package eu.simuline.util.sgml;
   
   import eu.simuline.util.ListMap;
   
   import java.io.Reader;
   import java.io.IOException;
   
  import org.xml.sax.ContentHandler;
  import org.xml.sax.Locator;
  import org.xml.sax.Attributes;
  import org.xml.sax.InputSource;
  import org.xml.sax.SAXException;
  import org.xml.sax.SAXParseException;
  import java.util.Locale;
  
  /**
   * A rudimentary <code>SGML</code> parser with something like a SAX-api. 
   *
   * @author <a href="mailto:ernst.reissner@simuline.eu">Ernst Reissner</a>
   * @version 1.0
   */
  public final class SGMLParser {
  
      private static final String QUOTE_DOT = "\". ";
      private static final char   SYMB_EQ   = '=';
      private static final char   SYMB_COMMENT = '-';
      private static final char   SYMB_TAG = '<';
  
      /* --------------------------------------------------------------------- *
       * inner classes                                                         *
       * --------------------------------------------------------------------- */
  
      /**
       * A <code>ContentHandler</code> which simply ignores all events.  
       * May be used for debugging. 
       */
      static class TrivialContentHandler implements ContentHandler {
  
  	/** <!-- api-docs inherited from interface implemented.  -->*/ 
  	public void setDocumentLocator(Locator locator) {
  	    // is empty. 
  	}
  
  	public void startDocument() throws SAXException {
  	    // is empty. 
  	}
  
  	public void endDocument() throws SAXException {
  	    // is empty. 
  	}
  
  	public void startPrefixMapping(String prefix,
  				       String uri)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void endPrefixMapping(String prefix)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void startElement(String namespaceURI,
  				 String localName,
  				 String qName,
  				 Attributes atts)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void endElement(String namespaceURI,
  			       String localName,
  			       String qName)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void characters(char[] chr,
  			       int start,
  			       int length)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void ignorableWhitespace(char[] chr,
  					int start,
  					int length)
  	    throws SAXException {
  	    // is empty. 
  	}
  
  	public void processingInstruction(String target,
  					  String data)
  	    throws SAXException {
  	    // is empty. 
  	}
  
 	public void skippedEntity(String name)
 	    throws SAXException {
 	    // is empty. 
 	}
     } // class TrivialContentHandler 
 
 
     /**
      * An **** partial **** implementation 
      * of the SAX-interface <code>Attributes</code> 
      * which allows to set name-value-pairs by method {@link #addAttribute}. 
      */
     class AttributesWrapper {
 
 	/* ----------------------------------------------------------------- *
 	 * fields                                                            *
 	 * ----------------------------------------------------------------- */
 
 	/**
 	 * See {@link AttributesImpl#name2value}. 
 	 */
 	private final ListMap<String, String> name2value;
 
 	/* ----------------------------------------------------------------- *
 	 * constructors                                                      *
 	 * ----------------------------------------------------------------- */
 
 	/**
 	 * Creates a new empty <code>AttributesWrapper</code> 
 	 * which represents an empty attribute list. 
 	 */
 	AttributesWrapper() {
 	    this.name2value = new ListMap<String, String>();
 	}
 
 	/* ----------------------------------------------------------------- *
 	 * methods                                                           *
 	 * ----------------------------------------------------------------- */
 
 
 	/**
 	 * Adds an attribute with the given name and value. 
 	 *
 	 * @param attName 
 	 *     the <code>String</code> representation 
 	 *     of the name of an attribute. 
 	 * @param attValue 
 	 *     the value of an attribute as a <code>String</code>. 
 	 *     If no value is provided, 
 	 *     this is {@link AttributesImpl#NO_VALUE}. 
 	 */
 	void addAttribute(String attName, String attValue) {
 	    String oldAttValue = this.name2value
 		.put(attName, attValue);
 	    if (oldAttValue != null) {
 		// Here, the attribute has occured before. 
 		SGMLParser.this.parseExceptionHandler
 		    .foundMultipleAttribute(attName,
 					    oldAttValue);
 	    }
  	}
 
 	Attributes getAttributes() {
 	    return new AttributesImpl(this.name2value);
 	}
     } // class AttributesWrapper 
 
 
     /**
      * Provides a single method which decides whether the given character 
      * passes a certain test. 
      */
     interface CharTester {
 
 	/**
 	 * Returns whether the given character <code>ch</code> 
 	 * passes the test given by this <code>CharTester</code>. 
 	 *
 	 * @param chr 
 	 *    an <code>int</code> value representing a character. 
 	 * @return the <code>boolean</code> 
 	 *    signifying whether the given character <code>ch</code> 
 	 *    passes the test given by this <code>CharTester</code>. 
 	 */
 	boolean testChar(char chr);
 
     } // interface CharTester 
 
     /**
      * Tests for blank, <code>/</code>, <code>&gt;</code>. 
      */
     private static final CharTester TEST_BLANK_GT_SLASH = new CharTester() {
 	public boolean testChar(char chr) {
 	    return Character.isWhitespace(chr) 
 		|| chr == '/' 
 		|| chr == '>';
 	}
     };
 
     /**
      * Tests for blank or <code>&gt;</code>. 
      */
     private static final CharTester TEST_BLANK_GT = new CharTester() {
 	public boolean testChar(char chr) {
 	    return Character.isWhitespace(chr) 
 		|| chr == '>';
 	}
     };
 
