xmlparser.java

好东西啊！你看看就知道了

     * <pre>
     * [41] Attribute ::= Name Eq AttValue
     * </pre>
     * @param name The name of the attribute's element.
     * @see SAXDriver#attribute
     */
    private void parseAttribute (String name)
    throws Exception
    {
	String aname;
	int type;
	String value;
	int flags = LIT_ATTRIBUTE |  LIT_ENTITY_REF;

	// Read the attribute name.
	aname = readNmtoken (true);
	type = getAttributeType (name, aname);

	// Parse '='
	parseEq ();

	// Read the value, normalizing whitespace
	// unless it is CDATA.
	if (type == ATTRIBUTE_CDATA || type == ATTRIBUTE_UNDECLARED) {
	    value = readLiteral (flags);
	} else {
	    value = readLiteral (flags | LIT_NORMALIZE);
	}

	// WFC: no duplicate attributes
	for (int i = 0; i < tagAttributePos; i++)
	    if (aname.equals (tagAttributes [i]))
		error ("duplicate attribute", aname, null);

	// Inform the handler about the
	// attribute.
	handler.attribute (aname, value, true);
	dataBufferPos = 0;

	// Note that the attribute has been
	// specified.
	if (tagAttributePos == tagAttributes.length) {
	    String newAttrib[] = new String [tagAttributes.length * 2];
	    System.arraycopy (tagAttributes, 0, newAttrib, 0, tagAttributePos);
	    tagAttributes = newAttrib;
	}
	tagAttributes [tagAttributePos++] = aname;
    }


    /**
     * Parse an equals sign surrounded by optional whitespace.
     * <pre>
     * [25] Eq ::= S? '=' S?
     * </pre>
     */
    private void parseEq ()
    throws SAXException, IOException
    {
	skipWhitespace ();
	require ('=');
	skipWhitespace ();
    }


    /**
     * Parse an end tag.
     * <pre>
     * [42] ETag ::= '</' Name S? '>'
     * </pre>
     * <p>NOTE: parseContent () chains to here, we already read the
     * "&lt;/".
     */
    private void parseETag ()
    throws Exception
    {
	require (currentElement);
	skipWhitespace ();
	require ('>');
	handler.endElement (currentElement);
	// not re-reporting any SAXException re bogus end tags,
	// even though that diagnostic might be clearer ...
    }


    /**
     * Parse the content of an element.
     * <pre>
     * [43] content ::= (element | CharData | Reference
     *		| CDSect | PI | Comment)*
     * [67] Reference ::= EntityRef | CharRef
     * </pre>
     * <p> NOTE: consumes ETtag.
     */
    private void parseContent ()
    throws Exception
    {
	String data;
	char c;

	while (true) {

	    switch (currentElementContent) {
	    case CONTENT_ANY:
	    case CONTENT_MIXED:
		parseCharData ();
		break;
	    case CONTENT_ELEMENTS:
		parseWhitespace ();
		break;
	    }

	    // Handle delimiters
	    c = readCh ();
	    switch (c) {

	    case '&': 			// Found "&"

		c = readCh ();
		if (c == '#') {
		    parseCharRef ();
		} else {
		    unread (c);
		    parseEntityRef (true);
		}
		break;

	      case '<': 			// Found "<"
		dataBufferFlush ();
		c = readCh ();
		switch (c) {
		  case '!': 			// Found "<!"
		    c = readCh ();
		    switch (c) {
		      case '-': 		// Found "<!-"
			require ('-');
			parseComment ();
			break;
		      case '[': 		// Found "<!["
			require ("CDATA[");
			handler.startCDATA ();
			inCDATA = true;
			parseCDSect ();
			inCDATA = false;
			handler.endCDATA ();
			break;
		      default:
			error ("expected comment or CDATA section", c, null);
			break;
		    }
		    break;

		  case '?': 		// Found "<?"
		    parsePI ();
		    break;

		  case '/': 		// Found "</"
		    parseETag ();
		    return;

		  default: 		// Found "<" followed by something else
		    unread (c);
		    parseElement ();
		    break;
		}
	    }
	}
    }


