// ---------------------------------------------------------------------------
// XMLASCIITranscoder: Implementation of the transcoder API
// ---------------------------------------------------------------------------
XMLSize_t
XMLASCIITranscoder::transcodeFrom( const XMLByte* const srcData
, const XMLSize_t srcCount
, XMLCh* const toFill
, const XMLSize_t maxChars
, XMLSize_t& bytesEaten
, unsigned char* const charSizes)
{
//
// Calculate the max chars we can do here. Its the lesser of the
// max output chars and the source byte count.
//
const XMLSize_t countToDo = srcCount < maxChars ? srcCount : maxChars;
//
// Now loop through that many source chars and just cast each one
// over to the XMLCh format. Check each source that its really a
// valid ASCI char.
//
const XMLByte* srcPtr = srcData;
XMLCh* outPtr = toFill;
XMLSize_t countDone = 0;
for (; countDone < countToDo; countDone++)
{
// Do the optimistic work up front
if (*srcPtr < 0x80)
{
*outPtr++ = XMLCh(*srcPtr++);
continue;
}
//
// We got non source encoding char. If we got more than 32 chars,
// the just break out. We'll come back here later to hit this again
// and give an error much closer to the real source position.
//
if (countDone > 32)
break;
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Set the bytes we ate
bytesEaten = countDone;
// Set the char sizes to the fixed size
memset(charSizes, 1, countDone);
// Return the chars we transcoded
return countDone;
}
unsigned int
XMLASCIITranscoder::transcodeTo(const XMLCh* const srcData
, const unsigned int srcCount
, XMLByte* const toFill
, const unsigned int maxBytes
, unsigned int& charsEaten
, const UnRepOpts options)
{
// If debugging, make sure that the block size is legal
#if defined(XERCES_DEBUG)
checkBlockSize(maxBytes);
#endif
//
// Calculate the max chars we can do here. Its the lesser of the
// max output chars and the source byte count.
//
const unsigned int countToDo = srcCount < maxBytes ? srcCount : maxBytes;
const XMLCh* srcPtr = srcData;
XMLByte* outPtr = toFill;
for (unsigned int index = 0; index < countToDo; index++)
{
// If its legal, do it and jump back to the top
if (*srcPtr < 0x80)
{
*outPtr++ = XMLByte(*srcPtr++);
continue;
}
//
// Its not representable so use a replacement char. According to
// the options, either throw or use the replacement.
//
if (options == UnRep_Throw)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Use the replacement char
*outPtr++ = 0x1A;
srcPtr++;
}
// Set the chars we ate
charsEaten = countToDo;
// Return the byte we transcoded
return countToDo;
}
XMLSize_t
XML88591Transcoder::transcodeTo(const XMLCh* const srcData
, const XMLSize_t srcCount
, XMLByte* const toFill
, const XMLSize_t maxBytes
, XMLSize_t& charsEaten
, const UnRepOpts options)
{
//
// Calculate the max chars we can do here. Its the lesser of the
// max output bytes and the number of chars in the source.
//
const XMLSize_t countToDo = srcCount < maxBytes ? srcCount : maxBytes;
//
// Loop through the bytes to do and convert over each byte. Its just
// a downcast of the wide char, checking for unrepresentable chars.
//
const XMLCh* srcPtr = srcData;
const XMLCh* srcEnd = srcPtr + countToDo;
XMLByte* destPtr = toFill;
while (srcPtr < srcEnd)
{
// If its legal, take it and jump back to top
if (*srcPtr < 256)
{
*destPtr++ = XMLByte(*srcPtr++);
continue;
}
//
// Its not representable so use a replacement char. According to
// the options, either throw or use the replacement.
//
if (options == UnRep_Throw)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
*destPtr++ = 0x1A;
srcPtr++;
}
// Set the chars eaten
charsEaten = countToDo;
// Return the bytes we transcoded
return countToDo;
}
// ---------------------------------------------------------------------------
// XMLASCIITranscoder390: Implementation of the transcoder API
// ---------------------------------------------------------------------------
unsigned int
XMLASCIITranscoder390::transcodeFrom( const XMLByte* const srcData
, const unsigned int srcCount
, XMLCh* const toFill
, const unsigned int maxChars
, unsigned int& bytesEaten
, unsigned char* const charSizes)
{
// If debugging, make sure that the block size is legal
#if defined(XERCES_DEBUG)
checkBlockSize(maxChars);
#endif
//
// Calculate the max chars we can do here. Its the lesser of the
// max output chars and the source byte count.
