HTMLCollectionImpl


1   /*
2    * The Apache Software License, Version 1.1
3    *
4    *
5    * Copyright (c) 1999,2000 The Apache Software Foundation.  All rights 
6    * reserved.
7    *
8    * Redistribution and use in source and binary forms, with or without
9    * modification, are permitted provided that the following conditions
10   * are met:
11   *
12   * 1. Redistributions of source code must retain the above copyright
13   *    notice, this list of conditions and the following disclaimer. 
14   *
15   * 2. Redistributions in binary form must reproduce the above copyright
16   *    notice, this list of conditions and the following disclaimer in
17   *    the documentation and/or other materials provided with the
18   *    distribution.
19   *
20   * 3. The end-user documentation included with the redistribution,
21   *    if any, must include the following acknowledgment:  
22   *       "This product includes software developed by the
23   *        Apache Software Foundation (http://www.apache.org/)."
24   *    Alternately, this acknowledgment may appear in the software itself,
25   *    if and wherever such third-party acknowledgments normally appear.
26   *
27   * 4. The names "Xerces" and "Apache Software Foundation" must
28   *    not be used to endorse or promote products derived from this
29   *    software without prior written permission. For written 
30   *    permission, please contact apache@apache.org.
31   *
32   * 5. Products derived from this software may not be called "Apache",
33   *    nor may "Apache" appear in their name, without prior written
34   *    permission of the Apache Software Foundation.
35   *
36   * THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESSED OR IMPLIED
37   * WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
38   * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
39   * DISCLAIMED.  IN NO EVENT SHALL THE APACHE SOFTWARE FOUNDATION OR
40   * ITS CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
41   * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
42   * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
43   * USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
44   * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
45   * OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
46   * OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
47   * SUCH DAMAGE.
48   * ====================================================================
49   *
50   * This software consists of voluntary contributions made by many
51   * individuals on behalf of the Apache Software Foundation and was
52   * originally based on software copyright (c) 1999, International
53   * Business Machines, Inc., http://www.apache.org.  For more
54   * information on the Apache Software Foundation, please see
55   * <http://www.apache.org/>.
56   */
57  package org.enhydra.xml.lazydom.html;
58  import org.w3c.dom.Element  ;
59  import org.w3c.dom.Node  ;
60  import org.w3c.dom.html.HTMLAnchorElement;
61  import org.w3c.dom.html.HTMLAppletElement;
62  import org.w3c.dom.html.HTMLAreaElement;
63  import org.w3c.dom.html.HTMLCollection;
64  import org.w3c.dom.html.HTMLElement;
65  import org.w3c.dom.html.HTMLFormElement;
66  import org.w3c.dom.html.HTMLImageElement;
67  import org.w3c.dom.html.HTMLObjectElement;
68  import org.w3c.dom.html.HTMLOptionElement;
69  import org.w3c.dom.html.HTMLTableCellElement;
70  import org.w3c.dom.html.HTMLTableRowElement;
71  import org.w3c.dom.html.HTMLTableSectionElement;
72  
73  
74  /**
75   * Implements {@link org.w3c.dom.html.HTMLCollection} to traverse any named
76   * elements on a {@link org.w3c.dom.html.HTMLDocument}. The elements type to
77   * look for is identified in the constructor by code. This collection is not
78   * optimized for traversing large trees.
79   * <p>
80   * The collection has to meet two requirements: it has to be live, and it has
81   * to traverse depth first and always return results in that order. As such,
82   * using an object container (such as {@link java.util.Vector}) is expensive on
83   * insert/remove operations. Instead, the collection has been implemented using
84   * three traversing functions. As a result, operations on large documents will
85   * result in traversal of the entire document tree and consume a considerable
86   * amount of time.
87   * <p>
88   * Note that synchronization on the traversed document cannot be achieved.
89   * The document itself cannot be locked, and locking each traversed node is
90   * likely to lead to a dead lock condition. Therefore, there is a chance of the
91   * document being changed as results are fetched; in all likelihood, the results
92   * might be out dated, but not erroneous.
93   * 
94   * 
95   * @version $Revision: 1.2 $ $Date: 2005/01/26 08:29:24 $
96   * @author <a HREF="mailto:arkin@exoffice.com">Assaf Arkin</a>
97   * @see org.w3c.dom.html.HTMLCollection
98   */
99  class HTMLCollectionImpl
100     implements HTMLCollection
101 {
102     
103 
104     /**
105      * Request collection of all anchors in document: &lt;A&gt; elements that
106      * have a <code>name</code> attribute.
107      */
108     static final short        ANCHOR = 1;
109     
110     
111     /**
112      * Request collection of all forms in document: &lt;FORM&gt; elements.
