scuffed-code/icu4c/source/common/ucnv.c

1116 lines
27 KiB
C
Raw Normal View History

1999-08-16 21:50:52 +00:00
/*
*******************************************************************************
*
* Copyright (C) 1998-1999, International Business Machines
* Corporation and others. All Rights Reserved.
*
*******************************************************************************
*
* ucnv.c:
* Implements APIs for the ICU's codeset conversion library
* mostly calls through internal functions created and maintained
* by Bertrand A. Damiba
*
* Modification History:
*
* Date Name Description
* 04/04/99 helena Fixed internal header inclusion.
*/
1999-08-16 21:50:52 +00:00
#include "umutex.h"
#include "unicode/ures.h"
1999-08-16 21:50:52 +00:00
#include "uhash.h"
#include "ucmp16.h"
#include "ucmp8.h"
#include "unicode/ucnv_bld.h"
1999-08-16 21:50:52 +00:00
#include "ucnv_io.h"
#include "unicode/ucnv_err.h"
1999-08-16 21:50:52 +00:00
#include "ucnv_cnv.h"
#include "ucnv_imp.h"
#include "unicode/ucnv.h"
1999-08-16 21:50:52 +00:00
#include "cmemory.h"
#include "cstring.h"
#include "unicode/ustring.h"
#include "unicode/uloc.h"
1999-08-16 21:50:52 +00:00
#define CHUNK_SIZE 5*1024
/* Internal function : begin */
static int32_t ucnv_getAmbiguousCCSID (const UConverter* cnv);
/* Internal function : end */
1999-08-16 21:50:52 +00:00
static void T_UConverter_fromCodepageToCodepage (UConverter * outConverter,
UConverter * inConverter,
char **target,
const char *targetLimit,
const char **source,
const char *sourceLimit,
int32_t* offsets,
bool_t flush,
UErrorCode * err);
const char* ucnv_getDefaultName ()
{
return ucnv_io_getDefaultConverterName();
1999-08-16 21:50:52 +00:00
}
void ucnv_setDefaultName (const char *converterName)
{
ucnv_io_setDefaultConverterName(converterName);
1999-08-16 21:50:52 +00:00
}
/*Calls through createConverter */
UConverter* ucnv_open (const char *name,
UErrorCode * err)
{
if (err == NULL || U_FAILURE (*err)) {
1999-08-16 21:50:52 +00:00
return NULL;
}
1999-08-16 21:50:52 +00:00
return createConverter (name, err);
1999-08-16 21:50:52 +00:00
}
/*Extracts the UChar* to a char* and calls through createConverter */
UConverter* ucnv_openU (const UChar * name,
UErrorCode * err)
{
char asciiName[UCNV_MAX_CONVERTER_NAME_LENGTH];
1999-08-16 21:50:52 +00:00
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return NULL;
if (name == NULL)
return ucnv_open (NULL, err);
if (u_strlen (name) > UCNV_MAX_CONVERTER_NAME_LENGTH)
1999-08-16 21:50:52 +00:00
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return NULL;
}
return ucnv_open (u_austrcpy (asciiName, name), err);
}
/*Assumes a $platform-#codepage.$CONVERTER_FILE_EXTENSION scheme and calls
*through createConverter*/
UConverter* ucnv_openCCSID (int32_t codepage,
UConverterPlatform platform,
1999-08-16 21:50:52 +00:00
UErrorCode * err)
{
char myName[UCNV_MAX_CONVERTER_NAME_LENGTH];
1999-08-16 21:50:52 +00:00
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return NULL;
copyPlatformString (myName, platform);
uprv_strcat (myName, "-");
T_CString_integerToString (myName + uprv_strlen (myName), codepage, 10);
1999-08-16 21:50:52 +00:00
return createConverter (myName, err);
}
/*Decreases the reference counter in the shared immutable section of the object
*and frees the mutable part*/
void ucnv_close (UConverter * converter)
{
if (converter == NULL)
