From: David Masover <ninja@...>
Date: 2008-08-11T04:02:42+09:00
Subject: Re: Hpricot and path of an elememt

On Sunday 10 August 2008 13:36:42 Li Chen wrote:

> I use hpricot to load a page. Then I  try to find the path for an
> element "font"(<font face="courier" color="black">) in the page.

So, you probably want:

(doc / 'font')

> doc.at("#header").xpath
> #=> "//div[@id='header']"

Right, that's searching for a tag that looks like this: <div id="header">

> here is my code:
> puts doc.at("#font").xpath

And that's searching for a tag that looks like this: <div id="font">

If you're following that example, you probably want:

puts doc.at('font').xpath

Now, first question: Why do you need the xpath? Usually, the idea is to try to 
find that element, and then do something with it. So, for example:


# To return all text:
(doc / 'font').text

# To loop over each font element:
(doc / 'font').each { |tag|
  puts tag.inner_text
}


Second question: Why is there a font tag on this page? If you had any hand in 
creating the page, shame on you -- go learn some CSS.

In fact, go learn some CSS anyway. Hpricot supports both CSS selectors and 
XPath, and it's usually much easier to use the selectors. Years later, I 
still remember, roughly, how selectors work -- but only a few months later, 
I've almost completely forgotten XPath.

There are things XPath can do that selectors can't. But until you encounter 
them, XPath is overkill.