* return to the conversion loop once more if the flush
|
|
|
|
* flag is set and the conversion function has not
|
|
|
|
* successfully processed the end of the input yet
|
|
|
|
*
|
|
|
|
* (continue converting by breaking out of only the inner loop)
|
|
|
|
*/
|
|
|
|
if(!converterSawEndOfInput) {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* reset the converter without calling the callback function */
|
2003-08-01 14:39:14 +00:00
|
|
|
_reset(cnv, UCNV_RESET_TO_UNICODE, FALSE);
|
2003-07-22 04:22:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* done successfully */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* U_FAILURE(*err) */
|
|
|
|
{
|
|
|
|
UErrorCode e;
|
|
|
|
|
|
|
|
if( calledCallback ||
|
|
|
|
(e=*err)==U_BUFFER_OVERFLOW_ERROR ||
|
|
|
|
(e!=U_INVALID_CHAR_FOUND &&
|
|
|
|
e!=U_ILLEGAL_CHAR_FOUND &&
|
2003-12-03 17:43:17 +00:00
|
|
|
e!=U_TRUNCATED_CHAR_FOUND &&
|
|
|
|
e!=U_ILLEGAL_ESCAPE_SEQUENCE &&
|
|
|
|
e!=U_UNSUPPORTED_ESCAPE_SEQUENCE)
|
2003-07-22 04:22:57 +00:00
|
|
|
) {
|
|
|
|
/*
|
|
|
|
* the callback did not or cannot resolve the error:
|
|
|
|
* set output pointers and return
|
|
|
|
*
|
|
|
|
* the check for buffer overflow is redundant but it is
|
|
|
|
* a high-runner case and hopefully documents the intent
|
|
|
|
* well
|
2003-10-25 00:29:13 +00:00
|
|
|
*
|
|
|
|
* if we were replaying, then the replay buffer must be
|
|
|
|
* copied back into the UConverter
|
|
|
|
* and the real arguments must be restored
|
2003-07-22 04:22:57 +00:00
|
|
|
*/
|
2003-10-25 00:29:13 +00:00
|
|
|
if(realSource!=NULL) {
|
|
|
|
int32_t length;
|
|
|
|
|
|
|
|
U_ASSERT(cnv->preToULength==0);
|
|
|
|
|
|
|
|
length=(int32_t)(pArgs->sourceLimit-pArgs->source);
|
|
|
|
if(length>0) {
|
|
|
|
uprv_memcpy(cnv->preToU, pArgs->source, length);
|
|
|
|
cnv->preToULength=(int8_t)-length;
|
|
|
|
}
|
|
|
|
|
|
|
|
pArgs->source=realSource;
|
|
|
|
pArgs->sourceLimit=realSourceLimit;
|
|
|
|
pArgs->flush=realFlush;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2003-08-01 14:39:14 +00:00
|
|
|
/* copy toUBytes[] to invalidCharBuffer[] */
|
|
|
|
errorInputLength=cnv->invalidCharLength=cnv->toULength;
|
|
|
|
if(errorInputLength>0) {
|
|
|
|
uprv_memcpy(cnv->invalidCharBuffer, cnv->toUBytes, errorInputLength);
|
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/* set the converter state to deal with the next character */
|
|
|
|
cnv->toULength=0;
|
|
|
|
|
|
|
|
/* call the callback function */
|
|
|
|
cnv->fromCharErrorBehaviour(cnv->toUContext, pArgs,
|
|
|
|
cnv->invalidCharBuffer, errorInputLength,
|
2003-12-03 17:43:17 +00:00
|
|
|
(*err==U_INVALID_CHAR_FOUND || *err==U_UNSUPPORTED_ESCAPE_SEQUENCE) ?
|
|
|
|
UCNV_UNASSIGNED : UCNV_ILLEGAL,
|
2003-07-22 04:22:57 +00:00
|
|
|
err);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* loop back to the offset handling
|
|
|
|
*
|
|
|
|
* this flag will indicate after offset handling
|
|
|
|
* that a callback was called;
|
|
|
|
* if the callback did not resolve the error, then we return
|
|
|
|
*/
|
|
|
|
calledCallback=TRUE;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* Output the toUnicode overflow buffer.
|
|
|
|
* Call this function if(cnv->UCharErrorBufferLength>0).
|
|
|
|
* @return TRUE if overflow
|
|
|
|
*/
|
|
|
|
static UBool
|
|
|
|
ucnv_outputOverflowToUnicode(UConverter *cnv,
|
|
|
|
UChar **target, const UChar *targetLimit,
|
|
|
|
int32_t **pOffsets,
|
|
|
|
UErrorCode *err) {
|
|
|
|
int32_t *offsets;
|
|
|
|
UChar *overflow, *t;
|
|
|
|
int32_t i, length;
|
|
|
|
|
|
|
|
t=*target;
|
|
|
|
if(pOffsets!=NULL) {
|
|
|
|
offsets=*pOffsets;
|
|
|
|
} else {
|
|
|
|
offsets=NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
overflow=cnv->UCharErrorBuffer;
|
|
|
|
length=cnv->UCharErrorBufferLength;
|
|
|
|
i=0;
|
|
|
|
while(i<length) {
|
|
|
|
if(t==targetLimit) {
|
|
|
|
/* the overflow buffer contains too much, keep the rest */
|
|
|
|
int32_t j=0;
|
|
|
|
|
|
|
|
do {
|
|
|
|
overflow[j++]=overflow[i++];
|
|
|
|
} while(i<length);
|
|
|
|
|
|
|
|
cnv->UCharErrorBufferLength=(int8_t)j;
|
|
|
|
*target=t;
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*pOffsets=offsets;
|
|
|
|
}
|
|
|
|
*err=U_BUFFER_OVERFLOW_ERROR;
|
|
|
|
return TRUE;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* copy the overflow contents to the target */
|
|
|
|
*t++=overflow[i++];
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*offsets++=-1; /* no source index available for old output */
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* the overflow buffer is completely copied to the target */
|
|
|
|
cnv->UCharErrorBufferLength=0;
|
|
|
|
*target=t;
|
|
|
|
if(offsets!=NULL) {
|
|
|
|
*pOffsets=offsets;
|
|
|
|
}
|
|
|
|
return FALSE;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_toUnicode(UConverter *cnv,
|
|
|
|
UChar **target, const UChar *targetLimit,
|
|
|
|
const char **source, const char *sourceLimit,
|
|
|
|
int32_t *offsets,
|
|
|
|
UBool flush,
|
|
|
|
UErrorCode *err) {
|
2001-10-04 01:04:28 +00:00
|
|
|
UConverterToUnicodeArgs args;
|
2003-07-22 04:22:57 +00:00
|
|
|
const char *s;
|
|
|
|
UChar *t;
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/* check parameters */
|
|
|
|
if(err==NULL || U_FAILURE(*err)) {
|
2001-10-04 01:04:28 +00:00
|
|
|
return;
|
|
|
|
}
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
if(cnv==NULL || target==NULL || source==NULL) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
2001-10-04 01:04:28 +00:00
|
|
|
return;
|
|
|
|
}
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
s=*source;
|
|
|
|
t=*target;
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2007-02-27 08:48:19 +00:00
|
|
|
if ((const void *)U_MAX_PTR(targetLimit) == (const void *)targetLimit) {
|
|
|
|
/*
|
|
|
|
Prevent code from going into an infinite loop in case we do hit this
|
|
|
|
limit. The limit pointer is expected to be on a UChar * boundary.
|
|
|
|
This also prevents the next argument check from failing.
|
|
|
|
*/
|
|
|
|
targetLimit = (const UChar *)(((const char *)targetLimit) - 1);
|
|
|
|
}
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
/*
|
2006-11-14 22:59:46 +00:00
|
|
|
* All these conditions should never happen.
|
|
|
|
*
|
|
|
|
* 1) Make sure that the limits are >= to the address source or target
|
|
|
|
*
|
|
|
|
* 2) Make sure that the buffer sizes do not exceed the number range for
|
2003-07-22 04:22:57 +00:00
|
|
|
* int32_t because some functions use the size (in units or bytes)
|
|
|
|
* rather than comparing pointers, and because offsets are int32_t values.
|
|
|
|
*
|
|
|
|
* size_t is guaranteed to be unsigned and large enough for the job.
|
|
|
|
*
|
|
|
|
* Return with an error instead of adjusting the limits because we would
|
|
|
|
* not be able to maintain the semantics that either the source must be
|
|
|
|
* consumed or the target filled (unless an error occurs).
|
|
|
|
* An adjustment would be sourceLimit=t+0x7fffffff; for example.
