How can I select links without id or class attributes using XPath?

I want to select number 4 or 5 links

i have tried using this expression:

//div/b[contains(text(),'Sida:')]/following-siblings::a[4]

but it didnt work

The code for this part of my HTML is:

<div><b>Sida: </b> <a>1</a> « <a>3</a> <a>4</a> <开发者_如何学编程a>5</a> » <a>814</a></div>

Here is how its looks

http://i.stack.imgur.com/rFmug.jpg

Sida:  1   2   3   4 »  814 

or

Sida:  1  «  3   4   5  »  814 

there are a long html page without any id or class name - so i need autopager in firefox to find the correct link <a> and automatically auto inert next page

i just need the expression to pinpoint the number 4 or 5 - but these number are changing when page number are changing

check the web page l-like.it

---

You almost did it yourself.

To get a node-set of two elements (which can be not what you want) you can use this:

//div/b[contains(text(),'Sida:')]/following-sibling::a[
                position() = 3 or position() = 4
                ]

Note: Use // only when schema is unknown.

To get a concatenated string value:

concat(
    //div/b[contains(text(),'Sida:')]/following-sibling::a[3],
    //div/b[contains(text(),'Sida:')]/following-sibling::a[4]
    )

Output with the input below will be 45.

XML test sample:

<t>
    <div>bla-bla</div>
    <div id="test">
        <div>
            <b>Sida: </b>
            <a>1</a> «
            <a>3</a>
            <a>4</a>
            <a>5</a> »
            <a>814</a>
        </div>
    </div>
    <div>bla-bla</div>
</t>