//
const unsigned int countToDo = srcCount < maxChars ? srcCount : maxChars;
//
// Now loop through that many source chars and just cast each one
// over to the XMLCh format. Check each source that its really a
// valid ASCI char.
//
const XMLByte* srcPtr = srcData;
XMLCh* outPtr = toFill;
unsigned int countDone = countToDo;
int flag = 0;
// if flag is set to 1, an non-ASCII character is encountered
TROTASC(srcPtr, toFill, &countDone, padding_temp.gFromTable, 0xFFFF, &flag);
if (flag == 1 && countDone < 32){
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}//end if
// Set the bytes we ate
bytesEaten = countDone;
// Set the char sizes to the fixed size
memset(charSizes, 1, countDone);
// Return the chars we transcoded
return countDone;
}
XMLSize_t
MacOSTranscoder::transcodeTo(const XMLCh* const srcData
, const XMLSize_t srcCount
, XMLByte* const toFill
, const XMLSize_t maxBytes
, XMLSize_t& charsEaten
, const UnRepOpts options)
{
// Reset the tec state (since we don't know that we're part of a
// larger run of text).
TECClearConverterContextInfo(mUnicodeToText);
// Do the conversion
ByteCount bytesConsumed = 0;
ByteCount bytesProduced = 0;
OSStatus status = TECConvertText(mUnicodeToText,
(ConstTextPtr) srcData,
srcCount * sizeof(XMLCh), // inputBufferLength
&bytesConsumed, // actualInputLength
(TextPtr) toFill, // outputBuffer
maxBytes, // outputBufferLength
&bytesProduced); // actualOutputLength
// Ignorable error codes
if( status == kTECUsedFallbacksStatus
|| status == kTECOutputBufferFullStatus
|| status == kTECPartialCharErr
)
status = noErr;
std::size_t charsConsumed = bytesConsumed / sizeof(XMLCh);
// Deal with errors
if (status != noErr)
{
if (status == kTECUnmappableElementErr && options == UnRep_Throw)
{
XMLCh tmpBuf[17];
XMLString::binToText(srcData[charsConsumed], tmpBuf, 16, 16);
ThrowXML2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
);
}
}
charsEaten = charsConsumed;
return bytesProduced;
}
unsigned int
CygwinTranscoder::transcodeTo(const XMLCh* const srcData
, const unsigned int srcCount
, XMLByte* const toFill
, const unsigned int maxBytes
, unsigned int& charsEaten
, const UnRepOpts options)
{
// Get pointers to the start and end of each buffer
const XMLCh* srcPtr = srcData;
const XMLCh* srcEnd = srcData + srcCount;
XMLByte* outPtr = toFill;
XMLByte* outEnd = toFill + maxBytes;
//
// Now loop until we either get our max chars, or cannot get a whole
// character from the input buffer.
//
// NOTE: We have to use a loop for this unfortunately because the
// conversion API is too dumb to tell us how many chars it converted if
// it couldn't do the whole source.
//
BOOL usedDef;
while ((outPtr < outEnd) && (srcPtr < srcEnd))
{
//
// Do one char and see if it made it.
const unsigned int bytesStored = ::WideCharToMultiByte
(
fWinCP
, WC_COMPOSITECHECK | WC_SEPCHARS
, (LPCWSTR)srcPtr
, 1
, (char*)outPtr
, outEnd - outPtr
, 0
, &usedDef
);
// If we didn't transcode anything, then we are done
if (!bytesStored)
break;
//
// If the defaault char was used and the options indicate that
// this isn't allowed, then throw.
//
if (usedDef && (options == UnRep_Throw))
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Update our pointers
outPtr += bytesStored;
srcPtr++;
}
// Update the chars eaten
charsEaten = srcPtr - srcData;
// And return the bytes we stored
return outPtr - toFill;
}
// ---------------------------------------------------------------------------
// CygwinTranscoder: The virtual transcoder API
// ---------------------------------------------------------------------------
unsigned int
CygwinTranscoder::transcodeFrom( const XMLByte* const srcData
, const unsigned int srcCount
, XMLCh* const toFill
, const unsigned int maxChars
, unsigned int& bytesEaten
, unsigned char* const charSizes)
{
// Get temp pointers to the in and out buffers, and the chars sizes one
XMLCh* outPtr = toFill;
const XMLByte* inPtr = srcData;
unsigned char* sizesPtr = charSizes;
// Calc end pointers for each of them
XMLCh* outEnd = toFill + maxChars;
const XMLByte* inEnd = srcData + srcCount;
//
// Now loop until we either get our max chars, or cannot get a whole
// character from the input buffer.