|
2006-11-14 22:59:46 +00:00
|
|
|
*
|
|
|
|
* 3) Make sure that the user didn't incorrectly cast a UChar * pointer
|
|
|
|
* to a char * pointer and provide an incomplete UChar code unit.
|
2003-07-22 04:22:57 +00:00
|
|
|
*/
|
2006-11-14 22:59:46 +00:00
|
|
|
if (sourceLimit<s || targetLimit<t ||
|
2003-07-22 04:22:57 +00:00
|
|
|
((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s) ||
|
2006-11-14 22:59:46 +00:00
|
|
|
((size_t)(targetLimit-t)>(size_t)0x3fffffff && targetLimit>t) ||
|
|
|
|
(((const char *)targetLimit-(const char *)t) & 1) != 0
|
2003-07-22 04:22:57 +00:00
|
|
|
) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/* output the target overflow buffer */
|
|
|
|
if( cnv->UCharErrorBufferLength>0 &&
|
|
|
|
ucnv_outputOverflowToUnicode(cnv, target, targetLimit, &offsets, err)
|
|
|
|
) {
|
|
|
|
/* U_BUFFER_OVERFLOW_ERROR */
|
|
|
|
return;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2007-02-06 05:24:14 +00:00
|
|
|
/* *target may have moved, therefore stop using t */
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-10-25 00:29:13 +00:00
|
|
|
if(!flush && s==sourceLimit && cnv->preToULength>=0) {
|
2003-04-18 23:11:39 +00:00
|
|
|
/* the overflow buffer is emptied and there is no new input: we are done */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/*
|
|
|
|
* Do not simply return with a buffer overflow error if
|
|
|
|
* !flush && t==targetLimit
|
|
|
|
* because it is possible that the source will not generate any output.
|
|
|
|
* For example, the skip callback may be called;
|
|
|
|
* it does not output anything.
|
|
|
|
*/
|
2000-02-08 23:41:16 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/* prepare the converter arguments */
|
|
|
|
args.converter=cnv;
|
|
|
|
args.flush=flush;
|
|
|
|
args.offsets=offsets;
|
|
|
|
args.source=s;
|
|
|
|
args.sourceLimit=sourceLimit;
|
2007-02-06 05:24:14 +00:00
|
|
|
args.target=*target;
|
2003-07-22 04:22:57 +00:00
|
|
|
args.targetLimit=targetLimit;
|
|
|
|
args.size=sizeof(args);
|
2000-08-11 01:27:17 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
_toUnicodeWithCallback(&args, err);
|
|
|
|
|
|
|
|
*source=args.source;
|
|
|
|
*target=args.target;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2003-08-01 14:39:14 +00:00
|
|
|
/* ucnv_to/fromUChars() ----------------------------------------------------- */
|
2003-07-22 04:22:57 +00:00
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
2001-09-28 22:22:59 +00:00
|
|
|
ucnv_fromUChars(UConverter *cnv,
|
|
|
|
char *dest, int32_t destCapacity,
|
|
|
|
const UChar *src, int32_t srcLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
const UChar *srcLimit;
|
|
|
|
char *originalDest, *destLimit;
|
|
|
|
int32_t destLength;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* check arguments */
|
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
|
|
|
return 0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
if( cnv==NULL ||
|
|
|
|
destCapacity<0 || (destCapacity>0 && dest==NULL) ||
|
|
|
|
srcLength<-1 || (srcLength!=0 && src==NULL)
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* initialize */
|
|
|
|
ucnv_resetFromUnicode(cnv);
|
|
|
|
originalDest=dest;
|
|
|
|
if(srcLength==-1) {
|
|
|
|
srcLength=u_strlen(src);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
2001-09-28 22:22:59 +00:00
|
|
|
if(srcLength>0) {
|
|
|
|
srcLimit=src+srcLength;
|
|
|
|
destLimit=dest+destCapacity;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* pin the destination limit to U_MAX_PTR; NULL check is for OS/400 */
|
|
|
|
if(destLimit<dest || (destLimit==NULL && dest!=NULL)) {
|
|
|
|
destLimit=(char *)U_MAX_PTR(dest);
|
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* perform the conversion */
|
|
|
|
ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
|
|
|
|
destLength=(int32_t)(dest-originalDest);
|
|
|
|
|
|
|
|
/* if an overflow occurs, then get the preflighting length */
|
|
|
|
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
|
|
|
|
char buffer[1024];
|
|
|
|
|
|
|
|
destLimit=buffer+sizeof(buffer);
|
|
|
|
do {
|
|
|
|
dest=buffer;
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
|
|
|
ucnv_fromUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
|
|
|
|
destLength+=(int32_t)(dest-buffer);
|
|
|
|
} while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
destLength=0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2007-08-31 19:39:14 +00:00
|
|
|
return u_terminateChars(originalDest, destCapacity, destLength, pErrorCode);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
2001-09-28 22:22:59 +00:00
|
|
|
ucnv_toUChars(UConverter *cnv,
|
|
|
|
UChar *dest, int32_t destCapacity,
|
|
|
|
const char *src, int32_t srcLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
const char *srcLimit;
|
|
|
|
UChar *originalDest, *destLimit;
|
|
|
|
int32_t destLength;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* check arguments */
|
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
2000-08-11 01:27:17 +00:00
|
|
|
return 0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
if( cnv==NULL ||
|
|
|
|
destCapacity<0 || (destCapacity>0 && dest==NULL) ||
|
2001-10-04 01:04:28 +00:00
|
|
|
srcLength<-1 || (srcLength!=0 && src==NULL))
|
|
|
|
{
|
2001-09-28 22:22:59 +00:00
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* initialize */
|
|
|
|
ucnv_resetToUnicode(cnv);
|
|
|
|
originalDest=dest;
|
|
|
|
if(srcLength==-1) {
|
2004-12-08 01:05:40 +00:00
|
|
|
srcLength=(int32_t)uprv_strlen(src);
|
2001-09-28 22:22:59 +00:00
|
|
|
}
|
|
|
|
if(srcLength>0) {
|
|
|
|
srcLimit=src+srcLength;
|
|
|
|
destLimit=dest+destCapacity;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* pin the destination limit to U_MAX_PTR; NULL check is for OS/400 */
|
|
|
|
if(destLimit<dest || (destLimit==NULL && dest!=NULL)) {
|
|
|
|
destLimit=(UChar *)U_MAX_PTR(dest);
|
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
/* perform the conversion */
|
|
|
|
ucnv_toUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
|
|
|
|
destLength=(int32_t)(dest-originalDest);
|
|
|
|
|
|
|
|
/* if an overflow occurs, then get the preflighting length */
|
2001-10-04 01:04:28 +00:00
|
|
|
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR)
|
|
|
|
{
|
2001-09-28 22:22:59 +00:00
|
|
|
UChar buffer[1024];
|
|
|
|
|
|
|
|
destLimit=buffer+sizeof(buffer)/U_SIZEOF_UCHAR;
|
|
|
|
do {
|
|
|
|
dest=buffer;
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
|
|
|
ucnv_toUnicode(cnv, &dest, destLimit, &src, srcLimit, 0, TRUE, pErrorCode);
|
|
|
|
destLength+=(int32_t)(dest-buffer);
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
|
|
|
while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
|
2001-09-28 22:22:59 +00:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
destLength=0;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-09-28 22:22:59 +00:00
|
|
|
return u_terminateUChars(originalDest, destCapacity, destLength, pErrorCode);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2003-08-01 14:39:14 +00:00
|
|
|
/* ucnv_getNextUChar() ------------------------------------------------------ */
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
U_CAPI UChar32 U_EXPORT2
|
|
|
|
ucnv_getNextUChar(UConverter *cnv,
|
|
|
|
const char **source, const char *sourceLimit,
|
|
|
|
UErrorCode *err) {
|
2001-10-04 01:04:28 +00:00
|
|
|
UConverterToUnicodeArgs args;
|
2003-08-01 14:39:14 +00:00
|
|
|
UChar buffer[U16_MAX_LENGTH];
|
2003-07-22 04:22:57 +00:00
|
|
|
const char *s;
|
|
|
|
UChar32 c;
|
|
|
|
int32_t i, length;
|
2000-07-10 20:51:54 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/* check parameters */
|
|
|
|
if(err==NULL || U_FAILURE(*err)) {
|
2001-10-04 01:04:28 +00:00
|
|
|
return 0xffff;
|
|
|
|
}
|
2000-07-21 23:58:31 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
if(cnv==NULL || source==NULL) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
2001-10-04 01:04:28 +00:00
|
|
|
return 0xffff;
|
|
|
|
}
|
2001-03-17 22:13:27 +00:00
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
s=*source;
|
|
|
|
if(sourceLimit<s) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0xffff;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Make sure that the buffer sizes do not exceed the number range for
|
|
|
|
* int32_t because some functions use the size (in units or bytes)
|
|
|
|
* rather than comparing pointers, and because offsets are int32_t values.