return;
if (converter->sharedData->impl->close != NULL) {
converter->sharedData->impl->close(converter);
}
1999-08-16 21:50:52 +00:00
if (converter->sharedData->referenceCounter != ~0) {
umtx_lock (NULL);
if (converter->sharedData->referenceCounter != 0) {
converter->sharedData->referenceCounter--;
}
umtx_unlock (NULL);
}
uprv_free (converter);
1999-08-16 21:50:52 +00:00
return;
}
/*Frees all shared immutable objects that aren't referred to (reference count = 0)
*/
int32_t ucnv_flushCache ()
{
UConverterSharedData *mySharedData = NULL;
int32_t pos = -1;
int32_t tableDeletedNum = 0;
const UHashElement *e;
1999-08-16 21:50:52 +00:00
/*if shared data hasn't even been lazy evaluated yet
* return 0
*/
if (SHARED_DATA_HASHTABLE == NULL)
return 0;
/*creates an enumeration to iterate through every element in the
*table
*/
umtx_lock (NULL);
while ((e = uhash_nextElement (SHARED_DATA_HASHTABLE, &pos)) != NULL)
1999-08-16 21:50:52 +00:00
{
mySharedData = (UConverterSharedData *) e->value;
1999-08-16 21:50:52 +00:00
/*deletes only if reference counter == 0 */
if (mySharedData->referenceCounter == 0)
{
tableDeletedNum++;
uhash_removeElement(SHARED_DATA_HASHTABLE, e);
deleteSharedConverterData (mySharedData);
}
1999-08-16 21:50:52 +00:00
}
umtx_unlock (NULL);
1999-08-16 21:50:52 +00:00
return tableDeletedNum;
}
/*returns a single Name from the list, will return NULL if out of bounds
1999-08-16 21:50:52 +00:00
*/
const char* ucnv_getAvailableName (int32_t n)
1999-08-16 21:50:52 +00:00
{
if (0 <= n && n <= 0xffff) {
UErrorCode err = U_ZERO_ERROR;
const char *name = ucnv_io_getAvailableConverter((uint16_t)n, &err);
if (U_SUCCESS(err)) {
return name;
}
}
return NULL;
1999-08-16 21:50:52 +00:00
}
int32_t ucnv_countAvailable ()
{
UErrorCode err = U_ZERO_ERROR;
return ucnv_io_countAvailableConverters(&err);
1999-08-16 21:50:52 +00:00
}
U_CAPI uint16_t
ucnv_countAliases(const char *alias, UErrorCode *pErrorCode) {
const char *p;
return ucnv_io_getAliases(alias, &p, pErrorCode);
}
U_CAPI const char *
ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode) {
return ucnv_io_getAlias(alias, n, pErrorCode);
}
U_CAPI void
ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode) {
const char *p;
uint16_t count=ucnv_io_getAliases(alias, &p, pErrorCode);
while(count>0) {
*aliases++=p;
/* skip a name, first the canonical converter name */
p+=uprv_strlen(p)+1;
--count;
}
}
1999-08-16 21:50:52 +00:00
void ucnv_getSubstChars (const UConverter * converter,
char *mySubChar,
int8_t * len,
UErrorCode * err)
{
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return;
if (*len < converter->subCharLen) /*not enough space in subChars */
{
*err = U_INDEX_OUTOFBOUNDS_ERROR;
1999-08-16 21:50:52 +00:00
return;
}
uprv_memcpy (mySubChar, converter->subChar, converter->subCharLen); /*fills in the subchars */
1999-08-16 21:50:52 +00:00
*len = converter->subCharLen; /*store # of bytes copied to buffer */
return;
}
void ucnv_setSubstChars (UConverter * converter,
const char *mySubChar,
int8_t len,
UErrorCode * err)
{
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return;
/*Makes sure that the subChar is within the codepages char length boundaries */
if ((len > converter->sharedData->maxBytesPerChar)
|| (len < converter->sharedData->minBytesPerChar))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return;
}
