xpath index text node

Question

I'm playing around with xpath syntax and got stuck on one problem:

Given the following xml

<root xmlns:foo="http://www.foo.org/" xmlns:bar="http://www.bar.org">
    <employees>
        <employee id="1">
            <name>
                Johnny Dapp
                <br />
                <br />
            </name>
        </employee>
        <employee id="2"><name>Al Pacino</name></employee>
        <employee id="3"><name>Robert De Niro</name></employee>
        <employee id="4"><name>Kevin Spacey</name></employee>
        <employee id="5"><name>Denzel Washington</name></employee>
        <notanemployee id="14"><name>HELLO WORLD</name></notanemployee>
    </employees>
    <foo:companies>
        <foo:company id="6">Tata Consultancy Services</foo:company>
        <foo:company id="7">Wipro</foo:company>
        <foo:company id="8">Infosys</foo:company>
        <foo:company id="9">Microsoft</foo:company>
        <foo:company id="10">IBM</foo:company>
        <foo:company id="11">Apple</foo:company>
        <foo:company id="12">Oracle</foo:company>
        <bar:company id="13">Google</bar:company>
    </foo:companies>
</root>

I first tried

//text()

which to my knowledge should select all the text nodes, and it does.

However I then tried

//text()[1]

expecting that it would select the first text node in the xml, which should be a text node consisting of a line break and then some white spaces (after the root opening tag and before the employees opening tag), but to my surprise it returned this

Text='
                Johnny Dapp
                '
Text='Al Pacino'
Text='Robert De Niro'
Text='Kevin Spacey'
Text='Denzel Washington'
Text='HELLO WORLD'
Text='
        '
Text='Tata Consultancy Services'
Text='Wipro'
Text='Infosys'
Text='Microsoft'
Text='IBM'
Text='Apple'
Text='Oracle'
Text='Google'

How does that even make sense? How did it even return a set of text nodes to begin with?

I then tried

(//text())[1]

and it did return the first text node of the xml. How is this different from

//text()[1]

Btw I'm using two online xpath testers:

https://www.freeformatter.com/xpath-tester.html#ad-output

https://www.webtoolkitonline.com/xml-xpath-tester.html?

JaSON JaSON · Accepted Answer · 2018-11-20T21:36:23

//text()[1] returns first text node of each element. (//text())[1] - returns the fisrt occurence of text node in document

For instance, if source code is

<a>
  foo
  <br>
  bar
</a>
<a>
  baz
</a>

//text()[1] should return "foo" and "baz" (first text nodes of each a node) while //text[2] will return "bar" (second text node)

xpath index text node

1 Answers