macosunicodeconverter.cpp

来自「IBM的解析xml的工具Xerces的源代码」· C++ 代码 · 共 1,049 行 · 第 1/3 页

bool MacOSUnicodeConverter::supportsSrcOfs() const
{
	// For now, we don't support source offsets
    return false;
}


void MacOSUnicodeConverter::upperCase(XMLCh* const toUpperCase) const
{
#if TARGET_API_MAC_CARBON

   // If we're targeting carbon, use the CFString conversion to uppercase
   int len = XMLString::stringLen(toUpperCase);
   CFMutableStringRef cfString = CFStringCreateMutableWithExternalCharactersNoCopy(
        kCFAllocatorDefault,
        (UniChar*)toUpperCase,
        len,		// length
        len,		// capacity
        kCFAllocatorNull);
   CFStringUppercase(cfString, NULL);
   CFRelease(cfString);

#elif defined(XML_METROWERKS) || (__GNUC__ >= 3 && _GLIBCPP_USE_WCHAR_T)

	// Use this if there's a reasonable c library available.
	// Metrowerks does this reasonably
	wchar_t c;
	for (XMLCh* p = (XMLCh*)toUpperCase; ((c = *p) != 0); )
		*p++ = std::towupper(c);

#else
	#error Sorry, no support for upperCase
#endif
}


void MacOSUnicodeConverter::lowerCase(XMLCh* const toLowerCase) const
{
#if TARGET_API_MAC_CARBON

   // If we're targeting carbon, use the CFString conversion to uppercase
   int len = XMLString::stringLen(toLowerCase);
   CFMutableStringRef cfString = CFStringCreateMutableWithExternalCharactersNoCopy(
        kCFAllocatorDefault,
        (UniChar*)toLowerCase,
        len,		// length
        len,		// capacity
        kCFAllocatorNull);
   CFStringLowercase(cfString, NULL);
   CFRelease(cfString);

#elif defined(XML_METROWERKS) || (__GNUC__ >= 3 && _GLIBCPP_USE_WCHAR_T)

	// Use this if there's a reasonable c library available.
	// Metrowerks does this reasonably
	wchar_t c;
	for (XMLCh* p = (XMLCh*)toLowerCase; ((c = *p) != 0); )
		*p++ = std::towlower(c);

#else
	#error Sorry, no support for lowerCase
#endif
}


void
MacOSUnicodeConverter::ConvertWideToNarrow(const XMLCh* wide, char* narrow, std::size_t maxChars)
{
	while (maxChars-- > 0)
		if ((*narrow++ = *wide++) == 0)
			break;
}


// ---------------------------------------------------------------------------
//  MacOSTransService: The protected virtual transcoding service API
// ---------------------------------------------------------------------------
XMLTranscoder*
MacOSUnicodeConverter::makeNewXMLTranscoder(const   XMLCh* const		encodingName
                                        ,       XMLTransService::Codes& resValue
                                        , const unsigned int			blockSize
                                        ,       MemoryManager* const    manager)
{
	XMLTranscoder* result = NULL;
	resValue = XMLTransService::Ok;
	
	TextToUnicodeInfo textToUnicodeInfo = NULL;
	UnicodeToTextInfo unicodeToTextInfo = NULL;

	//	Map the encoding to a Mac OS Encoding value
	Str255 pasEncodingName;
	char cEncodingName[256];
	ConvertWideToNarrow(encodingName, cEncodingName, sizeof(cEncodingName));
	CopyCStringToPascal(cEncodingName, pasEncodingName);
	
	TextEncoding textEncoding = 0;
	OSStatus status = TECGetTextEncodingFromInternetName (
							&textEncoding,
							pasEncodingName);
                            
    //  Make a transcoder for that encoding
	if (status == noErr)
		result = makeNewXMLTranscoder(encodingName, resValue, blockSize, textEncoding, manager);
	else
		resValue = XMLTransService::UnsupportedEncoding;
	
	return result;
}


XMLTranscoder*
MacOSUnicodeConverter::makeNewXMLTranscoder(const   XMLCh* const		encodingName
                                        ,       XMLTransService::Codes& resValue
                                        , const unsigned int			blockSize
										,		TextEncoding            textEncoding
                                        ,       MemoryManager* const    manager)
{
    XMLTranscoder* result = NULL;
	resValue = XMLTransService::Ok;
    OSStatus status = noErr;
    
    TECObjectRef textToUnicode = NULL;
    TECObjectRef unicodeToText = NULL;
    
    //  We convert to and from utf16
    TextEncoding utf16Encoding = CreateTextEncoding(kTextEncodingUnicodeDefault,
                                        kTextEncodingDefaultVariant,
                                        kUnicode16BitFormat);

    //  Create a TEC from our encoding to utf16
    if (status == noErr)
        status = TECCreateConverter(&textToUnicode, textEncoding, utf16Encoding);

