xmlparse.cpp

C++ class libraries for network-centric, portable applications, integrated perfectly with the C++ St

#define poolFinish(pool) ((pool)->start = (pool)->ptr)
#define poolAppendChar(pool, c) \
  (((pool)->ptr == (pool)->end && !poolGrow(pool)) \
   ? 0 \
   : ((*((pool)->ptr)++ = c), 1))

struct XML_ParserStruct {
  /* The first member must be userData so that the XML_GetUserData
     macro works. */
  void *m_userData;
  void *m_handlerArg;
  char *m_buffer;
  const XML_Memory_Handling_Suite m_mem;
  /* first character to be parsed */
  const char *m_bufferPtr;
  /* past last character to be parsed */
  char *m_bufferEnd;
  /* allocated end of buffer */
  const char *m_bufferLim;
  XML_Index m_parseEndByteIndex;
  const char *m_parseEndPtr;
  XML_Char *m_dataBuf;
  XML_Char *m_dataBufEnd;
  XML_StartElementHandler m_startElementHandler;
  XML_EndElementHandler m_endElementHandler;
  XML_CharacterDataHandler m_characterDataHandler;
  XML_ProcessingInstructionHandler m_processingInstructionHandler;
  XML_CommentHandler m_commentHandler;
  XML_StartCdataSectionHandler m_startCdataSectionHandler;
  XML_EndCdataSectionHandler m_endCdataSectionHandler;
  XML_DefaultHandler m_defaultHandler;
  XML_StartDoctypeDeclHandler m_startDoctypeDeclHandler;
  XML_EndDoctypeDeclHandler m_endDoctypeDeclHandler;
  XML_UnparsedEntityDeclHandler m_unparsedEntityDeclHandler;
  XML_NotationDeclHandler m_notationDeclHandler;
  XML_StartNamespaceDeclHandler m_startNamespaceDeclHandler;
  XML_EndNamespaceDeclHandler m_endNamespaceDeclHandler;
  XML_NotStandaloneHandler m_notStandaloneHandler;
  XML_ExternalEntityRefHandler m_externalEntityRefHandler;
  XML_Parser m_externalEntityRefHandlerArg;
  XML_SkippedEntityHandler m_skippedEntityHandler;
  XML_UnknownEncodingHandler m_unknownEncodingHandler;
  XML_ElementDeclHandler m_elementDeclHandler;
  XML_AttlistDeclHandler m_attlistDeclHandler;
  XML_EntityDeclHandler m_entityDeclHandler;
  XML_XmlDeclHandler m_xmlDeclHandler;
  const ENCODING *m_encoding;
  INIT_ENCODING m_initEncoding;
  const ENCODING *m_internalEncoding;
  const XML_Char *m_protocolEncodingName;
  XML_Bool m_ns;
  XML_Bool m_ns_triplets;
  void *m_unknownEncodingMem;
  void *m_unknownEncodingData;
  void *m_unknownEncodingHandlerData;
  void (XMLCALL *m_unknownEncodingRelease)(void *);
  PROLOG_STATE m_prologState;
  Processor *m_processor;
  enum XML_Error m_errorCode;
  const char *m_eventPtr;
  const char *m_eventEndPtr;
  const char *m_positionPtr;
  OPEN_INTERNAL_ENTITY *m_openInternalEntities;
  OPEN_INTERNAL_ENTITY *m_freeInternalEntities;
  XML_Bool m_defaultExpandInternalEntities;
  int m_tagLevel;
  ENTITY *m_declEntity;
  const XML_Char *m_doctypeName;
  const XML_Char *m_doctypeSysid;
  const XML_Char *m_doctypePubid;
  const XML_Char *m_declAttributeType;
  const XML_Char *m_declNotationName;
  const XML_Char *m_declNotationPublicId;
  ELEMENT_TYPE *m_declElementType;
  ATTRIBUTE_ID *m_declAttributeId;
  XML_Bool m_declAttributeIsCdata;
  XML_Bool m_declAttributeIsId;
  DTD *m_dtd;
  const XML_Char *m_curBase;
  TAG *m_tagStack;
  TAG *m_freeTagList;
  BINDING *m_inheritedBindings;
  BINDING *m_freeBindingList;
  int m_attsSize;
  int m_nSpecifiedAtts;
  int m_idAttIndex;
  ATTRIBUTE *m_atts;
  NS_ATT *m_nsAtts;
  unsigned long m_nsAttsVersion;
  unsigned char m_nsAttsPower;
  POSITION m_position;
  STRING_POOL m_tempPool;
  STRING_POOL m_temp2Pool;
  char *m_groupConnector;
  unsigned int m_groupSize;
  XML_Char m_namespaceSeparator;
  XML_Parser m_parentParser;
  XML_ParsingStatus m_parsingStatus;
#ifdef XML_DTD
  XML_Bool m_isParamEntity;
  XML_Bool m_useForeignDTD;
  enum XML_ParamEntityParsing m_paramEntityParsing;
#endif
};

