关于w3school提供的XPath教程的问题

Question

Here is the link http://www.w3schools.com/xpath/xpath_examples.asp

So clicked on the first try it yourself , you will see a fragment of code. And I don't know what this part means

while (result)
{
    document.write(result.childNodes[0].nodeValue);
    document.write("<br>");
    result=nodes.iterateNext();
}

The result there should be the least significant node there and should hava no childNodes. And according to the W3C standard, the first node should be denoted as node[1]. So I started to think that the atomic value is a node, and then made some modifications to the code.

I changed xpath expression and the while loop

  path="/bookstore/book";
  ...
  // original codes
  ...
  while (result)
  {
      document.write(result.childNodes[1].childNodes[0].nodeValue);
      document.write("<br>");
      result=nodes.iterateNext();
  }
  /* result
    Everyday Italian
    Harry Potter
    XQuery Kick Start
    Learning XML
  */

Seems it works well. Then another modification:

  path="/bookstore/book";
  ...
  // original codes
  ...
  while (result)
  {
      // the only change is here      **
      document.write(result.childNodes[2].childNodes[0].nodeValue);
      document.write("<br>");
      result=nodes.iterateNext();
  }
  /* result
    blank
  */

Seems strange now, I expected a list of authors, but what I have here is blank. Then another tiny modification:

      document.write(result.childNodes[3].childNodes[0].nodeValue);
 /* result
   Giada De Laurentiis
   J K. Rowling
   James McGovern
   Erik T. Ray
 */

The childNode[1] denotes the fisrt childNode of book - title, and the childNode[3] denoted the second - author? Then I was convinced that the childNode[2] should be the atomic value of title . But when I tried to display it using childNode[2] and failed, I won't show you more codes - already a load of codes here. So, can any one explain it to me?

And another question:

see link here http://www.w3schools.com/xpath/xpath_axes.asp

It introduced xpath axes and Location Path Expression , but no example is given within a code. I can hardly imagine how to use it, may anyone posts some examples and enlighten me? (I mainly use java)

Answer 1

You are confusing the DOM model and the XPath model. In the DOM model a node has a childNodes property where the index starts with 0 and where the childNodes contains all kind of nodes (element nodes, text nodes, CDATA section nodes, comment nodes, processing instruction nodes). The XPath data model is different and its path expressions are different from the DOM child nodes. Positional predicates in XPath start with 1 , not with 0 . Your sample uses XPath to address book elements in a DOM tree and then changes to address child nodes using the DOM API. That is often done in the browser environment as the browsers implement the W3C DOM but sometimes offer XPath selection over the DOM tree.

And you have to understand that in both the DOM and the XPath model with <foo>bar</foo> there is foo element node with a child node, a text node with contents bar . The examples you see are using XPath to select an element node, then sometimes use the childNodes collection to address a text child node and to extract its DOM nodeValue property. At least for the Mozilla, Opera, Chrome branch of the code that is not necessary, instead of foo.childNodes[0].nodeValue you could simply access foo.textContent . But all that is DOM and not XPath.

A tree for

<book category="COOKING">
  <title lang="en">Everyday Italian</title>
  <author>Giada De Laurentiis</author>
  <year>2005</year>
  <price>30.00</price>
</book>

can have a book element node which has 9 child nodes, the first (with index 0 in the DOM childNodes collection, with index 1 in XPath ./child::node() ) being a text node with white space, the second being the title element, the third being a text node with white space, the fourth (with DOM index 3 ) being the author element.

Of course with XPath, if you know you are only interested in element nodes you would not select node()[4] , instead you would select *[2] for the second element child or simply author for the author element child node.