/*
 * The Apache Software License, Version 1.1
 *
 *
 * Copyright (c) 1999,2000 The Apache Software Foundation.  All rights 
 * reserved.
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 *
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer. 
 *
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in
 *    the documentation and/or other materials provided with the
 *    distribution.
 *
 * 3. The end-user documentation included with the redistribution,
 *    if any, must include the following acknowledgment:  
 *       "This product includes software developed by the
 *        Apache Software Foundation (http://www.apache.org/)."
 *    Alternately, this acknowledgment may appear in the software itself,
 *    if and wherever such third-party acknowledgments normally appear.
 *
 * 4. The names "Xerces" and "Apache Software Foundation" must
 *    not be used to endorse or promote products derived from this
 *    software without prior written permission. For written 
 *    permission, please contact apache@apache.org.
 *
 * 5. Products derived from this software may not be called "Apache",
 *    nor may "Apache" appear in their name, without prior written
 *    permission of the Apache Software Foundation.
 *
 * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
 * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED.  IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR
 * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
 * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
 * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
 * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
 * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
 * SUCH DAMAGE.
 * ====================================================================
 *
 * This software consists of voluntary contributions made by many
 * individuals on behalf of the Apache Software Foundation and was
 * originally based on software copyright (c) 1999, International
 * Business Machines, Inc., http://www.apache.org.  For more
 * information on the Apache Software Foundation, please see
 * <http://www.apache.org/>.
 */
package org.apache.html.dom;


import org.w3c.dom.*;
import org.w3c.dom.html.*;


/**
 * Implements {@link org.w3c.dom.html.HTMLCollection} to traverse any named
 * elements on a {@link org.w3c.dom.html.HTMLDocument}. The elements type to
 * look for is identified in the constructor by code. This collection is not
 * optimized for traversing large trees.
 * <p>
 * The collection has to meet two requirements: it has to be live, and it has
 * to traverse depth first and always return results in that order. As such,
 * using an object container (such as {@link java.util.Vector}) is expensive on
 * insert/remove operations. Instead, the collection has been implemented using
 * three traversing functions. As a result, operations on large documents will
 * result in traversal of the entire document tree and consume a considerable
 * amount of time.
 * <p>
 * Note that synchronization on the traversed document cannot be achieved.
 * The document itself cannot be locked, and locking each traversed node is
 * likely to lead to a dead lock condition. Therefore, there is a chance of the
 * document being changed as results are fetched; in all likelihood, the results
 * might be out dated, but not erroneous.
 * 
 * 
 * @version $Revision$ $Date$
 * @author <a href="mailto:arkin@exoffice.com">Assaf Arkin</a>
 * @see org.w3c.dom.html.HTMLCollection
 */
final class HTMLCollectionImpl
    implements HTMLCollection
{
    

    /**
     * Request collection of all anchors in document: &lt;A&gt; elements that
     * have a <code>name</code> attribute.
     */
    static final short        ANCHOR = 1;
    
    
    /**
     * Request collection of all forms in document: &lt;FORM&gt; elements.
     */
    static final short        FORM = 2;
    
    
    /**
     * Request collection of all images in document: &lt;IMAGE&gt; elements.
     */
    static final short        IMAGE = 3;
    
    
    /**
     * Request collection of all Applets in document: &lt;APPLET&gt; and
     * &lt;OBJECT&gt; elements (&lt;OBJECT&gt; must contain an Applet).
     */
    static final short        APPLET = 4;
    
    
    /**
     * Request collection of all links in document: &lt;A&gt; and &lt;AREA&gt;
     * elements (must have a <code>href</code> attribute).
     */
    static final short        LINK = 5;
    
    
    /**
     * Request collection of all options in selection: &lt;OPTION&gt; elments in
     * &lt;SELECT&gt; or &lt;OPTGROUP&gt;.
     */
    static final short        OPTION = 6;
    
