Question

i try to extract data from a xml in rails application with the Nokogiri gem,

the xml :

<item>
    <description>
        <![CDATA[<img src="something" title="anothething">
        <p>text, bla bla...</p>]]>
    </description>
</item>

Actually i do something like this to extract data from the xml :

def test_content
    @return = Array.new
    site = 'http://www.les-encens.com/modules/feeder/rss.php?id_category=0'
    @doc = Nokogiri::XML(open(site, "UserAgent" => "Ruby-OpenURI"))
    @doc.xpath("//item").each do |n|
        @return << [
            n.xpath('description')
        ] 
   end
end

Could you show me how extract just the src attribute from the img tag ?

Edit :
I have replace the xml by the correct one.

Answer 1

The result of an xpath call made in Nokogiri is going to be a NodeSet, which is simply a list of Nokigiri Nodes

With this in mind we can just pull examples from the Nokogiri Documentation and adapt them.

To answer your question, “Could you show me how extract just the src attribute from the img tag ?”, here is one such way.

#the 'open' method here is part of the open-uri library
xml = Nokogiri::XML(open(your_url_here))

all_images = xml.xpath("//img") #returns NodeSet (list of Nokogiri Nodes)

image_sources = []

#iterate through each node
all_images.each() do |node|
  image_sources << node.get_attribute('src') #One method
  #image_sources << node['src'] #Another convention we could use
end

As Phrogz notes below, a more idomatic way of pulling the ‘src’ attribute from all of the images nodes is to map the ‘src’ attribute directly rather than iterating and pushing onto an array.

image_sources = all_images.map{ |node| node['src'] }