parseparam.cxx

SP是一个基于GNU C++编译器

	return 0;
      }
      gc.type = GroupConnector::seqGC;
      if (currentMarkup())
	currentMarkup()->addDelim(Syntax::dSEQ);
      return 1;
    case tokenOr:
      if (!allow.groupConnector(GroupConnector::orGC)) {
	groupConnectorInvalidToken(tokenOr, allow);
	return 0;
      }
      gc.type = GroupConnector::orGC;
      if (currentMarkup())
	currentMarkup()->addDelim(Syntax::dOR);
      return 1;
    case tokenDtgc:
      if (!allow.groupConnector(GroupConnector::dtgcGC)) {
	groupConnectorInvalidToken(tokenDtgc, allow);
	return 0;
      }
      gc.type = GroupConnector::dtgcGC;
      if (inputLevel() > groupInputLevel)
	message(ParserMessages::groupParameterEntityNotEnded);
      if (currentMarkup())
	currentMarkup()->addDelim(Syntax::dDTGC);
      return 1;
    case tokenGrpc:
      if (!allow.groupConnector(GroupConnector::grpcGC)) {
	groupConnectorInvalidToken(tokenGrpc, allow);
	return 0;
      }
      gc.type = GroupConnector::grpcGC;
      if (inputLevel() > groupInputLevel)
	message(ParserMessages::groupParameterEntityNotEnded);
      if (currentMarkup())
	currentMarkup()->addDelim(Syntax::dGRPC);
      return 1;
    default:
      groupConnectorInvalidToken(token, allow);
      return 0;
    }
  }
}

void Parser::groupConnectorInvalidToken(Token token,
					const AllowedGroupConnectors &allow)
{
  message(ParserMessages::connectorInvalidToken,
	  TokenMessageArg(token, grpMode, syntaxPointer(), sdPointer()),
	  AllowedGroupConnectorsMessageArg(allow, syntaxPointer()));
}

Boolean Parser::parseElementNameGroup(unsigned declInputLevel, Param &parm)
{
  if (!parseNameGroup(declInputLevel, parm))
    return 0;
  parm.elementVector.resize(parm.nameTokenVector.size());
  for (size_t i = 0; i < parm.nameTokenVector.size(); i++)
    parm.elementVector[i] = lookupCreateElement(parm.nameTokenVector[i].name);
  return 1;
}

Boolean Parser::parseEntityReferenceNameGroup(Boolean &ignore)
{
  Param parm;
  if (!parseNameGroup(inputLevel(), parm))
    return 0;
  if (inInstance()) {
    for (size_t i = 0; i < parm.nameTokenVector.size(); i++) {
      const Lpd *lpd = lookupLpd(parm.nameTokenVector[i].name).pointer();
      if (lpd && lpd->active()) {
	ignore = 0;
	return 1;
      }
    }
  }
  ignore = 1;
  return 1;
}

Boolean Parser::parseTagNameGroup(Boolean &active)
{
  Param parm;
  if (!parseNameGroup(inputLevel(), parm))
    return 0;
  active = 0;
  return 1;
}

Boolean Parser::parseNameGroup(unsigned declInputLevel, Param &parm)
{
  static AllowedGroupTokens allowName(GroupToken::name);
  return parseGroup(allowName, declInputLevel, parm);
}

Boolean Parser::parseNameTokenGroup(unsigned declInputLevel, Param &parm)
{
  static AllowedGroupTokens allowNameToken(GroupToken::nameToken);
  return parseGroup(allowNameToken, declInputLevel, parm);
}

static
Boolean groupContains(const Vector<NameToken> &vec, const StringC &str)
{
  for (size_t i = 0; i < vec.size(); i++)
    if (vec[i].name == str)
      return 1;
  return 0;
}

Boolean Parser::parseGroup(const AllowedGroupTokens &allowToken,
			   unsigned declInputLevel,
			   Param &parm)
{
  unsigned groupInputLevel = inputLevel();
  int nDuplicates = 0;
  Vector<NameToken> &vec = parm.nameTokenVector;
  vec.clear();
  GroupConnector::Type connector = GroupConnector::grpcGC;
  GroupToken gt;
  for (;;) {
    if (!parseGroupToken(allowToken, 0, declInputLevel, groupInputLevel, gt))
      return 0;
    if (groupContains(vec, gt.token)) {
      nDuplicates++;
      message(ParserMessages::duplicateGroupToken,
	      StringMessageArg(gt.token));
    }
    else {
      vec.resize(vec.size() + 1);
      gt.token.swap(vec.back().name);
      getCurrentToken(vec.back().origName);
      vec.back().loc = currentLocation();
    }
    GroupConnector gc;
    static AllowedGroupConnectors allowAnyConnectorGrpc(GroupConnector::orGC,
							GroupConnector::andGC,
							GroupConnector::seqGC,
							GroupConnector::grpcGC);
			    
