Ruby, Nokogiri Remove <ul> element selected by <class="foo"> - html

Question

I want to remove a Nokogiri node but I don't get it.
I got HTML code like this:
<div class="metis manual-toogle" id="tocList">...
<li id="tocElement-ebook_cs_1111111_11">...
<a data-content href="url" class=" "></a> <!-- only this urls I want -->
<ul class="foo">
<!-- the following content and urls I want to remove -->
<li class id="tocElement-ebook_cs_1111111_cs12">
<a data-content href="url" class=" "></a>
...
<a data-content href="url" class=" "></a>
</li>
</ul>
</li>
</div>
What I've tried so far:
document = Nokogiri::HTML.parse(html_input)
document.xpath('//ul[#class="foo"]').each {|x| x.remove}
document.xpath('//ul[#class="foo"]').children.map(:&remove)
What I'm doing wrong?
Edit:
I wan't to parse some URL's. I got the html structure like above. The URL I want is in the <li></li> block nested like <a data-content href="url"></a>. The problem is, inside the <ul></ul> is a <a data-content href="url"></a> too. I can extract every URL, but only need the main URL's.
It's a book with some chapters, I can download the chapter with the first link. Every sub-chapter (inside the <ul>) got a own pdf.
I can't use regular expression because the links are not build the same. For example in one book it's
chapter 1 pdf: ...-ch1.pdf (contains all sub-chapter)
chapter 1-1 pdf: ...-ch1-1.pdf
chapter 2 pdf: ...-923df2.pdf
chapter 3 pdf: ...-ch3.pdf
The HTML-Code is a mess. The easiest way is to remove the <ul> block itself.

Answer 1

You are not providing much context or detail here. But, the following code should remove the item you want if you are selecting it correctly. Please provide more details such as your output received, output expected, etc.
Given the limited information, you could try this bit:
UPDATE:
html.html
<div class="metis manual-toogle" id="tocList">...
<li id="tocElement-ebook_cs_1111111_11">...
<a data-content href="url" class=" "></a> <!-- only this urls I want -->
<ul class="foo">
<!-- the following content and urls I want to remove -->
<li class id="tocElement-ebook_cs_1111111_cs12">
<a data-content href="url" class=" "></a>
...
<a data-content href="url" class=" "></a>
</li>
</ul>
</li>
</div>
main.rb
require 'nokogiri'
require 'open-uri'
require 'pry'
doc = Nokogiri::HTML(open('html.html'))
doc.xpath('//ul[#class="foo"]').remove
doc.xpath('//a').each do |item|
puts item
end
Output:
~/code/projects/test ⌚ 8:28:32
$ ruby main.rb ‹2.6.1›
<a data-content href="urliwant" class=" "></a>
We worked this out through chat. Above example works. But, for his specific case we needed to do this because of the messy html:
document = Nokogiri::HTML(open('html.html'))
document.css('//ul//ul//ul').remove
document.css('ul .collapse').remove
links = document.xpath('//*[#id="toc"]//ul')
File.open("input.html", "a") do |output_txt|
links.each do |item|
output_txt.write(item)
end
end