113      */
114     static final short        FORM = 2;
115     
116     
117     /**
118      * Request collection of all images in document: &lt;IMAGE&gt; elements.
119      */
120     static final short        IMAGE = 3;
121     
122     
123     /**
124      * Request collection of all Applets in document: &lt;APPLET&gt; and
125      * &lt;OBJECT&gt; elements (&lt;OBJECT&gt; must contain an Applet).
126      */
127     static final short        APPLET = 4;
128     
129     
130     /**
131      * Request collection of all links in document: &lt;A&gt; and &lt;AREA&gt;
132      * elements (must have a <code>href</code> attribute).
133      */
134     static final short        LINK = 5;
135     
136     
137     /**
138      * Request collection of all options in selection: &lt;OPTION&gt; elments in
139      * &lt;SELECT&gt; or &lt;OPTGROUP&gt;.
140      */
141     static final short        OPTION = 6;
142     
143     
144     /**
145      * Request collection of all rows in table: &lt;TR&gt; elements in table or
146      * table section.
147      */
148     static final short        ROW = 7;
149 
150     
151     /**
152      * Request collection of all form elements: &lt;INPUT&gt;, &lt;BUTTON&gt;,
153      * &lt;SELECT&gt;, &lt;TEXT&gt; and &lt;TEXTAREA&gt; elements inside form
154      * &lt;FORM&gt;.
155      */
156     static final short        ELEMENT = 8;
157     
158     
159     /**
160      * Request collection of all areas in map: &lt;AREA&gt; element in &lt;MAP&gt;
161      * (non recursive).
162      */
163     static final short        AREA = -1;
164     
165 
166     /**
167      * Request collection of all table bodies in table: &lt;TBODY&gt; element in
168      * table &lt;TABLE&gt; (non recursive).
169      */
170     static final short        TBODY = -2;
171 
172     
173     /**
174      * Request collection of all cells in row: &lt;TD&gt; elements in &lt;TR&gt;
175      * (non recursive).
176      */
177     static final short        CELL = -3;
178 
179     
180     /**
181      * Indicates what this collection is looking for. Holds one of the enumerated
182      * values and used by {@link #collectionMatch}. Set by the constructor and
183      * determine the collection's use for its life time.
184      */
185     private short            _lookingFor;
186     
187     
188     /**
189      * This is the top level element underneath which the collection exists.
190      */
191     private Element              _topLevel;
192 
193 
194     /**
195      * Construct a new collection that retrieves element of the specific type
196      * (<code>lookingFor</code>) from the specific document portion
197      * (<code>topLevel</code>).
198      * 
199      * @param topLevel The element underneath which the collection exists
200      * @param lookingFor Code indicating what elements to look for
201      */
202     HTMLCollectionImpl( HTMLElement topLevel, short lookingFor )
203     {
204         if ( topLevel == null )
205             throw new NullPointerException  ( "HTM011 Argument 'topLevel' is null." );
206         _topLevel = topLevel;
207        _lookingFor = lookingFor;
208     }
209   
210   
211     /**
212      * Returns the length of the collection. This method might traverse the
213      * entire document tree.
214      * 
215      * @return Length of the collection
216      */
217     public final int getLength()
218     {
219         // Call recursive function on top-level element.
220         return getLength( _topLevel );
221     }
222 
223 
224     /**
225      * Retrieves the indexed node from the collection. Nodes are numbered in
226      * tree order - depth-first traversal order. This method might traverse
227      * the entire document tree.
228      * 
229      * @param index The index of the node to return
230      * @return The specified node or null if no such node found
231      */
232     public final Node   item( int index )
233     {
234         if ( index < 0 )
235             throw new IllegalArgumentException  ( "HTM012 Argument 'index' is negative." );
236         // Call recursive function on top-level element.
237         return item( _topLevel, new CollectionIndex( index ) );
238     }
239     
240     
241     /**
242      * Retrieves the named node from the collection. The name is matched case
243      * sensitive against the <TT>id</TT> attribute of each element in the
244      * collection, returning the first match. The tree is traversed in
245      * depth-first order. This method might traverse the entire document tree.
246      * 
247      * @param name The name of the node to return
248      * @return The specified node or null if no such node found
249      */
250     public final Node   namedItem( String   name )
251     {
252         if ( name == null )
253             throw new NullPointerException  ( "HTM013 Argument 'name' is null." );
254         // Call recursive function on top-level element.
255         return namedItem( _topLevel, name );
256     }
257     
258     
259     /**
260      * Recursive function returns the number of elements of a particular type
261      * that exist under the top level element. This is a recursive function
262      * and the top level element is passed along.
