saxhandler.java

一个用于搜索本地文件内容的小型搜索引擎

if (!inDTD) {
  if (inEntity) {
    getCurrentElement().setContent(factory.text(data));
  } else {
    getCurrentElement().addContent(factory.text(data));
}
*/

        if (previousCDATA) {
            factory.addContent(getCurrentElement(), factory.cdata(data));
        }
        else {
            factory.addContent(getCurrentElement(), factory.text(data));
        }

        previousCDATA = inCDATA;
    }

    /**
     * Indicates the end of an element
     * (<code>&lt;/[element name]&gt;</code>) is reached.  Note that
     * the parser does not distinguish between empty
     * elements and non-empty elements, so this will occur uniformly.
     *
     * @param namespaceURI <code>String</code> URI of namespace this
     *                     element is associated with
     * @param localName <code>String</code> name of element without prefix
     * @param qName <code>String</code> name of element in XML 1.0 form
     * @throws SAXException when things go wrong
     */
    public void endElement(String namespaceURI, String localName,
                           String qName) throws SAXException {

        if (suppress) return;

        flushCharacters();

        if (!atRoot) {
            Parent p = currentElement.getParent();
            if (p instanceof Document) {
               atRoot = true;
            }
            else {
                currentElement = (Element) p;
            }
        }
        else {
            throw new SAXException(
                "Ill-formed XML document (missing opening tag for " +
                localName + ")");
        }
    }

    /**
     * This will signify that a DTD is being parsed, and can be
     * used to ensure that comments and other lexical structures
     * in the DTD are not added to the JDOM <code>Document</code>
     * object.
     *
     * @param name <code>String</code> name of element listed in DTD
     * @param publicID <code>String</code> public ID of DTD
     * @param systemID <code>String</code> system ID of DTD
     */
    public void startDTD(String name, String publicID, String systemID)
        throws SAXException {

        flushCharacters(); // Is this needed here?

        factory.addContent(document, factory.docType(name, publicID, systemID));
        inDTD = true;
        inInternalSubset = true;
    }

    /**
     * This signifies that the reading of the DTD is complete.
     *
     * @throws SAXException
     */
    public void endDTD() throws SAXException {

        document.getDocType().setInternalSubset(internalSubset.toString());
        inDTD = false;
        inInternalSubset = false;
    }

    public void startEntity(String name) throws SAXException {
        entityDepth++;

        if (expand || entityDepth > 1) {
            // Short cut out if we're expanding or if we're nested
            return;
        }

        // A "[dtd]" entity indicates the beginning of the external subset
        if (name.equals("[dtd]")) {
            inInternalSubset = false;
            return;
        }

        // Ignore DTD references, and translate the standard 5
        if ((!inDTD) &&
            (!name.equals("amp")) &&
            (!name.equals("lt")) &&
            (!name.equals("gt")) &&
            (!name.equals("apos")) &&
            (!name.equals("quot"))) {

            if (!expand) {
                String pub = null;
                String sys = null;
                String[] ids = (String[]) externalEntities.get(name);
                if (ids != null) {
                  pub = ids[0];  // may be null, that's OK
                  sys = ids[1];  // may be null, that's OK
                }
                /**
                 * if no current element, this entity belongs to an attribute
                 * in these cases, it is an error on the part of the parser
                 * to call startEntity but this will help in some cases.
                 * See org/xml/sax/ext/LexicalHandler.html#startEntity(java.lang.String)
                 * for more information
                 */
                if (!atRoot) {
                    flushCharacters();
                    EntityRef entity = factory.entityRef(name, pub, sys);

                    // no way to tell if the entity was from an attribute or element so just assume element
                    factory.addContent(getCurrentElement(), entity);
                }
                suppress = true;
            }
        }
    }

    public void endEntity(String name) throws SAXException {
        entityDepth--;
        if (entityDepth == 0) {
            // No way are we suppressing if not in an entity,
            // regardless of the "expand" value
            suppress = false;
        }
        if (name.equals("[dtd]")) {
            inInternalSubset = true;
        }
    }

    /**
     * Report a CDATA section
     *
     * @throws SAXException
     */
    public void startCDATA() throws SAXException {
        if (suppress) return;

        inCDATA = true;
    }

    /**
     * Report a CDATA section
     */
    public void endCDATA() throws SAXException {
        if (suppress) return;

        previousCDATA = true;
        inCDATA = false;
    }

    /**
     * This reports that a comments is parsed.  If not in the
     * DTD, this comment is added to the current JDOM
     * <code>Element</code>, or the <code>Document</code> itself
     * if at that level.
     *
     * @param ch <code>ch[]</code> array of comment characters.
     * @param start <code>int</code> index to start reading from.
     * @param length <code>int</code> length of data.
     * @throws SAXException
     */
    public void comment(char[] ch, int start, int length)
        throws SAXException {

        if (suppress) return;

        flushCharacters();