//
bytesEaten = 0;
while ((outPtr < outEnd) && (inPtr < inEnd))
{
//
// If we are looking at a leading byte of a multibyte sequence,
// then we are going to eat 2 bytes, else 1.
//
const unsigned int toEat = ::IsDBCSLeadByteEx(fWinCP, *inPtr) ?
2 : 1;
// Make sure a whol char is in the source
if (inPtr + toEat > inEnd)
break;
// Try to translate this next char and check for an error
const unsigned int converted = ::MultiByteToWideChar
(
fWinCP
, MB_PRECOMPOSED | MB_ERR_INVALID_CHARS
, (const char*)inPtr
, toEat
, (LPWSTR)outPtr
, 1
);
if (converted != 1)
{
if (toEat == 1)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)(*inPtr), tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_BadSrcCP
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
else
{
ThrowXMLwithMemMgr(TranscodingException, XMLExcepts::Trans_BadSrcSeq, getMemoryManager());
}
}
// Update the char sizes array for this round
*sizesPtr++ = toEat;
// And update the bytes eaten count
bytesEaten += toEat;
// And update our in/out ptrs
inPtr += toEat;
outPtr++;
}
// Return the chars we output
return (outPtr - toFill);
}
unsigned int
XMLUTF8Transcoder::transcodeTo( const XMLCh* const srcData
, const unsigned int srcCount
, XMLByte* const toFill
, const unsigned int maxBytes
, unsigned int& charsEaten
, const UnRepOpts options)
{
// Watch for pathological scenario. Shouldn't happen, but...
if (!srcCount || !maxBytes)
return 0;
//
// Get pointers to our start and end points of the input and output
// buffers.
//
const XMLCh* srcPtr = srcData;
const XMLCh* srcEnd = srcPtr + srcCount;
XMLByte* outPtr = toFill;
XMLByte* outEnd = toFill + maxBytes;
while (srcPtr < srcEnd)
{
//
// Tentatively get the next char out. We have to get it into a
// 32 bit value, because it could be a surrogate pair.
//
XMLUInt32 curVal = *srcPtr;
//
// If its a leading surrogate, then lets see if we have the trailing
// available. If not, then give up now and leave it for next time.
//
unsigned int srcUsed = 1;
if ((curVal >= 0xD800) && (curVal <= 0xDBFF))
{
if (srcPtr + 1 >= srcEnd)
break;
// Create the composite surrogate pair
curVal = ((curVal - 0xD800) << 10)
+ ((*(srcPtr + 1) - 0xDC00) + 0x10000);
// And indicate that we ate another one
srcUsed++;
}
// Figure out how many bytes we need
unsigned int encodedBytes;
if (curVal < 0x80)
encodedBytes = 1;
else if (curVal < 0x800)
encodedBytes = 2;
else if (curVal < 0x10000)
encodedBytes = 3;
else if (curVal < 0x110000)
encodedBytes = 4;
else
{
// If the options say to throw, then throw
if (options == UnRep_Throw)
{
XMLCh tmpBuf[17];
XMLString::binToText(curVal, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Else, use the replacement character
*outPtr++ = chSpace;
srcPtr += srcUsed;
continue;
}
//
// If we cannot fully get this char into the output buffer,
// then leave it for the next time.
//
if (outPtr + encodedBytes > outEnd)
break;
// We can do it, so update the source index
srcPtr += srcUsed;
//
// And spit out the bytes. We spit them out in reverse order
// here, so bump up the output pointer and work down as we go.