    /**
     * Parse an element type declaration.
     * <pre>
     * [45] elementdecl ::= '&lt;!ELEMENT' S Name S contentspec S? '&gt;'
     * </pre>
     * <p> NOTE: the '&lt;!ELEMENT' has already been read.
     */
    private void parseElementdecl ()
    throws Exception
    {
	String name;

	requireWhitespace ();
	// Read the element type name.
	name = readNmtoken (true);

	requireWhitespace ();
	// Read the content model.
	parseContentspec (name);

	skipWhitespace ();
	require ('>');
    }


    /**
     * Content specification.
     * <pre>
     * [46] contentspec ::= 'EMPTY' | 'ANY' | Mixed | elements
     * </pre>
     */
    private void parseContentspec (String name)
    throws Exception
    {
	if (tryRead ("EMPTY")) {
	    setElement (name, CONTENT_EMPTY, null);
	    return;
	} else if (tryRead ("ANY")) {
	    setElement (name, CONTENT_ANY, null);
	    return;
	} else {
	    require ('(');
	    dataBufferAppend ('(');
	    skipWhitespace ();
	    if (tryRead ("#PCDATA")) {
		dataBufferAppend ("#PCDATA");
		parseMixed ();
		setElement (name, CONTENT_MIXED, dataBufferToString ());
	    } else {
		parseElements ();
		setElement (name, CONTENT_ELEMENTS, dataBufferToString ());
	    }
	}
    }


    /**
     * Parse an element-content model.
     * <pre>
     * [47] elements ::= (choice | seq) ('?' | '*' | '+')?
     * [49] choice ::= '(' S? cp (S? '|' S? cp)+ S? ')'
     * [50] seq ::= '(' S? cp (S? ',' S? cp)* S? ')'
     * </pre>
     *
     * <p> NOTE: the opening '(' and S have already been read.
     */
    private void parseElements ()
    throws Exception
    {
	char c;
	char sep;

	// Parse the first content particle
	skipWhitespace ();
	parseCp ();

	// Check for end or for a separator.
	skipWhitespace ();
	c = readCh ();
	switch (c) {
	case ')':
	    dataBufferAppend (')');
	    c = readCh ();
	    switch (c) {
	    case '*':
	    case '+':
	    case '?':
		dataBufferAppend (c);
		break;
	    default:
		unread (c);
	    }
	    return;
	case ',': 			// Register the separator.
	case '|':
	    sep = c;
	    dataBufferAppend (c);
	    break;
	default:
	    error ("bad separator in content model", c, null);
	    return;
	}

	// Parse the rest of the content model.
	while (true) {
	    skipWhitespace ();
	    parseCp ();
	    skipWhitespace ();
	    c = readCh ();
	    if (c == ')') {
		dataBufferAppend (')');
		break;
	    } else if (c != sep) {
		error ("bad separator in content model", c, null);
		return;
	    } else {
		dataBufferAppend (c);
	    }
	}

	// Check for the occurrence indicator.
	c = readCh ();
	switch (c) {
	case '?':
	case '*':
	case '+':
	    dataBufferAppend (c);
	    return;
	default:
	    unread (c);
	    return;
	}
    }


    /**
     * Parse a content particle.
     * <pre>
     * [48] cp ::= (Name | choice | seq) ('?' | '*' | '+')?
     * </pre>
     */
    private void parseCp ()
    throws Exception
    {
	char c;

	if (tryRead ('(')) {
	    dataBufferAppend ('(');
	    parseElements ();
	} else {
	    dataBufferAppend (readNmtoken (true));
	    c = readCh ();
	    switch (c) {
	    case '?':
	    case '*':
	    case '+':
		dataBufferAppend (c);
		break;
	    default:
		unread (c);
		break;
	    }
	}
    }


    /**
     * Parse mixed content.
     * <pre>
     * [51] Mixed ::= '(' S? ( '#PCDATA' (S? '|' S? Name)*) S? ')*'
     *	      | '(' S? ('#PCDATA') S? ')'
     * </pre>
     */
    private void parseMixed ()
    throws Exception
    {
	char c;

	// Check for PCDATA alone.
	skipWhitespace ();
	if (tryRead (')')) {
	    dataBufferAppend (")*");
	    tryRead ('*');
	    return;
	}