    if (!parseGroupConnector(allowAnyConnectorGrpc, declInputLevel,
			     groupInputLevel, gc))
      return 0;
    if (gc.type == GroupConnector::grpcGC)
      break;
    if (options().warnNameGroupNotOr) {
      if (gc.type != GroupConnector::orGC)
	message(ParserMessages::nameGroupNotOr);
    }
    else if (options().warnShould) {
      if (connector == GroupConnector::grpcGC)
	connector = gc.type;
      else if (gc.type != connector) {
	message(ParserMessages::mixedConnectors);
	connector = gc.type;
      }
    }
  }
  if (nDuplicates + vec.size() > syntax().grpcnt())
    message(ParserMessages::groupCount, NumberMessageArg(syntax().grpcnt()));
  return 1;
}

Boolean Parser::parseDataTagGroup(unsigned nestingLevel,
				  unsigned declInputLevel, GroupToken &result)
{
  if (nestingLevel - 1 == syntax().grplvl())
    message(ParserMessages::grplvl, NumberMessageArg(syntax().grplvl()));
  unsigned groupInputLevel = inputLevel();
  GroupToken gt;
  static AllowedGroupTokens allowName(GroupToken::name);
  if (!parseGroupToken(allowName, nestingLevel, declInputLevel,
		       groupInputLevel, gt))
    return 0;
  const ElementType *element = lookupCreateElement(gt.token);
  GroupConnector gc;
  static AllowedGroupConnectors allowSeq(GroupConnector::seqGC);
  if (!parseGroupConnector(allowSeq, declInputLevel, groupInputLevel, gc))
    return 0;
  static AllowedGroupTokens
    allowDataTagLiteralDataTagTemplateGroup(GroupToken::dataTagLiteral,
					    GroupToken::dataTagTemplateGroup);
  if (!parseGroupToken(allowDataTagLiteralDataTagTemplateGroup,
		       nestingLevel,
		       declInputLevel,
		       groupInputLevel,
		       gt))
    return 0;
  Vector<Text> templates;
  if (gt.type == GroupToken::dataTagTemplateGroup)
    gt.textVector.swap(templates);
  else {
    templates.resize(1);
    gt.text.swap(templates[0]);
  }
  static AllowedGroupConnectors allowSeqDtgc(GroupConnector::seqGC,
					     GroupConnector::dtgcGC);
  if (!parseGroupConnector(allowSeqDtgc, declInputLevel, groupInputLevel, gc))
    return 0;
  NCVector<Owner<ContentToken> > vec(2);
  vec[1] = new PcdataToken;
  if (gc.type != GroupConnector::dtgcGC) {
    static AllowedGroupTokens allowDataTagLiteral(GroupToken::dataTagLiteral);
    if (!parseGroupToken(allowDataTagLiteral,
			 nestingLevel,
			 declInputLevel,
			 groupInputLevel,
			 gt))
      return 0;
    vec[0] = new DataTagElementToken(element, templates, gt.text);
    static AllowedGroupConnectors allowDtgc(GroupConnector::dtgcGC);
    if (!parseGroupConnector(allowDtgc, declInputLevel, groupInputLevel, gc))
      return 0;
  }
  else
    vec[0] = new DataTagElementToken(element, templates);
  ContentToken::OccurrenceIndicator oi = getOccurrenceIndicator(grpMode);
  result.contentToken = new DataTagGroup(vec, oi);
  result.type = GroupToken::dataTagGroup;
  return 1;
}

Boolean Parser::parseDataTagTemplateGroup(unsigned nestingLevel,
					  unsigned declInputLevel,
					  GroupToken &result)
{
  if (nestingLevel - 1 == syntax().grplvl())
    message(ParserMessages::grplvl, NumberMessageArg(syntax().grplvl()));
  unsigned groupInputLevel = inputLevel();
  Vector<Text> &vec = result.textVector;
  for (;;) {
    GroupToken gt;
    static AllowedGroupTokens allowDataTagLiteral(GroupToken::dataTagLiteral);
    if (!parseGroupToken(allowDataTagLiteral,
			 nestingLevel,
			 declInputLevel,
			 groupInputLevel,
			 gt))
      return 0;
    if (vec.size() == syntax().grpcnt())
      message(ParserMessages::groupCount, NumberMessageArg(syntax().grpcnt()));
    vec.resize(vec.size() + 1);
    gt.text.swap(vec.back());
    static AllowedGroupConnectors allowOrGrpc(GroupConnector::orGC,
					      GroupConnector::grpcGC);
    GroupConnector gc;
    if (!parseGroupConnector(allowOrGrpc, declInputLevel, groupInputLevel, gc))
      return 0;
    if (gc.type == GroupConnector::grpcGC)
      break;
  }
  return 1;
}

