Question

I’m way new to working with XML but just had a need dropped in my lap. I have been given an usual (to me) XML format. There are colons within the tags.

<THING1:things type="Container">
  <PART1:Id type="Property">1234</PART1:Id>
  <PART1:Name type="Property">The Name</PART1:Name>
</THING1:things>

It is a large file and there is much more to it than this but I hope this format will be familiar to someone. Does anyone know a way to approach an XML document of this sort?

I’d rather not just write a brute-force way of parsing the text but I can’t seem to make any headway with REXML or Hpricot and I suspect it is due to these unusual tags.

my ruby code:

    require 'hpricot'
    xml = File.open( "myfile.xml" )

    doc = Hpricot::XML( xml )

   (doc/:things).each do |thg|
     [ 'Id', 'Name' ].each do |el|
       puts "#{el}: #{thg.at(el).innerHTML}"
     end
   end

…which is just lifted from: http://railstips.org/blog/archives/2006/12/09/parsing-xml-with-hpricot/

And I figured I would be able to figure some stuff out from here but this code returns nothing. It doens’t error. It just returns.

Answer 1

As @pguardiario mentioned, Nokogiri is the de facto XML and HTML parsing library. If you wanted to print out the Id and Name values in your example, here is how you would do it:

require 'nokogiri'

xml_str = <<EOF
<THING1:things type="Container">
  <PART1:Id type="Property">1234</PART1:Id>
  <PART1:Name type="Property">The Name</PART1:Name>
</THING1:things>
EOF

doc = Nokogiri::XML(xml_str)

thing = doc.at_xpath('//things')
puts "ID   = " + thing.at_xpath('//Id').content
puts "Name = " + thing.at_xpath('//Name').content

A few notes:

• at_xpath is for matching one thing. If you know you have multiple items, you want to use xpath instead.
• Depending on your document, namespaces can be problematic, so calling doc.remove_namespaces! can help (see this answer for a brief discussion).
• You can use the css methods instead of xpath if you’re more comfortable with those.
• Definitely play around with this in irb or pry to investigate methods.

Resources

• Parsing an HTML/XML document
• Getting started with Nokogiri

Update

To handle multiple items, you need a root element, and you need to remove the // in the xpath query.

require 'nokogiri'

xml_str = <<EOF
<root>
  <THING1:things type="Container">
    <PART1:Id type="Property">1234</PART1:Id>
    <PART1:Name type="Property">The Name1</PART1:Name>
  </THING1:things>
  <THING2:things type="Container">
    <PART2:Id type="Property">2234</PART2:Id>
    <PART2:Name type="Property">The Name2</PART2:Name>
  </THING2:things>
</root>
EOF

doc = Nokogiri::XML(xml_str)
doc.xpath('//things').each do |thing|
  puts "ID   = " + thing.at_xpath('Id').content
  puts "Name = " + thing.at_xpath('Name').content
end

This will give you:

Id   = 1234
Name = The Name1

ID   = 2234
Name = The Name2

If you are more familiar with CSS selectors, you can use this nearly identical bit of code:

doc.css('things').each do |thing|
  puts "ID   = " + thing.at_css('Id').content
  puts "Name = " + thing.at_css('Name').content
end