	// Parse mixed content.
	skipWhitespace ();
	while (!tryRead (")*")) {
	    require ('|');
	    dataBufferAppend ('|');
	    skipWhitespace ();
	    dataBufferAppend (readNmtoken (true));
	    skipWhitespace ();
	}
	dataBufferAppend (")*");
    }


    /**
     * Parse an attribute list declaration.
     * <pre>
     * [52] AttlistDecl ::= '&lt;!ATTLIST' S Name AttDef* S? '&gt;'
     * </pre>
     * <p>NOTE: the '&lt;!ATTLIST' has already been read.
     */
    private void parseAttlistDecl ()
    throws Exception
    {
	String elementName;

	requireWhitespace ();
	elementName = readNmtoken (true);
	boolean white = tryWhitespace ();
	while (!tryRead ('>')) {
	    if (!white)
		error ("whitespace required before attribute definition");
	    parseAttDef (elementName);
	    white = tryWhitespace ();
	}
    }


    /**
     * Parse a single attribute definition.
     * <pre>
     * [53] AttDef ::= S Name S AttType S DefaultDecl
     * </pre>
     */
    private void parseAttDef (String elementName)
    throws Exception
    {
	String name;
	int type;
	String enum = null;

	// Read the attribute name.
	name = readNmtoken (true);

	// Read the attribute type.
	requireWhitespace ();
	type = readAttType ();

	// Get the string of enumerated values
	// if necessary.
	if (type == ATTRIBUTE_ENUMERATED || type == ATTRIBUTE_NOTATION) {
	    enum = dataBufferToString ();
	}

	// Read the default value.
	requireWhitespace ();
	parseDefault (elementName, name, type, enum);
    }


    /**
     * Parse the attribute type.
     * <pre>
     * [54] AttType ::= StringType | TokenizedType | EnumeratedType
     * [55] StringType ::= 'CDATA'
     * [56] TokenizedType ::= 'ID' | 'IDREF' | 'IDREFS' | 'ENTITY'
     *		| 'ENTITIES' | 'NMTOKEN' | 'NMTOKENS'
     * [57] EnumeratedType ::= NotationType | Enumeration
     * </pre>
     */
    private int readAttType ()
    throws Exception
    {
	String typeString;
	Integer type;

	if (tryRead ('(')) {
	    parseEnumeration (false);
	    return ATTRIBUTE_ENUMERATED;
	} else {
	    typeString = readNmtoken (true);
	    if (typeString.equals ("NOTATION")) {
		parseNotationType ();
	    }
	    type = (Integer) attributeTypeHash.get (typeString);
	    if (type == null) {
		error ("illegal attribute type", typeString, null);
		return ATTRIBUTE_UNDECLARED;
	    } else {
		return type.intValue ();
	    }
	}
    }


    /**
     * Parse an enumeration.
     * <pre>
     * [59] Enumeration ::= '(' S? Nmtoken (S? '|' S? Nmtoken)* S? ')'
     * </pre>
     * <p>NOTE: the '(' has already been read.
     */
    private void parseEnumeration (boolean isNames)
    throws Exception
    {
	char c;

	dataBufferAppend ('(');

	// Read the first token.
	skipWhitespace ();
	dataBufferAppend (readNmtoken (isNames));
	// Read the remaining tokens.
	skipWhitespace ();
	while (!tryRead (')')) {
	    require ('|');
	    dataBufferAppend ('|');
	    skipWhitespace ();
	    dataBufferAppend (readNmtoken (isNames));
	    skipWhitespace ();
	}
	dataBufferAppend (')');
    }


    /**
     * Parse a notation type for an attribute.
     * <pre>
     * [58] NotationType ::= 'NOTATION' S '(' S? NameNtoks
     *		(S? '|' S? name)* S? ')'
     * </pre>
     * <p>NOTE: the 'NOTATION' has already been read
     */
    private void parseNotationType ()
    throws Exception
    {
	requireWhitespace ();
	require ('(');

	parseEnumeration (true);
    }


    /**
     * Parse the default value for an attribute.
     * <pre>
     * [60] DefaultDecl ::= '#REQUIRED' | '#IMPLIED'
     *		| (('#FIXED' S)? AttValue)
     * </pre>
     */
    private void parseDefault (
	String elementName,
	String name,
	int type,
	String enum
    ) throws Exception