132

Suppose I have the following XML:

<book category="CLASSICS">
  <title lang="it">Purgatorio</title>
  <author>Dante Alighieri</author>
  <year>1308</year>
  <price>30.00</price>
</book>

<book category="CLASSICS">
  <title lang="it">Inferno</title>
  <author>Dante Alighieri</author>
  <year>1308</year>
  <price>30.00</price>
</book>

<book category="CHILDREN">
  <title lang="en">Harry Potter</title>
  <author>J K. Rowling</author>
  <year>2005</year>
  <price>29.99</price>
</book>

<book category="WEB">
  <title lang="en">XQuery Kick Start</title>
  <author>James McGovern</author>
  <author>Per Bothner</author>
  <author>Kurt Cagle</author>
  <author>James Linn</author>
  <author>Vaidyanathan Nagarajan</author>
  <year>2003</year>
  <price>49.99</price>
</book>

<book category="WEB">
  <title lang="en">Learning XML</title>
  <author>Erik T. Ray</author>
  <year>2003</year>
  <price>39.95</price>
</book>

I would like to do an xpath that gets back all book nodes that have a title node with a language attribute of "it".

My attempt looked something like this:

//book[title[@lang='it']]

But that didn't work. I expect to get back the nodes:

<book category="CLASSICS">
  <title lang="it">Purgatorio</title>
  <author>Dante Alighieri</author>
  <year>1308</year>
  <price>30.00</price>
</book>

<book category="CLASSICS">
  <title lang="it">Inferno</title>
  <author>Dante Alighieri</author>
  <year>1308</year>
  <price>30.00</price>
</book>

Any hints?

Improve this question
1
  • What XPath implementation is this? Commented Sep 25, 2009 at 22:53

5 Answers 5

Reset to default
202

Try

//book[title/@lang = 'it']

This reads:

  • get all book elements
    • that have at least one title
      • which has an attribute lang
        • with a value of "it"

You may find this helpful — it's an article entitled "XPath in Five Paragraphs" by Ronald Bourret.

But in all honesty, //book[title[@lang='it']] and the above should be equivalent, unless your XPath engine has "issues." So it could be something in the code or sample XML that you're not showing us -- for example, your sample is an XML fragment. Could it be that the root element has a namespace, and you aren't counting for that in your query? And you only told us that it didn't work, but you didn't tell us what results you did get.

Improve this answer
Sign up to request clarification or add additional context in comments.

3 Comments

How to do the same if title is not a direct child of book, but somewhere deeper and we don't know where exactly? //book[/title/@lang = 'it'] doesn't seem to work?
Martin, you could use //book[.//title/@lang = 'it']. I believe the trick is the "." in the beginning of the condition.
Thanks for the link, excellent article. I've been using xPath for years but this really helped me understand the underlying logic!
64

Years later, but a useful option would be to utilize XPath Axes (https://www.w3schools.com/xml/xpath_axes.asp). More specifically, you are looking to use the descendants axes.

I believe this example would do the trick:

//book[descendant::title[@lang='it']]

This allows you to select all book elements that contain a child title element (regardless of how deep it is nested) containing language attribute value equal to 'it'.

I cannot say for sure whether or not this answer is relevant to the year 2009 as I am not 100% certain that the XPath Axes existed at that time. What I can confirm is that they do exist today and I have found them to be extremely useful in XPath navigation and I am sure you will as well.

Improve this answer

1 Comment

This is great, descendant or other axes can give us an opportunity to have locate node as its attribute,not hierarchical way as usual .
15 
//book[title[@lang='it']]

is actually equivalent to 

 //book[title/@lang = 'it']

I tried it using vtd-xml, both expressions spit out the same result... what xpath processing engine did you use? I guess it has conformance issue Below is the code

import com.ximpleware.*;
public class test1 {
  public static void main(String[] s) throws Exception{
      VTDGen vg = new VTDGen();
      if (vg.parseFile("c:/books.xml", true)){
          VTDNav vn = vg.getNav();
          AutoPilot ap = new AutoPilot(vn);
          ap.selectXPath("//book[title[@lang='it']]");
                  //ap.selectXPath("//book[title/@lang='it']");

          int i;
          while((i=ap.evalXPath())!=-1){
              System.out.println("index ==>"+i);
          }
          /*if (vn.endsWith(i, "< test")){
             System.out.println(" good ");  
          }else
              System.out.println(" bad ");*/

      }
  }
}
Improve this answer

3 Comments

+1 that its a compliance issue and that the syntax generates the same nodeset. Similar code in C# also works.
-1: Mr. Zhang, I was trying to do you a favor by removing code not relevant to the question. It allowed me to not downvote you, which I now feel I have to. Note that no other answer included code to call call the query.
+1: Because I can't figure out what Mr. Saunders is talking about - no other answer added ANY code, and this answer shows the code used so we can 1: validate his methods and 2: perform his test ourselves. The code is short and easy to read. I don't see the problem.
4

I would think your own suggestion is correct, however the xml is not quite valid. If you are running the //book[title[@lang='it']] on <root>[Your"XML"Here]</root> then the free online xPath testers such as one here will find the expected result.

Improve this answer

Comments

2

Try to use this xPath expression:

//book/title[@lang='it']/..

That should give you all book nodes in "it" lang

Improve this answer

2 Comments

the result of that expression is the title nodes, not the book nodes
That is not true. It will return book nodes (these two dots at the end are to aim at the upper node of title node).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.