//
outPtr += encodedBytes;
switch(encodedBytes)
{
case 6 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL);
curVal >>= 6;
case 5 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL);
curVal >>= 6;
case 4 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL);
curVal >>= 6;
case 3 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL);
curVal >>= 6;
case 2 : *--outPtr = XMLByte((curVal | 0x80UL) & 0xBFUL);
curVal >>= 6;
case 1 : *--outPtr = XMLByte
(
curVal | gFirstByteMark[encodedBytes]
);
}
// Add the encoded bytes back in again to indicate we've eaten them
outPtr += encodedBytes;
}
// Fill in the chars we ate
charsEaten = (srcPtr - srcData);
// And return the bytes we filled in
return (outPtr - toFill);
}
extern int
main(int argc, char* argv[]) {
const char *encoding = NULL;
const char *outputDir = NULL; /* NULL = no output directory, use current */
const char *inputDir = ".";
int tostdout = 0;
int prbom = 0;
const char *pname;
UResourceBundle *bundle = NULL;
UErrorCode status = U_ZERO_ERROR;
int32_t i = 0;
UConverter *converter;
const char* arg;
/* Get the name of tool. */
pname = uprv_strrchr(*argv, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (!pname) {
pname = uprv_strrchr(*argv, U_FILE_ALT_SEP_CHAR);
}
#endif
if (!pname) {
pname = *argv;
} else {
++pname;
}
/* error handling, printing usage message */
argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);
/* error handling, printing usage message */
if(argc<0) {
fprintf(stderr,
"%s: error in command line argument \"%s\"\n", pname,
argv[-argc]);
}
if(argc<0 || options[0].doesOccur || options[1].doesOccur) {
fprintf(argc < 0 ? stderr : stdout,
"%csage: %s [ -h, -?, --help ] [ -V, --version ]\n"
" [ -v, --verbose ] [ -e, --encoding encoding ] [ --bom ]\n"
" [ -t, --truncate [ size ] ]\n"
" [ -s, --sourcedir source ] [ -d, --destdir destination ]\n"
" [ -i, --icudatadir directory ] [ -c, --to-stdout ]\n"
" [ -A, --suppressAliases]\n"
" bundle ...\n", argc < 0 ? 'u' : 'U',
pname);
return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}
if(options[10].doesOccur) {
fprintf(stderr,
"%s version %s (ICU version %s).\n"
"%s\n",
pname, DERB_VERSION, U_ICU_VERSION, U_COPYRIGHT_STRING);
return U_ZERO_ERROR;
}
if(options[2].doesOccur) {
encoding = options[2].value;
}
if (options[3].doesOccur) {
tostdout = 1;
}
if(options[4].doesOccur) {
trunc = TRUE;
if(options[4].value != NULL) {
truncsize = atoi(options[4].value); /* user defined printable size */
} else {
truncsize = DERB_DEFAULT_TRUNC; /* we'll use default omitting size */
}
} else {
trunc = FALSE;
}
if(options[5].doesOccur) {
verbose = TRUE;
}
if (options[6].doesOccur) {
outputDir = options[6].value;
}
if(options[7].doesOccur) {
inputDir = options[7].value; /* we'll use users resources */
}
if (options[8].doesOccur) {
prbom = 1;
}
if (options[9].doesOccur) {
u_setDataDirectory(options[9].value);
}
if (options[11].doesOccur) {
suppressAliases = TRUE;
}
converter = ucnv_open(encoding, &status);
if (U_FAILURE(status)) {
fprintf(stderr, "%s: couldn't create %s converter for encoding\n", pname, encoding ? encoding : ucnv_getDefaultName());
return 2;
}
ucnv_setFromUCallBack(converter, UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_C, 0, 0, &status);
if (U_FAILURE(status)) {
fprintf(stderr, "%s: couldn't configure converter for encoding\n", pname);
return 3;
}
defaultConverter = ucnv_open(0, &status);
if (U_FAILURE(status)) {
fprintf(stderr, "%s: couldn't create %s converter for encoding\n", ucnv_getDefaultName(), pname);
return 2;
}
for (i = 1; i < argc; ++i) {
static const UChar sp[] = { 0x0020 }; /* " " */
char infile[4096]; /* XXX Sloppy. */
char locale[64];
const char *thename = 0, *p, *q;
UBool fromICUData = FALSE;
arg = getLongPathname(argv[i]);
if (verbose) {
printf("processing bundle \"%s\"\n", argv[i]);
}
p = uprv_strrchr(arg, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (p == NULL) {
p = uprv_strrchr(arg, U_FILE_ALT_SEP_CHAR);
}
#endif
if (!p) {
p = arg;
} else {
p++;
}
q = uprv_strrchr(p, '.');
if (!q) {
for (q = p; *q; ++q)
;
}
uprv_strncpy(locale, p, q - p);
locale[q - p] = 0;
if (!(fromICUData = !uprv_strcmp(inputDir, "-"))) {
UBool absfilename = *arg == U_FILE_SEP_CHAR;
#ifdef U_WINDOWS
if (!absfilename) {
absfilename = (uprv_strlen(arg) > 2 && isalpha(arg[0])
&& arg[1] == ':' && arg[2] == U_FILE_SEP_CHAR);
}
#endif
if (absfilename) {
thename = arg;
} else {
q = uprv_strrchr(arg, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (q == NULL) {