     /*
      * Tests for <code>/</code> or <code>&gt;</code>. 
      */
 /*
     private static final CharTester TEST_GT_SLASH = new CharTester() {
 	    public boolean testChar(char ch) {
 		return ch == '/' 
 		    || ch == '>';
 	    }
 	};
 */
 
     /**
      * Tests for <code>&lt;</code>. 
      */
     private static final CharTester TEST_LT = new CharTester() {
 	    public boolean testChar(char chr) {
 		return chr == '<';
 	    }
 	};
 
     /**
      * Tests for <code>&gt;</code>. 
      */
     private static final CharTester TEST_GT = new CharTester() {
 	    public boolean testChar(char chr) {
 		return chr == '>';
 	    }
 	};
 
     /**
      * Tests for <code>=</code> and for <code>&gt;</code>. 
      */
     private static final CharTester TEST_BLANK_EQUALS_GT = new CharTester() {
 	    public boolean testChar(char chr) {
 		return Character.isWhitespace(chr) 
 		    || chr == '='
 		    || chr == '>';
 	    }
 	};
 
     /**
      * Tests for whitespace. 
      */
     private static final CharTester TEST_NO_WHITESPACE = new CharTester() {
 	    public boolean testChar(char chr) {
 		return !Character.isWhitespace(chr);
 	    }
 	};
 
     /**
      * Tests for quote both for<code>'</code> and for <code>"</code>. 
      */
 /*
     private static final CharTester TEST_QUOTE = new CharTester() {
 	    public boolean testChar(char chr) {
 		return chr == '\'' 
 		    || chr == '"';
 	    }
 	};
 */
 
     /*
      * Tests for end of comment <code>--></code>. 
      * This tests for a sequence of characters 
      * and confirms after having read the last one. 
      */
     private static final CharTester TEST_END_OF_COMMENT = new CharTester() {
 
 	    /**
 	     * Contains the sequence <code>--></code> 
 	     * representing the end of a comment. 
 	     */
 	    static final String END_OF_COMMENT = "-->";
 
 	    /**
 	     * Contains the index in {@link #END_OF_COMMENT} 
 	     * which is to be compared next by {@link #testChar}. 
 	     */
 	    private int index = 0;
 
 	    /**
 	     * Returns whether the last characters tested 
 	     * are <code>--></code>. 
 	     *
 	     * @param chr 
 	     *    a <code>char</code>. 
 	     * @return
 	     *    whether the last characters tested 
 	     *    including <code>char</code> are <code>--></code>. 
 	     *    In particular, if less than three characters are read 
 	     *    this is <code>false</code>. 
 	     */
 	    public boolean testChar(char chr) {		
 		if (END_OF_COMMENT.charAt(index++) == chr) {
 		    if (this.index == END_OF_COMMENT.length() - 1) {
 			this.index = 0;
 			return true;
 		    } else {
 			return false;
 		    }
 		} else {
 		    this.index = 0;
 		    return false;
 		}
 	    }
 	}; // TEST_END_OF_COMMENT 
 
     /**
      * A <code>CharTester</code> which allows to specify 
      * the character which passes the test. 
      */
     static class SpecCharTester implements CharTester {
 
 	/**
 	 * The character which passes the test {@link #testChar}. 
 	 */
 	private char chr;
 
 	/**
 	 * Sets {@link #chr} to the specified character value. 
 	 *
 	 * @param chr 
 	 *    a <code>char</code> value. 
 	 */
 	void setChar(char chr) {
 	    this.chr = chr;
 	}
 
 	/**
 	 * Returns whether the given character coincides with {@link #chr}. 
 	 *
 	 * @param chr 
 	 *    a <code>char</code> value. 
 	 * @return 
 	 *    whether <code>ch</code> coincides with {@link #chr}. 
 	 */
 	public boolean testChar(char chr) {
 	    return chr == this.chr;
 	}
     } // SpecCharTester 
 
     /**
      * Tests for a specified character. 
      * This is used for quotes which allow the cases 
      * <code>'</code> and <code>"</code>. 
      *
      * @see XMLsGMLspecifica#parseAttribute
      */
     private static final SpecCharTester TEST_SPEC = new SpecCharTester();
 
     /**
      * Class which buffers the read stream. 
      */
     static class Buffer {
 
 	/* ----------------------------------------------------------------- *
 	 * fields                                                            *
 	 * ----------------------------------------------------------------- */
 
 	/**
 	 * The reader buffered. 
 	 */
 	private final Reader reader;
 
 	/**
 	 * The current buffer. 
 	 * The current parts to be read start with 
 	 * <code>bufferArray[{@link #start}]</code> and end with 
 	 * <code>bufferArray[{@link #end}]</code>, exclusively. 
 	 */
 	private final char[] bufferArray;
 
 	/**
 	 * The first index in {@link #bufferArray} 
 	 * read in from {@link #reader} but not returned 
 	 * by {@link #readArray} or {@link #readChar}. 
 	 */
 	private int start;
 
 	/**
 	 * Set by {@link #readArray} and read by {@link #getStartAndMove}. 
 	 * When invoking {@link #readArray} <code>newStart</code> 
 	 * is set to {@link #start} and increased 
 	 * by the number of read charactersincreases. 
 	 * Then {@link #getStartAndMove} updates {@link #start} 
 	 * according to <code>newStart</code>. 
 	 */
 	private int newStart;
 