263      * 
264      * @param topLevel Top level element from which to scan
265      * @return Number of elements
266      */
267     private int getLength( Element   topLevel )
268     {
269         int        length;
270         Node      node;
271     
272         synchronized ( topLevel )
273         {
274             // Always count from zero and traverse all the childs of the
275             // current element in the order they appear.
276             length = 0;
277             node = topLevel.getFirstChild();
278             while ( node != null )
279             {
280                 // If a particular node is an element (could be HTML or XML),
281         // do two things: if it's the one we're looking for, count
282         // another matched element; at any rate, traverse it's
283         // children as well.
284                 if ( node instanceof Element   )
285                 {
286                     if ( collectionMatch( (Element  ) node, null ) )
287                         ++ length;
288                     else if ( recurse() )
289                         length += getLength( (Element  ) node );
290                 }
291                 node = node.getNextSibling(); 
292             }
293         }
294         return length;
295     }
296     
297         
298     /**
299      * Recursive function returns the numbered element of a particular type
300      * that exist under the top level element. This is a recursive function
301      * and the top level element is passed along.
302      * <p>
303      * Note that this function must call itself with an index and get back both
304      * the element (if one was found) and the new index which is decremeneted
305      * for any like element found. Since integers are only passed by value,
306      * this function makes use of a separate class ({@link CollectionIndex})
307      * to hold that index.
308      * 
309      * @param topLevel Top level element from which to scan
310      * @param index The index of the item to retreive
311      * @return Number of elements
312      * @see CollectionIndex
313      */
314     private Node   item( Element   topLevel, CollectionIndex index )
315     {
316         Node      node;
317         Node      result;
318 
319         synchronized ( topLevel )
320         {
321             // Traverse all the childs of the current element in the order
322         // they appear. Count from the index backwards until you reach
323         // matching element with an index of zero. Return that element.
324             node = topLevel.getFirstChild();
325             while ( node != null )
326             {
327                 // If a particular node is an element (could be HTML or XML),
328         // do two things: if it's the one we're looking for, decrease
329         // the index and if zero, return this node; at any rate,
330         // traverse it's children as well.
331                 if ( node instanceof Element   )
332                 {
333                     if ( collectionMatch( (Element  ) node, null ) )
334                     {
335                         if ( index.isZero() )
336                             return node;
337                         index.decrement();
338                     } else if ( recurse() )
339                     {
340                         result = item( (Element  ) node, index );
341                         if ( result != null )
342                             return result;
343                     }
344                 }
345                 node = node.getNextSibling(); 
346             }
347         }
348         return null;
349     }
350     
351     
352     /**
353      * Recursive function returns an element of a particular type with the
354      * specified name (<TT>id</TT> attribute).
355      * 
356      * @param topLevel Top level element from which to scan
357      * @param name The named element to look for
358      * @return The first named element found
359      */
360     private  Node   namedItem( Element   topLevel, String   name )
361     {
362         Node      node;
363         Node      result;
364 
365         synchronized ( topLevel )
366         {
367             // Traverse all the childs of the current element in the order
368         // they appear.
369             node = topLevel.getFirstChild();
370             while ( node != null )
371             {
372                 // If a particular node is an element (could be HTML or XML),
373                 // do two things: if it's the one we're looking for, and the
374         // name (id attribute) attribute is the one we're looking for,
375         // return this element; otherwise, traverse it's children.
376                 if ( node instanceof Element   )
377                 {
378                     if ( collectionMatch( (Element  ) node, name ) )
379                         return node;
380                     else if ( recurse() )
381                     {
382                         result = namedItem( (Element  ) node, name );
383                         if ( result != null )
384                             return result;
385                     }
386                 }
387                 node = node.getNextSibling(); 
388             }
389             return node;
390         }
391     }
392     
393     
394     /**
395      * Returns true if scanning methods should iterate through the collection.
396      * When looking for elements in the document, recursing is needed to traverse
397      * the full document tree. When looking inside a specific element (e.g. for a
398      * cell inside a row), recursing can lead to erroneous results.
399      * 
400      * @return True if methods should recurse to traverse entire tree
401      */
402     protected boolean recurse()
403     {
404         return _lookingFor > 0;
405     }
406     
407 
408     /**
409      * Determines if current element matches based on what we're looking for.
410      * The element is passed along with an optional identifier name. If the
411      * element is the one we're looking for, return true. If the name is also
412      * specified, the name must match the <code>id</code> attribute
413      * (match <code>name</code> first for anchors).
414      * 
415      * @param elem The current element
416      * @param name The identifier name or null
417      * @return The element matches what we're looking for
418      */
419     protected boolean collectionMatch( Element   elem, String   name )
420     {
421         boolean    match;
422         
423         synchronized ( elem )
424         {
425             // Begin with no matching. Depending on what we're looking for,
426             // attempt to match based on the element type. This is the quickest
427             // way to match involving only a cast. Do the expensive string
428             // comparison later on.
