Question

I am trying to extract text out of a html.

doc = Nokogiri::HTML('<B> <A href="http://www.asl.com/foo/bar"> Status :</A></B> REGISTERED <BR>')

puts doc.search('//b').first.text
puts doc.search('//b[contains(text(),"Status")]/following-sibling::text()[1]').first.text

the first puts returns Status :
But the second puts throws an exception undefined method 'text' for nil:NilClass

Why the contains doesn’t search properly ?
or am I doing something wrong ?

Answer 1

I think you have the wrong idea of the text function in XPath. Unlike the DOM function it does not return a concatenated string of all text sub-nodes. Instead it selects individual text nodes.

In your example //text() would select three text nodes:

 [" ", " Status :", " REGISTERED "]

What you might want is this XPath expression:

//b/a[contains(text(),"Status")]/../following-sibling::text()[1]

Essentially it finds the a element having the correct text node, than walks up to the parent element (b) and then gets its sibling text node.