igxmlscanner.cpp

来自「IBM的解析xml的工具Xerces的源代码」· C++ 代码 · 共 1,855 行 · 第 1/5 页

                );
                toFill.addElement(curPair);
            }
             else
            {
                curPair = toFill.elementAt(attCount);
                curPair->set(fAttNameBuf.getRawBuffer(), fAttValueBuf.getRawBuffer());
            }

            // And bump the count of attributes we've gotten
            attCount++;

            // And go to the top again for another attribute
            continue;
        }

        //  It was some special case character so do all of the checks and
        //  deal with it.
        if (!nextCh)
            ThrowXMLwithMemMgr(UnexpectedEOFException, XMLExcepts::Gen_UnexpectedEOF, fMemoryManager);

        if (nextCh == chForwardSlash)
        {
            fReaderMgr.getNextChar();
            isEmpty = true;
            if (!fReaderMgr.skippedChar(chCloseAngle))
                emitError(XMLErrs::UnterminatedStartTag, elemName);
            break;
        }
        else if (nextCh == chCloseAngle)
        {
            fReaderMgr.getNextChar();
            break;
        }
        else if (nextCh == chOpenAngle)
        {
            //  Check for this one specially, since its going to be common
            //  and it is kind of auto-recovering since we've already hit the
            //  next open bracket, which is what we would have seeked to (and
            //  skipped this whole tag.)
            emitError(XMLErrs::UnterminatedStartTag, elemName);
            break;
        }
        else if ((nextCh == chSingleQuote) || (nextCh == chDoubleQuote))
        {
            //  Check for this one specially, which is probably a missing
            //  attribute name, e.g. ="value". Just issue expected name
            //  error and eat the quoted string, then jump back to the
            //  top again.
            emitError(XMLErrs::ExpectedAttrName);
            fReaderMgr.getNextChar();
            fReaderMgr.skipQuotedString(nextCh);
            fReaderMgr.skipPastSpaces();
            continue;
        }
    }

    return attCount;
}


//  This method will kick off the scanning of the primary content of the
//  document, i.e. the elements.
bool IGXMLScanner::scanContent()
{
    //  Go into a loop until we hit the end of the root element, or we fall
    //  out because there is no root element.
    //
    //  We have to do kind of a deeply nested double loop here in order to
    //  avoid doing the setup/teardown of the exception handler on each
    //  round. Doing it this way we only do it when an exception actually
    //  occurs.
    bool gotData = true;
    bool inMarkup = false;
    while (gotData)
    {
        try
        {
            while (gotData)
            {
                //  Sense what the next top level token is. According to what
                //  this tells us, we will call something to handle that kind
                //  of thing.
                unsigned int orgReader;
                const XMLTokens curToken = senseNextToken(orgReader);

                //  Handle character data and end of file specially. Char data
                //  is not markup so we don't want to handle it in the loop
                //  below.
                if (curToken == Token_CharData)
                {
                    //  Scan the character data and call appropriate events. Let
                    //  him use our local character data buffer for efficiency.
                    scanCharData(fCDataBuf);
                    continue;
                }
                else if (curToken == Token_EOF)
                {
                    //  The element stack better be empty at this point or we
                    //  ended prematurely before all elements were closed.
                    if (!fElemStack.isEmpty())
                    {
                        const ElemStack::StackElem* topElem = fElemStack.popTop();
                        emitError
                        (
                            XMLErrs::EndedWithTagsOnStack
                            , topElem->fThisElement->getFullName()
                        );
                    }

                    // Its the end of file, so clear the got data flag
                    gotData = false;
                    continue;
                }

                // We are in some sort of markup now
                inMarkup = true;