|
|
|
|
*
|
|
|
|
* size_t is guaranteed to be unsigned and large enough for the job.
|
|
|
|
*
|
|
|
|
* Return with an error instead of adjusting the limits because we would
|
|
|
|
* not be able to maintain the semantics that either the source must be
|
|
|
|
* consumed or the target filled (unless an error occurs).
|
|
|
|
* An adjustment would be sourceLimit=t+0x7fffffff; for example.
|
|
|
|
*/
|
|
|
|
if(((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s)) {
|
|
|
|
*err=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0xffff;
|
|
|
|
}
|
|
|
|
|
|
|
|
c=U_SENTINEL;
|
|
|
|
|
|
|
|
/* flush the target overflow buffer */
|
|
|
|
if(cnv->UCharErrorBufferLength>0) {
|
|
|
|
UChar *overflow;
|
|
|
|
|
|
|
|
overflow=cnv->UCharErrorBuffer;
|
|
|
|
i=0;
|
|
|
|
length=cnv->UCharErrorBufferLength;
|
|
|
|
U16_NEXT(overflow, i, length, c);
|
|
|
|
|
|
|
|
/* move the remaining overflow contents up to the beginning */
|
|
|
|
if((cnv->UCharErrorBufferLength=(int8_t)(length-i))>0) {
|
|
|
|
uprv_memmove(cnv->UCharErrorBuffer, cnv->UCharErrorBuffer+i,
|
|
|
|
cnv->UCharErrorBufferLength*U_SIZEOF_UCHAR);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(!U16_IS_LEAD(c) || i<length) {
|
|
|
|
return c;
|
|
|
|
}
|
|
|
|
/*
|
|
|
|
* Continue if the overflow buffer contained only a lead surrogate,
|
|
|
|
* in case the converter outputs single surrogates from complete
|
|
|
|
* input sequences.
|
|
|
|
*/
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* flush==TRUE is implied for ucnv_getNextUChar()
|
|
|
|
*
|
|
|
|
* do not simply return even if s==sourceLimit because the converter may
|
|
|
|
* not have seen flush==TRUE before
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* prepare the converter arguments */
|
|
|
|
args.converter=cnv;
|
|
|
|
args.flush=TRUE;
|
|
|
|
args.offsets=NULL;
|
|
|
|
args.source=s;
|
|
|
|
args.sourceLimit=sourceLimit;
|
|
|
|
args.target=buffer;
|
|
|
|
args.targetLimit=buffer+1;
|
|
|
|
args.size=sizeof(args);
|
|
|
|
|
|
|
|
if(c<0) {
|
2003-08-01 14:39:14 +00:00
|
|
|
/*
|
|
|
|
* call the native getNextUChar() implementation if we are
|
|
|
|
* at a character boundary (toULength==0)
|
|
|
|
*
|
|
|
|
* unlike with _toUnicode(), getNextUChar() implementations must set
|
|
|
|
* U_TRUNCATED_CHAR_FOUND for truncated input,
|
|
|
|
* in addition to setting toULength/toUBytes[]
|
|
|
|
*/
|
|
|
|
if(cnv->toULength==0 && cnv->sharedData->impl->getNextUChar!=NULL) {
|
2003-07-22 04:22:57 +00:00
|
|
|
c=cnv->sharedData->impl->getNextUChar(&args, err);
|
2003-08-01 14:39:14 +00:00
|
|
|
*source=s=args.source;
|
|
|
|
if(*err==U_INDEX_OUTOFBOUNDS_ERROR) {
|
2003-10-31 04:59:36 +00:00
|
|
|
/* reset the converter without calling the callback function */
|
|
|
|
_reset(cnv, UCNV_RESET_TO_UNICODE, FALSE);
|
|
|
|
return 0xffff; /* no output */
|
2003-08-01 14:39:14 +00:00
|
|
|
} else if(U_SUCCESS(*err) && c>=0) {
|
|
|
|
return c;
|
|
|
|
/*
|
|
|
|
* else fall through to use _toUnicode() because
|
|
|
|
* UCNV_GET_NEXT_UCHAR_USE_TO_U: the native function did not want to handle it after all
|
|
|
|
* U_FAILURE: call _toUnicode() for callback handling (do not output c)
|
|
|
|
*/
|
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
}
|
|
|
|
|
2003-08-01 14:39:14 +00:00
|
|
|
/* convert to one UChar in buffer[0], or handle getNextUChar() errors */
|
2003-07-22 04:22:57 +00:00
|
|
|
_toUnicodeWithCallback(&args, err);
|
2003-08-01 14:39:14 +00:00
|
|
|
|
|
|
|
if(*err==U_BUFFER_OVERFLOW_ERROR) {
|
|
|
|
*err=U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
|
|
|
|
i=0;
|
|
|
|
length=(int32_t)(args.target-buffer);
|
2001-10-04 01:04:28 +00:00
|
|
|
} else {
|
2003-07-22 04:22:57 +00:00
|
|
|
/* write the lead surrogate from the overflow buffer */
|
|
|
|
buffer[0]=(UChar)c;
|
|
|
|
args.target=buffer+1;
|
2003-08-01 14:39:14 +00:00
|
|
|
i=0;
|
|
|
|
length=1;
|
2003-07-22 04:22:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* buffer contents starts at i and ends before length */
|
|
|
|
|
|
|
|
if(U_FAILURE(*err)) {
|
|
|
|
c=0xffff; /* no output */
|
|
|
|
} else if(length==0) {
|
|
|
|
/* no input or only state changes */
|
|
|
|
*err=U_INDEX_OUTOFBOUNDS_ERROR;
|
2003-10-31 16:36:32 +00:00
|
|
|
/* no need to reset explicitly because _toUnicodeWithCallback() did it */
|
2003-07-22 04:22:57 +00:00
|
|
|
c=0xffff; /* no output */
|
|
|
|
} else {
|
|
|
|
c=buffer[0];
|
|
|
|
i=1;
|
|
|
|
if(!U16_IS_LEAD(c)) {
|
|
|
|
/* consume c=buffer[0], done */
|
|
|
|
} else {
|
|
|
|
/* got a lead surrogate, see if a trail surrogate follows */
|
|
|
|
UChar c2;
|
|
|
|
|
|
|
|
if(cnv->UCharErrorBufferLength>0) {
|
|
|
|
/* got overflow output from the conversion */
|
|
|
|
if(U16_IS_TRAIL(c2=cnv->UCharErrorBuffer[0])) {
|
|
|
|
/* got a trail surrogate, too */
|
|
|
|
c=U16_GET_SUPPLEMENTARY(c, c2);
|
|
|
|
|
|
|
|
/* move the remaining overflow contents up to the beginning */
|
|
|
|
if((--cnv->UCharErrorBufferLength)>0) {
|
|
|
|
uprv_memmove(cnv->UCharErrorBuffer, cnv->UCharErrorBuffer+1,
|
|
|
|
cnv->UCharErrorBufferLength*U_SIZEOF_UCHAR);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/* c is an unpaired lead surrogate, just return it */
|
|
|
|
}
|
|
|
|
} else if(args.source<sourceLimit) {
|
|
|
|
/* convert once more, to buffer[1] */
|
|
|
|
args.targetLimit=buffer+2;
|
|
|
|
_toUnicodeWithCallback(&args, err);
|
2003-09-15 21:09:21 +00:00
|
|
|
if(*err==U_BUFFER_OVERFLOW_ERROR) {
|
2003-07-22 04:22:57 +00:00
|
|
|
*err=U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
|
|
|
|
length=(int32_t)(args.target-buffer);
|
|
|
|
if(U_SUCCESS(*err) && length==2 && U16_IS_TRAIL(c2=buffer[1])) {
|
|
|
|
/* got a trail surrogate, too */
|
|
|
|
c=U16_GET_SUPPLEMENTARY(c, c2);
|
|
|
|
i=2;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2001-10-04 01:04:28 +00:00
|
|
|
}
|
2003-07-22 04:22:57 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* move leftover output from buffer[i..length[
|
|
|
|
* into the beginning of the overflow buffer
|
|
|
|
*/
|
|
|
|
if(i<length) {
|
|
|
|
/* move further overflow back */
|
|
|
|
int32_t delta=length-i;
|
|
|
|
if((length=cnv->UCharErrorBufferLength)>0) {
|
|
|
|
uprv_memmove(cnv->UCharErrorBuffer+delta, cnv->UCharErrorBuffer,
|
|
|
|
length*U_SIZEOF_UCHAR);
|
|
|
|
}
|
|
|
|
cnv->UCharErrorBufferLength=(int8_t)(length+delta);
|
|
|
|
|
|
|
|
cnv->UCharErrorBuffer[0]=buffer[i++];
|
|
|
|
if(delta>1) {
|
|
|
|
cnv->UCharErrorBuffer[1]=buffer[i];
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
*source=args.source;
|
|
|
|
return c;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2003-07-22 04:22:57 +00:00
|
|
|
/* ucnv_convert() and siblings ---------------------------------------------- */
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
|
|
|
|
char **target, const char *targetLimit,
|
|
|
|
const char **source, const char *sourceLimit,
|
|
|
|
UChar *pivotStart, UChar **pivotSource,
|
|
|
|
UChar **pivotTarget, const UChar *pivotLimit,
|
|
|
|
UBool reset, UBool flush,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
UChar pivotBuffer[CHUNK_SIZE];
|
2007-02-06 05:24:14 +00:00
|
|
|
const UChar *myPivotSource;
|
|
|
|
UChar *myPivotTarget;
|
|
|
|
const char *s;
|
|
|
|
char *t;
|
|
|
|
|
|
|
|
UConverterToUnicodeArgs toUArgs;
|
|
|
|
UConverterFromUnicodeArgs fromUArgs;
|
|
|
|
UConverterConvert convert;
|
2003-04-18 23:11:39 +00:00
|
|
|
|
|
|
|
/* error checking */
|
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if( targetCnv==NULL || sourceCnv==NULL ||
|
|
|
|
source==NULL || *source==NULL ||
|
|
|
|
target==NULL || *target==NULL || targetLimit==NULL
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
s=*source;
|
|
|
|
t=*target;
|
|
|
|
if((sourceLimit!=NULL && sourceLimit<s) || targetLimit<t) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Make sure that the buffer sizes do not exceed the number range for
|
|
|
|
* int32_t. See ucnv_toUnicode() for a more detailed comment.
