Nokogiri Xpath Double Looping

Question

What I'm trying to do is pul the code block that contains the td with the class default. This works perfectly fine. But then I need to sort out the different parts of the code block. When I try to do this with the second xpath call what it does is each time it prints all the comheads in each of the blocks

   def HeaderProcessor(doc)         
        doc.xpath("//td[@class='default']").each do |block|             
            puts block.xpath("//span[@class='comhead']").text
        end
    end

When I just print out block each block prints out once and contains the comment header and the comment. When I try to run the xpath it prints out EVERY comhead found in doc and seems to be ignoring the block variable.

Any ideas on how I can make this work? What am I miss understanding about xpath?

UPDATE:

<td class="default">
<div style="margin-top:2px; margin-bottom:-10px; ">
<span class="comhead">
#some data        
</span></div>
<br><span class="comment"><font color="#000000">#some more data</span>
</td>

Answer 1

You're telling Nokogiri to search from the root when you say //span[@class='comhead'], you just want */span[@class='comhead']:

doc.xpath("//td[@class='default']").each do |block|
    block.xpath("*/span[@class='comhead']").each do |span|
        puts span.text
    end
end

or even just this:

doc.xpath('//td[@class="default"]/*/span[@class="comhead"]').each do |span|
    puts span.text
end

if you don't need to do anything with the <td> elements.