uprv_memcpy (converter->subChar, mySubChar, len); /*copies the subchars */
1999-08-16 21:50:52 +00:00
converter->subCharLen = len; /*sets the new len */
return;
}
int32_t ucnv_getDisplayName (const UConverter * converter,
const char *displayLocale,
UChar * displayName,
int32_t displayNameCapacity,
UErrorCode * err)
{
UChar stringToWriteBuffer[UCNV_MAX_CONVERTER_NAME_LENGTH];
1999-08-16 21:50:52 +00:00
UChar const *stringToWrite;
int32_t stringToWriteLength;
UResourceBundle *rb = NULL;
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return 0;
/*create an RB, init the fill-in string, gets it from the RB */
rb = ures_open (NULL, displayLocale, err);
stringToWrite = ures_get (rb,
converter->sharedData->name,
err);
if (rb)
ures_close (rb);
if (U_SUCCESS (*err))
1999-08-16 21:50:52 +00:00
stringToWriteLength = u_strlen (stringToWrite);
else
{
/*Error While creating or getting resource from the resource bundle
*use the internal name instead
*
*sets stringToWriteLength (which accounts for a NULL terminator)
*and stringToWrite
*/
stringToWriteLength = uprv_strlen (converter->sharedData->name) + 1;
1999-08-16 21:50:52 +00:00
stringToWrite = u_uastrcpy (stringToWriteBuffer, converter->sharedData->name);
/*Hides the fallback to the internal name from the user */
if (*err == U_MISSING_RESOURCE_ERROR)
*err = U_ZERO_ERROR;
1999-08-16 21:50:52 +00:00
}
/*At this point we have a displayName and its length
*we want to see if it fits in the user provided params
*/
if (stringToWriteLength <= displayNameCapacity)
{
/*it fits */
u_strcpy (displayName, stringToWrite);
}
else
{
/*it doesn't fit */
*err = U_BUFFER_OVERFLOW_ERROR;
1999-08-16 21:50:52 +00:00
u_strncpy (displayName,
stringToWrite,
displayNameCapacity);
/*Zero terminates the string */
if (displayNameCapacity > 0)
displayName[displayNameCapacity - 1] = 0x0000;
}
/*if the user provided us with a with an outputLength
*buffer we'll store in it the theoretical size of the
*displayString
*/
return stringToWriteLength;
}
/*resets the internal states of a converter
*goal : have the same behaviour than a freshly created converter
*/
void ucnv_reset (UConverter * converter)
{
converter->toUnicodeStatus = converter->sharedData->defaultConverterValues.toUnicodeStatus;
converter->fromUnicodeStatus = 0;
converter->UCharErrorBufferLength = 0;
converter->charErrorBufferLength = 0;
if (converter->sharedData->impl->reset != NULL) {
converter->sharedData->impl->reset(converter);
} else {
converter->mode = UCNV_SI;
}
1999-08-16 21:50:52 +00:00
return;
}
int8_t ucnv_getMaxCharSize (const UConverter * converter)
{
return converter->sharedData->maxBytesPerChar;
}
int8_t ucnv_getMinCharSize (const UConverter * converter)
{
return converter->sharedData->minBytesPerChar;
}
const char* ucnv_getName (const UConverter * converter, UErrorCode * err)
{
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return NULL;
return converter->sharedData->name;
}
int32_t ucnv_getCCSID (const UConverter * converter,
UErrorCode * err)
{
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return -1;
return converter->sharedData->codepage;
}
UConverterPlatform ucnv_getPlatform (const UConverter * converter,
1999-08-16 21:50:52 +00:00
UErrorCode * err)
{
if (U_FAILURE (*err))
return UCNV_UNKNOWN;
1999-08-16 21:50:52 +00:00
return converter->sharedData->platform;
}
UConverterToUCallback ucnv_getToUCallBack (const UConverter * converter)