|
|
|
|
*/
|
|
|
|
if(
|
|
|
|
(sourceLimit!=NULL && ((size_t)(sourceLimit-s)>(size_t)0x7fffffff && sourceLimit>s)) ||
|
|
|
|
((size_t)(targetLimit-t)>(size_t)0x7fffffff && targetLimit>t)
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
if(pivotStart==NULL) {
|
2006-06-16 21:55:52 +00:00
|
|
|
if(!flush) {
|
|
|
|
/* streaming conversion requires an explicit pivot buffer */
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
/* use the stack pivot buffer */
|
2007-02-06 05:24:14 +00:00
|
|
|
myPivotSource=myPivotTarget=pivotStart=pivotBuffer;
|
|
|
|
pivotSource=(UChar **)&myPivotSource;
|
2003-04-18 23:11:39 +00:00
|
|
|
pivotTarget=&myPivotTarget;
|
|
|
|
pivotLimit=pivotBuffer+CHUNK_SIZE;
|
|
|
|
} else if( pivotStart>=pivotLimit ||
|
|
|
|
pivotSource==NULL || *pivotSource==NULL ||
|
|
|
|
pivotTarget==NULL || *pivotTarget==NULL ||
|
|
|
|
pivotLimit==NULL
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(sourceLimit==NULL) {
|
|
|
|
/* get limit of single-byte-NUL-terminated source string */
|
|
|
|
sourceLimit=uprv_strchr(*source, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
if(reset) {
|
|
|
|
ucnv_resetToUnicode(sourceCnv);
|
|
|
|
ucnv_resetFromUnicode(targetCnv);
|
2007-02-06 05:24:14 +00:00
|
|
|
*pivotSource=*pivotTarget=pivotStart;
|
|
|
|
} else if(targetCnv->charErrorBufferLength>0) {
|
|
|
|
/* output the targetCnv overflow buffer */
|
|
|
|
if(ucnv_outputOverflowFromUnicode(targetCnv, target, targetLimit, NULL, pErrorCode)) {
|
|
|
|
/* U_BUFFER_OVERFLOW_ERROR */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
/* *target has moved, therefore stop using t */
|
|
|
|
|
|
|
|
if( !flush &&
|
|
|
|
targetCnv->preFromULength>=0 && *pivotSource==*pivotTarget &&
|
|
|
|
sourceCnv->UCharErrorBufferLength==0 && sourceCnv->preToULength>=0 && s==sourceLimit
|
|
|
|
) {
|
|
|
|
/* the fromUnicode overflow buffer is emptied and there is no new input: we are done */
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Is direct-UTF-8 conversion available? */
|
|
|
|
if( sourceCnv->sharedData->staticData->conversionType==UCNV_UTF8 &&
|
|
|
|
targetCnv->sharedData->impl->fromUTF8!=NULL
|
|
|
|
) {
|
|
|
|
convert=targetCnv->sharedData->impl->fromUTF8;
|
|
|
|
} else if( targetCnv->sharedData->staticData->conversionType==UCNV_UTF8 &&
|
|
|
|
sourceCnv->sharedData->impl->toUTF8!=NULL
|
|
|
|
) {
|
|
|
|
convert=sourceCnv->sharedData->impl->toUTF8;
|
|
|
|
} else {
|
|
|
|
convert=NULL;
|
2003-04-18 23:11:39 +00:00
|
|
|
}
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* If direct-UTF-8 conversion is available, then we use a smaller
|
|
|
|
* pivot buffer for error handling and partial matches
|
|
|
|
* so that we quickly return to direct conversion.
|
|
|
|
*
|
|
|
|
* 32 is large enough for UCNV_EXT_MAX_UCHARS and UCNV_ERROR_BUFFER_LENGTH.
|
|
|
|
*
|
|
|
|
* We could reduce the pivot buffer size further, at the cost of
|
|
|
|
* buffer overflows from callbacks.
|
|
|
|
* The pivot buffer should not be smaller than the maximum number of
|
|
|
|
* fromUnicode extension table input UChars
|
|
|
|
* (for m:n conversion, see
|
|
|
|
* targetCnv->sharedData->mbcs.extIndexes[UCNV_EXT_COUNT_UCHARS])
|
|
|
|
* or 2 for surrogate pairs.
|
|
|
|
*
|
|
|
|
* Too small a buffer can cause thrashing between pivoting and direct
|
|
|
|
* conversion, with function call overhead outweighing the benefits
|
|
|
|
* of direct conversion.
|
|
|
|
*/
|
|
|
|
if(convert!=NULL && (pivotLimit-pivotStart)>32) {
|
|
|
|
pivotLimit=pivotStart+32;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* prepare the converter arguments */
|
|
|
|
fromUArgs.converter=targetCnv;
|
|
|
|
fromUArgs.flush=FALSE;
|
|
|
|
fromUArgs.offsets=NULL;
|
|
|
|
fromUArgs.target=*target;
|
|
|
|
fromUArgs.targetLimit=targetLimit;
|
|
|
|
fromUArgs.size=sizeof(fromUArgs);
|
|
|
|
|
|
|
|
toUArgs.converter=sourceCnv;
|
|
|
|
toUArgs.flush=flush;
|
|
|
|
toUArgs.offsets=NULL;
|
|
|
|
toUArgs.source=s;
|
|
|
|
toUArgs.sourceLimit=sourceLimit;
|
|
|
|
toUArgs.targetLimit=pivotLimit;
|
|
|
|
toUArgs.size=sizeof(toUArgs);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* TODO: Consider separating this function into two functions,
|
|
|
|
* extracting exactly the conversion loop,
|
|
|
|
* for readability and to reduce the set of visible variables.
|
|
|
|
*
|
|
|
|
* Otherwise stop using s and t from here on.
|
|
|
|
*/
|
|
|
|
s=t=NULL;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* conversion loop
|
|
|
|
*
|
|
|
|
* The sequence of steps in the loop may appear backward,
|
|
|
|
* but the principle is simple:
|
|
|
|
* In the chain of
|
|
|
|
* source - sourceCnv overflow - pivot - targetCnv overflow - target
|
|
|
|
* empty out later buffers before refilling them from earlier ones.
|
|
|
|
*
|
|
|
|
* The targetCnv overflow buffer is flushed out only once before the loop.
|
|
|
|
*/
|
2003-04-18 23:11:39 +00:00
|
|
|
for(;;) {
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* if(pivot not empty or error or replay or flush fromUnicode) {
|
|
|
|
* fromUnicode(pivot -> target);
|
|
|
|
* }
|
|
|
|
*
|
|
|
|
* For pivoting conversion; and for direct conversion for
|
|
|
|
* error callback handling and flushing the replay buffer.
