coredocumentimpl.java

来自「JAVA 所有包」· Java 代码 · 共 1,776 行 · 第 1/5 页

/*
 * Copyright 1999-2002,2004,2005 The Apache Software Foundation.
 * 
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 * 
 *      http://www.apache.org/licenses/LICENSE-2.0
 * 
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package com.sun.org.apache.xerces.internal.dom;

import java.lang.reflect.Constructor;
import java.util.Enumeration;
import java.util.Hashtable;
import com.sun.org.apache.xerces.internal.util.URI;

import org.w3c.dom.DOMConfiguration;
import org.w3c.dom.UserDataHandler;
import com.sun.org.apache.xerces.internal.util.XMLChar;
import com.sun.org.apache.xerces.internal.util.XML11Char;
import com.sun.org.apache.xerces.internal.xni.NamespaceContext;
import org.w3c.dom.Attr;
import org.w3c.dom.CDATASection;
import org.w3c.dom.Comment;
import org.w3c.dom.DOMException;
import org.w3c.dom.DOMImplementation;
import org.w3c.dom.Document;
import org.w3c.dom.DocumentFragment;
import org.w3c.dom.DocumentType;
import org.w3c.dom.Element;
import org.w3c.dom.Entity;
import org.w3c.dom.EntityReference;
import org.w3c.dom.NamedNodeMap;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import org.w3c.dom.Notation;
import org.w3c.dom.ProcessingInstruction;
import org.w3c.dom.Text;
import org.w3c.dom.events.Event;
import org.w3c.dom.events.EventListener;
import org.w3c.dom.ls.DOMImplementationLS;
import org.w3c.dom.ls.LSSerializer;

/**
 * The Document interface represents the entire HTML or XML document.
 * Conceptually, it is the root of the document tree, and provides the
 * primary access to the document's data.
 * <P>
 * Since elements, text nodes, comments, processing instructions,
 * etc. cannot exist outside the context of a Document, the Document
 * interface also contains the factory methods needed to create these
 * objects. The Node objects created have a ownerDocument attribute
 * which associates them with the Document within whose context they
 * were created.
 * <p>
 * The CoreDocumentImpl class only implements the DOM Core. Additional modules
 * are supported by the more complete DocumentImpl subclass.
 * <p>
 * <b>Note:</b> When any node in the document is serialized, the
 * entire document is serialized along with it.
 * 
 * @xerces.internal 
 *
 * @author Arnaud  Le Hors, IBM
 * @author Joe Kesselman, IBM
 * @author Andy Clark, IBM
 * @author Ralf Pfeiffer, IBM
 * @version $Id: CoreDocumentImpl.java,v 1.3 2005/09/26 13:02:11 sunithareddy Exp $
 * @since  PR-DOM-Level-1-19980818.
 */


public class CoreDocumentImpl
extends ParentNode implements Document  {

	/**TODO::
	 * 1. Change XML11Char method names similar to XMLChar. That will prevent lot
	 * of dirty version checking code.
	 *
	 * 2. IMO during cloneNode qname/isXMLName check should not be made.
	 */
    //
    // Constants
    //

    /** Serialization version. */
    static final long serialVersionUID = 0;

    //
    // Data
    //

    // document information

    /** Document type. */
    protected DocumentTypeImpl docType;

    /** Document element. */
    protected ElementImpl docElement;

    /** NodeListCache free list */
    transient NodeListCache fFreeNLCache;

    /**Experimental DOM Level 3 feature: Document encoding */
    protected String encoding;

    /**Experimental DOM Level 3 feature: Document actualEncoding */
    protected String actualEncoding;

    /**Experimental DOM Level 3 feature: Document version */
    protected String version;

    /**Experimental DOM Level 3 feature: Document standalone */
    protected boolean standalone;

    /**Experimental DOM Level 3 feature: documentURI */
    protected String fDocumentURI;

	//Revisit :: change to a better data structure.
    /** Table for user data attached to this document nodes. */
    protected Hashtable userData;


    /** Identifiers. */
    protected Hashtable identifiers;

    // DOM Level 3: normalizeDocument
    transient DOMNormalizer domNormalizer = null;
    transient DOMConfigurationImpl fConfiguration = null;
    
    // support of XPath API   
    transient Object fXPathEvaluator = null;

    /** Table for quick check of child insertion. */
    private final static int[] kidOK;

    /**
     * Number of alterations made to this document since its creation.
     * Serves as a "dirty bit" so that live objects such as NodeList can
     * recognize when an alteration has been made and discard its cached
     * state information.
     * <p>
     * Any method that alters the tree structure MUST cause or be
     * accompanied by a call to changed(), to inform it that any outstanding
     * NodeLists may have to be updated.
     * <p>
     * (Required because NodeList is simultaneously "live" and integer-
     * indexed -- a bad decision in the DOM's design.)
     * <p>
     * Note that changes which do not affect the tree's structure -- changing
     * the node's name, for example -- do _not_ have to call changed().
     * <p>
     * Alternative implementation would be to use a cryptographic
     * Digest value rather than a count. This would have the advantage that
     * "harmless" changes (those producing equal() trees) would not force
     * NodeList to resynchronize. Disadvantage is that it's slightly more prone
     * to "false negatives", though that's the difference between "wildly
     * unlikely" and "absurdly unlikely". IF we start maintaining digests,
     * we should consider taking advantage of them.
     *
     * Note: This used to be done a node basis, so that we knew what
     * subtree changed. But since only DeepNodeList really use this today,
     * the gain appears to be really small compared to the cost of having
     * an int on every (parent) node plus having to walk up the tree all the
     * way to the root to mark the branch as changed everytime a node is
     * changed.
     * So we now have a single counter global to the document. It means that
     * some objects may flush their cache more often than necessary, but this
     * makes nodes smaller and only the document needs to be marked as changed.
     */
    protected int changes = 0;