1999-08-16 21:50:52 +00:00
{
return converter->fromCharErrorBehaviour;
}
UConverterFromUCallback ucnv_getFromUCallBack (const UConverter * converter)
1999-08-16 21:50:52 +00:00
{
return converter->fromUCharErrorBehaviour;
}
UConverterToUCallback ucnv_setToUCallBack (UConverter * converter,
UConverterToUCallback action,
1999-08-16 21:50:52 +00:00
UErrorCode * err)
{
UConverterToUCallback myReturn = NULL;
1999-08-16 21:50:52 +00:00
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return NULL;
myReturn = converter->fromCharErrorBehaviour;
converter->fromCharErrorBehaviour = action;
return myReturn;
}
UConverterFromUCallback ucnv_setFromUCallBack (UConverter * converter,
UConverterFromUCallback action,
1999-08-16 21:50:52 +00:00
UErrorCode * err)
{
UConverterFromUCallback myReturn = NULL;
1999-08-16 21:50:52 +00:00
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return NULL;
myReturn = converter->fromUCharErrorBehaviour;
converter->fromUCharErrorBehaviour = action;
return myReturn;
}
void ucnv_fromUnicode (UConverter * _this,
char **target,
const char *targetLimit,
const UChar ** source,
const UChar * sourceLimit,
int32_t* offsets,
bool_t flush,
UErrorCode * err)
{
/*
* Check parameters in for all conversions
*/
if (U_FAILURE (*err)) return;
1999-08-16 21:50:52 +00:00
if ((_this == NULL) || ((char *) targetLimit < *target) || (sourceLimit < *source))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return;
}
/*
* Deal with stored carry over data. This is done in the common location
* to avoid doing it for each conversion.
*/
if (_this->charErrorBufferLength > 0)
{
int32_t myTargetIndex = 0;
flushInternalCharBuffer (_this,
(char *) *target,
&myTargetIndex,
targetLimit - *target,
offsets?&offsets:NULL,
err);
*target += myTargetIndex;
if (U_FAILURE (*err)) return;
1999-08-16 21:50:52 +00:00
}
if (offsets) {
if (_this->sharedData->impl->fromUnicodeWithOffsets != NULL) {
_this->sharedData->impl->fromUnicodeWithOffsets(_this,
target,
targetLimit,
source,
sourceLimit,
offsets,
flush,
err);
return;
} else {
/* all code points are of the same length */
int32_t targetSize = targetLimit - *target;
int32_t i, bytesPerChar = _this->sharedData->maxBytesPerChar;
if(bytesPerChar == 1) {
for (i=0; i<targetSize; i++) {
offsets[i] = i;
}
} else if(bytesPerChar == 2) {
for (i=0; i<targetSize; i++) {
offsets[i] = i>>1;
}
} else {
int32_t j = 0, k = bytesPerChar;
for (i=0; i<targetSize; i++) {
/* offsets[i] = i/bytesPerChar; -- without division */
offsets[i] = j;
if(--k == 0) {
k = bytesPerChar;
++j;
}
}
}
1999-08-16 21:50:52 +00:00
}
}
1999-08-16 21:50:52 +00:00
/*calls the specific conversion routines */
_this->sharedData->impl->fromUnicode(_this,
1999-08-16 21:50:52 +00:00
target,
targetLimit,
source,
sourceLimit,
offsets,
flush,
err);
return;
}
void ucnv_toUnicode (UConverter * _this,
UChar ** target,
const UChar * targetLimit,
const char **source,
const char *sourceLimit,
int32_t* offsets,
bool_t flush,
UErrorCode * err)
{
/*
* Check parameters in for all conversions
*/
if (U_FAILURE (*err)) return;
1999-08-16 21:50:52 +00:00
if ((_this == NULL) || ((UChar *) targetLimit < *target) || (sourceLimit < *source))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return;
}
/*
* Deal with stored carry over data. This is done in the common location
* to avoid doing it for each conversion.