    
    /**
     * Request collection of all rows in table: &lt;TR&gt; elements in table or
     * table section.
     */
    static final short        ROW = 7;

    
    /**
     * Request collection of all form elements: &lt;INPUT&gt;, &lt;BUTTON&gt;,
     * &lt;SELECT&gt;, &lt;TEXT&gt; and &lt;TEXTAREA&gt; elements inside form
     * &lt;FORM&gt;.
     */
    static final short        ELEMENT = 8;
    
    
    /**
     * Request collection of all areas in map: &lt;AREA&gt; element in &lt;MAP&gt;
     * (non recursive).
     */
    static final short        AREA = -1;
    

    /**
     * Request collection of all table bodies in table: &lt;TBODY&gt; element in
     * table &lt;TABLE&gt; (non recursive).
     */
    static final short        TBODY = -2;

    
    /**
     * Request collection of all cells in row: &lt;TD&gt; elements in &lt;TR&gt;
     * (non recursive).
     */
    static final short        CELL = -3;

    
    /**
     * Indicates what this collection is looking for. Holds one of the enumerated
     * values and used by {@link #collectionMatch}. Set by the constructor and
     * determine the collection's use for its life time.
     */
    private short            _lookingFor;
    
    
    /**
     * This is the top level element underneath which the collection exists.
     */
    private Element            _topLevel;


    /**
     * Construct a new collection that retrieves element of the specific type
     * (<code>lookingFor</code>) from the specific document portion
     * (<code>topLevel</code>).
     * 
     * @param topLevel The element underneath which the collection exists
     * @param lookingFor Code indicating what elements to look for
     */
    HTMLCollectionImpl( HTMLElement topLevel, short lookingFor )
    {
        if ( topLevel == null )
            throw new NullPointerException( "HTM011 Argument 'topLevel' is null." );
        _topLevel = topLevel;
       _lookingFor = lookingFor;
    }
  
  
    /**
     * Returns the length of the collection. This method might traverse the
     * entire document tree.
     * 
     * @return Length of the collection
     */
    public final int getLength()
    {
        // Call recursive function on top-level element.
        return getLength( _topLevel );
    }


    /**
     * Retrieves the indexed node from the collection. Nodes are numbered in
     * tree order - depth-first traversal order. This method might traverse
     * the entire document tree.
     * 
     * @param index The index of the node to return
     * @return The specified node or null if no such node found
     */
    public final Node item( int index )
    {
        if ( index < 0 )
            throw new IllegalArgumentException( "HTM012 Argument 'index' is negative." );
        // Call recursive function on top-level element.
        return item( _topLevel, new CollectionIndex( index ) );
    }
    
    
    /**
     * Retrieves the named node from the collection. The name is matched case
     * sensitive against the <TT>id</TT> attribute of each element in the
     * collection, returning the first match. The tree is traversed in
     * depth-first order. This method might traverse the entire document tree.
     * 
     * @param name The name of the node to return
     * @return The specified node or null if no such node found
     */
    public final Node namedItem( String name )
    {
        if ( name == null )
            throw new NullPointerException( "HTM013 Argument 'name' is null." );
        // Call recursive function on top-level element.
        return namedItem( _topLevel, name );
    }
    
    
    /**
     * Recursive function returns the number of elements of a particular type
     * that exist under the top level element. This is a recursive function
     * and the top level element is passed along.
     * 
     * @param topLevel Top level element from which to scan
     * @return Number of elements
     */
    private int getLength( Element topLevel )
    {
        int        length;
        Node    node;
    
        synchronized ( topLevel )
        {
            // Always count from zero and traverse all the childs of the
            // current element in the order they appear.
            length = 0;
            node = topLevel.getFirstChild();
            while ( node != null )
            {
                // If a particular node is an element (could be HTML or XML),
		// do two things: if it's the one we're looking for, count
		// another matched element; at any rate, traverse it's
		// children as well.
                if ( node instanceof Element )
                {
                    if ( collectionMatch( (Element) node, null ) )
                        ++ length;
                    else if ( recurse() )
                        length += getLength( (Element) node );
                }
                node = node.getNextSibling(); 
            }
        }
        return length;
    }
    