    //  Create a TEC from utf16 to our encoding
    if (status == noErr)
        status = TECCreateConverter(&unicodeToText, utf16Encoding, textEncoding);

	if (status != noErr)
	{
        //  Clean up on error
		if (textToUnicode != NULL)
            TECDisposeConverter(textToUnicode);
			
		if (unicodeToText != NULL)
            TECDisposeConverter(unicodeToText);

		resValue = XMLTransService::UnsupportedEncoding;
	}
	else
    {
        //  Create our transcoder, passing in the converters
		result = new (manager) MacOSTranscoder(encodingName, textToUnicode, unicodeToText, blockSize, manager);
    }
	
    return result;
}


// ---------------------------------------------------------------------------
//  IsMacOSUnicodeConverterSupported
// ---------------------------------------------------------------------------
bool
MacOSUnicodeConverter::IsMacOSUnicodeConverterSupported(void)
{
    return UpgradeScriptInfoToTextEncoding != (void*)kUnresolvedCFragSymbolAddress
        && CreateTextToUnicodeInfoByEncoding != (void*)kUnresolvedCFragSymbolAddress
        ;
}


// ---------------------------------------------------------------------------
//  MacOSTranscoder: Constructors and Destructor
// ---------------------------------------------------------------------------
MacOSTranscoder::MacOSTranscoder(const  XMLCh* const    encodingName
								, TECObjectRef          textToUnicode
								, TECObjectRef          unicodeToText
                                , const unsigned int    blockSize
                                , MemoryManager* const  manager) :
    XMLTranscoder(encodingName, blockSize, manager),
    mTextToUnicode(textToUnicode),
    mUnicodeToText(unicodeToText)
{
}


MacOSTranscoder::~MacOSTranscoder()
{
	//	Dispose our text encoding converters
	TECDisposeConverter(mTextToUnicode);
	TECDisposeConverter(mUnicodeToText);
}


// ---------------------------------------------------------------------------
//  MacOSTranscoder: The virtual transcoder API
// ---------------------------------------------------------------------------

unsigned int
MacOSTranscoder::transcodeFrom(  const  XMLByte* const			srcData
                                , const unsigned int			srcCount
                                ,       XMLCh* const			toFill
                                , const unsigned int			maxChars
                                ,       unsigned int&			bytesEaten
                                ,       unsigned char* const	charSizes)
{
	//  Reset the tec state (since we don't know that we're part of a
	//  larger run of text).
	TECClearConverterContextInfo(mTextToUnicode);
	
    //  Do the conversion
    ByteCount bytesConsumed = 0;
    ByteCount bytesProduced = 0;
    OSStatus status = TECConvertText(mTextToUnicode,
                (ConstTextPtr) srcData,
                srcCount,                   // inputBufferLength
                &bytesConsumed,				// actualInputLength
                (TextPtr) toFill,           // outputBuffer
                maxChars * sizeof(XMLCh),	// outputBufferLength
                &bytesProduced);			// actualOutputLength

    //  Ignorable error codes
    if(    status == kTECUsedFallbacksStatus
        || status == kTECOutputBufferFullStatus
        || status == kTECPartialCharErr
		)
        status = noErr;
    	
    if (status != noErr)
        ThrowXML(TranscodingException, XMLExcepts::Trans_BadSrcSeq);
	
	std::size_t charsProduced = bytesProduced / sizeof(XMLCh);
	
    bytesEaten = bytesConsumed;
    return charsProduced;
}


unsigned int
MacOSTranscoder::transcodeTo(const  XMLCh* const    srcData
                            , const unsigned int    srcCount
                            ,       XMLByte* const  toFill
                            , const unsigned int    maxBytes
                            ,       unsigned int&   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((unsigned int)&srcData[charsConsumed], tmpBuf, 16, 16);
            ThrowXML2
            (
                TranscodingException
                , XMLExcepts::Trans_Unrepresentable
                , tmpBuf
                , getEncodingName()
            );
    	}
    }
	
    charsEaten = charsConsumed;
    return bytesProduced;
}


bool
MacOSTranscoder::canTranscodeTo(const unsigned int toCheck) const
{
	//
    //  If the passed value is really a surrogate embedded together, then
    //  we need to break it out into its two chars. Else just one.
    //
    unsigned int    srcCnt = 0;
    UniChar         srcBuf[2];

    if (toCheck & 0xFFFF0000)
    {
        srcBuf[srcCnt++] = XMLCh(toCheck >> 10)   + 0xD800;
        srcBuf[srcCnt++] = XMLCh(toCheck & 0x3FF) + 0xDC00;
    }
    else
    {
        srcBuf[srcCnt++] = XMLCh(toCheck);
    }

	//  Clear the converter state: we're in a new run of text
	TECClearConverterContextInfo(mUnicodeToText);

    //
    //  Use a local temp buffer that would hold any sane multi-byte char
    //  sequence and try to transcode this guy into it.
    //
    char tmpBuf[64];

    ByteCount bytesConsumed = 0;
    ByteCount bytesProduced = 0;
    OSStatus status = TECConvertText(mUnicodeToText,
                (ConstTextPtr) srcBuf,
                srcCnt * sizeof(XMLCh),     // inputBufferLength
                &bytesConsumed,				// actualInputLength
                (TextPtr) tmpBuf,           // outputBuffer
                sizeof(tmpBuf),             // outputBufferLength
                &bytesProduced);			// actualOutputLength

    std::size_t charsConsumed = bytesConsumed / sizeof(XMLCh);
	
	//	Return true if we transcoded the character(s)
	//	successfully
	return status == noErr && charsConsumed == srcCnt;
}


// ---------------------------------------------------------------------------
//  MacOSLCPTranscoder: Constructors and Destructor
// ---------------------------------------------------------------------------
MacOSLCPTranscoder::MacOSLCPTranscoder(XMLTranscoder* const transcoder, MemoryManager* const manager)
 : mTranscoder(transcoder),
   mManager(manager)
{
}