q = uprv_strrchr(arg, U_FILE_ALT_SEP_CHAR);
}
#endif
uprv_strcpy(infile, inputDir);
if(q != NULL) {
uprv_strcat(infile, U_FILE_SEP_STRING);
strncat(infile, arg, q-arg);
}
thename = infile;
}
}
status = U_ZERO_ERROR;
if (thename) {
bundle = ures_openDirect(thename, locale, &status);
} else {
bundle = ures_open(fromICUData ? 0 : inputDir, locale, &status);
}
if (status == U_ZERO_ERROR) {
FILE *out;
const char *filename = 0;
const char *ext = 0;
if (!locale || !tostdout) {
filename = uprv_strrchr(arg, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (!filename) {
filename = uprv_strrchr(arg, U_FILE_ALT_SEP_CHAR);
}
#endif
if (!filename) {
filename = arg;
} else {
++filename;
}
ext = uprv_strrchr(arg, '.');
if (!ext) {
ext = filename + uprv_strlen(filename);
}
}
if (tostdout) {
out = stdout;
#if defined(U_WINDOWS) || defined(U_CYGWIN)
if (setmode(fileno(out), O_BINARY) == -1) {
fprintf(stderr, "%s: couldn't set standard output to binary mode\n", pname);
return 4;
}
#endif
} else {
char thefile[4096], *tp;
int32_t len;
if (outputDir) {
uprv_strcpy(thefile, outputDir);
uprv_strcat(thefile, U_FILE_SEP_STRING);
} else {
*thefile = 0;
}
uprv_strcat(thefile, filename);
tp = thefile + uprv_strlen(thefile);
len = (int32_t)uprv_strlen(ext);
if (len) {
tp -= len - 1;
} else {
*tp++ = '.';
}
uprv_strcpy(tp, "txt");
out = fopen(thefile, "w");
if (!out) {
fprintf(stderr, "%s: couldn't create %s\n", pname, thefile);
return 4;
}
}
if (prbom) { /* XXX: Should be done only for UTFs */
static const UChar bom[] = { 0xFEFF };
printString(out, converter, bom, (int32_t)(sizeof(bom)/sizeof(*bom)));
}
printCString(out, converter, "// -*- Coding: ", -1);
printCString(out, converter, encoding ? encoding : getEncodingName(ucnv_getDefaultName()), -1);
printCString(out, converter, "; -*-\n//\n", -1);
printCString(out, converter, "// This file was dumped by derb(8) from ", -1);
if (thename) {
printCString(out, converter, thename, -1);
} else if (fromICUData) {
printCString(out, converter, "the ICU internal ", -1);
printCString(out, converter, locale, -1);
printCString(out, converter, " locale", -1);
}
printCString(out, converter, "\n// derb(8) by Vladimir Weinstein and Yves Arrouye\n\n", -1);
if (locale) {
printCString(out, converter, locale, -1);
} else {
printCString(out, converter, filename, (int32_t)(ext - filename));
printString(out, converter, sp, (int32_t)(sizeof(sp)/sizeof(*sp)));
}
printOutBundle(out, converter, bundle, 0, pname, &status);
if (out != stdout) {
fclose(out);
}
}
else {
reportError(pname, &status, "opening resource file");
}
ures_close(bundle);
}
ucnv_close(defaultConverter);
ucnv_close(converter);
return 0;
}
unsigned int
XML256TableTranscoder390::transcodeTo( const XMLCh* const srcData
, const unsigned int srcCount
, XMLByte* const toFill
, const unsigned int maxBytes
, unsigned int& charsEaten
, const UnRepOpts options)
{
// If debugging, make sure that the block size is legal
#if defined(XERCES_DEBUG)
checkBlockSize(maxBytes);
#endif
//
// Calculate the max chars we can do here. Its the lesser of the
// max output chars and the number of chars in the source.
//
const unsigned int countToDo = srcCount < maxBytes ? srcCount : maxBytes;
//
// Loop through the count we have to do and map each char via the
// lookup table.
//
const XMLCh* srcPtr = srcData;
const XMLCh* endPtr = (srcPtr + countToDo);
XMLByte* outPtr = toFill;
XMLByte nextOut;
while (srcPtr < endPtr)
{
//
// Get the next src char out to a temp, then do a binary search
// of the 'to' table for this entry.
//
if ((nextOut = xlatOneTo(*srcPtr)))
{
*outPtr++ = nextOut;
srcPtr++;
continue;
}
//
// Its not representable so, according to the options, either
// throw or use the replacement.
//
if (options == UnRep_Throw)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*srcPtr, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Eat the source char and use the replacement char
srcPtr++;
*outPtr++ = 0x3F;
}
// Set the chars eaten
charsEaten = countToDo;
// Return the bytes we transcoded
return countToDo;
}
unsigned int
ICUTranscoder::transcodeTo( const XMLCh* const srcData
, const unsigned int srcCount
, XMLByte* const toFill
, const unsigned int maxBytes
, unsigned int& charsEaten
, const UnRepOpts options)
{
//
// Get a pointer to the buffer to transcode. If UChar and XMLCh are
// the same size here, then use the original. Else, create a temp
// one and put a janitor on it.
