Can XPath match on parts of an element's name?

Question

I want to do this:

//*fu

which returns all nodes whose name ends in fu, such as <tarfu /> and <snafu />, but not <fubar />

Solution 1

Do something like:

//*[ends-with(name(), 'fu')]

For a good XPath reference, check out W3Schools.

Solution 2

This answer is for XPath 1.0 where there is no equivalent of the XPath 2.0 standard ends-with() function.

The following XPath 1.0 expression selects all elements in the xml document, whose names end with the string "fu":

//*[substring(name(),string-length(name())-1) = 'fu']

Solution 3

I struggled with Dimitre Novatchev's answer, it wouldn't return matches. I knew your XPath must have a section telling that "fu" has length 2.

It's advised to have a string-length('fu') to determine what to substring.

For those who aren't able to get results with his answer and they require solution with xpath 1.0:

//*[substring(name(), string-length(name()) - string-length('fu') +1) = 'fu']

Finds matches of elements ending with "fu"

or

//*[substring(name(), string-length(name()) - string-length('Position') +1) = 'Position']

Finds matches to elements ending with "Position"

Comments

[Wayne]

This is an XPath 2.0 solution only.

[LostNomad311]

note that you can use local-name() to omit namespaces in complex XML

[user766308]

Thanks for this helpful comment. I was trying to do this with lxml, which supports ends-with, but not on element names. Your answer was exactly what I needed!