|
|
|
|
*/
|
|
|
|
if( *pivotSource<*pivotTarget ||
|
|
|
|
U_FAILURE(*pErrorCode) ||
|
|
|
|
targetCnv->preFromULength<0 ||
|
|
|
|
fromUArgs.flush
|
|
|
|
) {
|
|
|
|
fromUArgs.source=*pivotSource;
|
|
|
|
fromUArgs.sourceLimit=*pivotTarget;
|
|
|
|
_fromUnicodeWithCallback(&fromUArgs, pErrorCode);
|
|
|
|
if(U_FAILURE(*pErrorCode)) {
|
|
|
|
/* target overflow, or conversion error */
|
|
|
|
*pivotSource=(UChar *)fromUArgs.source;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
/*
|
2007-02-06 05:24:14 +00:00
|
|
|
* _fromUnicodeWithCallback() must have consumed the pivot contents
|
|
|
|
* (*pivotSource==*pivotTarget) since it returned with U_SUCCESS()
|
2003-04-18 23:11:39 +00:00
|
|
|
*/
|
2007-02-06 05:24:14 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* The pivot buffer is empty; reset it so we start at pivotStart. */
|
|
|
|
*pivotSource=*pivotTarget=pivotStart;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* if(sourceCnv overflow buffer not empty) {
|
|
|
|
* move(sourceCnv overflow buffer -> pivot);
|
|
|
|
* continue;
|
|
|
|
* }
|
|
|
|
*/
|
|
|
|
/* output the sourceCnv overflow buffer */
|
|
|
|
if(sourceCnv->UCharErrorBufferLength>0) {
|
|
|
|
if(ucnv_outputOverflowToUnicode(sourceCnv, pivotTarget, pivotLimit, NULL, pErrorCode)) {
|
|
|
|
/* U_BUFFER_OVERFLOW_ERROR */
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* check for end of input and break if done
|
|
|
|
*
|
|
|
|
* Checking both flush and fromUArgs.flush ensures that the converters
|
|
|
|
* have been called with the flush flag set if the ucnv_convertEx()
|
|
|
|
* caller set it.
|
|
|
|
*/
|
|
|
|
if( toUArgs.source==sourceLimit &&
|
|
|
|
sourceCnv->preToULength>=0 && sourceCnv->toULength==0 &&
|
|
|
|
(!flush || fromUArgs.flush)
|
|
|
|
) {
|
|
|
|
/* done successfully */
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* use direct conversion if available
|
|
|
|
* but not if continuing a partial match
|
|
|
|
* or flushing the toUnicode replay buffer
|
|
|
|
*/
|
|
|
|
if(convert!=NULL && targetCnv->preFromUFirstCP<0 && sourceCnv->preToULength==0) {
|
|
|
|
if(*pErrorCode==U_USING_DEFAULT_WARNING) {
|
|
|
|
/* remove a warning that may be set by this function */
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
convert(&fromUArgs, &toUArgs, pErrorCode);
|
|
|
|
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
|
|
|
|
break;
|
|
|
|
} else if(U_FAILURE(*pErrorCode)) {
|
|
|
|
if(sourceCnv->toULength>0) {
|
|
|
|
/*
|
|
|
|
* Fall through to calling _toUnicodeWithCallback()
|
|
|
|
* for callback handling.
|
|
|
|
*
|
|
|
|
* The pivot buffer will be reset with
|
|
|
|
* *pivotSource=*pivotTarget=pivotStart;
|
|
|
|
* which indicates a toUnicode error to the caller
|
|
|
|
* (*pivotSource==pivotStart shows no pivot UChars consumed).
|
|
|
|
*/
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Indicate a fromUnicode error to the caller
|
|
|
|
* (*pivotSource>pivotStart shows some pivot UChars consumed).
|
|
|
|
*/
|
|
|
|
*pivotSource=*pivotTarget=pivotStart+1;
|
|
|
|
/*
|
|
|
|
* Loop around to calling _fromUnicodeWithCallbacks()
|
|
|
|
* for callback handling.
|
|
|
|
*/
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
} else if(*pErrorCode==U_USING_DEFAULT_WARNING) {
|
|
|
|
/*
|
|
|
|
* No error, but the implementation requested to temporarily
|
|
|
|
* fall back to pivoting.
|
|
|
|
*/
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
2003-04-18 23:11:39 +00:00
|
|
|
/*
|
2007-02-06 05:24:14 +00:00
|
|
|
* The following else branches are almost identical to the end-of-input
|
|
|
|
* handling in _toUnicodeWithCallback().
|
|
|
|
* Avoid calling it just for the end of input.
|
2003-04-18 23:11:39 +00:00
|
|
|
*/
|
2007-02-06 05:24:14 +00:00
|
|
|
} else if(flush && sourceCnv->toULength>0) { /* flush==toUArgs.flush */
|
|
|
|
/*
|
|
|
|
* the entire input stream is consumed
|
|
|
|
* and there is a partial, truncated input sequence left
|
|
|
|
*/
|
|
|
|
|
|
|
|
/* inject an error and continue with callback handling */
|
|
|
|
*pErrorCode=U_TRUNCATED_CHAR_FOUND;
|
|
|
|
} else {
|
|
|
|
/* input consumed */
|
|
|
|
if(flush) {
|
|
|
|
/* reset the converters without calling the callback functions */
|
|
|
|
_reset(sourceCnv, UCNV_RESET_TO_UNICODE, FALSE);
|
|
|
|
_reset(targetCnv, UCNV_RESET_FROM_UNICODE, FALSE);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* done successfully */
|
2003-04-18 23:11:39 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2007-02-06 05:24:14 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* toUnicode(source -> pivot);
|
|
|
|
*
|
|
|
|
* For pivoting conversion; and for direct conversion for
|
|
|
|
* error callback handling, continuing partial matches
|
|
|
|
* and flushing the replay buffer.
|
|
|
|
*
|
|
|
|
* The pivot buffer is empty and reset.
|
|
|
|
*/
|
|
|
|
toUArgs.target=pivotStart; /* ==*pivotTarget */
|
|
|
|
/* toUArgs.targetLimit=pivotLimit; already set before the loop */
|
|
|
|
_toUnicodeWithCallback(&toUArgs, pErrorCode);
|
|
|
|
*pivotTarget=toUArgs.target;
|
2003-04-18 23:11:39 +00:00
|
|
|
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR) {
|
|
|
|
/* pivot overflow: continue with the conversion loop */
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
2007-02-06 05:24:14 +00:00
|
|
|
} else if(U_FAILURE(*pErrorCode) || (!flush && *pivotTarget==pivotStart)) {
|
2003-04-18 23:11:39 +00:00
|
|
|
/* conversion error, or there was nothing left to convert */
|
|
|
|
break;
|
|
|
|
}
|
2007-02-06 05:24:14 +00:00
|
|
|
/*
|
|
|
|
* else:
|
|
|
|
* _toUnicodeWithCallback() wrote into the pivot buffer,
|
|
|
|
* continue with fromUnicode conversion.
|
|
|
|
*
|
|
|
|
* Set the fromUnicode flush flag if we flush and if toUnicode has
|
|
|
|
* processed the end of the input.