                //  According to the token we got, call the appropriate
                //  scanning method.
                switch(curToken)
                {
                    case Token_CData :
                        // Make sure we are within content
                        if (fElemStack.isEmpty())
                            emitError(XMLErrs::CDATAOutsideOfContent);
                        scanCDSection();
                        break;

                    case Token_Comment :
                        scanComment();
                        break;

                    case Token_EndTag :
                        scanEndTag(gotData);
                        break;

                    case Token_PI :
                        scanPI();
                        break;

                    case Token_StartTag :
                        if (fDoNamespaces)
                            scanStartTagNS(gotData);
                        else
                            scanStartTag(gotData);
                        break;

                    default :
                        fReaderMgr.skipToChar(chOpenAngle);
                        break;
                }

                if (orgReader != fReaderMgr.getCurrentReaderNum())
                    emitError(XMLErrs::PartialMarkupInEntity);

                // And we are back out of markup again
                inMarkup = false;
            }
        }
        catch(const EndOfEntityException& toCatch)
        {
            //  If we were in some markup when this happened, then its a
            //  partial markup error.
            if (inMarkup)
                emitError(XMLErrs::PartialMarkupInEntity);

            // Send an end of entity reference event
            if (fDocHandler)
                fDocHandler->endEntityReference(toCatch.getEntity());

            inMarkup = false;
        }
    }

    // It went ok, so return success
    return true;
}


void IGXMLScanner::scanEndTag(bool& gotData)
{
    //  Assume we will still have data until proven otherwise. It will only
    //  ever be false if this is the end of the root element.
    gotData = true;

    //  Check if the element stack is empty. If so, then this is an unbalanced
    //  element (i.e. more ends than starts, perhaps because of bad text
    //  causing one to be skipped.)
    if (fElemStack.isEmpty())
    {
        emitError(XMLErrs::MoreEndThanStartTags);
        fReaderMgr.skipPastChar(chCloseAngle);
        ThrowXMLwithMemMgr(RuntimeException, XMLExcepts::Scan_UnbalancedStartEnd, fMemoryManager);
    }

    //  Pop the stack of the element we are supposed to be ending. Remember
    //  that we don't own this. The stack just keeps them and reuses them.
    unsigned int uriId = (fDoNamespaces)
        ? fElemStack.getCurrentURI() : fEmptyNamespaceId;

    // these get initialized below
    const ElemStack::StackElem* topElem = 0;
    XMLElementDecl *tempElement = 0;
    XMLCh *elemName = 0;

    // Make sure that its the end of the element that we expect
    // special case for schema validation, whose element decls,
    // obviously don't contain prefix information
    if(fGrammarType == Grammar::SchemaGrammarType)
    {
        elemName = fElemStack.getCurrentSchemaElemName();
        topElem = fElemStack.popTop(); 
        tempElement = topElem->fThisElement; 
    }
    else
    {
        topElem = fElemStack.popTop(); 
        tempElement = topElem->fThisElement;
        elemName = (XMLCh *)tempElement->getFullName();
    }
    if (!fReaderMgr.skippedString(elemName))
    {
        emitError
        (
            XMLErrs::ExpectedEndOfTagX
            , elemName
        );
        fReaderMgr.skipPastChar(chCloseAngle);
        return;
    }

    // See if it was the root element, to avoid multiple calls below
    const bool isRoot = fElemStack.isEmpty();

    // Make sure we are back on the same reader as where we started
    if (topElem->fReaderNum != fReaderMgr.getCurrentReaderNum())
        emitError(XMLErrs::PartialTagMarkupError);

    // Skip optional whitespace
    fReaderMgr.skipPastSpaces();

