wfxmlscanner.cpp
来自「IBM的解析xml的工具Xerces的源代码」· C++ 代码 · 共 1,883 行 · 第 1/5 页
CPP
1,883 行
fElements->addElement(elemDecl); } elemDecl->setElementName(qnameRawBuf, fEmptyNamespaceId); fElementLookup->put((void*)elemDecl->getFullName(), elemDecl); fElementIndex++; } // Expand the element stack and add the new element fElemStack.addLevel(elemDecl, fReaderMgr.getCurrentReaderNum()); // reset NS attribute list fAttrNSList->removeAllElements(); // We loop until we either see a /> or >, handling attribute/value // pairs until we get there. unsigned int attCount = 0; unsigned int curAttListSize = fAttrList->size(); while (true) { // And get the next non-space character XMLCh nextCh = fReaderMgr.peekNextChar(); // If the next character is not a slash or closed angle bracket, // then it must be whitespace, since whitespace is required // between the end of the last attribute and the name of the next // one. if (attCount) { if ((nextCh != chForwardSlash) && (nextCh != chCloseAngle)) { if (fReaderMgr.getCurrentReader()->isWhitespace(nextCh)) { // Ok, skip by them and peek another char fReaderMgr.skipPastSpaces(); nextCh = fReaderMgr.peekNextChar(); } else { // Emit the error but keep on going emitError(XMLErrs::ExpectedWhitespace); } } } // Ok, here we first check for any of the special case characters. // If its not one, then we do the normal case processing, which // assumes that we've hit an attribute value, Otherwise, we do all // the special case checks. if (!fReaderMgr.getCurrentReader()->isSpecialStartTagChar(nextCh)) { // Assume its going to be an attribute, so get a name from // the input. if (!fReaderMgr.getName(fAttNameBuf)) { emitError(XMLErrs::ExpectedAttrName); fReaderMgr.skipPastChar(chCloseAngle); return false; } // And next must be an equal sign if (!scanEq()) { static const XMLCh tmpList[] = { chSingleQuote, chDoubleQuote, chCloseAngle , chOpenAngle, chForwardSlash, chNull }; emitError(XMLErrs::ExpectedEqSign); // Try to sync back up by skipping forward until we either // hit something meaningful. const XMLCh chFound = fReaderMgr.skipUntilInOrWS(tmpList); if ((chFound == chCloseAngle) || (chFound == chForwardSlash)) { // Jump back to top for normal processing of these continue; } else if ((chFound == chSingleQuote) || (chFound == chDoubleQuote) || fReaderMgr.getCurrentReader()->isWhitespace(chFound)) { // Just fall through assuming that the value is to follow } else if (chFound == chOpenAngle) { // Assume a malformed tag and that new one is starting emitError(XMLErrs::UnterminatedStartTag, qnameRawBuf); return false; } else { // Something went really wrong return false; } } // See if this attribute is declared more than one for this element. const XMLCh* attNameRawBuf = fAttNameBuf.getRawBuffer(); unsigned int attNameHash = XMLString::hash(attNameRawBuf, 109, fMemoryManager); if (attCount) { for (unsigned int k=0; k < attCount; k++) { if (fAttrNameHashList->elementAt(k) == attNameHash) { if (XMLString::equals( fAttrList->elementAt(k)->getQName() , attNameRawBuf)) { emitError ( XMLErrs::AttrAlreadyUsedInSTag , attNameRawBuf , qnameRawBuf ); break; } } } } // Skip any whitespace before the value and then scan the att // value. This will come back normalized with entity refs and // char refs expanded. fReaderMgr.skipPastSpaces(); if (!scanAttValue(attNameRawBuf, fAttValueBuf)) { static const XMLCh tmpList[] = { chCloseAngle, chOpenAngle, chForwardSlash, chNull }; emitError(XMLErrs::ExpectedAttrValue); // It failed, so lets try to get synced back up. We skip // forward until we find some whitespace or one of the // chars in our list. const XMLCh chFound = fReaderMgr.skipUntilInOrWS(tmpList); if ((chFound == chCloseAngle) || (chFound == chForwardSlash) || fReaderMgr.getCurrentReader()->isWhitespace(chFound)) { // Just fall through and process this attribute, though // the value will be "". } else if (chFound == chOpenAngle) { // Assume a malformed tag and that new one is starting emitError(XMLErrs::UnterminatedStartTag, qnameRawBuf); return false; } else { // Something went really wrong return false; } } // Add this attribute to the attribute list that we use to // pass them to the handler. We reuse its existing elements // but expand it as required. const XMLCh* attValueRawBuf = fAttValueBuf.getRawBuffer(); XMLAttr* curAtt = 0; if (attCount >= curAttListSize) { curAtt = new (fMemoryManager) XMLAttr ( fEmptyNamespaceId , attNameRawBuf , attValueRawBuf , XMLAttDef::CData , true , fMemoryManager ); fAttrList->addElement(curAtt); fAttrNameHashList->addElement(attNameHash); } else { curAtt = fAttrList->elementAt(attCount); curAtt->set ( fEmptyNamespaceId , attNameRawBuf , attValueRawBuf ); curAtt->setSpecified(true); fAttrNameHashList->setElementAt(attNameHash, attCount); } // Make sure that the name is basically well formed for namespace // enabled rules. It either has no colons, or it has one which // is neither the first or last char. const int colonFirst = XMLString::indexOf(attNameRawBuf, chColon); if (colonFirst != -1) { const int colonLast = XMLString::lastIndexOf(attNameRawBuf, chColon); if (colonFirst != colonLast) { emitError(XMLErrs::TooManyColonsInName); continue; } else if ((colonFirst == 0) || (colonLast == (int)fAttNameBuf.getLen() - 1)) { emitError(XMLErrs::InvalidColonPos); continue; } } // Map prefix to namespace const XMLCh* attPrefix = curAtt->getPrefix(); const XMLCh* attLocalName = curAtt->getName(); const XMLCh* namespaceURI = fAttValueBuf.getRawBuffer(); if (attPrefix && *attPrefix) { if (XMLString::equals(attPrefix, XMLUni::fgXMLString)) { curAtt->setURIId(fXMLNamespaceId); } else if (XMLString::equals(attPrefix, XMLUni::fgXMLNSString)) { if (XMLString::equals(attLocalName, XMLUni::fgXMLNSString)) emitError(XMLErrs::NoUseOfxmlnsAsPrefix); else if (XMLString::equals(attLocalName, XMLUni::fgXMLString)) { if (!XMLString::equals(namespaceURI, XMLUni::fgXMLURIName)) emitError(XMLErrs::PrefixXMLNotMatchXMLURI); } if (!namespaceURI) emitError(XMLErrs::NoEmptyStrNamespace, attNameRawBuf); else if(!*namespaceURI && fXMLVersion == XMLReader::XMLV1_0) emitError(XMLErrs::NoEmptyStrNamespace, attNameRawBuf); fElemStack.addPrefix ( attLocalName , fURIStringPool->addOrFind(namespaceURI) ); curAtt->setURIId(fXMLNSNamespaceId); } else { fAttrNSList->addElement(curAtt); } } else { if (XMLString::equals(XMLUni::fgXMLNSString, attLocalName)) { if (XMLString::equals(namespaceURI, XMLUni::fgXMLNSURIName)) emitError(XMLErrs::NoUseOfxmlnsURI); else if (XMLString::equals(namespaceURI, XMLUni::fgXMLURIName)) emitError(XMLErrs::XMLURINotMatchXMLPrefix); fElemStack.addPrefix ( XMLUni::fgZeroLenString , fURIStringPool->addOrFind(namespaceURI) ); } } // increment attribute count attCount++; // And jump back to the top of the loop 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, qnameRawBuf); 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, qnameRawBuf); 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; } } // Handle provided attributes that we did not map their prefixes for (unsigned int i=0; i < fAttrNSList->size(); i++) { XMLAttr* providedAttr = fAttrNSList->elementAt(i); providedAttr->setURIId ( resolvePrefix ( providedAttr->getPrefix(), ElemStack::Mode_Attribute ) ); } if(attCount) { // // Decide if to use hash table to do duplicate checking // bool toUseHashTable = false; setAttrDupChkRegistry(attCount, toUseHashTable); // check for duplicate namespace attributes: // by checking for qualified names with the same local part and with prefixes // which have been bound to namespace names that are identical. XMLAttr* loopAttr; XMLAttr* curAtt; for (unsigned int attrIndex=0; attrIndex < attCount-1; attrIndex++) { loopAttr = fAttrList->elementAt(attrIndex); if (!toUseHashTable) { for (unsigned int curAttrIndex = attrIndex+1; curAttrIndex < attCount; curAttrIndex++) { curAtt = fAttrList->elementAt(curAttrIndex); if (curAtt->getURIId() == loopAttr->getURIId() && XMLString::equals(curAtt->getName(), loopAttr->getName())) { emitError ( XMLErrs::AttrAlreadyUsedInSTag , curAtt->getName() , elemDecl->getFullName() ); } } } else { if (fAttrDupChkRegistry->containsKey((void*)loopAttr->getName(), loopAttr->getURIId())) { emitError ( XMLErrs::AttrAlreadyUsedInSTag , loopAttr->getName() , elemDecl->getFullName() ); } fAttrDupChkRegistry->put((void*)loopAttr->getName(), loopAttr->getURIId(), loopAttr); } } } // Resolve the qualified name to a URI. unsigned int uriId = resolvePrefix
⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?