Boolean Parser::parseModelGroup(unsigned nestingLevel, unsigned declInputLevel,
				ModelGroup *&group, Mode oiMode)
{
  if (nestingLevel - 1 == syntax().grplvl())
    message(ParserMessages::grplvl, NumberMessageArg(syntax().grplvl()));
  unsigned groupInputLevel = inputLevel();
  GroupToken gt;
  NCVector<Owner<ContentToken> > tokenVector;
  GroupConnector::Type connector = GroupConnector::grpcGC;
  
  static AllowedGroupTokens allowContentToken(GroupToken::pcdata,
					      GroupToken::dataTagGroup,
					      GroupToken::elementToken,
					      GroupToken::modelGroup);
  static AllowedGroupConnectors allowAnyConnectorGrpc(GroupConnector::orGC,
						      GroupConnector::andGC,
						      GroupConnector::seqGC,
						      GroupConnector::grpcGC);
			    
  static AllowedGroupConnectors allowOrGrpc(GroupConnector::orGC,
					    GroupConnector::grpcGC);
  static AllowedGroupConnectors allowAndGrpc(GroupConnector::andGC,
					     GroupConnector::grpcGC);
  static AllowedGroupConnectors allowSeqGrpc(GroupConnector::seqGC,
					     GroupConnector::grpcGC);
  const AllowedGroupConnectors *connectorp = &allowAnyConnectorGrpc;

  GroupConnector gc;
  Boolean pcdataCheck = 0;
  do {
    if (!parseGroupToken(allowContentToken, nestingLevel, declInputLevel,
			 groupInputLevel, gt))
      return 0;
    ContentToken *contentToken;
    if (gt.type == GroupToken::modelGroup)
      contentToken = gt.model.extract();
    else
      contentToken = gt.contentToken.extract();
    if (tokenVector.size() == syntax().grpcnt())
      message(ParserMessages::groupCount, NumberMessageArg(syntax().grpcnt()));
    tokenVector.resize(tokenVector.size() + 1);
    tokenVector.back() = contentToken;
    if (!parseGroupConnector(*connectorp, declInputLevel, groupInputLevel, gc))
      return 0;
    if (options().warnMixedContentRepOrGroup && gt.type == GroupToken::pcdata) {
      if (tokenVector.size() != 1)
	message(ParserMessages::pcdataNotFirstInGroup);
      else if (gc.type == GroupConnector::seqGC)
	message(ParserMessages::pcdataInSeqGroup);
      else
	pcdataCheck = 1;
      if (nestingLevel != 1)
	message(ParserMessages::pcdataInNestedModelGroup);
    }
    else if (pcdataCheck) {
      if (gt.type == GroupToken::modelGroup)
	message(ParserMessages::pcdataGroupMemberModelGroup);
      if (contentToken->occurrenceIndicator() != ContentToken::none)
	message(ParserMessages::pcdataGroupMemberOccurrenceIndicator);
    }
    if (tokenVector.size() == 1) {
      connector = gc.type;
      switch (gc.type) {
      case GroupConnector::orGC:
	connectorp = &allowOrGrpc;
	break;
      case GroupConnector::seqGC:
	connectorp = &allowSeqGrpc;
	break;
      case GroupConnector::andGC:
	connectorp = &allowAndGrpc;
	if (options().warnAndGroup)
	  message(ParserMessages::andGroup);
	break;
      default:
	break;
      }
    }
  } while (gc.type != GroupConnector::grpcGC);
  ContentToken::OccurrenceIndicator oi
    = getOccurrenceIndicator(oiMode);
  switch (connector) {
  case GroupConnector::orGC:
    group = new OrModelGroup(tokenVector, oi);
    if (pcdataCheck && oi != ContentToken::rep)
      message(ParserMessages::pcdataGroupNotRep);
    break;
  case GroupConnector::grpcGC:
    if (pcdataCheck && oi != ContentToken::rep && oi != ContentToken::none)
      message(ParserMessages::pcdataGroupNotRep);
    // fall through
  case GroupConnector::seqGC:
    group = new SeqModelGroup(tokenVector, oi);
    break;
  case GroupConnector::andGC:
    group = new AndModelGroup(tokenVector, oi);
    break;
  default:
    break;
  }
  return 1;
}

ContentToken::OccurrenceIndicator
Parser::getOccurrenceIndicator(Mode oiMode)
{
  Token token = getToken(oiMode);
  switch (token) {
  case tokenPlus:
    if (currentMarkup())
      currentMarkup()->addDelim(Syntax::dPLUS);
    return ContentToken::plus;
  case tokenOpt: 
    if (currentMarkup())
      currentMarkup()->addDelim(Syntax::dOPT);
   return ContentToken::opt;
  case tokenRep:
    if (currentMarkup())
      currentMarkup()->addDelim(Syntax::dREP);
    return ContentToken::rep;
  default:
    currentInput()->ungetToken();
    return ContentToken::none;
  }
}