 	/**
 	 * The first index in {@link #bufferArray} not read 
 	 * from {@link #reader} 
 	 * or <code>-1</code> if the end of the stream is reached. 
 	 * This means that <code>bufferArray[end]</code> 
 	 * either does not exist or at least is not significant. 
 	 */
 	private int end;
 
 	/* ----------------------------------------------------------------- *
 	 * constructors                                                      *
 	 * ----------------------------------------------------------------- */
 
 	/**
 	 * Creates a new <code>Buffer</code> from the given reader 
 	 * with the given size. 
 	 *
 	 * @param reader 
 	 *    the <code>Reader</code> to be buffered. 
 	 * @param length 
 	 *    the length of the buffer. 
 	 * @exception IOException 
 	 *    if an error occurs 
 	 */
 	Buffer(Reader reader, int length) throws IOException {
 	    this.reader = reader;
 	    this.bufferArray = new char[length];
 	    this.start = 0;
 	    this.end = this.start; // signifies: reading necessary. 
 	}
 
 	/* ----------------------------------------------------------------- *
 	 * methods                                                           *
 	 * ----------------------------------------------------------------- */
 
 
 	/**
 	 * Returns whether this buffer is currently empty. 
 	 * When this is the case and someone tries to read further characters 
 	 * this will lead to a trial 
 	 * to read further pieces from {@link #reader}. 
 	 *
 	 * @return a <code>boolean</code> value 
 	 *    signifying whether this buffer is currently empty. 
 	 */
 	boolean isEmpty() {
 	    return this.end == this.start;
 	}
 
 	/**
 	 * Returns whether the end of the stream is reached. 
 	 *
 	 * @return 
 	 *    a <code>boolean</code> specifying 
 	 *    whether the end of the stream is reached. 
 	 */
 	boolean reachedEOS() {
 	    return this.end == -1;
 	}
 
 	/**
 	 * Reads a single <code>char</code> and returns it. 
 	 *
 	 * @return 
 	 *    an <code>int</code> value 
 	 *    which is either the next <code>char</code> read in 
 	 *    or <code>-1</code> which signifies the end of the stream. 
 	 * @exception IOException 
 	 *    if an error occurs
 	 */
 	int readChar() throws IOException {
 	    if (reachedEOS()) {
 		return -1;
 	    }
 	    if (isEmpty()) {
 		this.start = 0;
 		this.end = this.reader.read(this.bufferArray);
 		if (reachedEOS()) {
 		    return -1;
 		}
 	    }
 	    return this.bufferArray[this.start++];
 	}
 
 	/**
 	 * Reads an array from {@link #reader}. 
 	 * As a side effect, writes the field {@link #newStart}. 
 	 * Also, if the portion of {@link #bufferArray} 
 	 * to be read, i.e. between {@link #start} and {@link #end}, 
 	 * is empty, a new portion is buffered. 
 	 *
 	 * @param charTester 
 	 *    a <code>CharTester</code> which signifies 
 	 *    when to end reading from the buffer. 
 	 * @return 
 	 *    an <code>int</code> signifying the number of <code>char</code>s 
 	 *    read or <code>-1</code> which signifies the end of the stream. 
 	 *    It is read to the next &lt; or, if there is none, 
 	 *    to the end of the stream. 
 	 *    Thus there is a difference between the return values 
 	 *    <code>-1</code> and <code>0</code>. 
 	 * @exception IOException 
 	 *    if an error occurs
 	 */
 	int readArray(CharTester charTester) throws IOException {
 	    if (reachedEOS()) {
 		return -1;
 	    }
 	    if (isEmpty()) {
 		this.start = 0;
 		this.end = this.reader.read(this.bufferArray);
 //System.out.println("read: "+this.end);
 		if (reachedEOS()) {
 		    return -1;
 		}
 	    }
 
 	    for (int i = this.start; i < this.end; i++) {
 		if (charTester.testChar(this.bufferArray[i])) {
 		    // found match described by charTester 
 		    this.newStart = i;
 		    return this.newStart - this.start;
 		}
 	    }
             // Here, the test always failed. 
 	    this.newStart = this.end;
 	    return this.end - this.start;
 	}
 
 	/**
 	 * Describe <code>readStringBuffer</code> method here.
 	 *
 	 * @param charTester 
 	 *    a <code>CharTester</code> which determines 
 	 *    the first character not read 
 	 *    into the resulting <code>StringBuffer</code>. 
 	 * @param elementName
 	 *    a <code>String</code> which determines 
 	 *    the element under consideration. 
 	 *    This is only used for generating the message of a 
 	 *    <code>SAXParseException</code>. 
 	 *    <p>
 	 *    Allowed values: {@link #START_TAG}, {@link #END_TAG}, 
 	 *    {@link #PROC_INSTR}, 
 	 *    {@link #ATTR_NAME}, {@link #WHITESP_IN_ATTR} 
 	 *    and {@link #ATTR_VALUE}. ****** comment and &lt;!element missing. 
 	 * @return 
 	 *    a <code>StringBuffer</code> containing characters 
 	 *    starting with the current one until one 
 	 *    <code>charTester</code> returns <code>true</code>. 
 	 * @exception IOException 
 	 *    if an io-error occurs
 	 * @exception SAXParseException 
 	 *    if the parser faces the end of the stream 
 	 *    while scanning the current element. 
 	 */
 	StringBuffer readStringBuffer(CharTester charTester, 
 				      String elementName) 
 	    throws IOException, SAXParseException {
 
