domwriterimpl.cpp

来自「IBM的解析xml的工具Xerces的源代码」· C++ 代码 · 共 1,852 行 · 第 1/5 页

    return (DOMNodeFilter::FilterAction) fFilter->acceptNode(node);
}


bool DOMWriterImpl::checkFeature(const XMLCh* const featName
                               , bool               toThrow
                               , int&               featureId) const
{
    // check for null and/or empty feature name
    if (!featName || !*featName)
    {
        if (toThrow)
            throw DOMException(DOMException::NOT_FOUND_ERR, 0, fMemoryManager);

        return false;
    }

    featureId = INVALID_FEATURE_ID;

    if (XMLString::equals(featName, XMLUni::fgDOMWRTCanonicalForm))
        featureId = CANONICAL_FORM_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTDiscardDefaultContent))
        featureId = DISCARD_DEFAULT_CONTENT_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTEntities))
        featureId = ENTITIES_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTFormatPrettyPrint))
        featureId = FORMAT_PRETTY_PRINT_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTNormalizeCharacters))
        featureId = NORMALIZE_CHARACTERS_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTSplitCdataSections))
        featureId = SPLIT_CDATA_SECTIONS_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTValidation))
        featureId = VALIDATION_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTWhitespaceInElementContent))
        featureId = WHITESPACE_IN_ELEMENT_CONTENT_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMWRTBOM))
        featureId = BYTE_ORDER_MARK_ID;
    else if (XMLString::equals(featName, XMLUni::fgDOMXMLDeclaration))
        featureId = XML_DECLARATION;


    //feature name not resolvable
    if (featureId == INVALID_FEATURE_ID)
    {
        if (toThrow)
            throw DOMException(DOMException::NOT_FOUND_ERR, featName, fMemoryManager);

        return false;
    }

    return true;
}

bool DOMWriterImpl::reportError(const DOMNode* const    errorNode
                              , DOMError::ErrorSeverity errorType
                              , const XMLCh*   const    errorMsg)
{
    bool toContinueProcess = true;   // default value for no error handler

    if (fErrorHandler)
    {
        DOMLocatorImpl  locator(0, 0, (DOMNode* const) errorNode, 0, 0);
        DOMErrorImpl    domError(errorType , errorMsg, &locator);
        toContinueProcess = fErrorHandler->handleError(domError);
    }

    if (errorType != DOMError::DOM_SEVERITY_WARNING)
        fErrorCount++;

    return toContinueProcess;
}

bool DOMWriterImpl::reportError(const DOMNode* const    errorNode
                              , DOMError::ErrorSeverity errorType
                              , XMLDOMMsg::Codes        toEmit)
{
    const unsigned int msgSize = 1023;
    XMLCh errText[msgSize + 1];

    DOMImplementationImpl::getMsgLoader4DOM()->loadMsg(toEmit, errText, msgSize);

    bool toContinueProcess = true;   // default value for no error handler

    if (fErrorHandler)
    {
        DOMLocatorImpl  locator(0, 0, (DOMNode* const) errorNode, 0, 0);
        DOMErrorImpl    domError(errorType , errText, &locator);
        toContinueProcess = fErrorHandler->handleError(domError);
    }

    if (errorType != DOMError::DOM_SEVERITY_WARNING)
        fErrorCount++;

    if (errorType == DOMError::DOM_SEVERITY_FATAL_ERROR || !toContinueProcess)
        throw toEmit;

    return toContinueProcess;
}

//
//
//
void DOMWriterImpl::procCdataSection(const XMLCh*   const nodeValue
                                   , const DOMNode* const nodeToWrite
                                   , int level)
{
    /***
     * Append a ']]>' at the end
     */
    int len = XMLString::stringLen(nodeValue);
    XMLCh* repNodeValue = (XMLCh*) fMemoryManager->allocate
    (
        (len + offset + 1) * sizeof(XMLCh)
    );//new XMLCh [len + offset + 1];
    XMLString::copyString(repNodeValue, nodeValue);
    XMLString::catString(repNodeValue, gEndCDATA);
    ArrayJanitor<XMLCh>  jName(repNodeValue, fMemoryManager);

    XMLCh* curPtr  = (XMLCh*) repNodeValue;
    XMLCh* nextPtr = 0;
    int    endTagPos = -1;

    bool   endTagFound = true;

    while (endTagFound)
    {
        endTagPos = XMLString::patternMatch(curPtr, gEndCDATA);
        if (endTagPos != -1)
        {
            nextPtr = curPtr + endTagPos + offset;  // skip the ']]>'
            *(curPtr + endTagPos) = chNull;         //nullify the first ']'
            if (endTagPos != len)
                reportError(nodeToWrite, DOMError::DOM_SEVERITY_WARNING, XMLDOMMsg::Writer_NestedCDATA);
            len = len - endTagPos - offset;
        }
        else
        {
            endTagFound = false;
        }

        /***
            to check ]]>]]>
        ***/
        if (endTagPos == 0)
        {
            printNewLine();
            printIndent(level);
            TRY_CATCH_THROW
            (
                *fFormatter << XMLFormatter::NoEscapes << gStartCDATA << gEndCDATA;
                , true
            )
        }
        else
        {
            procUnrepCharInCdataSection(curPtr, nodeToWrite, level);
        }

        if (endTagFound)
        {
            *(nextPtr - offset) = chCloseSquare;   //restore the first ']'
            curPtr = nextPtr;
        }
    }

    return;
}

//
//
//
void DOMWriterImpl::procUnrepCharInCdataSection(const XMLCh*   const nodeValue
                                              , const DOMNode* const nodeToWrite
                                              , int level)
{
    //
    //  We have to check each character and see if it could be represented.
    //  As long as it can, we just keep up with where we started and how
    //  many chars we've checked. When we hit an unrepresentable one, we
    //  stop, transcode everything we've collected, then start handling
    //  the unrepresentables via char refs. We repeat this until we get all
    //  the chars done.
    //
    const XMLCh*    srcPtr = nodeValue;
    const XMLCh*    endPtr = nodeValue +  XMLString::stringLen(nodeValue);

