Finding elements containing specific text

Find all elements with certain text

Imagine the following XML:

<root>
    <element>hello</element>
    <another>
        hello
    </another>
    <example>Hello, <nested> I am an example </nested>.</example>
</root>

The following XPath expression:

//*[text() = 'hello']

will return the <element>hello</element> element, but not the <another> element. This is because the <another> element contains whitespace surrounding the hello text.

To retrieve both <element> and <another>, one could use:

//*[normalize-space(text()) = 'hello']

or

//*[normalize-space() = 'hello']

which will trim the surrounding whitespace before doing the comparison. Here we can see that the text() node specifier is optional when using normalize-space.

To find an element containing specific text, you can use the contains function. The following expression will return the <example> element:

//example[contains(text(), 'Hello')]

If you want to find text that spans multiple children/text nodes, then you can use . instead of text(). . refers to the entire text content of the element and it’s children.

//example[. = 'Hello,  I am an example .']

To see the multiple text nodes, you can use:

//example//text()

which will return:

• “Hello, ”
• ” I am an example ”
• ”.”

And to more clearly see the entire text content of an element, one can use the string function:

string(//example[1])

or just

string(//example)

Hello, I am an example .

The latter works because, if a node-set is passed into functions like string, XPath 1.0 just looks at the first node (in document order) in that node-set, and ignores the rest.

so:

string(/root/*)

would return:

hello