How can you identify multiple elements with the same name in XPath?

Question

What if I had a document that had multiple elements with the same name -- how would I retrieve, for instance, the second element?

<doc>
...
 <element name="same">foo</element>
...
 <element name="same">bar</element>
...
 <element name="same">baz</element>
...
</doc>

I'd expect something like //elem[@name='same'][2] to work.

Additionally, how would I find the second from last element in xpath with a variable number of elements with the same name

Answer 1

[ ] has higher priority than // (and "//" is actually only an abbreviation, not an operator). This is so, because according to the XPath 1.0 Spec,

"// is short for /descendant-or-self::node()/"

and later:

"NOTE: The location path //para[1] does not mean the same as the location path /descendant::para[1]. The latter selects the first descendant para element; the former selects all descendant para elements that are the first para children of their parents."

Therefore, the XPath expression:

     //element[@name='same'][2]

means:

Select any element in the document, that is named "element", has an attribute "name" with value "same", and this element is the second such child of its parent.

What you want is:

     (//element[@name='same'])[2]

Note the brackets, which override the higher precedence of [].

Similarly, the last but one such node is selected by the following XPath expression:

     (//element[@name='same'])[last()-1]

Finally, a necessary warning: The use of the "//" abbreviation is very expensive as it causes the whole (sub)tree to be traversed. Whenever the structure of the document is known, it is recommended to use more specific constructs (location paths).