*/
if (_this->UCharErrorBufferLength > 0)
{
int32_t myTargetIndex = 0;
flushInternalUnicodeBuffer (_this,
*target,
&myTargetIndex,
targetLimit - *target,
offsets?&offsets:NULL,
err);
*target += myTargetIndex;
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return;
}
if (offsets) {
if (_this->sharedData->impl->toUnicodeWithOffsets != NULL) {
_this->sharedData->impl->toUnicodeWithOffsets(_this,
target,
targetLimit,
source,
sourceLimit,
offsets,
flush,
err);
return;
} else {
/* all code points are of the same length */
1999-08-16 21:50:52 +00:00
int32_t targetSize = targetLimit - *target;
int32_t i, bytesPerChar = _this->sharedData->maxBytesPerChar;
if(bytesPerChar == 1) {
for (i=0; i<targetSize; i++) {
offsets[i] = i;
}
} else if(bytesPerChar == 2) {
for (i=0; i<targetSize; i++) {
offsets[i] = i<<1;
}
} else {
for (i=0; i<targetSize; i++) {
offsets[i] = i*bytesPerChar;
}
}
1999-08-16 21:50:52 +00:00
}
}
1999-08-16 21:50:52 +00:00
/*calls the specific conversion routines */
_this->sharedData->impl->toUnicode(_this,
1999-08-16 21:50:52 +00:00
target,
targetLimit,
source,
sourceLimit,
offsets,
flush,
err);
return;
}
int32_t ucnv_fromUChars (const UConverter * converter,
char *target,
int32_t targetSize,
const UChar * source,
int32_t sourceSize,
1999-08-16 21:50:52 +00:00
UErrorCode * err)
{
const UChar *mySource = source;
const UChar *mySource_limit;
int32_t mySourceLength = sourceSize;
1999-08-16 21:50:52 +00:00
UConverter myConverter;
char *myTarget = target;
char *myTarget_limit;
1999-08-16 21:50:52 +00:00
int32_t targetCapacity = 0;
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return 0;
if ((converter == NULL) || (targetSize < 0))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return 0;
}
/*makes a local copy of the UConverter */
myConverter = *converter;
/*Removes all state info on the UConverter */
ucnv_reset (&myConverter);
/*if the source is empty we return immediately */
if (sourceSize == -1) {
mySourceLength = u_strlen (source);
}
1999-08-16 21:50:52 +00:00
if (mySourceLength == 0)
{
/*for consistency we still need to
*store 0 in the targetCapacity
*if the user requires it
*/
return 0;
}
mySource_limit = mySource + mySourceLength;
myTarget_limit = target + targetSize;
/* Pin the limit to U_MAX_PTR. NULL check is for AS/400. */
if((myTarget_limit < target) || (myTarget_limit == NULL)) {
myTarget_limit = (char *)U_MAX_PTR;
}
1999-08-16 21:50:52 +00:00
if (targetSize > 0)
{
ucnv_fromUnicode (&myConverter,
&myTarget,
myTarget_limit,
1999-08-16 21:50:52 +00:00
&mySource,
mySource_limit,
NULL,
TRUE,
err);
targetCapacity = myTarget - target;
}
/*Updates targetCapacity to contain the number of bytes written to target */
if (targetSize == 0)
{
*err = U_INDEX_OUTOFBOUNDS_ERROR;
1999-08-16 21:50:52 +00:00
}
/* If the output buffer is exhausted, we need to stop writing
* to it but continue the conversion in order to store in targetSize
* the number of bytes that was required*/
if (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
char target2[CHUNK_SIZE];
char *target2_alias = target2;
const char *target2_limit = target2 + CHUNK_SIZE;
/*We use a stack allocated buffer around which we loop
*(in case the output is greater than CHUNK_SIZE)
*/
while (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
*err = U_ZERO_ERROR;
1999-08-16 21:50:52 +00:00
target2_alias = target2;
ucnv_fromUnicode (&myConverter,
&target2_alias,
target2_limit,
&mySource,
mySource_limit,
NULL,
TRUE,
err);
/*updates the output parameter to contain the number of char required */
targetCapacity += (target2_alias - target2) + 1;
}