 	    StringBuffer qName = new StringBuffer();
 	    int numRead = 0;
 	    do {
 		numRead = readArray(charTester);
 		if (numRead == -1) {
 		    throw new SAXParseException
 			("End of stream while scanning " 
 			 + elementName + ". " 
 			 + "Read so far: \"" 
 			 + qName + QUOTE_DOT, null);
 		}
 		qName.append(getChars(),
 			     getStartAndMove(),
 			     numRead);
 	    } while (isEmpty());
 
 	    return qName;
 	}
 
 	/**
 	 * Returns the buffer of <code>char</code>s. 
 	 *
 	 * @return 
 	 *    the <code>char[]</code> {@link #bufferArray}. 
 	 */
 	char[] getChars() {
 	    return this.bufferArray;
 	}
 
 	/**
 	 * Moves {@link #newStart} to {@link #start} 
 	 * and returns the old value of {@link #start}. 
 	 *
 	 * @return 
 	 *    the old <code>int</code> value of {@link #start}. 
 	 */
 	int getStartAndMove() {
 	    int ret = this.start;
 	    this.start = this.newStart;
 	    return ret;
 	}
 
 	/**
 	 * Get method for {@link #start}. 
 	 *
 	 * @return {@link #start}
 	 */
 	int getStart() {
 	    return this.start;
 	}
 
 	/**
 	 * Get method for {@link #end}. 
 	 *
 	 * @return {@link #end}
 	 */
 	int getEnd() {
 	    return this.end;
 	}
     } // class Buffer
 
     /**
      * Provides a bunch of methods fpr parsing 
      * with implementations specific to xml and sgml. 
      */
     interface XMLsGMLspecifica {
 	// **** SGMLParser.this.currChar must be the character 
 	// after the attribute list. 
 	/**
 	 * Parses one attribute and adds it to the given attribute list. 
 	 *
 	 * @param attributes 
 	 *    an <code>AttributesImpl</code> 
 	 *    to which the attribute parsed is added. 
 	 * @exception IOException 
 	 *    if an io-error occurs
 	 * @exception SAXException 
 	 *    if a syntactical error occurs
 	 */
 	void parseAttribute(AttributesWrapper attributes) 
 	    throws IOException, SAXException;
 
 	/**
 	 * Parses a comment or any declaration 
 	 * starting with <code>&lt;!...</code> and notifying the handler. 
 	 *
 	 * @exception IOException 
 	 *    if an io-error occurs
 	 * @exception SAXException 
 	 *    if a syntactical error occurs
 	 */
 	void parseCommentElemTypeDecl() throws IOException, SAXException;
 
 	/**
 	 * Parses a processing instruction or any declaration 
 	 * starting with <code>&lt;?...</code> and notifying the handler. 
 	 *
 	 * @exception IOException 
 	 *    if an io-error occurs
 	 * @exception SAXException 
 	 *    if a syntactical error occurs
 	 */
 	void parseExtProcessingInstruction() throws IOException, SAXException;
 
     } // interface XML_SGML_Specifica 
 
     /**
      * Contains the <code>HTML</code>-specific part of the parser. 
      */
     private final XMLsGMLspecifica htmlAttributeParser = 
     new XMLsGMLspecifica() {
 
 	public void parseAttribute(AttributesWrapper attributes) 
 	    throws IOException, SAXException {
 	    String attName;
 	    String attValue;
 	    StringBuffer qName;
 
 	    // Parse attribute name 
 	    qName = SGMLParser.this.buffer.
 		readStringBuffer(TEST_BLANK_EQUALS_GT, ATTR_NAME);
 	    qName.insert(0, (char) SGMLParser.this.currChar);
 	    attName = qName.toString().toLowerCase(Locale.ENGLISH);
 //System.out.println("attName: |"+attName+"|");
 
 	    // Here, the attribute may have a value or not. 
 
 	    // Skip whitespace either after having parsed the attribute 
 	    // or between its name and its value. 
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 	    if (Character.isWhitespace((char) SGMLParser.this.currChar)) {
 		qName = SGMLParser.this.buffer.
 		    readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 		SGMLParser.this.currChar = 
 		    SGMLParser.this.buffer.readChar(); //NOPMD
 	    }
 
 	    // Here is the decision whether a value is provided or not. 
 	    if (SGMLParser.this.currChar != SYMB_EQ) {
 		// Here, no value may be given 
 		attributes.addAttribute(attName, AttributesImpl.NO_VALUE);
 //System.out.println("attName: |"+attName+"|");
 //System.out.println("noValue@@"+(char)SGMLParser.this.currChar+"|");
 		return;
 	    }
 	    // Here, clearly a value must follow 
 