|
|
|
|
*/
|
|
|
|
if( flush && toUArgs.source==sourceLimit &&
|
|
|
|
sourceCnv->preToULength>=0 &&
|
|
|
|
sourceCnv->UCharErrorBufferLength==0
|
|
|
|
) {
|
|
|
|
fromUArgs.flush=TRUE;
|
|
|
|
}
|
2003-04-18 23:11:39 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* The conversion loop is exited when one of the following is true:
|
|
|
|
* - the entire source text has been converted successfully to the target buffer
|
|
|
|
* - a target buffer overflow occurred
|
|
|
|
* - a conversion error occurred
|
|
|
|
*/
|
|
|
|
|
2007-02-06 05:24:14 +00:00
|
|
|
*source=toUArgs.source;
|
|
|
|
*target=fromUArgs.target;
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
/* terminate the target buffer if possible */
|
|
|
|
if(flush && U_SUCCESS(*pErrorCode)) {
|
|
|
|
if(*target!=targetLimit) {
|
|
|
|
**target=0;
|
|
|
|
if(*pErrorCode==U_STRING_NOT_TERMINATED_WARNING) {
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
*pErrorCode=U_STRING_NOT_TERMINATED_WARNING;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
/* internal implementation of ucnv_convert() etc. with preflighting */
|
|
|
|
static int32_t
|
|
|
|
ucnv_internalConvert(UConverter *outConverter, UConverter *inConverter,
|
|
|
|
char *target, int32_t targetCapacity,
|
|
|
|
const char *source, int32_t sourceLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
2001-05-19 00:55:13 +00:00
|
|
|
UChar pivotBuffer[CHUNK_SIZE];
|
|
|
|
UChar *pivot, *pivot2;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-10-04 00:23:09 +00:00
|
|
|
char *myTarget;
|
2001-05-21 22:58:19 +00:00
|
|
|
const char *sourceLimit;
|
2001-05-19 00:55:13 +00:00
|
|
|
const char *targetLimit;
|
2003-04-21 02:11:54 +00:00
|
|
|
int32_t targetLength=0;
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
/* set up */
|
|
|
|
if(sourceLength<0) {
|
|
|
|
sourceLimit=uprv_strchr(source, 0);
|
|
|
|
} else {
|
|
|
|
sourceLimit=source+sourceLength;
|
2001-05-19 00:55:13 +00:00
|
|
|
}
|
1999-08-16 21:50:52 +00:00
|
|
|
|
2001-05-19 00:55:13 +00:00
|
|
|
/* if there is no input data, we're done */
|
2003-04-21 02:11:54 +00:00
|
|
|
if(source==sourceLimit) {
|
2007-08-31 19:39:14 +00:00
|
|
|
return u_terminateChars(target, targetCapacity, 0, pErrorCode);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-05-19 00:55:13 +00:00
|
|
|
pivot=pivot2=pivotBuffer;
|
2001-10-04 00:23:09 +00:00
|
|
|
myTarget=target;
|
2003-04-21 02:11:54 +00:00
|
|
|
targetLength=0;
|
2001-05-19 00:55:13 +00:00
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
if(targetCapacity>0) {
|
2001-05-19 00:55:13 +00:00
|
|
|
/* perform real conversion */
|
2003-04-21 02:11:54 +00:00
|
|
|
targetLimit=target+targetCapacity;
|
2003-04-18 23:11:39 +00:00
|
|
|
ucnv_convertEx(outConverter, inConverter,
|
|
|
|
&myTarget, targetLimit,
|
|
|
|
&source, sourceLimit,
|
|
|
|
pivotBuffer, &pivot, &pivot2, pivotBuffer+CHUNK_SIZE,
|
|
|
|
FALSE,
|
|
|
|
TRUE,
|
|
|
|
pErrorCode);
|
2004-12-08 01:05:40 +00:00
|
|
|
targetLength=(int32_t)(myTarget-target);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-05-19 00:55:13 +00:00
|
|
|
/*
|
|
|
|
* If the output buffer is exhausted (or we are only "preflighting"), we need to stop writing
|
2003-04-21 02:11:54 +00:00
|
|
|
* to it but continue the conversion in order to store in targetCapacity
|
2001-05-19 00:55:13 +00:00
|
|
|
* the number of bytes that was required.
|
|
|
|
*/
|
2003-04-21 02:11:54 +00:00
|
|
|
if(*pErrorCode==U_BUFFER_OVERFLOW_ERROR || targetCapacity==0)
|
2001-10-04 01:04:28 +00:00
|
|
|
{
|
2001-05-19 00:55:13 +00:00
|
|
|
char targetBuffer[CHUNK_SIZE];
|
|
|
|
|
|
|
|
targetLimit=targetBuffer+CHUNK_SIZE;
|
|
|
|
do {
|
|
|
|
*pErrorCode=U_ZERO_ERROR;
|
2003-04-18 23:11:39 +00:00
|
|
|
myTarget=targetBuffer;
|
|
|
|
ucnv_convertEx(outConverter, inConverter,
|
|
|
|
&myTarget, targetLimit,
|
2001-05-19 00:55:13 +00:00
|
|
|
&source, sourceLimit,
|
2003-04-18 23:11:39 +00:00
|
|
|
pivotBuffer, &pivot, &pivot2, pivotBuffer+CHUNK_SIZE,
|
|
|
|
FALSE,
|
2001-05-19 00:55:13 +00:00
|
|
|
TRUE,
|
|
|
|
pErrorCode);
|
2004-12-08 01:05:40 +00:00
|
|
|
targetLength+=(int32_t)(myTarget-targetBuffer);
|
2003-04-18 23:11:39 +00:00
|
|
|
} while(*pErrorCode==U_BUFFER_OVERFLOW_ERROR);
|
|
|
|
|
2003-04-21 02:11:54 +00:00
|
|
|
/* done with preflighting, set warnings and errors as appropriate */
|
2007-08-31 19:39:14 +00:00
|
|
|
return u_terminateChars(target, targetCapacity, targetLength, pErrorCode);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2003-04-18 23:11:39 +00:00
|
|
|
/* no need to call u_terminateChars() because ucnv_convertEx() took care of that */
|
2003-04-21 02:11:54 +00:00
|
|
|
return targetLength;
|
|
|
|
}
|
|
|
|
|
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_convert(const char *toConverterName, const char *fromConverterName,
|
|
|
|
char *target, int32_t targetCapacity,
|
|
|
|
const char *source, int32_t sourceLength,
|
2007-08-31 19:39:14 +00:00
|
|
|
UErrorCode *pErrorCode) {
|
2003-04-21 02:11:54 +00:00
|
|
|
UConverter in, out; /* stack-allocated */
|
2007-08-31 19:39:14 +00:00
|
|
|
UConverter *inConverter, *outConverter;
|
|
|
|
int32_t targetLength;
|
2003-04-21 02:11:54 +00:00
|
|
|
|
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if( source==NULL || sourceLength<-1 ||
|
|
|
|
targetCapacity<0 || (targetCapacity>0 && target==NULL)
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* if there is no input data, we're done */
|
|
|
|
if(sourceLength==0 || (sourceLength<0 && *source==0)) {
|
2007-08-31 19:39:14 +00:00
|
|
|
return u_terminateChars(target, targetCapacity, 0, pErrorCode);
|
2003-04-21 02:11:54 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* create the converters */
|
|
|
|
inConverter=ucnv_createConverter(&in, fromConverterName, pErrorCode);
|
|
|
|
if(U_FAILURE(*pErrorCode)) {
|
2007-08-31 19:39:14 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
outConverter=ucnv_createConverter(&out, toConverterName, pErrorCode);
|
|
|
|
if(U_FAILURE(*pErrorCode)) {
|
|
|
|
ucnv_close(inConverter);
|
|
|
|
return 0;
|
2003-04-21 02:11:54 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
targetLength=ucnv_internalConvert(outConverter, inConverter,
|
|
|
|
target, targetCapacity,
|
|
|
|
source, sourceLength,
|
|
|
|
pErrorCode);
|
|
|
|
|
|
|
|
ucnv_close(inConverter);
|
|
|
|
ucnv_close(outConverter);
|
|
|
|
|
|
|
|
return targetLength;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* @internal */
|
|
|
|
static int32_t
|
|
|
|
ucnv_convertAlgorithmic(UBool convertToAlgorithmic,
|
|
|
|
UConverterType algorithmicType,
|
|
|
|
UConverter *cnv,
|
|
|
|
char *target, int32_t targetCapacity,
|
|
|
|
const char *source, int32_t sourceLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
UConverter algoConverterStatic; /* stack-allocated */