//
const UChar* srcPtr;
UChar* tmpBufPtr = 0;
if (sizeof(XMLCh) == sizeof(UChar))
{
srcPtr = (const UChar*)srcData;
}
else
{
tmpBufPtr = convertToUChar(srcData, srcCount, getMemoryManager());
srcPtr = tmpBufPtr;
}
ArrayJanitor<UChar> janTmpBuf(tmpBufPtr, getMemoryManager());
//
// Set the appropriate callback so that it will either fail or use
// the rep char. Remember the old one so we can put it back.
//
UErrorCode err = U_ZERO_ERROR;
UConverterFromUCallback oldCB = NULL;
#if (U_ICU_VERSION_MAJOR_NUM < 2)
void* orgContent;
#else
const void* orgContent;
#endif
ucnv_setFromUCallBack
(
fConverter
, (options == UnRep_Throw) ? UCNV_FROM_U_CALLBACK_STOP
: UCNV_FROM_U_CALLBACK_SUBSTITUTE
, NULL
, &oldCB
, &orgContent
, &err
);
//
// Ok, lets transcode as many chars as we we can in one shot. The
// ICU API gives enough info not to have to do this one char by char.
//
XMLByte* startTarget = toFill;
const UChar* startSrc = srcPtr;
err = U_ZERO_ERROR;
ucnv_fromUnicode
(
fConverter
, (char**)&startTarget
, (char*)(startTarget + maxBytes)
, &startSrc
, srcPtr + srcCount
, 0
, false
, &err
);
// Rememember the status before we possibly overite the error code
const bool res = (err == U_ZERO_ERROR);
// Put the old handler back
err = U_ZERO_ERROR;
UConverterFromUCallback orgAction = NULL;
ucnv_setFromUCallBack(fConverter, oldCB, NULL, &orgAction, &orgContent, &err);
if (!res)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)*startSrc, tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_Unrepresentable
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
// Fill in the chars we ate from the input
charsEaten = startSrc - srcPtr;
// Return the chars we stored
return startTarget - toFill;
}
// ---------------------------------------------------------------------------
// ICUTranscoder: The virtual transcoder API
// ---------------------------------------------------------------------------
unsigned int
ICUTranscoder::transcodeFrom(const XMLByte* const srcData
, const unsigned int srcCount
, XMLCh* const toFill
, const unsigned int maxChars
, unsigned int& bytesEaten
, unsigned char* const charSizes)
{
// If debugging, insure the block size is legal
#if defined(XERCES_DEBUG)
checkBlockSize(maxChars);
#endif
// Set up pointers to the start and end of the source buffer
const XMLByte* startSrc = srcData;
const XMLByte* endSrc = srcData + srcCount;
//
// And now do the target buffer. This works differently according to
// whether XMLCh and UChar are the same size or not.
//
UChar* startTarget;
if (sizeof(XMLCh) == sizeof(UChar))
startTarget = (UChar*)toFill;
else
startTarget = (UChar*) getMemoryManager()->allocate
(
maxChars * sizeof(UChar)
);//new UChar[maxChars];
UChar* orgTarget = startTarget;
//
// Transoode the buffer. Buffer overflow errors are normal, occuring
// when the raw input buffer holds more characters than will fit in
// the Unicode output buffer.
//
UErrorCode err = U_ZERO_ERROR;
ucnv_toUnicode
(
fConverter
, &startTarget
, startTarget + maxChars
, (const char**)&startSrc
, (const char*)endSrc
, (fFixed ? 0 : (int32_t*)fSrcOffsets)
, false
, &err
);
if ((err != U_ZERO_ERROR) && (err != U_BUFFER_OVERFLOW_ERROR))
{
if (orgTarget != (UChar*)toFill)
getMemoryManager()->deallocate(orgTarget);//delete [] orgTarget;
if (fFixed)
{
XMLCh tmpBuf[17];
XMLString::binToText((unsigned int)(*startTarget), tmpBuf, 16, 16, getMemoryManager());
ThrowXMLwithMemMgr2
(
TranscodingException
, XMLExcepts::Trans_BadSrcCP
, tmpBuf
, getEncodingName()
, getMemoryManager()
);
}
else
{
ThrowXMLwithMemMgr(TranscodingException, XMLExcepts::Trans_BadSrcSeq, getMemoryManager());
}
}
// Calculate the bytes eaten and store in caller's param
bytesEaten = startSrc - srcData;
// And the characters decoded
const unsigned int charsDecoded = startTarget - orgTarget;
//
// Translate the array of char offsets into an array of character
// sizes, which is what the transcoder interface semantics requires.
