morten 01/12/18 07:55:16
Modified: java/xdocs/sources/xsltc xsltc_iterators.xml
Log:
Completed the node iterator design document.
Submitted by: [EMAIL PROTECTED]
Revision Changes Path
1.2 +58 -4 xml-xalan/java/xdocs/sources/xsltc/xsltc_iterators.xml
Index: xsltc_iterators.xml
===================================================================
RCS file: /home/cvs/xml-xalan/java/xdocs/sources/xsltc/xsltc_iterators.xml,v
retrieving revision 1.1
retrieving revision 1.2
diff -u -r1.1 -r1.2
--- xsltc_iterators.xml 2001/12/18 14:53:48 1.1
+++ xsltc_iterators.xml 2001/12/18 15:55:16 1.2
@@ -321,21 +321,75 @@
</s3>
<s3 title="Iterators for Filtering/Predicates">
+
+ <p>The <code>org.apache.xalan.xsltc.dom</code> package contains a few
+ iterators that are used to implement predicates and filters. Such iterators
+ are normally placed on top of another iterator, and return only those nodes
+ that match a specific node value, position, etc.
+ These iterators include:</p>
+
<ul>
+ <li>NthIterator</li>
<li>NodeValueIterator</li>
<li>FilteredStepIterator</li>
<li>CurrentNodeListIterator</li>
- <li>NthIterator</li>
</ul>
+
+ <p>The last one is the most interesting. This iterator is used to implement
+ chained predicates, such as:</p><source>
+ <xsl:value-of select="foo[@blob='boo'][2]"></source>
+
+ <p>The first predicate reduces the node set from containing all
+ <code><foo></code> elements, to containing only those elements that
+ have a <i>"blob"</i> attribute with the value 'boo'. The
+ <code>CurrentNodeListIterator</code> is used to contain this reduced
+ node-set. The iterator is constructed by passing it a source iterator (in
+ this case an iterator that contains all <code><foo></code> elements)
+ and a filter that implements the predicate (<code>@blob = 'boo'</code>).</p>
+
</s3>
- <s3 title="SortingIterator"></s3>
+ <s3 title="SortingIterator">
+ <p>The sorting iterator is one of the main functional components behind the
+ implementation of the <code><xsl:sort></code> element. This element,
+ including the sorting iterator, is described in detail in the
+ <code><xsl:sort></code>
+ <link idref="xsl_sort_design">design document</link>.</p>
+
+ </s3>
+
<s3 title="SingletonIterator"></s3>
+
+ <p>The singleton iterator is a wrapper for a single node. The node passed
+ in to the <code>setStartNode()</code> method is the only node that will be
+ returned by the <code>next()</code> method. The singleton iterator is used
+ mainly for node to node-set type conversions.</p>
+
+ <s3 title="UnionIterator">
+
+ <p>The union iterator is used to contain unions of node-sets contained in
+ other iterators. Some of the methods in this iterator are unnecessary
+ comlicated. The <code>next()</code> method contains an algorithm for
+ ensuring that the union node-set is returned in document order. We might be
+ better off by simply wrapping the union iterator inside a duplicate filter
+ iterator, but there could be some performance implications. Worth checking.
+ </p>
- <s3 title="UnionIterator"></s3>
+ </s3>
+
+ <s3 title="KeyIndex">
- <s3 title="KeyIndex"></s3>
+ <p>This is not just an node iterator. An index used for keys and ids will
+ return a set of nodes that are contained within the named index and that
+ share a certain property. The <code>KeyIndex</code> implements the node
+ iterator interface, so that these nodes can be returned and handled just
+ like any other node set. See the
+ <link idref="xsl_key_design">design document</link> for
+ <code><xsl:key></code>, <code>key()</code> and <code>id()</code>
+ for further details.</p>
+
+ </s3>
</s2>
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