|
|
|
|
UConverter *algoConverter, *to, *from;
|
|
|
|
int32_t targetLength;
|
|
|
|
|
|
|
|
if(pErrorCode==NULL || U_FAILURE(*pErrorCode)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if( cnv==NULL || source==NULL || sourceLength<-1 ||
|
|
|
|
targetCapacity<0 || (targetCapacity>0 && target==NULL)
|
|
|
|
) {
|
|
|
|
*pErrorCode=U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* if there is no input data, we're done */
|
|
|
|
if(sourceLength==0 || (sourceLength<0 && *source==0)) {
|
2007-08-31 19:39:14 +00:00
|
|
|
return u_terminateChars(target, targetCapacity, 0, pErrorCode);
|
2003-04-21 02:11:54 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* create the algorithmic converter */
|
|
|
|
algoConverter=ucnv_createAlgorithmicConverter(&algoConverterStatic, algorithmicType,
|
|
|
|
"", 0, pErrorCode);
|
|
|
|
if(U_FAILURE(*pErrorCode)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* reset the other converter */
|
|
|
|
if(convertToAlgorithmic) {
|
|
|
|
/* cnv->Unicode->algo */
|
|
|
|
ucnv_resetToUnicode(cnv);
|
|
|
|
to=algoConverter;
|
|
|
|
from=cnv;
|
|
|
|
} else {
|
|
|
|
/* algo->Unicode->cnv */
|
|
|
|
ucnv_resetFromUnicode(cnv);
|
|
|
|
from=algoConverter;
|
|
|
|
to=cnv;
|
|
|
|
}
|
|
|
|
|
|
|
|
targetLength=ucnv_internalConvert(to, from,
|
|
|
|
target, targetCapacity,
|
|
|
|
source, sourceLength,
|
|
|
|
pErrorCode);
|
|
|
|
|
|
|
|
ucnv_close(algoConverter);
|
|
|
|
|
|
|
|
return targetLength;
|
|
|
|
}
|
|
|
|
|
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_toAlgorithmic(UConverterType algorithmicType,
|
|
|
|
UConverter *cnv,
|
|
|
|
char *target, int32_t targetCapacity,
|
|
|
|
const char *source, int32_t sourceLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
return ucnv_convertAlgorithmic(TRUE, algorithmicType, cnv,
|
|
|
|
target, targetCapacity,
|
|
|
|
source, sourceLength,
|
|
|
|
pErrorCode);
|
|
|
|
}
|
|
|
|
|
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_fromAlgorithmic(UConverter *cnv,
|
|
|
|
UConverterType algorithmicType,
|
|
|
|
char *target, int32_t targetCapacity,
|
|
|
|
const char *source, int32_t sourceLength,
|
|
|
|
UErrorCode *pErrorCode) {
|
|
|
|
return ucnv_convertAlgorithmic(FALSE, algorithmicType, cnv,
|
|
|
|
target, targetCapacity,
|
|
|
|
source, sourceLength,
|
|
|
|
pErrorCode);
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UConverterType U_EXPORT2
|
|
|
|
ucnv_getType(const UConverter* converter)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2000-12-20 20:34:23 +00:00
|
|
|
int8_t type = converter->sharedData->staticData->conversionType;
|
2002-09-20 01:54:48 +00:00
|
|
|
#if !UCONFIG_NO_LEGACY_CONVERSION
|
2000-12-20 20:34:23 +00:00
|
|
|
if(type == UCNV_MBCS) {
|
2004-09-16 04:00:12 +00:00
|
|
|
return ucnv_MBCSGetType(converter);
|
2000-12-20 20:34:23 +00:00
|
|
|
}
|
2002-09-20 01:54:48 +00:00
|
|
|
#endif
|
2000-12-20 20:34:23 +00:00
|
|
|
return (UConverterType)type;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_getStarters(const UConverter* converter,
|
|
|
|
UBool starters[256],
|
|
|
|
UErrorCode* err)
|
1999-08-16 21:50:52 +00:00
|
|
|
{
|
2000-02-09 19:15:17 +00:00
|
|
|
if (err == NULL || U_FAILURE(*err)) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(converter->sharedData->impl->getStarters != NULL) {
|
|
|
|
converter->sharedData->impl->getStarters(converter, starters, err);
|
|
|
|
} else {
|
|
|
|
*err = U_ILLEGAL_ARGUMENT_ERROR;
|
1999-08-16 21:50:52 +00:00
|
|
|
}
|
|
|
|
}
|
2000-01-08 00:51:44 +00:00
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
static const UAmbiguousConverter *ucnv_getAmbiguous(const UConverter *cnv)
|
|
|
|
{
|
2001-03-16 20:49:41 +00:00
|
|
|
UErrorCode errorCode;
|
|
|
|
const char *name;
|
|
|
|
int32_t i;
|
|
|
|
|
|
|
|
if(cnv==NULL) {
|
|
|
|
return NULL;
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
2001-03-16 20:49:41 +00:00
|
|
|
|
|
|
|
errorCode=U_ZERO_ERROR;
|
|
|
|
name=ucnv_getName(cnv, &errorCode);
|
|
|
|
if(U_FAILURE(errorCode)) {
|
|
|
|
return NULL;
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
2001-03-16 20:49:41 +00:00
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
for(i=0; i<(int32_t)(sizeof(ambiguousConverters)/sizeof(UAmbiguousConverter)); ++i)
|
|
|
|
{
|
|
|
|
if(0==uprv_strcmp(name, ambiguousConverters[i].name))
|
|
|
|
{
|
2001-03-16 20:49:41 +00:00
|
|
|
return ambiguousConverters+i;
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
|
|
|
}
|
2001-03-16 20:49:41 +00:00
|
|
|
|
|
|
|
return NULL;
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_fixFileSeparator(const UConverter *cnv,
|
|
|
|
UChar* source,
|
|
|
|
int32_t sourceLength) {
|
2001-03-16 20:49:41 +00:00
|
|
|
const UAmbiguousConverter *a;
|
|
|
|
int32_t i;
|
|
|
|
UChar variant5c;
|
|
|
|
|
2001-10-04 01:04:28 +00:00
|
|
|
if(cnv==NULL || source==NULL || sourceLength<=0 || (a=ucnv_getAmbiguous(cnv))==NULL)
|
|
|
|
{
|
2000-01-08 00:51:44 +00:00
|
|
|
return;
|
|
|
|
}
|
2001-03-16 20:49:41 +00:00
|
|
|
|
|
|
|
variant5c=a->variant5c;
|
|
|
|
for(i=0; i<sourceLength; ++i) {
|
|
|
|
if(source[i]==variant5c) {
|
|
|
|
source[i]=0x5c;
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UBool U_EXPORT2
|
|
|
|
ucnv_isAmbiguous(const UConverter *cnv) {
|
2001-03-16 20:49:41 +00:00
|
|
|
return (UBool)(ucnv_getAmbiguous(cnv)!=NULL);
|
2000-01-08 00:51:44 +00:00
|
|
|
}
|
2000-05-12 19:59:03 +00:00
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
|
|
|
ucnv_setFallback(UConverter *cnv, UBool usesFallback)
|
2000-05-12 19:59:03 +00:00
|
|
|
{
|
|
|
|
cnv->useFallback = usesFallback;
|
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI UBool U_EXPORT2
|
|
|
|
ucnv_usesFallback(const UConverter *cnv)
|
2000-05-12 19:59:03 +00:00
|
|
|
{
|
|
|
|
return cnv->useFallback;
|
|
|
|
}
|
2000-07-13 15:25:04 +00:00
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
2000-07-13 15:25:04 +00:00
|
|
|
ucnv_getInvalidChars (const UConverter * converter,
|
2000-08-11 01:27:17 +00:00
|
|
|
char *errBytes,
|
|
|
|
int8_t * len,
|
|
|
|
UErrorCode * err)
|
2000-07-13 15:25:04 +00:00
|
|
|
{
|
2000-07-15 01:11:05 +00:00
|
|
|
if (err == NULL || U_FAILURE(*err))
|
|
|
|
{
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (len == NULL || errBytes == NULL || converter == NULL)
|
|
|
|
{
|
|
|
|
*err = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
2000-07-15 20:09:58 +00:00
|
|
|
if (*len < converter->invalidCharLength)
|
2000-07-15 01:11:05 +00:00
|
|
|
{
|
|
|
|
*err = U_INDEX_OUTOFBOUNDS_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
2000-07-15 20:09:58 +00:00
|
|
|
if ((*len = converter->invalidCharLength) > 0)
|
2000-07-15 01:11:05 +00:00
|
|
|
{
|
2000-07-15 20:09:58 +00:00
|
|
|
uprv_memcpy (errBytes, converter->invalidCharBuffer, *len);
|
2000-07-15 01:11:05 +00:00