    // Make sure we find the closing bracket
    if (!fReaderMgr.skippedChar(chCloseAngle))
    {
        emitError
        (
            XMLErrs::UnterminatedEndTag
            , topElem->fThisElement->getFullName()
        );
    }

    if (fGrammarType == Grammar::SchemaGrammarType)
    {
        // reset error occurred
        fPSVIElemContext.fErrorOccurred = fErrorStack->pop();
        if (fValidate && topElem->fThisElement->isDeclared())
        {
            fPSVIElemContext.fCurrentTypeInfo = ((SchemaValidator*) fValidator)->getCurrentTypeInfo();
            if(!fPSVIElemContext.fCurrentTypeInfo)
                fPSVIElemContext.fCurrentDV = ((SchemaValidator*) fValidator)->getCurrentDatatypeValidator();
            else
                fPSVIElemContext.fCurrentDV = 0;
            if(fPSVIHandler)
            {
                fPSVIElemContext.fNormalizedValue = ((SchemaValidator*) fValidator)->getNormalizedValue();

                if (XMLString::equals(fPSVIElemContext.fNormalizedValue, XMLUni::fgZeroLenString))
                    fPSVIElemContext.fNormalizedValue = 0;
            }
        }
        else
        {
            fPSVIElemContext.fCurrentDV = 0;
            fPSVIElemContext.fCurrentTypeInfo = 0;
            fPSVIElemContext.fNormalizedValue = 0;
        }
    }

    //  If validation is enabled, then lets pass him the list of children and
    //  this element and let him validate it.
    DatatypeValidator* psviMemberType = 0;
    if (fValidate)
    {

       //
       // XML1.0-3rd
       // Validity Constraint: 
       // The declaration matches EMPTY and the element has no content (not even 
       // entity references, comments, PIs or white space).
       //
       if ( (fGrammarType == Grammar::DTDGrammarType) &&
            (topElem->fCommentOrPISeen)               &&
            (((DTDElementDecl*) topElem->fThisElement)->getModelType() == DTDElementDecl::Empty))
       {
           fValidator->emitError
               (
               XMLValid::EmptyElemHasContent
               , topElem->fThisElement->getFullName()
               );
       }

       //
       // XML1.0-3rd
       // Validity Constraint: 
       // 
       // The declaration matches children and the sequence of child elements 
       // belongs to the language generated by the regular expression in the 
       // content model, with optional white space, comments and PIs 
       // (i.e. markup matching production [27] Misc) between the start-tag and 
       // the first child element, between child elements, or between the last 
       // child element and the end-tag. 
       //
       // Note that 
       //    a CDATA section containing only white space or 
       //    a reference to an entity whose replacement text is character references 
       //       expanding to white space do not match the nonterminal S, and hence 
       //       cannot appear in these positions; however,
       //    a reference to an internal entity with a literal value consisting 
       //       of character references expanding to white space does match S, 
       //       since its replacement text is the white space resulting from expansion 
       //       of the character references.
       //
       if ( (fGrammarType == Grammar::DTDGrammarType)  &&
            (topElem->fReferenceEscaped)               &&
            (((DTDElementDecl*) topElem->fThisElement)->getModelType() == DTDElementDecl::Children))
       {
           fValidator->emitError
               (
               XMLValid::ElemChildrenHasInvalidWS
               , topElem->fThisElement->getFullName()
               );
       }
       
        int res = fValidator->checkContent
        (
            topElem->fThisElement
            , topElem->fChildren
            , topElem->fChildCount
        );

        if (res >= 0)
        {
            //  One of the elements is not valid for the content. NOTE that
            //  if no children were provided but the content model requires
            //  them, it comes back with a zero value. But we cannot use that
            //  to index the child array in this case, and have to put out a
            //  special message.
            if (!topElem->fChildCount)
            {
                fValidator->emitError
                (
                    XMLValid::EmptyNotValidForContent
                    , topElem->fThisElement->getFormattedContentModel()
                );
            }
            else if ((unsigned int)res >= topElem->fChildCount)
            {
                fValidator->emitError
                (
                    XMLValid::NotEnoughElemsForCM
                    , topElem->fThisElement->getFormattedContentModel()
                );
            }
            else
            {
                fValidator->emitError
                (
                    XMLValid::ElementNotValidForContent
                    , topElem->fChildren[res]->getRawName()
                    , topElem->fThisElement->getFormattedContentModel()