// file: $isip/class/mmedia/XMLParser/xp_03.cc
// version: $Id: xp_03.cc 10284 2005-10-27 20:52:14Z wholland $

#if defined(HAVE_EXPAT)

// ISIP include files
//
#include "XMLParser.h"

// method: OnPostCreate
//
// arguments: none
//
// return: none
//
// Invoked by Create
//
void XMLParser::OnPostCreate () {
  
  // Enable all the event routines we want
  //
  EnableStartElementHandler ();
  EnableEndElementHandler ();
  
  // Note: EnableElementHandler will do both start and end
  //
  EnableCharacterDataHandler ();
}

// method: OnStartElement
//
// arguments:
//  XML_Char* pszName: (input) a character string containing
//    the value of the XML tag
//  XML_Char** papszAttrs: (input) a character string array
//    containing the XML tag's attributes, in alternating
//    name,value,name,value, fashion.
//
// return: none
//
// This method is called every time an XML start tag is encountered while
// parsing.
//
void XMLParser::OnStartElement (const XML_Char *psz_name,
				const XML_Char **papsz_attrs) {
  
  // store the start element in a String
  //
  SysString sysstart_element((byte8*)psz_name);
  String start_element;
  start_element.assign(sysstart_element);
  
  XMLToken temp_token;
  
  // store the element's type, value, and attributes
  //
  temp_token.init(XMLToken::START_TAG, start_element,
		   (const byte8**)papsz_attrs);
  
  // add to vector
  //
  addToken(temp_token);
  
  return;
}

// method: OnEndElement
//
// arguments:
//  XML_Char* psz_name: (input) a character string containing
//    the value of the XML tag
//
// return: none
//
// This method is called every time an XML end tag is encountered while
// parsing.
//
void XMLParser::OnEndElement (const XML_Char *psz_name) {
  
  // store the end element in a String
  //
  SysString sysend_element((byte8*)psz_name);
  String end_element;
  end_element.assign(sysend_element);

  // declare a token to store the parsed tag
  //
  XMLToken temp_token;
  
  // store the element's type and  value
  //
  temp_token.init(XMLToken::END_TAG, end_element);
  
  // add the token to the vector
  //
  addToken(temp_token);
  
  return;
}

// method: OnCharacterData
//
// arguments:
//   XML_Char* parsedData: (input) a string containing the character data found.
//     this also frequently contains unvalid characters following
//     the character data.
//   int nlength: (input) an integer specifying the number of valid characters
//     in parsedData.
//
// return: none
//
// this method is called whenever parsing encounters a piece of
// character data not within < > delimiters. It creates an XMLToken
// to store the character data, and adds the token to the current rule
//
//
void XMLParser::OnCharacterData (const XML_Char *parsed_data, int nLength) {
  
  // store the character data element in a String
  //
  SysString syschar_data((byte8*)parsed_data);
  String char_data;
  char_data.assign(syschar_data);

  // declare a token to store the parsed tag
  //
  XMLToken temp_token;
  
  // declare a string to hold the token's value
  //
  String string_token;
  
  // declare a buffer to store only the fraction of the
  // parsed_data that we want to use. the buffer parsed_data
  // contains not only this character data, but sometimes tags
  // following it as well.
  //
  byte8 pszData[Sof::BUFFER_SIZE];
  
  // declare a variable to store the position of a character
  // in a string while tokenizing the string.
  //
  int32 pos = 0;
  
  // get only the part of this buffer that is character data
  // as specified by nlength, (the buffer
  // sometimes has an end tag in it)
  //
  char_data.getBuffer(pszData, nLength+1, SysChar::ENCODE_UTF8);
  char_data.assign(pszData);
  
  // do not generate tokens for plain white space
  // (sometimes CDATA handler is called for "\n"
  // chars between tags)
  //
  if(char_data.firstNotSpace() == Integral::NO_POS) {
    return;
  }
  
  // normalize CDATA white space according to SRGS 1.0
  //       
  // remove internal space characters, and replace with a single
  // space. internal end line characters are removed already due
  // to the line-by-line parsing.
  //
  // here, multiple sequential tokens in a single cdata parse
  // will be seperated into seperate tokens
  // (e.g. "a b c" => "a" "b" "c")
  //
  // tokenize parses the next token starting from pos, and stores
  // it in string_token
  //
  while(char_data.tokenize(string_token, pos)) {
    
    // store the element's type, value, and attributes
    // 
    temp_token.init(XMLToken::CDATA, string_token);
    
    // add the token to the vector
    //
    addToken(temp_token);
    
  } // end looping over tokens within single cdata
  
  return;
}

#endif