CODEWITHSUNDEEP
rubyxmlnokogiri
Dubb
Dubb

Reputation: 45

How to get siblings' child according to specific defined sibling content

I need to find the best method to gather writer and artist information from the following XML data. The comic node appears multiple times and includes data for a single comic book.

I can't grab the appropriate person according to their job function, writer, artist, etc. There are sometimes multiple writers and artists of each comic book. My plan is to add/append each to a List.

So, for this single comic book, I need to get all the writers' and artists' display name, but the job function (e.g. writer) is a sibling of the persons name.

Here is what I have, but doesn't work:

writer = []
penciler = []
doc.xpath('//comic').each do |main_element|
 main_element.xpath("mainsection/credits/credit/role[@id='dfWriter']").each do |n|
    writer << n.xpath('person/displayname').text
  end
  main_element.xpath("mainsection/credits/credit/role[@id='dfPenciler']").each do |n|
    penciler << n.xpath('person/displayname').text
  end
end

p "Writer(s): ",writer
p "Penciler(s): ",penciler

This is the XML file/data:

<comic>
  <id>3398</id>
  <index>195</index>
  <mainsection>
    <title>Mind Games</title>
    <myrating>0</myrating>
    <myrating>
      <displayname>0</displayname>
      <sortname>0</sortname>
    </myrating>
    <pagecount>32</pagecount>
    <credits>
      <credit>
        <role id="dfWriter">Writer</role>
        <roleid>dfWriter</roleid>
        <person>
          <displayname>Will Pfeifer</displayname>
          <sortname>Pfeifer, Will</sortname>
          <lastname>Pfeifer</lastname>
          <firstname>Will</firstname>
        </person>
      </credit>
      <credit>
        <role id="dfWriter">Writer</role>
        <roleid>dfWriter</roleid>
        <person>
          <displayname>John Byrne</displayname>
          <sortname>Byrne, John</sortname>
          <lastname>Byrne</lastname>
          <firstname>John</firstname>
        </person>
      </credit>
      <credit>
        <role id="dfPenciler">Penciller</role>
        <roleid>dfPenciler</roleid>
        <person>
          <displayname>John Byrne</displayname>
          <sortname>Byrne, John</sortname>
          <lastname>Byrne</lastname>
          <firstname>John</firstname>
        </person>
      </credit>
    </credits>
  </mainsection>
</comic>

The code I have does not give me the desired results. I found "Getting the siblings of a node with Nokogiri" but I need to iterate and grab each sibling.

I can either search by <roleid>dfWriter</roleid> or <role id="dfWriter">Writer</role> as they are the same.

My expected output would be:

Writer(s): Will Pfeifer, John Byrne 
Penciler(s): John Byrne

Upvotes: 2

Views: 445

Answers (2)

the Tin Man
the Tin Man

Reputation: 160631

Here's how I'd go about doing this:

require 'nokogiri'

XML = <<EOT
<comic>
  <mainsection>
    <credits>
      <credit>
        <role id="dfWriter">Writer</role>
        <person>
          <displayname>Will Pfeifer</displayname>
        </person>
      </credit>
      <credit>
        <role id="dfWriter">Writer</role>
        <person>
          <displayname>John Byrne</displayname>
        </person>
      </credit>
      <credit>
        <role id="dfPenciler">Penciller</role>
        <person>
          <displayname>John Byrne</displayname>
        </person>
      </credit>
    </credits>
  </mainsection>
</comic>
EOT

doc = Nokogiri::XML(XML)

writers = doc.search("credits role[id='dfWriter']").map { |w| w.parent.at('displayname').text }
pencilers = doc.search("credits role[id='dfPenciler']").map { |n| n.parent.at('displayname').text }

puts "Writer(s): %s" % writers.join(', ')
puts "Penciler(s): %s" % pencilers.join(', ')

# >> Writer(s): Will Pfeifer, John Byrne
# >> Penciler(s): John Byrne

Which, when run, outputs:

# >> Writer(s): Will Pfeifer, John Byrne
# >> Penciler(s): John Byrne

This:

writers = doc.search("credits role[id='dfWriter']").map { |w| w.parent.at('displayname').text }
pencilers = doc.search("credits role[id='dfPenciler']").map { |n| n.parent.at('displayname').text }

could be DRY'd to:

writers, pencilers = %w(dfWriter dfPenciler).map { |s|
  doc.search("credits role[id='#{s}']").map { |w| w.parent.at('displayname').text }
}

I used CSS for readability, and at, which returns a Node, when I want the text instead of xpath, which returns a NodeSet.

That distinction between using text on a NodeSet versus a Node is extremely important. Consider this:

require 'nokogiri'

xml = <<EOT
<root>
  <displayname>Will Pfeifer</displayname>
  <displayname>John Byrne</displayname>
  <displayname>John Byrne</displayname>
</root>
EOT

doc = Nokogiri::XML(xml)
doc.search('displayname').class # => Nokogiri::XML::NodeSet
doc.search('displayname').text # => "Will PfeiferJohn ByrneJohn Byrne"

doc.at('displayname').class # => Nokogiri::XML::Element
doc.at('displayname').text # => "Will Pfeifer"

If you want all the text for a NodeSet in a easily usable form then extract it from each node:

doc.search('displayname').map(&:text) # => ["Will Pfeifer", "John Byrne", "John Byrne"]

Upvotes: 0

har07
har07

Reputation: 89335

You can use XPath following-sibling axis for this purpose assuming the target element always located after role :

doc.xpath('//comic').each do |main_element|
 main_element.xpath("mainsection/credits/credit/role[@id='dfWriter']").each do |n|
    writer << n.xpath('following-sibling::person/displayname').text
  end
  main_element.xpath("mainsection/credits/credit/role[@id='dfPenciler']").each do |n|
    penciler << n.xpath('following-sibling::person/displayname').text
  end
end

Or you can just iterate through credit instead of role in the first place :

doc.xpath('//comic').each do |main_element|
 main_element.xpath("mainsection/credits/credit[role/@id='dfWriter']").each do |n|
    writer << n.xpath('person/displayname').text
  end
  main_element.xpath("mainsection/credits/credit[role/@id='dfPenciler']").each do |n|
    penciler << n.xpath('person/displayname').text
  end
end

Upvotes: 1

Related Questions