    // experimental

    /** Allow grammar access. */
    protected boolean allowGrammarAccess;

    /** Bypass error checking. */
    protected boolean errorChecking = true;

    //Did version change at any point when the document was created ?
    //this field helps us to optimize when normalizingDocument.
    protected boolean xmlVersionChanged = false ;

    /** The following are required for compareDocumentPosition
     */
    // Document number.   Documents are ordered across the implementation using
    // positive integer values.  Documents are assigned numbers on demand.
    private int documentNumber=0;
    // Node counter and table.  Used to assign numbers to nodes for this
    // document.  Node number values are negative integers.  Nodes are
    // assigned numbers on demand.
    private int nodeCounter = 0;
    private Hashtable nodeTable;
    private boolean xml11Version = false; //by default 1.0
    //
    // Static initialization
    //

    static {

        kidOK = new int[13];

        kidOK[DOCUMENT_NODE] =
        1 << ELEMENT_NODE | 1 << PROCESSING_INSTRUCTION_NODE |
        1 << COMMENT_NODE | 1 << DOCUMENT_TYPE_NODE;

        kidOK[DOCUMENT_FRAGMENT_NODE] =
        kidOK[ENTITY_NODE] =
        kidOK[ENTITY_REFERENCE_NODE] =
        kidOK[ELEMENT_NODE] =
        1 << ELEMENT_NODE | 1 << PROCESSING_INSTRUCTION_NODE |
        1 << COMMENT_NODE | 1 << TEXT_NODE |
        1 << CDATA_SECTION_NODE | 1 << ENTITY_REFERENCE_NODE ;


        kidOK[ATTRIBUTE_NODE] =
        1 << TEXT_NODE | 1 << ENTITY_REFERENCE_NODE;

        kidOK[DOCUMENT_TYPE_NODE] =
        kidOK[PROCESSING_INSTRUCTION_NODE] =
        kidOK[COMMENT_NODE] =
        kidOK[TEXT_NODE] =
        kidOK[CDATA_SECTION_NODE] =
        kidOK[NOTATION_NODE] =
        0;

    } // static

    //
    // Constructors
    //

    /**
     * NON-DOM: Actually creating a Document is outside the DOM's spec,
     * since it has to operate in terms of a particular implementation.
     */
    public CoreDocumentImpl() {
        this(false);
    }

    /** Constructor. */
    public CoreDocumentImpl(boolean grammarAccess) {
        super(null);
        ownerDocument = this;
        allowGrammarAccess = grammarAccess;
    }

    /**
     * For DOM2 support.
     * The createDocument factory method is in DOMImplementation.
     */
    public CoreDocumentImpl(DocumentType doctype) {
        this(doctype, false);
    }

    /** For DOM2 support. */
    public CoreDocumentImpl(DocumentType doctype, boolean grammarAccess) {
        this(grammarAccess);
        if (doctype != null) {
            DocumentTypeImpl doctypeImpl;
            try {
                doctypeImpl = (DocumentTypeImpl) doctype;
            } catch (ClassCastException e) {
                String msg = DOMMessageFormatter.formatMessage(DOMMessageFormatter.DOM_DOMAIN, "WRONG_DOCUMENT_ERR", null);
                throw new DOMException(DOMException.WRONG_DOCUMENT_ERR, msg);
            }
            doctypeImpl.ownerDocument = this;
            appendChild(doctype);
        }
    }

    //
    // Node methods
    //

    // even though ownerDocument refers to this in this implementation
    // the DOM Level 2 spec says it must be null, so make it appear so
    final public Document getOwnerDocument() {
        return null;
    }

    /** Returns the node type. */
    public short getNodeType() {
        return Node.DOCUMENT_NODE;
    }

    /** Returns the node name. */
    public String getNodeName() {
        return "#document";
    }

    /**
     * Deep-clone a document, including fixing ownerDoc for the cloned
     * children. Note that this requires bypassing the WRONG_DOCUMENT_ERR
     * protection. I've chosen to implement it by calling importNode
     * which is DOM Level 2.
     *
     * @return org.w3c.dom.Node
     * @param deep boolean, iff true replicate children
     */
    public Node cloneNode(boolean deep) {

        CoreDocumentImpl newdoc = new CoreDocumentImpl();
        callUserDataHandlers(this, newdoc, UserDataHandler.NODE_CLONED);
        cloneNode(newdoc, deep);

        return newdoc;

    } // cloneNode(boolean):Node


    /**
     * internal method to share code with subclass
     **/
    protected void cloneNode(CoreDocumentImpl newdoc, boolean deep) {

        // clone the children by importing them
        if (needsSyncChildren()) {
            synchronizeChildren();
        }

        if (deep) {
            Hashtable reversedIdentifiers = null;

            if (identifiers != null) {
                // Build a reverse mapping from element to identifier.
                reversedIdentifiers = new Hashtable();
                Enumeration elementIds = identifiers.keys();
                while (elementIds.hasMoreElements()) {
                    Object elementId = elementIds.nextElement();
                    reversedIdentifiers.put(identifiers.get(elementId),
                    elementId);
                }
            }

            // Copy children into new document.
            for (ChildNode kid = firstChild; kid != null;
            kid = kid.nextSibling) {
                newdoc.appendChild(newdoc.importNode(kid, true, true,
                reversedIdentifiers));
            }
        }

        // experimental
        newdoc.allowGrammarAccess = allowGrammarAccess;
        newdoc.errorChecking = errorChecking;

    } // cloneNode(CoreDocumentImpl,boolean):void