dgxmlscanner.cpp

来自「IBM的解析xml的工具Xerces的源代码」· C++ 代码 · 共 1,902 行 · 第 1/5 页

                }

                // That went ok, so scan for any miscellaneous stuff
                scanMiscellaneous();

                if (fDocHandler)
                    fDocHandler->endDocument();
            }
        }
    }
    //  NOTE:
    //
    //  In all of the error processing below, the emitError() call MUST come
    //  before the flush of the reader mgr, or it will fail because it tries
    //  to find out the position in the XML source of the error.
    catch(const XMLErrs::Codes)
    {
        // This is a 'first failure' exception, so reset and return failure
        fReaderMgr.reset();
        return false;
    }
    catch(const XMLValid::Codes)
    {
        // This is a 'first fatal error' type exit, so reset and reuturn failure
        fReaderMgr.reset();
        return false;
    }
    catch(const XMLException& excToCatch)
    {
        //  Emit the error and catch any user exception thrown from here. Make
        //  sure in all cases we flush the reader manager.
        fInException = true;
        try
        {
            if (excToCatch.getErrorType() == XMLErrorReporter::ErrType_Warning)
                emitError
                (
                    XMLErrs::XMLException_Warning
                    , excToCatch.getType()
                    , excToCatch.getMessage()
                );
            else if (excToCatch.getErrorType() >= XMLErrorReporter::ErrType_Fatal)
                emitError
                (
                    XMLErrs::XMLException_Fatal
                    , excToCatch.getType()
                    , excToCatch.getMessage()
                );
            else
                emitError
                (
                    XMLErrs::XMLException_Error
                    , excToCatch.getType()
                    , excToCatch.getMessage()
                );
        }
        catch(const OutOfMemoryException&)
        {
            throw;
        }
        catch(...)
        {
            // Reset and rethrow user error
            fReaderMgr.reset();
            throw;
        }

        // Reset and return failure
        fReaderMgr.reset();
        return false;
    }
    catch(const OutOfMemoryException&)
    {
        throw;
    }
    catch(...)
    {
        // Reset and rethrow original error
        fReaderMgr.reset();
        throw;
    }

    // If we hit the end, then flush the reader manager
    if (!retVal)
        fReaderMgr.reset();

    return retVal;
}


// ---------------------------------------------------------------------------
//  DGXMLScanner: Private scanning methods
// ---------------------------------------------------------------------------

//  This method will kick off the scanning of the primary content of the
//  document, i.e. the elements.
bool DGXMLScanner::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 :
                        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 DGXMLScanner::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;

    //  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.
    const ElemStack::StackElem* topElem = fElemStack.popTop();
    XMLElementDecl *tempElement = topElem->fThisElement;

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

    // Make sure that its the end of the element that we expect
    if (!fReaderMgr.skippedString(tempElement->getFullName()))
    {
        emitError
        (
            XMLErrs::ExpectedEndOfTagX
            , tempElement->getFullName()
        );
        fReaderMgr.skipPastChar(chCloseAngle);
        return;
    }

    // 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 validation is enabled, then lets pass him the list of children and
    //  this element and let him validate it.
    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 ( (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 ( (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()
                );
            }
        }
    }

    // If we have a doc handler, tell it about the end tag
    if (fDocHandler)
    {
        fDocHandler->endElement
        (
            *topElem->fThisElement
            , uriId
            , isRoot
            , (fDoNamespaces)
                ? topElem->fThisElement->getElementName()->getPrefix()
                : XMLUni::fgZeroLenString
        );
    }

    // If this was the root, then done with content
    gotData = !isRoot;
}