    // Set up the common part of the buffer that we build char refs into
    XMLCh tmpBuf[32];
    tmpBuf[0] = chAmpersand;
    tmpBuf[1] = chPound;
    tmpBuf[2] = chLatin_x;

    while (srcPtr < endPtr)
    {
        const XMLCh* tmpPtr = srcPtr;
        while (tmpPtr < endPtr)
        {
            if (fFormatter->getTranscoder()->canTranscodeTo(*tmpPtr))
                tmpPtr++;
            else
                break;
        }

        if (tmpPtr > srcPtr)
        {
            printNewLine();
            printIndent(level);
            TRY_CATCH_THROW
           (
                *fFormatter << XMLFormatter::NoEscapes << gStartCDATA;
                , true
            )

            // We got at least some chars that can be done normally
            fFormatter->formatBuf
            (
                srcPtr
                , tmpPtr - srcPtr
                , XMLFormatter::NoEscapes
                , XMLFormatter::UnRep_Fail
            );

            TRY_CATCH_THROW
            (
                *fFormatter << XMLFormatter::NoEscapes << gEndCDATA;
                , true
            )

            // Update the source pointer to our new spot
            srcPtr = tmpPtr;
        }
        else
        {
            //
            //  We hit something unrepresentable. So continue forward doing
            //  char refs until we hit something representable again or the
            //  end of input.
            //

            // one warning for consective unrep chars
            reportError(nodeToWrite, DOMError::DOM_SEVERITY_WARNING, XMLDOMMsg::Writer_NotRepresentChar);

            while (srcPtr < endPtr)
            {
                // Build a char ref for the current char
                XMLString::binToText(*srcPtr, &tmpBuf[3], 8, 16, fMemoryManager);
                const unsigned int bufLen = XMLString::stringLen(tmpBuf);
                tmpBuf[bufLen] = chSemiColon;
                tmpBuf[bufLen+1] = chNull;

                // And now call recursively back to our caller to format this
                fFormatter->formatBuf
                (
                    tmpBuf
                    , bufLen + 1
                    , XMLFormatter::NoEscapes
                    , XMLFormatter::UnRep_Fail
                );

                // Move up the source pointer and break out if needed
                srcPtr++;
                if (fFormatter->getTranscoder()->canTranscodeTo(*srcPtr))
                    break;
            }
        }
    }
}

void DOMWriterImpl::processNode(const DOMNode* const nodeToWrite)
{
    processNode(nodeToWrite, 0);
}

bool DOMWriterImpl::canSetFeature(const int featureId
                                       , bool      val) const
{
    return featuresSupported[2*featureId + (val? 0: 1)];
}

void DOMWriterImpl::printNewLine()
{
    if (getFeature(FORMAT_PRETTY_PRINT_ID))
    {
        fCurrentLine++;
        *fFormatter << fNewLineUsed;
    }
}

void DOMWriterImpl::printIndent(int level) const
{
    if (getFeature(FORMAT_PRETTY_PRINT_ID))
    {
        if (lastWhiteSpaceInTextNode)
        {
            level -= lastWhiteSpaceInTextNode/2; // two chSpaces equals one indent level
            lastWhiteSpaceInTextNode = 0;
            // if lastWhiteSpaceInTextNode/2 is greater than level, then
            // it means too many spaces have been written to the
            // output stream and we can no longer indent properly
        }

        for(int i = 0; i < level; i++)
            *fFormatter << chSpace << chSpace;
    }
}

void DOMWriterImpl::release()
{
    DOMWriterImpl* writer = (DOMWriterImpl*) this;
    delete writer;
}

void DOMWriterImpl::processBOM()
{
    // if the feature is not set, don't output bom
    if (!getFeature(BYTE_ORDER_MARK_ID))
        return;

    if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16LEncodingString)  == 0) ||
        (XMLString::compareIString(fEncoding, XMLUni::fgUTF16LEncodingString2) == 0)  )
    {
        fFormatter->writeBOM(BOM_utf16le, 2);
    }
    else if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16BEncodingString)  == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUTF16BEncodingString2) == 0)  )
    {
        fFormatter->writeBOM(BOM_utf16be, 2);
    }
    else if ((XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString)  == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString2) == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString3) == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString4) == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUTF16EncodingString5) == 0)  ) 
    {
#if defined(ENDIANMODE_LITTLE)
            fFormatter->writeBOM(BOM_utf16le, 2);
#elif defined(ENDIANMODE_BIG)
            fFormatter->writeBOM(BOM_utf16be, 2);
#endif
    }
    else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4LEncodingString)  == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUCS4LEncodingString2) == 0)  )
    {
        fFormatter->writeBOM(BOM_ucs4le, 4);
    }
    else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4BEncodingString)  == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUCS4BEncodingString2) == 0)  )
    {
        fFormatter->writeBOM(BOM_ucs4be, 4);
    }
    else if ((XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString)  == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString2) == 0) ||
             (XMLString::compareIString(fEncoding, XMLUni::fgUCS4EncodingString3) == 0)  )
    {
#if defined(ENDIANMODE_LITTLE)
        fFormatter->writeBOM(BOM_ucs4le, 4);
#elif defined(ENDIANMODE_BIG)
        fFormatter->writeBOM(BOM_ucs4be, 4);
#endif
    }

    return;

}

XERCES_CPP_NAMESPACE_END