Boolean Parser::parseMinimumLiteral(Boolean lita, Text &text)
{
  return parseLiteral(lita ? mlitaMode : mlitMode, mlitMode,
		      Syntax::referenceQuantity(Syntax::qLITLEN),
		      ParserMessages::minimumLiteralLength,
		      literalSingleSpace|literalMinimumData
		      |(eventsWanted().wantPrologMarkup()
			? literalDelimInfo
			: 0),
		      text);
}

Boolean Parser::parseSystemIdentifier(Boolean lita, Text &text)
{
  return parseLiteral(lita ? slitaMode : slitMode, slitMode, syntax().litlen(),
		      ParserMessages::systemIdentifierLength,
		      (eventsWanted().wantPrologMarkup()
			? literalDelimInfo
			: 0), text);
}

Boolean Parser::parseParameterLiteral(Boolean lita, Text &text)
{
  return parseLiteral(lita ? plitaMode : plitMode, pliteMode, syntax().litlen(), 
		      ParserMessages::parameterLiteralLength,
		      (eventsWanted().wantPrologMarkup()
		       ? literalDelimInfo
		       : 0),
		      text);
}

Boolean Parser::parseDataTagParameterLiteral(Boolean lita, Text &text)
{
  return parseLiteral(lita ? plitaMode : plitMode, pliteMode,
		      syntax().dtemplen(), 
		      ParserMessages::dataTagPatternLiteralLength,
		      literalDataTag
		      | (eventsWanted().wantPrologMarkup()
			 ? literalDelimInfo
			 : 0),
		      text);
}

Boolean Parser::parseIndicatedReservedName(const AllowedParams &allow,
					   Param &parm)
{
  Syntax::ReservedName rn;
  if (!getIndicatedReservedName(&rn))
    return 0;
  if (!allow.reservedName(rn)) {
    message(ParserMessages::invalidReservedName,
	    StringMessageArg(currentToken()));
    return 0;
  }
  parm.type = Param::indicatedReservedName + rn;
  return 1;
}

Boolean Parser::parseReservedName(const AllowedParams &allow,
				  Param &parm)
{
  Syntax::ReservedName rn;
  if (!getReservedName(&rn))
    return 0;
  if (!allow.reservedName(rn)) {
    message(ParserMessages::invalidReservedName,
	    StringMessageArg(syntax().reservedName(rn)));
    return 0;
  }
  parm.type = Param::reservedName + rn;
  return 1;
}


Boolean Parser::parseAttributeValueParam(Param &parm)
{
  extendNameToken(syntax().litlen() > syntax().normsep()
		  ? syntax().litlen() - syntax().normsep()
		  : 0,
		  ParserMessages::attributeValueLength);
  parm.type = Param::attributeValue;
  Text text;
  text.addChars(currentInput()->currentTokenStart(),
		currentInput()->currentTokenLength(),
		currentLocation());
  text.swap(parm.literalText);
  if (currentMarkup())
    currentMarkup()->addAttributeValue(currentInput());
  return 1;
}

Boolean Parser::getIndicatedReservedName(Syntax::ReservedName *result)
{
  if (currentMarkup())
    currentMarkup()->addDelim(Syntax::dRNI);
  InputSource *in = currentInput();
  in->startToken();
  if (!syntax().isNameStartCharacter(in->tokenChar(messenger()))) {
    message(ParserMessages::rniNameStart);
    return 0;
  }
  extendNameToken(syntax().namelen(), ParserMessages::nameLength);
  StringC &buffer = nameBuffer();
  getCurrentToken(syntax().generalSubstTable(), buffer);
  if (!syntax().lookupReservedName(buffer, result)) {
    message(ParserMessages::noSuchReservedName, StringMessageArg(buffer));
    return 0;
  }
  if (currentMarkup())
    currentMarkup()->addReservedName(*result, currentInput());
  return 1;
}

Boolean Parser::getReservedName(Syntax::ReservedName *result)
{
  extendNameToken(syntax().namelen(), ParserMessages::nameLength);
  StringC &buffer = nameBuffer();
  getCurrentToken(syntax().generalSubstTable(), buffer);
  if (!syntax().lookupReservedName(buffer, result)) {
    message(ParserMessages::noSuchReservedName, StringMessageArg(buffer));
    return 0;
  }
  if (currentMarkup())
    currentMarkup()->addReservedName(*result, currentInput());
  return 1;
}


#ifdef SP_NAMESPACE
}
#endif