 	    // Skip whitespaces 
 	    qName = SGMLParser.this.buffer.
 		readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 
 	    // Parse the attribute value. 
 	    switch (SGMLParser.this.currChar) {
 		case '\'':
 		    // fall through 
 		case '"':
 		    // the attribute value is quoted. 
 		    char quote = (char) SGMLParser.this.currChar;
 		    TEST_SPEC.setChar(quote);
 		    //SGMLParser.this.currChar = 
 		    //	SGMLParser.this.buffer.readChar();
 
 //System.out.println("quote@@"+SGMLParser.this.currChar);
 		    qName = new StringBuffer();
 		    while (true) {
 			qName.append(SGMLParser.this.buffer.
 				     readStringBuffer(TEST_SPEC, ATTR_VALUE));
 			if (qName.length() != 0 
 			    && qName.charAt(qName.length() - 1) == '\\') {
 			    qName.setCharAt(qName.length() - 1, quote);
 			} else {
 			    // read the quote 
 			    SGMLParser.this.currChar = //NOPMD
 				SGMLParser.this.buffer.readChar();
 			    break;
 			}
 		    }
 		    break;
 		default:
 //System.out.println("no quote@@"+SGMLParser.this.currChar);
 		    // the attribute value is not quoted. 
 		    qName = SGMLParser.this.buffer.
 			readStringBuffer(TEST_BLANK_GT, ATTR_VALUE);
 		    qName.insert(0, (char) SGMLParser.this.currChar);
 		    break;
 	    }
 	    // read the character after the attribute value 
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 
 	    attValue = qName.toString();
 	    attributes.addAttribute(attName, attValue);
 //System.out.println("attName: |"+attName+"|");
 //System.out.println("attValue: |"+attValue+"|");
 	}
 
 	public void parseCommentElemTypeDecl() 
 	    throws IOException, SAXException {
 //System.out.println("comment?");
 
 	    SGMLParser.this.currChar = //NOPMD
 		SGMLParser.this.buffer.readChar();
 	    if (SGMLParser.this.currChar != SYMB_COMMENT) {
 		//int numRead = 
 		SGMLParser.this.buffer.readArray(TEST_GT);
 		SGMLParser.this.buffer.getStartAndMove();
 		return;
 	    }
 	    // Here, object starts with "<!-....."
 
 	    SGMLParser.this.currChar = //NOPMD
 		SGMLParser.this.buffer.readChar();
 	    if (SGMLParser.this.currChar != SYMB_COMMENT) {
 		throw new SAXParseException
 		    ("Comments must start with \"<!--\" but found " 
 		     + "\"<!-" + (char) SGMLParser.this.currChar + QUOTE_DOT, 
 		     null);
 	    }
 //System.out.println("comment!");
 
 	    int numRead = 0;
 	    do {
 		numRead = SGMLParser.this.buffer
 		    .readArray(TEST_END_OF_COMMENT);
 		if (numRead == -1) {
 		    StringBuffer qName = new StringBuffer();
 		    qName.append(SGMLParser.this.buffer.getChars(),
 				 SGMLParser.this.buffer.getStartAndMove(),
 				 numRead);
 		    throw new SAXParseException
 			("End of stream while scanning comment. " 
 			 + "Recently read: \"" + qName + QUOTE_DOT, 
 			 null);
 		}
 
 		SGMLParser.this.buffer.getStartAndMove();
 	    } while (SGMLParser.this.buffer.isEmpty());
 /*
 	    StringBuffer qName = new StringBuffer();
 	    qName.append(SGMLParser.this.buffer.getChars(),
 			 SGMLParser.this.buffer.getStartAndMove(),
 			 numRead);
 
 	    System.out.println("read so far: |"+qName+"|");
 */
 
 	    SGMLParser.this.buffer.getStartAndMove();	    
 	    // NO NOTIFY!!
 	}
 
 	public void parseExtProcessingInstruction() 
 	    throws IOException, SAXException {
 	    parseStartOrStartEndTag();
 	}
     }; // htmlXML_SGML_Specifica
 
     /**
      * Contains the <code>XML</code>-specific part of the parser. 
      */
     private final XMLsGMLspecifica xmlAttributeParser = 
     new XMLsGMLspecifica() {
 
 	public void parseAttribute(AttributesWrapper attributes) 
 	    throws IOException, SAXException {
 	    String attName;
 	    String attValue;
 	    StringBuffer qName;
 
 	    // Parse attribute name 
 	    qName = SGMLParser.this.buffer.
 		readStringBuffer(TEST_BLANK_EQUALS_GT, ATTR_NAME);
 	    qName.insert(0, (char) SGMLParser.this.currChar);
 	    attName = qName.toString();
 //System.out.println("attName: |"+attName+"|");
 
 	    // Here, the attribute may have a value or not. 
 
