Getting certain attributes based on other attributes in Nokogiri? - ruby

Question

Here's the XML I'm working with:
<order xmlns="http://example.com/schemas/1.0">
<link type="application/xml" rel="http://example.com/rel/self" href="https://example.com/orders/1631"/>
<link type="application/xml" rel="http://example.com/rel/order/history" href="http://example.com/orders/1631/history"/>
<link type="application/xml" rel="http://example.com/rel/order/transition/release" href="https://example.com/orders/1631/release"/>
<link type="application/xml" rel="http://example.com/rel/order/transition/cancel" href="https://example.com/orders/1631/cancel"/>
<state>hold</state>
<order-number>123-456-789</order-number>
<survey-title>Testing</survey-title>
<survey-url>http://example.com/s/123456</survey-url>
<number-of-questions>6</number-of-questions>
<number-of-completes>100</number-of-completes>
<target-group>
<country>
<id>US</id>
<name>United States</name>
</country>
<min-age>15</min-age>
</target-group>
<quote>319.00</quote>
<currency>USD</currency>
</order>
What I need to do is get the href attribute, from the link that has a rel of http://example.com/rel/order/transition/release
So, how can I do that using Nokogiri?

Answer 1

Easy-peasy:
require 'nokogiri'
doc = Nokogiri::XML(<<EOT)
<order xmlns="http://example.com/schemas/1.0">
<link type="application/xml" rel="http://example.com/rel/self" href="https://example.com/orders/1631"/>
<link type="application/xml" rel="http://example.com/rel/order/history" href="http://example.com/orders/1631/history"/>
<link type="application/xml" rel="http://example.com/rel/order/transition/release" href="https://example.com/orders/1631/release"/>
<link type="application/xml" rel="http://example.com/rel/order/transition/cancel" href="https://example.com/orders/1631/cancel"/>
<state>hold</state>
<order-number>123-456-789</order-number>
<survey-title>Testing</survey-title>
<survey-url>http://example.com/s/123456</survey-url>
<number-of-questions>6</number-of-questions>
<number-of-completes>100</number-of-completes>
<target-group>
<country>
<id>US</id>
<name>United States</name>
</country>
<min-age>15</min-age>
</target-group>
<quote>319.00</quote>
<currency>USD</currency>
</order>
EOT
href = doc.at('link[rel="http://example.com/rel/order/transition/release"]')['href']
=> "https://example.com/orders/1631/release"
This is using Nokogiri's ability to use CSS accessors. Sometimes it's easier (or the only way) to use XPath, but I prefer CSS because they tend to be more readable.
Nokogiri::Node.at can take a CSS accessor or XPath, and will return the first node matching that pattern. If you need to iterate over all the matches, use search instead, which returns a NodeSet, which you can treat as an array. Nokogiri also supports at_xpath and at_css along with css and xpath for at and search symmetry.

Answer 2

That's a one-liner:
#doc.xpath('//xmlns:link[#rel = "http://example.com/rel/order/transition/release"]').attr('href')