Question

I am brand new to Ruby and Xpath. I need to extract the System features from the table at

http://h10010.www1.hp.com/wwpc/ie/en/ho/WF06b/321957-321957-3329742-89318-89318-5186820-5231694.html?dnr=1

So far I have tried targeting all of the td tags, the page doesn’t use CSS ids so I cant target that way.

I tried the following code

doc.xpath('//tr/th/span[normalize-space(text())="System features"]/..')

but it returns nothing ;(

Does anyone have any idea the best way to approach this?

Answer 1

That expression should work fine on the given source, but it’s not really idiomatic. You probably want to use something more like this:

//tr/th[span[normalize-space()='System features']

1. normalize-space expects a string argument. Passing the node-set returned by text() forces conversion to a string by taking the first text node in document order. This doesn’t really matter in your document, because there’s only one child text node, but you should be aware that this is what’s happening.
2. You don’t need to backtrack up the tree using /.. at the end of the expression. You can test for the presence of the child span using a nested predicate and thereby select the desired th directly.

If you want to exploit the fact that the target th contains only the one child span node, you could write this simplified expression:

//tr/th[normalize-space(span)='System features']

So, why isn’t working? Hard to tell, but it could be because the tool you’re using to parse the document is creating a structure that differs from how it appears in the literal source (e.g. because the input isn’t really well-formed XML). Try a slightly different expression:

//*[span[@class='themebody' and normalize-space()='System features']]

Or maybe you should first verify that you can retrieve the span itself, then build the expression up from that:

//span[@class='themebody' and normalize-space()='System features']