429             match = false;
430             switch ( _lookingFor )
431             {
432             case ANCHOR:
433                 // Anchor is an <A> element with a 'name' attribute. Otherwise, it's
434                 // just a link.
435                 match = ( elem instanceof HTMLAnchorElement ) &&
436                         elem.getAttribute( "name" ).length() > 0;
437                 break;
438             case FORM:
439                 // Any <FORM> element.
440                 match = ( elem instanceof HTMLFormElement );
441                 break;
442             case IMAGE:
443                 // Any <IMG> element. <OBJECT> elements with images are not returned.
444                 match = ( elem instanceof HTMLImageElement );
445                 break;
446             case APPLET:
447                 // Any <APPLET> element, and any <OBJECT> element which represents an
448                 // Applet. This is determined by 'codetype' attribute being
449                 // 'application/java' or 'classid' attribute starting with 'java:'.
450                 match = ( elem instanceof HTMLAppletElement ) ||
451                         ( elem instanceof HTMLObjectElement &&
452                           ( "application/java".equals( elem.getAttribute( "codetype" ) ) ||
453                             elem.getAttribute( "classid" ).startsWith( "java:" ) ) );
454                 break;
455             case ELEMENT:
456                 // All form elements implement HTMLFormControl for easy identification.
457                 match = ( elem instanceof HTMLFormControl );
458                 break;
459             case LINK:
460                 // Any <A> element, and any <AREA> elements with an 'href' attribute.
461                 match = ( ( elem instanceof HTMLAnchorElement ||
462                             elem instanceof HTMLAreaElement ) &&
463                           elem.getAttribute( "href" ).length() > 0 );
464                 break;
465             case AREA:
466                 // Any <AREA> element.
467                 match = ( elem instanceof HTMLAreaElement );
468                 break;
469             case OPTION:
470                 // Any <OPTION> element.
471                 match = ( elem instanceof HTMLOptionElement );
472                 break;
473             case ROW:
474                 // Any <TR> element.
475                 match = ( elem instanceof HTMLTableRowElement );
476                 break;
477             case TBODY:
478                 // Any <TBODY> element (one of three table section types).
479                 match = ( elem instanceof HTMLTableSectionElement &&
480                           elem.getTagName().equals( "tbody" ) );
481                 break;
482             case CELL:
483                 // Any <TD> element.
484                 match = ( elem instanceof HTMLTableCellElement );
485                 break;
486             }
487         
488             // If element type was matched and a name was specified, must also match
489             // the name against either the 'id' or the 'name' attribute. The 'name'
490             // attribute is relevant only for <A> elements for backward compatibility.
491             if ( match && name != null )
492             {
493                 // If an anchor and 'name' attribute matches, return true. Otherwise,
494                 // try 'id' attribute.
495                 if ( elem instanceof HTMLAnchorElement &&
496                      name.equals( elem.getAttribute( "name" ) ) )
497                     return true;
498                 match = name.equals( elem.getAttribute( "id" ) );
499             }
500         }
501         return match;
502     }
503 
504     
505 }
506 
507 
508 /**
509  * {@link CollectionImpl#item} must traverse down the tree and decrement the
510  * index until it matches an element who's index is zero. Since integers are
511  * passed by value, this class servers to pass the index into each recursion
512  * by reference. It encompasses all the operations that need be performed on
513  * the index, although direct access is possible.
514  * 
515  * @see CollectionImpl#item
516  */
517 class CollectionIndex
518 {
519     
520     
521     /**
522      * Returns the current index.
523      * 
524      * @return Current index
525      */
526     int getIndex()
527     {
528         return _index;
529     }
530     
531     
532     /**
533      * Decrements the index by one.
534      */
535     void decrement()
536     {
537         -- _index;
538     }
539     
540     
541     /**
542      * Returns true if index is zero (or negative).
543      * 
544      * @return True if index is zero
545      */
546     boolean isZero()
547     {
548         return _index <= 0;
549     }
550     
551     
552     /**
553      * Constructs a new index with the specified initial value. The index will
554      * then be decremeneted until it reaches zero.
555      * 
556      * @param index The initial value
557      */
558     CollectionIndex( int index )
559     {
560         _index = index;
561     }
562     
563     
564     /**
565      * Holds the actual value that is passed by reference using this class.
566      */
567     private int        _index;
568     
569 
570 }
571
A to Z: JavaDoc & Examples Daily Java News & Articles Open Source Projects Open Source Codes Free Computer Books Remove Frame
Popular Tags