#define MALLOC(s) (parser->m_mem.malloc_fcn((s)))
#define REALLOC(p,s) (parser->m_mem.realloc_fcn((p),(s)))
#define FREE(p) (parser->m_mem.free_fcn((p)))

#define userData (parser->m_userData)
#define handlerArg (parser->m_handlerArg)
#define startElementHandler (parser->m_startElementHandler)
#define endElementHandler (parser->m_endElementHandler)
#define characterDataHandler (parser->m_characterDataHandler)
#define processingInstructionHandler \
        (parser->m_processingInstructionHandler)
#define commentHandler (parser->m_commentHandler)
#define startCdataSectionHandler \
        (parser->m_startCdataSectionHandler)
#define endCdataSectionHandler (parser->m_endCdataSectionHandler)
#define defaultHandler (parser->m_defaultHandler)
#define startDoctypeDeclHandler (parser->m_startDoctypeDeclHandler)
#define endDoctypeDeclHandler (parser->m_endDoctypeDeclHandler)
#define unparsedEntityDeclHandler \
        (parser->m_unparsedEntityDeclHandler)
#define notationDeclHandler (parser->m_notationDeclHandler)
#define startNamespaceDeclHandler \
        (parser->m_startNamespaceDeclHandler)
#define endNamespaceDeclHandler (parser->m_endNamespaceDeclHandler)
#define notStandaloneHandler (parser->m_notStandaloneHandler)
#define externalEntityRefHandler \
        (parser->m_externalEntityRefHandler)
#define externalEntityRefHandlerArg \
        (parser->m_externalEntityRefHandlerArg)
#define internalEntityRefHandler \
        (parser->m_internalEntityRefHandler)
#define skippedEntityHandler (parser->m_skippedEntityHandler)
#define unknownEncodingHandler (parser->m_unknownEncodingHandler)
#define elementDeclHandler (parser->m_elementDeclHandler)
#define attlistDeclHandler (parser->m_attlistDeclHandler)
#define entityDeclHandler (parser->m_entityDeclHandler)
#define xmlDeclHandler (parser->m_xmlDeclHandler)
#define encoding (parser->m_encoding)
#define initEncoding (parser->m_initEncoding)
#define internalEncoding (parser->m_internalEncoding)
#define unknownEncodingMem (parser->m_unknownEncodingMem)
#define unknownEncodingData (parser->m_unknownEncodingData)
#define unknownEncodingHandlerData \
  (parser->m_unknownEncodingHandlerData)
#define unknownEncodingRelease (parser->m_unknownEncodingRelease)
#define protocolEncodingName (parser->m_protocolEncodingName)
#define ns (parser->m_ns)
#define ns_triplets (parser->m_ns_triplets)
#define prologState (parser->m_prologState)
#define processor (parser->m_processor)
#define errorCode (parser->m_errorCode)
#define eventPtr (parser->m_eventPtr)
#define eventEndPtr (parser->m_eventEndPtr)
#define positionPtr (parser->m_positionPtr)
#define position (parser->m_position)
#define openInternalEntities (parser->m_openInternalEntities)
#define freeInternalEntities (parser->m_freeInternalEntities)
#define defaultExpandInternalEntities \
        (parser->m_defaultExpandInternalEntities)
#define tagLevel (parser->m_tagLevel)
#define buffer (parser->m_buffer)
#define bufferPtr (parser->m_bufferPtr)
#define bufferEnd (parser->m_bufferEnd)
#define parseEndByteIndex (parser->m_parseEndByteIndex)
#define parseEndPtr (parser->m_parseEndPtr)
#define bufferLim (parser->m_bufferLim)
#define dataBuf (parser->m_dataBuf)
#define dataBufEnd (parser->m_dataBufEnd)
#define _dtd (parser->m_dtd)
#define curBase (parser->m_curBase)
#define declEntity (parser->m_declEntity)
#define doctypeName (parser->m_doctypeName)
#define doctypeSysid (parser->m_doctypeSysid)
#define doctypePubid (parser->m_doctypePubid)
#define declAttributeType (parser->m_declAttributeType)
#define declNotationName (parser->m_declNotationName)
#define declNotationPublicId (parser->m_declNotationPublicId)
#define declElementType (parser->m_declElementType)
#define declAttributeId (parser->m_declAttributeId)
#define declAttributeIsCdata (parser->m_declAttributeIsCdata)
#define declAttributeIsId (parser->m_declAttributeIsId)
#define freeTagList (parser->m_freeTagList)
#define freeBindingList (parser->m_freeBindingList)
#define inheritedBindings (parser->m_inheritedBindings)
#define tagStack (parser->m_tagStack)
#define atts (parser->m_atts)
#define attsSize (parser->m_attsSize)
#define nSpecifiedAtts (parser->m_nSpecifiedAtts)
#define idAttIndex (parser->m_idAttIndex)
#define nsAtts (parser->m_nsAtts)
#define nsAttsVersion (parser->m_nsAttsVersion)
#define nsAttsPower (parser->m_nsAttsPower)
#define tempPool (parser->m_tempPool)
#define temp2Pool (parser->m_temp2Pool)
#define groupConnector (parser->m_groupConnector)
#define groupSize (parser->m_groupSize)
#define namespaceSeparator (parser->m_namespaceSeparator)
#define parentParser (parser->m_parentParser)
#define ps_parsing (parser->m_parsingStatus.parsing)
#define ps_finalBuffer (parser->m_parsingStatus.finalBuffer)
#ifdef XML_DTD
#define isParamEntity (parser->m_isParamEntity)
#define useForeignDTD (parser->m_useForeignDTD)
#define paramEntityParsing (parser->m_paramEntityParsing)
#endif /* XML_DTD */