// If its fixed, then we can optimize it.
//
if (fFixed)
{
const unsigned char fillSize = (unsigned char)ucnv_getMaxCharSize(fConverter);
memset(charSizes, fillSize, maxChars);
}
else
{
//
// We have to convert the series of offsets into a series of
// sizes. If just one char was decoded, then its the total bytes
// eaten. Otherwise, do a loop and subtract out each element from
// its previous element.
//
if (charsDecoded == 1)
{
charSizes[0] = (unsigned char)bytesEaten;
}
else
{
// ICU does not return an extra element to allow us to figure
// out the last char size, so we have to compute it from the
// total bytes used.
unsigned int index;
for (index = 0; index < charsDecoded - 1; index++)
{
charSizes[index] = (unsigned char)(fSrcOffsets[index + 1]
- fSrcOffsets[index]);
}
if( charsDecoded > 0 ) {
charSizes[charsDecoded - 1] = (unsigned char)(bytesEaten
- fSrcOffsets[charsDecoded - 1]);
}
}
}
//
// If XMLCh and UChar are not the same size, then we need to copy over
// the temp buffer to the new one.
//
if (sizeof(UChar) != sizeof(XMLCh))
{
XMLCh* outPtr = toFill;
startTarget = orgTarget;
for (unsigned int index = 0; index < charsDecoded; index++)
*outPtr++ = XMLCh(*startTarget++);
// And delete the temp buffer
getMemoryManager()->deallocate(orgTarget);//delete [] orgTarget;
}
// Return the chars we put into the target buffer
return charsDecoded;
}
extern int
main(int argc, char* argv[]) {
const char *encoding = NULL;
const char *outputDir = NULL; /* NULL = no output directory, use current */
const char *inputDir = ".";
int tostdout = 0;
int prbom = 0;
const char *pname;
UResourceBundle *bundle = NULL;
int32_t i = 0;
const char* arg;
/* Get the name of tool. */
pname = uprv_strrchr(*argv, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (!pname) {
pname = uprv_strrchr(*argv, U_FILE_ALT_SEP_CHAR);
}
#endif
if (!pname) {
pname = *argv;
} else {
++pname;
}
/* error handling, printing usage message */
argc=u_parseArgs(argc, argv, UPRV_LENGTHOF(options), options);
/* error handling, printing usage message */
if(argc<0) {
fprintf(stderr,
"%s: error in command line argument \"%s\"\n", pname,
argv[-argc]);
}
if(argc<0 || options[0].doesOccur || options[1].doesOccur) {
fprintf(argc < 0 ? stderr : stdout,
"%csage: %s [ -h, -?, --help ] [ -V, --version ]\n"
" [ -v, --verbose ] [ -e, --encoding encoding ] [ --bom ]\n"
" [ -t, --truncate [ size ] ]\n"
" [ -s, --sourcedir source ] [ -d, --destdir destination ]\n"
" [ -i, --icudatadir directory ] [ -c, --to-stdout ]\n"
" [ -A, --suppressAliases]\n"
" bundle ...\n", argc < 0 ? 'u' : 'U',
pname);
return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}
if(options[10].doesOccur) {
fprintf(stderr,
"%s version %s (ICU version %s).\n"
"%s\n",
pname, DERB_VERSION, U_ICU_VERSION, U_COPYRIGHT_STRING);
return U_ZERO_ERROR;
}
if(options[2].doesOccur) {
encoding = options[2].value;
}
if (options[3].doesOccur) {
if(options[2].doesOccur) {
fprintf(stderr, "%s: Error: don't specify an encoding (-e) when writing to stdout (-c).\n", pname);
return 3;
}
tostdout = 1;
}
if(options[4].doesOccur) {
opt_truncate = TRUE;
if(options[4].value != NULL) {
truncsize = atoi(options[4].value); /* user defined printable size */
} else {
truncsize = DERB_DEFAULT_TRUNC; /* we'll use default omitting size */
}
} else {
opt_truncate = FALSE;
}
if(options[5].doesOccur) {
verbose = TRUE;
}
if (options[6].doesOccur) {
outputDir = options[6].value;
}
if(options[7].doesOccur) {
inputDir = options[7].value; /* we'll use users resources */
}
if (options[8].doesOccur) {
prbom = 1;
}
if (options[9].doesOccur) {
u_setDataDirectory(options[9].value);
}
if (options[11].doesOccur) {
suppressAliases = TRUE;
}
fflush(stderr); // use ustderr now.
