|
Traversing the DOM using XPATH is much easier than writing all of the
node traversal logic yourself - there is less code and therefore fewer
bugs. On the other hand, the DOM is faster than XPATH so only use it
when it will replace a lot of code.
See the selectNodes() method on XMLNode
Example I
Here is an email from an Oracle engineer:
Using the method requires providing a pattern and a NSResolver
(namespace resolver). I'm not making heavy use of namespaces, so I believe
you can pass the element itself as the namespace resolver.
Attached is a set of helper methods I use on top
of selectNodes() which eventually should get pushed down into the V2
parser implementation for making this feature even easier to use:
With my helper methods, I can just do:
NodeList nl = XSLT.select("this/that[@theother='foo']");
or
System.out.println("This is the value of "
+XSLT.valueOf("/something//like[@this='5']"));
Quite powerful when you use the XPath expression document('foo.xml')
in the expression to selectNodes()...
Example II
public static final String ACTION_DELETE = "*/*[@rep_action=" + "\'" + "delete" + "\']";
public static void processMessage(String xmlRequestDoc)
{
boolean success = false;
// Parse the xmlRequestDoc string to a DOM
DOMParser parser = new DOMParser();
byte [] aByteArr = xmlRequestDoc.getBytes();
ByteArrayInputStream bais = new ByteArrayInputStream(aByteArr , 0,
aByteArr.length);
parser.parse(bais);
XMLDocument xmlDoc = (XMLDocument) parser.getDocument();
org.w3c.dom.NodeList nl = xmlDoc.selectNodes(ACTION_DELETE, null);
for (int i = 0; i < nl.getLength(); i++)
{
org.w3c.dom.Node node = nl.item(i);
String elementType = node.getNodeName();
}
}
|