XML_Parser XMLCALL
XML_ParserCreate(const XML_Char *encodingName)
{
  return XML_ParserCreate_MM(encodingName, NULL, NULL);
}

XML_Parser XMLCALL
XML_ParserCreateNS(const XML_Char *encodingName, XML_Char nsSep)
{
  XML_Char tmp[2];
  *tmp = nsSep;
  return XML_ParserCreate_MM(encodingName, NULL, tmp);
}

static const XML_Char implicitContext[] = {
  'x', 'm', 'l', '=', 'h', 't', 't', 'p', ':', '/', '/',
  'w', 'w', 'w', '.', 'w', '3', '.', 'o', 'r', 'g', '/',
  'X', 'M', 'L', '/', '1', '9', '9', '8', '/',
  'n', 'a', 'm', 'e', 's', 'p', 'a', 'c', 'e', '\0'
};

XML_Parser XMLCALL
XML_ParserCreate_MM(const XML_Char *encodingName,
                    const XML_Memory_Handling_Suite *memsuite,
                    const XML_Char *nameSep)
{
  XML_Parser parser = parserCreate(encodingName, memsuite, nameSep, NULL);
  if (parser != NULL && ns) {
    /* implicit context only set for root parser, since child
       parsers (i.e. external entity parsers) will inherit it
    */
    if (!setContext(parser, implicitContext)) {
      XML_ParserFree(parser);
      return NULL;
    }
  }
  return parser;
}

static XML_Parser
parserCreate(const XML_Char *encodingName,
             const XML_Memory_Handling_Suite *memsuite,
             const XML_Char *nameSep,
             DTD *dtd)
{
  XML_Parser parser;

  if (memsuite) {
    XML_Memory_Handling_Suite *mtemp;
    parser = (XML_Parser)
      memsuite->malloc_fcn(sizeof(struct XML_ParserStruct));
    if (parser != NULL) {
      mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem);
      mtemp->malloc_fcn = memsuite->malloc_fcn;
      mtemp->realloc_fcn = memsuite->realloc_fcn;
      mtemp->free_fcn = memsuite->free_fcn;
    }
  }
  else {
    XML_Memory_Handling_Suite *mtemp;
    parser = (XML_Parser)malloc(sizeof(struct XML_ParserStruct));
    if (parser != NULL) {
      mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem);
      mtemp->malloc_fcn = malloc;
      mtemp->realloc_fcn = realloc;
      mtemp->free_fcn = free;
    }
  }

  if (!parser)
    return parser;

  buffer = NULL;
  bufferLim = NULL;

  attsSize = INIT_ATTS_SIZE;
  atts = (ATTRIBUTE *)MALLOC(attsSize * sizeof(ATTRIBUTE));
  if (atts == NULL) {
    FREE(parser);
    return NULL;
  }
  dataBuf = (XML_Char *)MALLOC(INIT_DATA_BUF_SIZE * sizeof(XML_Char));
  if (dataBuf == NULL) {
    FREE(atts);
    FREE(parser);
    return NULL;
  }
  dataBufEnd = dataBuf + INIT_DATA_BUF_SIZE;

