Update XML document using Java

Question

On the internet I am finding a lot of examples of how to achieve this functionality, however none of them seem to work in my case. I have an string containing XML data, and I know what the XPath of the result should look like. Say for example I have an xml structure looking like this:

   <?xml version="1.0"?>
     <m:parent xmlns:m="http://www.somescheme">
      <doc:header xmlns:doc="http://www.somescheme">
        <doc:documentId>137</doc:documentId>
        <doc:documentDescription>Some Description</doc:documentDescription>  
         <doc:task>
           <doc:id>49</doc:id>
           <doc:name>Some Name</doc:name>
           <doc:description>Some Task Description</doc:description>
           <doc:outcome/>
           <doc:dueDate/>
           <doc:priority>50</doc:priority>
          </doc:task>
        </doc:header>
        <m:otherinfo>234324</m:otherInfo>
      </m:parent>

Say for example I need to access and update otherInfo the XPath would be 'm:parent/m:otherInfo' which I should be able to read and set. For doc:outcome it would be 'm:parent/doc:header/doc:task/doc:outcome' This is a value I need to set at runtime.

Document doc = DocumentBuilderFactory.newInstance().newDocumentBuilder().parse(
        new InputSource("data.xml"));

    XPath xpath = XPathFactory.newInstance().newXPath();
    NodeList nodes = (NodeList) xpath.evaluate("//m:parent/doc:header/doc:task/doc:outcome", doc,
        XPathConstants.NODESET);

The code should be something like above, however I am not getting results,

Can you point out what I am missing ?

Kind Regards Max

Answer 1

You have to do two additional things before you can evaluate XPATH expressions with namespaces.

1. The DocumentBuilderFactory instance is not namespace aware by default. So you have to call setNamespaceAware(true) .
2. Create your own implementation of the interface NamespaceContext and pass it to your XPath instance.

Here a simple NamespaceContext implementation as a static inner class. Only the getNamespaceURI() method is implemented here. ( See this IBM developerWorks article for more details and alternative implementations: Using the Java language NamespaceContext object with XPath )

static class MyNamespaceContext implements NamespaceContext {
  @Override
  public String getNamespaceURI(String prefix) {
    if(prefix == null) {
      throw new IllegalArgumentException();
    } else if(prefix.equals(XMLConstants.DEFAULT_NS_PREFIX)) {
      return XMLConstants.NULL_NS_URI;
    } else if(prefix.equals(XMLConstants.XML_NS_PREFIX)) {
      return XMLConstants.XML_NS_URI;
    } else if(prefix.equals(XMLConstants.XMLNS_ATTRIBUTE)) {
      return XMLConstants.XMLNS_ATTRIBUTE_NS_URI;
    } else if(prefix.equals("m")) {
      return "http://www.somescheme";
    } else if(prefix.equals("doc")) {
      return "http://www.somescheme";
    } else {
      return XMLConstants.NULL_NS_URI;
    }
  }

  @Override
  public String getPrefix(String namespaceURI) {
    throw new UnsupportedOperationException();
  }

  @Override
  public Iterator getPrefixes(String namespaceURI) {
    throw new UnsupportedOperationException();
  }
}

Here the part with the DocumentBuilderFactory :

DocumentBuilderFactory documentBuilderFactory = 
    DocumentBuilderFactory.newInstance();
documentBuilderFactory.setNamespaceAware(true);
Document doc = documentBuilderFactory.newDocumentBuilder().parse(
    new InputSource("data.xml"));

XPath xpath = XPathFactory.newInstance().newXPath();
xpath.setNamespaceContext(new MyNamespaceContext());
NodeList nodes = (NodeList) xpath.evaluate(
    "//m:parent/doc:header/doc:task/doc:outcome", doc,
    XPathConstants.NODESET);

The two namespace prefixes are actually linked to the same namespace URI. In fact you can use only one prefix in your XPATH expression like "//doc:parent/doc:header/doc:task/doc:outcome" and you don't need to declare both in your NamespaceContext implementation.