        String commentText = new String(ch, start, length);
        if (inDTD && inInternalSubset && (expand == false)) {
            internalSubset.append("  <!--")
                  .append(commentText)
                  .append("-->\n");
            return;
        }
        if ((!inDTD) && (!commentText.equals(""))) {
            if (atRoot) {
                factory.addContent(document, factory.comment(commentText));
            } else {
                factory.addContent(getCurrentElement(), factory.comment(commentText));
            }
        }
    }

    /**
     * Handle the declaration of a Notation in a DTD
     *
     * @param name name of the notation
     * @param publicID the public ID of the notation
     * @param systemID the system ID of the notation
     */
    public void notationDecl(String name, String publicID, String systemID)
        throws SAXException {

        if (!inInternalSubset) return;

        internalSubset.append("  <!NOTATION ")
              .append(name);
        appendExternalId(publicID, systemID);
        internalSubset.append(">\n");
    }

    /**
     * Handler for unparsed entity declarations in the DTD
     *
     * @param name <code>String</code> of the unparsed entity decl
     * @param publicID <code>String</code> of the unparsed entity decl
     * @param systemID <code>String</code> of the unparsed entity decl
     * @param notationName <code>String</code> of the unparsed entity decl
     */
    public void unparsedEntityDecl(String name, String publicID,
                                   String systemID, String notationName)
        throws SAXException {

        if (!inInternalSubset) return;

        internalSubset.append("  <!ENTITY ")
              .append(name);
        appendExternalId(publicID, systemID);
        internalSubset.append(" NDATA ")
              .append(notationName);
        internalSubset.append(">\n");
    }

    /**
     * Appends an external ID to the internal subset buffer. Either publicID
     * or systemID may be null, but not both.
     *
     * @param publicID the public ID
     * @param systemID the system ID
     */
    private void appendExternalId(String publicID, String systemID) {
        if (publicID != null) {
            internalSubset.append(" PUBLIC \"")
                  .append(publicID)
                  .append('\"');
        }
        if (systemID != null) {
            if (publicID == null) {
                internalSubset.append(" SYSTEM ");
            }
            else {
                internalSubset.append(' ');
            }
            internalSubset.append('\"')
                  .append(systemID)
                  .append('\"');
        }
    }

    /**
     * Returns the being-parsed element.
     *
     * @return <code>Element</code> - element being built.
     * @throws SAXException
     */
    public Element getCurrentElement() throws SAXException {
        if (currentElement == null) {
            throw new SAXException(
                "Ill-formed XML document (multiple root elements detected)");
        }
        return currentElement;
    }

    /**
     * Returns the the JDOM Attribute type value from the SAX 2.0
     * attribute type string provided by the parser.
     *
     * @param typeName <code>String</code> the SAX 2.0 attribute
     * type string.
     *
     * @return <code>int</code> the JDOM attribute type.
     *
     * @see Attribute#setAttributeType
     * @see Attributes#getType
     */
    private static int getAttributeType(String typeName) {
        Integer type = (Integer)(attrNameToTypeMap.get(typeName));
        if (type == null) {
            if (typeName != null && typeName.length() > 0 &&
                typeName.charAt(0) == '(') {
                // Xerces 1.4.X reports attributes of enumerated type with
                // a type string equals to the enumeration definition, i.e.
                // starting with a parenthesis.
                return Attribute.ENUMERATED_TYPE;
            }
            else {
                return Attribute.UNDECLARED_TYPE;
            }
        } else {
            return type.intValue();
        }
    }

    /**
     * Receives an object for locating the origin of SAX document
     * events.  This method is invoked by the SAX parser.
     * <p>
     * {@link org.jdom.JDOMFactory} implementations can use the
     * {@link #getDocumentLocator} method to get access to the
     * {@link Locator} during parse.
     * </p>
     *
     * @param locator <code>Locator</code> an object that can return
     * the location of any SAX document event.
     */
    public void setDocumentLocator(Locator locator) {
        this.locator = locator;
    }

    /**
     * Provides access to the {@link Locator} object provided by the
     * SAX parser.
     *
     * @return <code>Locator</code> an object that can return
     * the location of any SAX document event.
     */
    public Locator getDocumentLocator() {
        return locator;
    }
}