Question

My problem is that i need to retrieve original html from XPath expressions designed to get text parts of html.
For example if i got the current html :

<div class='text'>
    <p class='nointerrest'> blabla </p>
    <p class='goodstuff'> blablabla </p>
    <p class='goodstuff'> blablabla </p>
</div>

Then my XPath expression is the string :

    string-join(//p[@class='goodstuff'],' ') 

which give me :

        "blablabla blablabla"

is there an XPath expression which can return me :

<p class='goodstuff'> blablabla </p>
<p class='goodstuff'> blablabla </p>

as a string ? (If possible with an XPath expression the most similar to my original XPath expression )

Thanks in advance

Answer 1

You’ll need more than just XPath to get the desired output. XPath is a document query language, not an XML formatting language, so you should think of this in distinct steps.

1. First, use XPath to query the document for a NodeList:

   NodeList list = (NodeList) xpath.evaluate(
       "//p[@class='goodstuff']", document, XPathConstants.NODESET);
   

2. Then create a Transformer:

   javax.xml.transform.TransformerFactory tfactory = 
       TransformerFactory.newInstance();
   javax.xml.transform.Transformer xform = tfactory.newTransformer();
   

3. And use it to output each node:

   for (int i = 0; i < list.getLength(); i++) {
       Node node = list.item(i);
       javax.xml.transform.Source src = new DOMSource(node);
       java.io.StringWriter writer = new StringWriter();
       javax.xml.transform.Result result = new   
       javax.xml.transform.stream.StreamResult(writer);
       xform.setOutputProperty(
       javax.xml.transform.OutputKeys.OMIT_XML_DECLARATION, "yes");
       xform.transform(src, result);
       System.out.println(writer.toString());
   }