/*We will set the erro code to U_BUFFER_OVERFLOW_ERROR only if
1999-08-16 21:50:52 +00:00
*nothing graver happened in the previous loop*/
(targetCapacity)--;
if (U_SUCCESS (*err))
*err = U_BUFFER_OVERFLOW_ERROR;
1999-08-16 21:50:52 +00:00
}
return targetCapacity;
}
int32_t ucnv_toUChars (const UConverter * converter,
UChar * target,
int32_t targetSize,
const char *source,
int32_t sourceSize,
UErrorCode * err)
{
const char *mySource = source;
const char *mySource_limit = source + sourceSize;
UConverter myConverter;
UChar *myTarget = target;
UChar *myTarget_limit;
1999-08-16 21:50:52 +00:00
int32_t targetCapacity;
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return 0;
if ((converter == NULL) || (targetSize < 0) || (sourceSize < 0))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return 0;
}
/*Means there is no work to be done */
if (sourceSize == 0)
{
/*for consistency we still need to
*store 0 in the targetCapacity
*if the user requires it
*/
if (targetSize >= 1)
{
target[0] = 0x0000;
return 1;
}
else
return 0;
}
/*makes a local copy of the UConverter */
myConverter = *converter;
/*Removes all state info on the UConverter */
ucnv_reset (&myConverter);
if (targetSize > 0)
{
myTarget_limit = target + targetSize - 1;
/* Pin the limit to U_MAX_PTR. NULL check is for AS/400. */
if ((myTarget_limit == NULL) || (myTarget_limit < target)) {
myTarget_limit = ((UChar*)U_MAX_PTR) - 1;
}
/*Not in pure pre-flight mode */
1999-08-16 21:50:52 +00:00
ucnv_toUnicode (&myConverter,
&myTarget,
myTarget_limit, /*Save a spot for the Null terminator */
1999-08-16 21:50:52 +00:00
&mySource,
mySource_limit,
NULL,
TRUE,
err);
/*Null terminates the string */
*(myTarget) = 0x0000;
}
/*Rigs targetCapacity to have at least one cell for zero termination */
/*Updates targetCapacity to contain the number of bytes written to target */
targetCapacity = 1;
targetCapacity += myTarget - target;
if (targetSize == 0)
{
*err = U_INDEX_OUTOFBOUNDS_ERROR;
1999-08-16 21:50:52 +00:00
}
/* If the output buffer is exhausted, we need to stop writing
* to it but if the input buffer is not exhausted,
* we need to continue the conversion in order to store in targetSize
* the number of bytes that was required
*/
if (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
UChar target2[CHUNK_SIZE];
UChar *target2_alias = target2;
const UChar *target2_limit = target2 + CHUNK_SIZE;
/*We use a stack allocated buffer around which we loop
(in case the output is greater than CHUNK_SIZE) */
while (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
*err = U_ZERO_ERROR;
1999-08-16 21:50:52 +00:00
target2_alias = target2;
ucnv_toUnicode (&myConverter,
&target2_alias,
target2_limit,
&mySource,
mySource_limit,
NULL,
TRUE,
err);
/*updates the output parameter to contain the number of char required */
targetCapacity += target2_alias - target2 + 1;
}
(targetCapacity)--; /*adjust for last one */
if (U_SUCCESS (*err))
*err = U_BUFFER_OVERFLOW_ERROR;
1999-08-16 21:50:52 +00:00
}
return targetCapacity;
}
UChar32 ucnv_getNextUChar (UConverter * converter,
1999-08-16 21:50:52 +00:00
const char **source,
const char *sourceLimit,
UErrorCode * err)
{
/* In case internal data had been stored
* we return the first UChar32 in the internal buffer,
1999-08-16 21:50:52 +00:00
* and update the internal state accordingly
*/
if (converter->UCharErrorBufferLength > 0)
{
UTextOffset i = 0;
UChar32 myUChar;
UTF_NEXT_CHAR(converter->UCharErrorBuffer, i, sizeof(converter->UCharErrorBuffer), myUChar);
1999-08-16 21:50:52 +00:00
/*In this memmove we update the internal buffer by
*popping the first character.
