org.apache.xalan.transformer
Class NodeSorter

java.lang.Object
  org.apache.xalan.transformer.NodeSorter

public class NodeSorter
extends java.lang.Object

This class can sort vectors of DOM nodes according to a select pattern.

Nested Class Summary

(package private) class

NodeSorter.NodeCompareElem This class holds the value(s) from executing the given node against the sort key(s).

Field Summary

(package private) XPathContext	m_execContext Current XPath context
(package private) java.util.Vector	m_keys Vector of NodeSortKeys

Constructor Summary

NodeSorter(XPathContext p)
Construct a NodeSorter, passing in the XSL TransformerFactory so it can know how to get the node data according to the proper whitespace rules.

Method Summary

(package private) int	compare(NodeSorter.NodeCompareElem n1, NodeSorter.NodeCompareElem n2, int kIndex, XPathContext support) Return the results of a compare of two nodes.
(package private) void	mergesort(java.util.Vector a, java.util.Vector b, int l, int r, XPathContext support) This implements a standard Mergesort, as described in Robert Sedgewick's Algorithms book.
void	sort(DTMIterator v, java.util.Vector keys, XPathContext support) Given a vector of nodes, sort each node according to the criteria in the keys.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

m_execContext

XPathContext m_execContext

Current XPath context

m_keys

java.util.Vector m_keys

Vector of NodeSortKeys

Constructor Detail

NodeSorter

public NodeSorter(XPathContext p)

Construct a NodeSorter, passing in the XSL TransformerFactory so it can know how to get the node data according to the proper whitespace rules.

Parameters:

p - Xpath context to use

Method Detail

sort

public void sort(DTMIterator v,
                 java.util.Vector keys,
                 XPathContext support)
          throws javax.xml.transform.TransformerException

Given a vector of nodes, sort each node according to the criteria in the keys.

Parameters:

v - an vector of Nodes.

keys - a vector of NodeSortKeys.

support - XPath context to use

Throws:

javax.xml.transform.TransformerException

compare

int compare(NodeSorter.NodeCompareElem n1,
            NodeSorter.NodeCompareElem n2,
            int kIndex,
            XPathContext support)
      throws javax.xml.transform.TransformerException

Return the results of a compare of two nodes. TODO: Optimize compare -- cache the getStringExpr results, key by m_selectPat + hash of node.

Parameters:

n1 - First node to use in compare

n2 - Second node to use in compare

kIndex - Index of NodeSortKey to use for sort

support - XPath context to use

Returns:

The results of the compare of the two nodes.

Throws:

javax.xml.transform.TransformerException

mergesort

void mergesort(java.util.Vector a,
               java.util.Vector b,
               int l,
               int r,
               XPathContext support)
         throws javax.xml.transform.TransformerException

This implements a standard Mergesort, as described in Robert Sedgewick's Algorithms book. This is a better sort for our purpose than the Quicksort because it maintains the original document order of the input if the order isn't changed by the sort.

Parameters:

a - First vector of nodes to compare

b - Second vector of nodes to compare

l - Left boundary of partition

r - Right boundary of partition

support - XPath context to use

Throws:

javax.xml.transform.TransformerException