        
    /**
     * Recursive function returns the numbered element of a particular type
     * that exist under the top level element. This is a recursive function
     * and the top level element is passed along.
     * <p>
     * Note that this function must call itself with an index and get back both
     * the element (if one was found) and the new index which is decremeneted
     * for any like element found. Since integers are only passed by value,
     * this function makes use of a separate class ({@link CollectionIndex})
     * to hold that index.
     * 
     * @param topLevel Top level element from which to scan
     * @param index The index of the item to retreive
     * @return Number of elements
     * @see CollectionIndex
     */
    private Node item( Element topLevel, CollectionIndex index )
    {
        Node    node;
        Node    result;

        synchronized ( topLevel )
        {
            // Traverse all the childs of the current element in the order
	    // they appear. Count from the index backwards until you reach
	    // matching element with an index of zero. Return that element.
            node = topLevel.getFirstChild();
            while ( node != null )
            {
                // If a particular node is an element (could be HTML or XML),
		// do two things: if it's the one we're looking for, decrease
		// the index and if zero, return this node; at any rate,
		// traverse it's children as well.
                if ( node instanceof Element )
                {
                    if ( collectionMatch( (Element) node, null ) )
                    {
                        if ( index.isZero() )
                            return node;
                        index.decrement();
                    } else if ( recurse() )
                    {
                        result = item( (Element) node, index );
                        if ( result != null )
                            return result;
                    }
                }
                node = node.getNextSibling(); 
            }
        }
        return null;
    }
    
    
    /**
     * Recursive function returns an element of a particular type with the
     * specified name (<TT>id</TT> attribute).
     * 
     * @param topLevel Top level element from which to scan
     * @param name The named element to look for
     * @return The first named element found
     */
    private  Node namedItem( Element topLevel, String name )
    {
        Node    node;
        Node    result;

        synchronized ( topLevel )
        {
            // Traverse all the childs of the current element in the order
	    // they appear.
            node = topLevel.getFirstChild();
            while ( node != null )
            {
                // If a particular node is an element (could be HTML or XML),
                // do two things: if it's the one we're looking for, and the
		// name (id attribute) attribute is the one we're looking for,
		// return this element; otherwise, traverse it's children.
                if ( node instanceof Element )
                {
                    if ( collectionMatch( (Element) node, name ) )
                        return node;
                    else if ( recurse() )
                    {
                        result = namedItem( (Element) node, name );
                        if ( result != null )
                            return result;
                    }
                }
                node = node.getNextSibling(); 
            }
            return node;
        }
    }
    
    
    /**
     * Returns true if scanning methods should iterate through the collection.
     * When looking for elements in the document, recursing is needed to traverse
     * the full document tree. When looking inside a specific element (e.g. for a
     * cell inside a row), recursing can lead to erroneous results.
     * 
     * @return True if methods should recurse to traverse entire tree
     */
    protected boolean recurse()
    {
        return _lookingFor > 0;
    }
    

    /**
     * Determines if current element matches based on what we're looking for.
     * The element is passed along with an optional identifier name. If the
     * element is the one we're looking for, return true. If the name is also
     * specified, the name must match the <code>id</code> attribute
     * (match <code>name</code> first for anchors).
     * 
     * @param elem The current element
     * @param name The identifier name or null
     * @return The element matches what we're looking for
     */
    protected boolean collectionMatch( Element elem, String name )
    {
        boolean    match;
        