*Note that in the call itself we decrement
*UCharErrorBufferLength
*/
uprv_memmove (converter->UCharErrorBuffer,
converter->UCharErrorBuffer + i,
(converter->UCharErrorBufferLength - i) * sizeof (UChar));
converter->UCharErrorBufferLength -= (int8_t)i;
1999-08-16 21:50:52 +00:00
return myUChar;
}
/*calls the specific conversion routines */
/*as dictated in a code review, avoids a switch statement */
return converter->sharedData->impl->getNextUChar(converter,
1999-08-16 21:50:52 +00:00
source,
sourceLimit,
err);
}
/**************************
* Will convert a sequence of bytes from one codepage to another.
* @param toConverterName: The name of the converter that will be used to encode the output buffer
* @param fromConverterName: The name of the converter that will be used to decode the input buffer
* @param target: Pointer to the output buffer* written
* @param targetLength: on input contains the capacity of target, on output the number of bytes copied to target
* @param source: Pointer to the input buffer
* @param sourceLength: on input contains the capacity of source, on output the number of bytes processed in "source"
* @param internal: used internally to store store state data across calls
* @param err: fills in an error status
*/
void
T_UConverter_fromCodepageToCodepage (UConverter * outConverter,
UConverter * inConverter,
char **target,
const char *targetLimit,
const char **source,
const char *sourceLimit,
int32_t* offsets,
bool_t flush,
UErrorCode * err)
{
UChar out_chunk[CHUNK_SIZE];
const UChar *out_chunk_limit = out_chunk + CHUNK_SIZE;
UChar *out_chunk_alias;
UChar const *out_chunk_alias2;
if (U_FAILURE (*err)) return;
1999-08-16 21:50:52 +00:00
/*loops until the input buffer is completely consumed
*or if an error has be encountered
*first we convert from inConverter codepage to Unicode
*then from Unicode to outConverter codepage
*/
while ((*source != sourceLimit) && U_SUCCESS (*err))
1999-08-16 21:50:52 +00:00
{
out_chunk_alias = out_chunk;
ucnv_toUnicode (inConverter,
&out_chunk_alias,
out_chunk_limit,
source,
sourceLimit,
NULL,
flush,
err);
/*U_INDEX_OUTOFBOUNDS_ERROR means that the output "CHUNK" is full
1999-08-16 21:50:52 +00:00
*we will require at least another loop (it's a recoverable error)
*/
if (U_SUCCESS (*err) || (*err == U_INDEX_OUTOFBOUNDS_ERROR))
1999-08-16 21:50:52 +00:00
{
*err = U_ZERO_ERROR;
1999-08-16 21:50:52 +00:00
out_chunk_alias2 = out_chunk;
while ((out_chunk_alias2 != out_chunk_alias) && U_SUCCESS (*err))
1999-08-16 21:50:52 +00:00
{
ucnv_fromUnicode (outConverter,
target,
targetLimit,
&out_chunk_alias2,
out_chunk_alias,
NULL,
TRUE,
err);
}
}
else
break;
}
return;
}
int32_t ucnv_convert(const char *toConverterName,
const char *fromConverterName,
char *target,
int32_t targetSize,
const char *source,
int32_t sourceSize,
UErrorCode * err)
{
const char *mySource = source;
const char *mySource_limit = source + sourceSize;
UConverter *inConverter;
UConverter *outConverter;
char *myTarget = target;
int32_t targetCapacity = 0;
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
return 0;
if ((targetSize < 0) || (sourceSize < 0))
{
*err = U_ILLEGAL_ARGUMENT_ERROR;
