使用Nokogiri进行快速XML解析

Question

I'm parsing an XML file with Ruby and Nokogiri, but I have some problems with execution time. It's taking one minute for 4000 lines - it is too much. I'm trying to find some ideas for code optimisation.

This is an example of the code:

input = Nokogiri::XML(File.open(file_name))
input.xpath('//Transition').each do |transition_record|
  from = transition_record.xpath('@From')
  to = transition_record.xpath('@To')
  name = transition_record.xpath('@Name')
  Transition.create do |transition|
    transition.from_id = from.to_s
    transition.to_id = to.to_s
    transition.name = name.to_s

    if transition_record.at_xpath('Condition').at_xpath('@Type')
      transition.condition = transition_record.at_xpath('Condition').at_xpath('@Type').to_s
    end
  end
end

Transition isn't fixed place in place in the XML and can be at any level. Also, I'm parsing three other tags in the some way.

Does anybody have any ideas how I can improve it?

Answer 1

You can reorganize your code so that you are creating an array of hashes with Nokogiri, then you can use them directly in create . This will allow you to time each part separately. I think you'll find that Nokogiri isn't the problem.