 	    // Skip whitespace either after having parsed the attribute 
 	    // or between its name and its value. 
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 	    if (Character.isWhitespace((char) SGMLParser.this.currChar)) {
 		qName = SGMLParser.this.buffer.
 		    readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 		SGMLParser.this.currChar = 
 		    SGMLParser.this.buffer.readChar(); //NOPMD
 	    }
 
 	    // Here is the decision whether a value is provided or not. 
 	    if (SGMLParser.this.currChar != SYMB_EQ) {
 		// Here, a value is missing. 
 		throw new SAXParseException
 		    ("Missing value for attribute \"" 
 		     + attName + QUOTE_DOT, null);
 	    }
 	    // Here, clearly a value must follow ****
 
 
 	    // Skip whitespaces 
 	    qName = SGMLParser.this.buffer.
 		readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 
 
 	    // Parse the attribute value. 
 	    switch (SGMLParser.this.currChar) {
 		case '\'':
 		    // fall through 
 		case '"':
 		    // the attribute value is quoted. 
 		    char quote = (char) SGMLParser.this.currChar;
 		    TEST_SPEC.setChar(quote);
 		    //SGMLParser.this.currChar = 
 		    //	SGMLParser.this.buffer.readChar();
 
 //System.out.println("quote@@"+SGMLParser.this.currChar);
 		    qName = new StringBuffer();
 		    while (true) {
 			qName.append(SGMLParser.this.buffer.
 				     readStringBuffer(TEST_SPEC, ATTR_VALUE));
 			if (qName.length() != 0 
 			    && qName.charAt(qName.length() - 1) == '\\') {
 			    qName.setCharAt(qName.length() - 1, quote);
 			} else {
 			    // read the quote 
 			    SGMLParser.this.currChar = //NOPMD
 				SGMLParser.this.buffer.readChar();
 			    break;
 			}
 		    }
 		    break;
 		default:
 		    throw new SAXParseException
 			("Value of attribute \"" + attName + 
 			 "\" is not quoted. ",
 			 null);
 	    }
 	    // read the character after the attribute value 
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 
 	    attValue = qName.toString();
 	    attributes.addAttribute(attName, attValue);
 //System.out.println("attName: |"+attName+"|");
 //System.out.println("attValue: |"+attValue+"|");
 	}
 
 	public void parseCommentElemTypeDecl() 
 	    throws IOException, SAXException {
 	    // ******** comments will not work that way!!!*****
 
 	    SGMLParser.this.buffer.readStringBuffer(TEST_GT, PROC_INSTR);
 	    //**** comment
 	    // Here, also the empty processing instruction or comment 
 	    // would be possible. 
 	
 	    //this.buffer.getStart();
 //System.out.println("-qName: |"+qName+"|");
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 	    //assert this.currChar == '>';
 	    SGMLParser.this.contentHandler.processingInstruction(null, null);
 	}
 
 
 	public void parseExtProcessingInstruction() 
 	    throws IOException, SAXException {
 
 	    //StringBuffer qName = 
 	    SGMLParser.this.buffer.readStringBuffer(TEST_GT, PROC_INSTR);
 	    // Here, also the empty processing instruction would be possible. 
 	
 	    //this.buffer.getStart();
 //System.out.println("-qName: |"+qName+"|");
 	    SGMLParser.this.currChar = 
 		SGMLParser.this.buffer.readChar(); //NOPMD
 	    //assert this.currChar == '>';
 	    SGMLParser.this.contentHandler.processingInstruction(null, null);
 	}
 
     }; // xmlXML_SGML_Specifica
 
     /* --------------------------------------------------------------------- *
      * class constants                                                       *
      * --------------------------------------------------------------------- */
 
 
     /**
      * The size of the buffer used internally. 
      * This must be at least <code>1</code>. 
      * I found no significant difference in speed when increasing this number. 
      * The buffer coming from a stream from a URL seems to hav maximal size 
      * of <code>1448</code> whereas for file streams there seems no bound. 
      * In the cases considered, the file is read in as a whole. 
      */
     private static final int BUFFER_SIZE = 999999;
 
     // for notification of a sax parse exception with Buffer.readStringBuffer. 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String START_TAG = "start tag";
 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String END_TAG = "end tag";
 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String PROC_INSTR = "processing instruction";
 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String ATTR_NAME = "attribute name";
 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String WHITESP_IN_ATTR = "whitespace in attribute";
 
     /**
      * Short string representation of the object currently parsed. 
      * Contains the specific part of the message of the exception 
      * that may be thrown by {@link SGMLParser.Buffer#readStringBuffer}. 
      */
     private static final String ATTR_VALUE = "attribute value";
 
     /* --------------------------------------------------------------------- *
      * fields                                                                *
      * --------------------------------------------------------------------- */
 
     /**
      * Contains class with methods specific for xml and sgml, respectively. 
      */
     private XMLsGMLspecifica xmlSgmlSpecifica = htmlAttributeParser;
 
     /**
      * The current character or <code>-1</code> 
      * to signfy the end of the stream. 
      */
     private int currChar;
 
     /**
      * The <code>ContentHandler</code> registered. 
      */
     private ContentHandler contentHandler;
 
     /**
      * The <code>ParseExceptionHandler</code> registered. 
      */
     private ParseExceptionHandler parseExceptionHandler;
 
     /**
      * The buffer of the input stream. 
      */
     private Buffer buffer;
 
     /* --------------------------------------------------------------------- *
      * constructors                                                          *
      * --------------------------------------------------------------------- */
 
     /**
      * Creates a new <code>SGMLParser</code> 
      * with the default handlers for content and exceptions. 
      */
     @SuppressWarnings("checkstyle:nowhitespaceafter")
     public SGMLParser() {
 	this.       contentHandler = new TrivialContentHandler();
 	this.parseExceptionHandler = new ParseExceptionHandler.Impl();
     }
 
     /* --------------------------------------------------------------------- *
      * methods                                                               *
      * --------------------------------------------------------------------- */
 