        synchronized ( elem )
        {
            // Begin with no matching. Depending on what we're looking for,
            // attempt to match based on the element type. This is the quickest
            // way to match involving only a cast. Do the expensive string
            // comparison later on.
            match = false;
            switch ( _lookingFor )
            {
            case ANCHOR:
                // Anchor is an <A> element with a 'name' attribute. Otherwise, it's
                // just a link.
                match = ( elem instanceof HTMLAnchorElement ) &&
                        elem.getAttribute( "name" ).length() > 0;
                break;
            case FORM:
                // Any <FORM> element.
                match = ( elem instanceof HTMLFormElement );
                break;
            case IMAGE:
                // Any <IMG> element. <OBJECT> elements with images are not returned.
                match = ( elem instanceof HTMLImageElement );
                break;
            case APPLET:
                // Any <APPLET> element, and any <OBJECT> element which represents an
                // Applet. This is determined by 'codetype' attribute being
                // 'application/java' or 'classid' attribute starting with 'java:'.
                match = ( elem instanceof HTMLAppletElement ) ||
                        ( elem instanceof HTMLObjectElement &&
                          ( "application/java".equals( elem.getAttribute( "codetype" ) ) ||
                            elem.getAttribute( "classid" ).startsWith( "java:" ) ) );
                break;
            case ELEMENT:
                // All form elements implement HTMLFormControl for easy identification.
                match = ( elem instanceof HTMLFormControl );
                break;
            case LINK:
                // Any <A> element, and any <AREA> elements with an 'href' attribute.
                match = ( ( elem instanceof HTMLAnchorElement ||
                            elem instanceof HTMLAreaElement ) &&
                          elem.getAttribute( "href" ).length() > 0 );
                break;
            case AREA:
                // Any <AREA> element.
                match = ( elem instanceof HTMLAreaElement );
                break;
            case OPTION:
                // Any <OPTION> element.
                match = ( elem instanceof HTMLOptionElement );
                break;
            case ROW:
                // Any <TR> element.
                match = ( elem instanceof HTMLTableRowElement );
                break;
            case TBODY:
                // Any <TBODY> element (one of three table section types).
                match = ( elem instanceof HTMLTableSectionElement &&
                          elem.getTagName().equals( "tbody" ) );
                break;
            case CELL:
                // Any <TD> element.
                match = ( elem instanceof HTMLTableCellElement );
                break;
            }
        
            // If element type was matched and a name was specified, must also match
            // the name against either the 'id' or the 'name' attribute. The 'name'
            // attribute is relevant only for <A> elements for backward compatibility.
            if ( match && name != null )
            {
                // If an anchor and 'name' attribute matches, return true. Otherwise,
                // try 'id' attribute.
                if ( elem instanceof HTMLAnchorElement &&
                     name.equals( elem.getAttribute( "name" ) ) )
                    return true;
                match = name.equals( elem.getAttribute( "id" ) );
            }
        }
        return match;
    }

    
}


/**
 * {@link CollectionImpl#item} must traverse down the tree and decrement the
 * index until it matches an element who's index is zero. Since integers are
 * passed by value, this class servers to pass the index into each recursion
 * by reference. It encompasses all the operations that need be performed on
 * the index, although direct access is possible.
 * 
 * @see CollectionImpl#item
 */
class CollectionIndex
{
    
    
    /**
     * Returns the current index.
     * 
     * @return Current index
     */
    int getIndex()
    {
        return _index;
    }
    
    
    /**
     * Decrements the index by one.
     */
    void decrement()
    {
        -- _index;
    }
    
    
    /**
     * Returns true if index is zero (or negative).
     * 
     * @return True if index is zero
     */
    boolean isZero()
    {
        return _index <= 0;
    }
    
    
    /**
     * Constructs a new index with the specified initial value. The index will
     * then be decremeneted until it reaches zero.
     * 
     * @param index The initial value
     */
    CollectionIndex( int index )
    {
        _index = index;
    }
    
    
    /**
     * Holds the actual value that is passed by reference using this class.
     */
    private int        _index;
    

}