1999-08-16 21:50:52 +00:00
return 0;
}
/*if there is no input data, we're done */
if (sourceSize == 0)
{
/*in case the caller passed an output ptr
*we update it
*/
return 0;
}
/*create the converters */
inConverter = ucnv_open (fromConverterName, err);
if (U_FAILURE (*err)) return 0;
1999-08-16 21:50:52 +00:00
outConverter = ucnv_open (toConverterName, err);
if (U_FAILURE (*err))
1999-08-16 21:50:52 +00:00
{
ucnv_close (inConverter);
return 0;
}
if (targetSize > 0)
{
T_UConverter_fromCodepageToCodepage (outConverter,
inConverter,
&myTarget,
target + targetSize,
&mySource,
mySource_limit,
NULL,
TRUE,
err);
}
/*Updates targetCapacity to contain the number of bytes written to target */
targetCapacity = myTarget - target;
if (targetSize == 0)
{
*err = U_INDEX_OUTOFBOUNDS_ERROR;
1999-08-16 21:50:52 +00:00
}
/* If the output buffer is exhausted, we need to stop writing
* to it but continue the conversion in order to store in targetSize
* the number of bytes that was required*/
if (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
char target2[CHUNK_SIZE];
char *target2_alias = target2;
const char *target2_limit = target2 + CHUNK_SIZE;
/*We use a stack allocated buffer around which we loop
*(in case the output is greater than CHUNK_SIZE)
*/
while (*err == U_INDEX_OUTOFBOUNDS_ERROR)
1999-08-16 21:50:52 +00:00
{
*err = U_ZERO_ERROR;
1999-08-16 21:50:52 +00:00
target2_alias = target2;
T_UConverter_fromCodepageToCodepage (outConverter,
inConverter,
&target2_alias,
target2_limit,
&mySource,
mySource_limit,
NULL,
TRUE,
err);
/*updates the output parameter to contain the number of char required */
targetCapacity += (target2_alias - target2) + 1;
}
/*We will set the erro code to U_BUFFER_OVERFLOW_ERROR only if
1999-08-16 21:50:52 +00:00
*nothing graver happened in the previous loop*/
(targetCapacity)--;
if (U_SUCCESS (*err))
*err = U_BUFFER_OVERFLOW_ERROR;
1999-08-16 21:50:52 +00:00
}
ucnv_close (inConverter);
ucnv_close (outConverter);
return targetCapacity;
}
UConverterType ucnv_getType(const UConverter* converter)
1999-08-16 21:50:52 +00:00
{
return converter->sharedData->conversionType;
}
void ucnv_getStarters(const UConverter* converter,
bool_t starters[256],
UErrorCode* err)
{
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
}
}
int32_t ucnv_getAmbiguousCCSID(const UConverter *cnv)
{
UErrorCode status = U_ZERO_ERROR;
int32_t i = 0;
int32_t ccsid = 0;
if (cnv == NULL)
{
return -1;
}
ccsid = ucnv_getCCSID(cnv, &status);
if (U_FAILURE(status))
{
return -1;
}
for (i = 0; i < UCNV_MAX_AMBIGUOUSCCSIDS; i++) {
if (ccsid == UCNV_AMBIGUOUSCONVERTERS[i].ccsid)
{
return i;
}
}
return -1;
}
void ucnv_fixFileSeparator(const UConverter *cnv,
UChar* source,
int32_t sourceLength)
{
int32_t i = 0;
int32_t offset = 0;
if ((source == NULL) || (cnv == NULL))
{
return;
}
if ((offset = ucnv_getAmbiguousCCSID(cnv)) != -1)
{
for (i = 0; i < sourceLength; i++)
{
if (source[i] == UCNV_AMBIGUOUSCONVERTERS[offset].mismapped)
{
source[i] = UCNV_AMBIGUOUSCONVERTERS[offset].replacement;
}
}
}
}
bool_t ucnv_isAmbiguous(const UConverter *cnv)
{
return (ucnv_getAmbiguousCCSID(cnv) == -1 ? FALSE : TRUE);
}