  if (dtd)
    _dtd = dtd;
  else {
    _dtd = dtdCreate(&parser->m_mem);
    if (_dtd == NULL) {
      FREE(dataBuf);
      FREE(atts);
      FREE(parser);
      return NULL;
    }
  }

  freeBindingList = NULL;
  freeTagList = NULL;
  freeInternalEntities = NULL;

  groupSize = 0;
  groupConnector = NULL;

  unknownEncodingHandler = NULL;
  unknownEncodingHandlerData = NULL;

  namespaceSeparator = '!';
  ns = XML_FALSE;
  ns_triplets = XML_FALSE;

  nsAtts = NULL;
  nsAttsVersion = 0;
  nsAttsPower = 0;

  poolInit(&tempPool, &(parser->m_mem));
  poolInit(&temp2Pool, &(parser->m_mem));
  parserInit(parser, encodingName);

  if (encodingName && !protocolEncodingName) {
    XML_ParserFree(parser);
    return NULL;
  }

  if (nameSep) {
    ns = XML_TRUE;
    internalEncoding = XmlGetInternalEncodingNS();
    namespaceSeparator = *nameSep;
  }
  else {
    internalEncoding = XmlGetInternalEncoding();
  }

  return parser;
}

static void
parserInit(XML_Parser parser, const XML_Char *encodingName)
{
  processor = prologInitProcessor;
  XmlPrologStateInit(&prologState);
  protocolEncodingName = (encodingName != NULL
                          ? poolCopyString(&tempPool, encodingName)
                          : NULL);
  curBase = NULL;
  XmlInitEncoding(&initEncoding, &encoding, 0);
  userData = NULL;
  handlerArg = NULL;
  startElementHandler = NULL;
  endElementHandler = NULL;
  characterDataHandler = NULL;
  processingInstructionHandler = NULL;
  commentHandler = NULL;
  startCdataSectionHandler = NULL;
  endCdataSectionHandler = NULL;
  defaultHandler = NULL;
  startDoctypeDeclHandler = NULL;
  endDoctypeDeclHandler = NULL;
  unparsedEntityDeclHandler = NULL;
  notationDeclHandler = NULL;
  startNamespaceDeclHandler = NULL;
  endNamespaceDeclHandler = NULL;
  notStandaloneHandler = NULL;
  externalEntityRefHandler = NULL;
  externalEntityRefHandlerArg = parser;
  skippedEntityHandler = NULL;
  elementDeclHandler = NULL;
  attlistDeclHandler = NULL;
  entityDeclHandler = NULL;
  xmlDeclHandler = NULL;
  bufferPtr = buffer;
  bufferEnd = buffer;
  parseEndByteIndex = 0;
  parseEndPtr = NULL;
  declElementType = NULL;
  declAttributeId = NULL;
  declEntity = NULL;
  doctypeName = NULL;
  doctypeSysid = NULL;
  doctypePubid = NULL;
  declAttributeType = NULL;
  declNotationName = NULL;
  declNotationPublicId = NULL;
  declAttributeIsCdata = XML_FALSE;
  declAttributeIsId = XML_FALSE;
  memset(&position, 0, sizeof(POSITION));
  errorCode = XML_ERROR_NONE;
  eventPtr = NULL;
  eventEndPtr = NULL;
  positionPtr = NULL;
  openInternalEntities = NULL;
  defaultExpandInternalEntities = XML_TRUE;
  tagLevel = 0;
  tagStack = NULL;
  inheritedBindings = NULL;
  nSpecifiedAtts = 0;
  unknownEncodingMem = NULL;
  unknownEncodingRelease = NULL;
  unknownEncodingData = NULL;
  parentParser = NULL;
  ps_parsing = XML_INITIALIZED;
#ifdef XML_DTD
  isParamEntity = XML_FALSE;
  useForeignDTD = XML_FALSE;
  paramEntityParsing = XML_PARAM_ENTITY_PARSING_NEVER;
#endif
}

/* moves list of bindings to freeBindingList */
static void FASTCALL
moveToFreeBindingList(XML_Parser parser, BINDING *bindings)
{
  while (bindings) {
    BINDING *b = bindings;
    bindings = bindings->nextTagBinding;
    b->nextTagBinding = freeBindingList;
    freeBindingList = b;
  }
}

XML_Bool XMLCALL
XML_ParserReset(XML_Parser parser, const XML_Char *encodingName)
{
  TAG *tStk;
  OPEN_INTERNAL_ENTITY *openEntityList;
  if (parentParser)