 
     /**
      * Parses the <code>InputSource</code> given 
      * but delegates everything inside a tag or a processing instruction 
      * to {@link #parseTagOrPI}. 
      *
      * @param src 
      *    an <code>InputSource</code>. 
      * @exception IOException if an error occurs
      * @exception SAXException if an error occurs
      */
     void parse(InputSource src) throws IOException, SAXException {
 	parse(src.getCharacterStream());
     }
 
     /**
      * Parses the given <code>InputStream</code>. 
      *
      * @param reader 
      *     an <code>Reader</code> sequentializing an SGML document. 
      * @exception IOException 
      *     if an error reading the stream occurs. 
      * @exception SAXException 
      *    if an error with the sgml-syntax occurs. 
      */
     public void parse(Reader reader) throws IOException, SAXException {
 
 	this.buffer = new Buffer(reader, BUFFER_SIZE);
 	int numRead = this.buffer.readArray(TEST_LT);
 	// notify handler that first part of document was successfully read. 
 	this.contentHandler.startDocument();
 	while (numRead != -1) {
 	    this.currChar = this.buffer.readChar(); // the '<' char? 
 	    if (this.currChar == SYMB_TAG) {
 		// a tag or a PI. 
 		numRead = parseTagOrPI();
 	    } else {
 		// either characters or ignoreableWhitespace
 		numRead = parseText();
 	    }
 	}
 	// Here, the document is finished. 
 	this.contentHandler.endDocument();
     }
 
     /**
      * Parses everything outside a tag, a processing instruction, ... 
      * everything within brackets <code>&lt;</code> and <code>&gt;</code>. 
      * ***** Missing: distinction between notification 
      * of characters and whitespace. ****
      *
      * @exception IOException 
      *     if an error reading the stream occurs. 
      * @exception SAXException 
      *    if an error with the sgml-syntax occurs. 
      * @see #parseTagOrPI
      */
     private int parseText() throws IOException, SAXException {
 	int numRead = this.buffer.readArray(TEST_LT);
 	if (numRead != -1) {
 /*
   System.out.println("text: |"+new String(buffer.getChars(),
   buffer.getStartAndMove(),
   numRead)+"|");
 */
 	    this.contentHandler.characters(this.buffer.getChars(),
 					   this.buffer.getStartAndMove(),
 					   numRead);
 	}
 	
 //buffer.getStartAndMove();
 	return numRead;
     }
 
     /**
      * Parses an end-tag notifying the underlying handler. 
      *
      * @exception IOException 
      *     if an error reading the stream occurs. 
      * @exception SAXException 
      *    if an error with the sgml-syntax occurs. 
      */
     void parseEndTag() throws IOException, SAXException {
 	StringBuffer qName = this.buffer.readStringBuffer(TEST_GT, END_TAG);
 	// Here, also the empty tag would be possible. 
 	
 	//this.buffer.getStart();
 //System.out.println("end tag: |"+qName+"|");
 	this.currChar = this.buffer.readChar();
 	//assert this.currChar == '>';
 	this.contentHandler.endElement(null,
 				       null,
 				       qName.toString());
 	this.currChar = this.buffer.readChar();
     }
 /*
   public void parseCommentElemTypeDecl() 
   throws IOException, SAXException {
   // ******** comments will not work that way!!!*****
 
   StringBuffer qName = this.buffer
   .readStringBuffer(TEST_GT, PROC_INSTR);//**** comment
   // Here, also the empty processing instruction or comment 
   // would be possible. 
 	
   //this.buffer.getStart();
 //System.out.println("-qName: |"+qName+"|");
 this.currChar = this.buffer.readChar();
 //assert this.currChar == '>';
 this.handler.processingInstruction(null, null);
 }
 */
 
     /**
      * Parses a start-tag or, for xml, an empty tag. 
      *
      * @exception IOException 
      *     if an error reading the stream occurs. 
      * @exception SAXException 
      *    if an error with the sgml-syntax occurs. 
      */
     void parseStartOrStartEndTag() throws IOException, SAXException {
 
 	// ***** Better read the name of the tag and 
 	// then single out problems with chars by a handler
 	if (!Character.isLetter((char) this.currChar)) {
 	    this.parseExceptionHandler
 		.foundIllegalCharInTag((char) this.currChar);
 	    // Ignore the previously read char. 
 	    this.currChar = this.buffer.readChar();
 	}
 
 
 	StringBuffer qName = this.buffer
 	    .readStringBuffer(TEST_BLANK_GT_SLASH, START_TAG);
 	qName.insert(0, (char) this.currChar);
 	// Here, also the empty tag would be possible. 
 //System.out.println("start tag: |"+qName+"|");
 
 	// Skip whitespaces 
 	this.currChar = this.buffer.readChar();
 	while (Character.isWhitespace((char) this.currChar)) {
 	    this.buffer.
 		readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 	    this.currChar = this.buffer.readChar();
 	}
 
 	AttributesWrapper attributesWrapper = new AttributesWrapper();
 	// Here, either /, > or an attribute occurs
 //System.out.println("this.currChar: |"+(char)this.currChar+"|");
 	while (this.currChar != '/' && this.currChar != '>') {
 	    // parse the following attribute list
 	    this.xmlSgmlSpecifica.parseAttribute(attributesWrapper);
 