|
|
|
}
|
2000-07-13 15:25:04 +00:00
|
|
|
}
|
|
|
|
|
2001-11-21 01:02:11 +00:00
|
|
|
U_CAPI void U_EXPORT2
|
2000-07-13 15:25:04 +00:00
|
|
|
ucnv_getInvalidUChars (const UConverter * converter,
|
2000-08-11 01:27:17 +00:00
|
|
|
UChar *errChars,
|
|
|
|
int8_t * len,
|
|
|
|
UErrorCode * err)
|
2000-07-13 15:25:04 +00:00
|
|
|
{
|
2000-07-15 01:11:05 +00:00
|
|
|
if (err == NULL || U_FAILURE(*err))
|
|
|
|
{
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (len == NULL || errChars == NULL || converter == NULL)
|
|
|
|
{
|
|
|
|
*err = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
2000-07-15 20:09:58 +00:00
|
|
|
if (*len < converter->invalidUCharLength)
|
2000-07-15 01:11:05 +00:00
|
|
|
{
|
|
|
|
*err = U_INDEX_OUTOFBOUNDS_ERROR;
|
|
|
|
return;
|
|
|
|
}
|
2000-07-15 20:09:58 +00:00
|
|
|
if ((*len = converter->invalidUCharLength) > 0)
|
2000-07-15 01:11:05 +00:00
|
|
|
{
|
2000-07-15 20:09:58 +00:00
|
|
|
uprv_memcpy (errChars, converter->invalidUCharBuffer, sizeof(UChar) * (*len));
|
2000-07-15 01:11:05 +00:00
|
|
|
}
|
2000-07-13 15:25:04 +00:00
|
|
|
}
|
2000-08-10 01:32:29 +00:00
|
|
|
|
2002-11-08 01:28:14 +00:00
|
|
|
#define SIG_MAX_LEN 5
|
2002-01-08 01:05:57 +00:00
|
|
|
|
|
|
|
U_CAPI const char* U_EXPORT2
|
|
|
|
ucnv_detectUnicodeSignature( const char* source,
|
|
|
|
int32_t sourceLength,
|
|
|
|
int32_t* signatureLength,
|
2002-11-08 01:28:14 +00:00
|
|
|
UErrorCode* pErrorCode) {
|
|
|
|
int32_t dummy;
|
|
|
|
|
|
|
|
/* initial 0xa5 bytes: make sure that if we read <SIG_MAX_LEN
|
2002-01-08 01:05:57 +00:00
|
|
|
* bytes we don't misdetect something
|
|
|
|
*/
|
2002-11-08 01:28:14 +00:00
|
|
|
char start[SIG_MAX_LEN]={ '\xa5', '\xa5', '\xa5', '\xa5', '\xa5' };
|
2002-01-08 01:05:57 +00:00
|
|
|
int i = 0;
|
|
|
|
|
|
|
|
if((pErrorCode==NULL) || U_FAILURE(*pErrorCode)){
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2002-11-08 01:28:14 +00:00
|
|
|
if(source == NULL || sourceLength < -1){
|
2002-01-08 01:05:57 +00:00
|
|
|
*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2002-11-08 01:28:14 +00:00
|
|
|
if(signatureLength == NULL) {
|
|
|
|
signatureLength = &dummy;
|
|
|
|
}
|
|
|
|
|
2002-01-08 01:05:57 +00:00
|
|
|
if(sourceLength==-1){
|
2004-12-08 01:05:40 +00:00
|
|
|
sourceLength=(int32_t)uprv_strlen(source);
|
2002-01-08 01:05:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
while(i<sourceLength&& i<SIG_MAX_LEN){
|
|
|
|
start[i]=source[i];
|
|
|
|
i++;
|
|
|
|
}
|
2002-11-08 01:28:14 +00:00
|
|
|
|
2002-01-08 01:05:57 +00:00
|
|
|
if(start[0] == '\xFE' && start[1] == '\xFF') {
|
|
|
|
*signatureLength=2;
|
|
|
|
return "UTF-16BE";
|
|
|
|
} else if(start[0] == '\xFF' && start[1] == '\xFE') {
|
2002-04-17 20:36:53 +00:00
|
|
|
if(start[2] == '\x00' && start[3] =='\x00') {
|
2002-01-08 01:05:57 +00:00
|
|
|
*signatureLength=4;
|
|
|
|
return "UTF-32LE";
|
|
|
|
} else {
|
|
|
|
*signatureLength=2;
|
|
|
|
return "UTF-16LE";
|
|
|
|
}
|
|
|
|
} else if(start[0] == '\xEF' && start[1] == '\xBB' && start[2] == '\xBF') {
|
|
|
|
*signatureLength=3;
|
|
|
|
return "UTF-8";
|
2002-04-17 20:36:53 +00:00
|
|
|
} else if(start[0] == '\x00' && start[1] == '\x00' &&
|
|
|
|
start[2] == '\xFE' && start[3]=='\xFF') {
|
2002-01-08 01:05:57 +00:00
|
|
|
*signatureLength=4;
|
|
|
|
return "UTF-32BE";
|
2002-04-17 20:36:53 +00:00
|
|
|
} else if(start[0] == '\x0E' && start[1] == '\xFE' && start[2] == '\xFF') {
|
|
|
|
*signatureLength=3;
|
|
|
|
return "SCSU";
|
2002-05-25 00:32:25 +00:00
|
|
|
} else if(start[0] == '\xFB' && start[1] == '\xEE' && start[2] == '\x28') {
|
2002-04-17 20:36:53 +00:00
|
|
|
*signatureLength=3;
|
|
|
|
return "BOCU-1";
|
2002-11-08 01:28:14 +00:00
|
|
|
} else if(start[0] == '\x2B' && start[1] == '\x2F' && start[2] == '\x76') {
|
|
|
|
/*
|
|
|
|
* UTF-7: Initial U+FEFF is encoded as +/v8 or +/v9 or +/v+ or +/v/
|
|
|
|
* depending on the second UTF-16 code unit.
|
|
|
|
* Detect the entire, closed Unicode mode sequence +/v8- for only U+FEFF
|
|
|
|
* if it occurs.
|
|
|
|
*
|
|
|
|
* So far we have +/v
|
|
|
|
*/
|
|
|
|
if(start[3] == '\x38' && start[4] == '\x2D') {
|
|
|
|
/* 5 bytes +/v8- */
|
|
|
|
*signatureLength=5;
|
|
|
|
return "UTF-7";
|
|
|
|
} else if(start[3] == '\x38' || start[3] == '\x39' || start[3] == '\x2B' || start[3] == '\x2F') {
|
|
|
|
/* 4 bytes +/v8 or +/v9 or +/v+ or +/v/ */
|
|
|
|
*signatureLength=4;
|
|
|
|
return "UTF-7";
|
|
|
|
}
|
2004-07-16 21:48:33 +00:00
|
|
|
}else if(start[0]=='\xDD' && start[1]== '\x73'&& start[2]=='\x66' && start[3]=='\x73'){
|
|
|
|
*signatureLength=4;
|
|
|
|
return "UTF-EBCDIC";
|
2002-01-08 01:05:57 +00:00
|
|
|
}
|
2002-11-08 01:28:14 +00:00
|
|
|
|
2004-07-16 21:48:33 +00:00
|
|
|
|
2002-11-08 01:28:14 +00:00
|
|
|
/* no known Unicode signature byte sequence recognized */
|
|
|
|
*signatureLength=0;
|
|
|
|
return NULL;
|
2002-01-08 01:05:57 +00:00
|
|
|
}
|
|
|
|
|
2007-08-19 08:59:28 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
|
|
|
ucnv_fromUCountPending(const UConverter* cnv, UErrorCode* status)
|
|
|
|
{
|
2005-02-01 01:06:48 +00:00
|
|
|
if(status == NULL || U_FAILURE(*status)){
|
2005-01-28 02:45:30 +00:00
|
|
|
return -1;
|
|
|
|
}
|
2005-02-01 01:06:48 +00:00
|
|
|
if(cnv == NULL){
|
2005-01-28 02:45:30 +00:00
|
|
|
*status = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(cnv->preFromULength > 0){
|
|
|
|
return U16_LENGTH(cnv->preFromUFirstCP)+cnv->preFromULength ;
|
|
|
|
}else if(cnv->preFromULength < 0){
|
|
|
|
return -cnv->preFromULength ;
|
|
|
|
}else if(cnv->fromUChar32 > 0){
|
|
|
|
return 1;
|
|
|
|
}else if(cnv->preFromUFirstCP >0){
|
|
|
|
return U16_LENGTH(cnv->preFromUFirstCP);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
|
2007-08-19 08:59:28 +00:00
|
|
|
}
|
2005-01-28 02:45:30 +00:00
|
|
|
|
2007-08-19 08:59:28 +00:00
|
|
|
U_CAPI int32_t U_EXPORT2
|
2005-02-01 01:06:48 +00:00
|
|
|
ucnv_toUCountPending(const UConverter* cnv, UErrorCode* status){
|
2005-01-28 02:45:30 +00:00
|
|
|
|
2005-02-01 01:06:48 +00:00
|
|
|
if(status == NULL || U_FAILURE(*status)){
|
2005-01-28 02:45:30 +00:00
|
|
|
return -1;
|
|
|
|
}
|
2005-02-01 01:06:48 +00:00
|
|
|
if(cnv == NULL){
|
2005-01-28 02:45:30 +00:00
|
|
|
*status = U_ILLEGAL_ARGUMENT_ERROR;
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if(cnv->preToULength > 0){
|
|
|
|
return cnv->preToULength ;
|
|
|
|
}else if(cnv->preToULength < 0){
|
|
|
|
return -cnv->preToULength;
|
|
|
|
}else if(cnv->toULength > 0){
|
|
|
|
return cnv->toULength;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
2004-08-26 22:51:40 +00:00
|
|
|
#endif
|
|
|
|
|
2000-08-10 01:32:29 +00:00
|
|
|
/*
|
|
|
|
* Hey, Emacs, please set the following:
|
|
|
|
*
|
|
|
|
* Local Variables:
|
|
|
|
* indent-tabs-mode: nil
|
|
|
|
* End:
|
|
|
|
*
|
|
|
|
*/
|