ustderr = u_finit(stderr, NULL, NULL);
for (i = 1; i < argc; ++i) {
static const UChar sp[] = { 0x0020 }; /* " " */
arg = getLongPathname(argv[i]);
if (verbose) {
u_fprintf(ustderr, "processing bundle \"%s\"\n", argv[i]);
}
icu::CharString locale;
UErrorCode status = U_ZERO_ERROR;
{
const char *p = findBasename(arg);
const char *q = uprv_strrchr(p, '.');
if (q == NULL) {
locale.append(p, status);
} else {
locale.append(p, (int32_t)(q - p), status);
}
}
if (U_FAILURE(status)) {
return status;
}
icu::CharString infile;
const char *thename = NULL;
UBool fromICUData = !uprv_strcmp(inputDir, "-");
if (!fromICUData) {
UBool absfilename = *arg == U_FILE_SEP_CHAR;
#if U_PLATFORM_HAS_WIN32_API
if (!absfilename) {
absfilename = (uprv_strlen(arg) > 2 && isalpha(arg[0])
&& arg[1] == ':' && arg[2] == U_FILE_SEP_CHAR);
}
#endif
if (absfilename) {
thename = arg;
} else {
const char *q = uprv_strrchr(arg, U_FILE_SEP_CHAR);
#if U_FILE_SEP_CHAR != U_FILE_ALT_SEP_CHAR
if (q == NULL) {
q = uprv_strrchr(arg, U_FILE_ALT_SEP_CHAR);
}
#endif
infile.append(inputDir, status);
if(q != NULL) {
infile.appendPathPart(icu::StringPiece(arg, (int32_t)(q - arg)), status);
}
if (U_FAILURE(status)) {
return status;
}
thename = infile.data();
}
}
if (thename) {
bundle = ures_openDirect(thename, locale.data(), &status);
} else {
bundle = ures_open(fromICUData ? 0 : inputDir, locale.data(), &status);
}
if (U_SUCCESS(status)) {
UFILE *out = NULL;
const char *filename = 0;
const char *ext = 0;
if (locale.isEmpty() || !tostdout) {
filename = findBasename(arg);
ext = uprv_strrchr(filename, '.');
if (!ext) {
ext = uprv_strchr(filename, 0);
}
}
if (tostdout) {
out = u_get_stdout();
} else {
icu::CharString thefile;
if (outputDir) {
thefile.append(outputDir, status);
}
thefile.appendPathPart(filename, status);
if (*ext) {
thefile.truncate(thefile.length() - (int32_t)uprv_strlen(ext));
}
thefile.append(".txt", status);
if (U_FAILURE(status)) {
return status;
}
out = u_fopen(thefile.data(), "w", NULL, encoding);
if (!out) {
u_fprintf(ustderr, "%s: couldn't create %s\n", pname, thefile.data());
u_fclose(ustderr);
return 4;
}
}
// now, set the callback.
ucnv_setFromUCallBack(u_fgetConverter(out), UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_C, 0, 0, &status);
if (U_FAILURE(status)) {
u_fprintf(ustderr, "%s: couldn't configure converter for encoding\n", pname);
u_fclose(ustderr);
if(!tostdout) {
u_fclose(out);
}
return 3;
}
if (prbom) { /* XXX: Should be done only for UTFs */
u_fputc(0xFEFF, out);
}
u_fprintf(out, "// -*- Coding: %s; -*-\n//\n", encoding ? encoding : getEncodingName(ucnv_getDefaultName()));
u_fprintf(out, "// This file was dumped by derb(8) from ");
if (thename) {
u_fprintf(out, "%s", thename);
} else if (fromICUData) {
u_fprintf(out, "the ICU internal %s locale", locale.data());
}
u_fprintf(out, "\n// derb(8) by Vladimir Weinstein and Yves Arrouye\n\n");
if (!locale.isEmpty()) {
u_fprintf(out, "%s", locale.data());
} else {
u_fprintf(out, "%.*s%.*S", (int32_t)(ext - filename), filename, UPRV_LENGTHOF(sp), sp);
}
printOutBundle(out, bundle, 0, pname, &status);
if (!tostdout) {
u_fclose(out);
}
}
else {
reportError(pname, &status, "opening resource file");
}
ures_close(bundle);
}
return 0;
}
std::string ImageFormat::toString() const
{
std::stringstream stream;
stream << getWidth() << "x" << getHeight() << " pixels, " << getEncodingName() << " encoding";
return stream.str();
}