 	    // Skip whitespaces 
 	    while (Character.isWhitespace((char) this.currChar)) {
 		this.buffer
 		    .readStringBuffer(TEST_NO_WHITESPACE, WHITESP_IN_ATTR);
 		this.currChar = this.buffer.readChar();
 	    }
 	} // end parsing attribute list 
 //System.out.println("-this.currChar: |"+(char)this.currChar+"|");
 
 
 	Attributes attributes = attributesWrapper.getAttributes();
 	switch (this.currChar) {
 	    case '/':
 		// start-end-tag called empty tag
 
 		// skip illegal characters between "/" and ">" **** 
 	skipped: while (true) { //NOPMD
 		    this.currChar = this.buffer.readChar();
 		    switch (this.currChar) {
 			case '>':
 			    break skipped;
 			case -1:
 			    this.parseExceptionHandler
 				.foundUnexpectedEndOfDocument();
 				break;
 			default:
 			    this.parseExceptionHandler
 				.foundCharAfterEndOfEndTag
 				((char) this.currChar);
 			    break;
 		    } // switch
 		}
 
 		this.contentHandler.startElement(null,
 						 null,
 						 qName.toString(),
 						 attributes);
 		this.contentHandler.endElement(null,
 					       null,
 					       qName.toString());
 		break;
 	    case '>':
 		this.contentHandler.startElement(null,
 						 null,
 						 qName.toString(),
 						 attributes);
 		break;
 	    default:
 		throw new SAXParseException
 		    ("Expected finishing tag \"" + qName 
 		     + "\" with character '/' or '>' " 
 		     + "but found '" + (char) this.currChar + "'. ", null);
 	}
     }
 
     /**
      * Parses everything within a tag, a processing instruction, ... 
      * everything within brackets <code>&lt;</code> and <code>&gt;</code>. 
      *
      * @see #parseText
      */
     private int parseTagOrPI() throws IOException, SAXException {
 //System.out.println("parseTagOrPI");
 	
 	this.currChar = this.buffer.readChar();
 	//this.currChar = this.reader.read();
 	switch (this.currChar) {
 	    case '/':
 		// parsing an end-tag 
 		parseEndTag();
 		//this.currChar = this.reader.read();
 		break;
 	    case '!':
 		// parsing no tag at all: 
 		// a processing instruction or a comment
 		this.xmlSgmlSpecifica.parseCommentElemTypeDecl();
 		//this.currChar = this.reader.read();
 		break;
 	    case '?':
 		// parsing no tag at all: 
 		// a processing instruction or a comment
 		this.xmlSgmlSpecifica.parseExtProcessingInstruction();
 		//this.currChar = this.reader.read();
 		break;
 	    default:
 		// parsing a start-tag or an empty-element-tag 
 		parseStartOrStartEndTag();
 		break;
 	} // end of switch ()
 	//this.currChar = this.buffer.readChar();
 	// Here, the buffer is ready 
 	// to read the first character. after the generalized tag. 
 
 //System.out.println("read last: |"+(char)this.currChar+"|");
 //System.out.println("read last: |"+      this.currChar+"|");
 
 	return 1;
     }
 
     /**
      * Sets {@link #contentHandler}. 
      *
      * @param contentHandler 
      *    a <code>ContentHandler</code>. 
      */
     public void setContentHandler(ContentHandler contentHandler) {
 	if (isXMLParser()) {
 	    this.contentHandler = contentHandler;
 	} else {
 	    this.contentHandler = new SGMLFilter(contentHandler);
 	}
     }
 
     /**
      * Returns {@link #contentHandler}. 
      *
      * @return
      *    the <code>ContentHandler</code> {@link #contentHandler}. 
      */
     public ContentHandler getContentHandler() {
 	if (isXMLParser()) {
 	    return this.contentHandler;
 	} else {
 	    return ((SGMLFilter) this.contentHandler).getWrapped();
 	}
     }
 
     /**
      * Sets {@link #parseExceptionHandler}. 
      *
      * @param peHandler 
      *    a <code>ParseExceptionHandler</code>. 
      */
     public void setExceptionHandler(ParseExceptionHandler peHandler) {
 	this.parseExceptionHandler = peHandler;
     }
 
     /**
      * Returns {@link #parseExceptionHandler}. 
      *
      * @return
      *    the <code>ContentHandler</code> {@link #parseExceptionHandler}. 
      */
     public ParseExceptionHandler getExceptionHandler() {
 	return this.parseExceptionHandler;
     }
 
     /**
      * Sets whether this parser is used as an xml-parser. 
      * If this is false, which is the default, 
      * it s an html-parser. 
      *
      * @param xml 
      *    a <code>boolean</code> value signifying 
      *    whether this parser will be used as an xml-parser in the sequel. 
      * @return 
      *    a <code>boolean</code> value signifying 
      *    whether before invoking this method 
      *    this parser was used as an xml-parser
      */
     public boolean parseXML(boolean xml) {
 	boolean result = this.xmlSgmlSpecifica == xmlAttributeParser;
 	this.xmlSgmlSpecifica = xml 
 	    ? xmlAttributeParser 
 	    : htmlAttributeParser;
 	return result;
     }
 
     public boolean isXMLParser() {
 	return this.xmlSgmlSpecifica == xmlAttributeParser;
     }
 }