Comments on: How to test a XML builder view

By: Mislav

Mislav — Wed, 31 Oct 2007 08:32:00 +0000

Great article. How about making your life easier with a helper?

def hpricot

  yield Hpricot.XML(response.body.to_s)

end

  it "should have the high average as values of the first graph" do

    render "/averages/index.xml.builder"

    hpricot do |doc|

      (doc/:graph/:value).first.inner_html.should == "74.5"

    end

  end

Also, in places where you use @inner_html@ you should maybe use @inner_text@ if you want to test text content of a specific element, not markup.

By: Mislav

Mislav — Wed, 31 Oct 2007 08:32:00 +0000

I hate Textile, it always mangles what I wanted to say

By: Josh Knowles

Josh Knowles — Wed, 31 Oct 2007 08:32:00 +0000

Cool use of hpricot, but why not just use the built in have_tag matcher? There is nothing unique about the xml that couldn’t be tested the same as html.

By: Matt Aimonetti

Matt Aimonetti — Wed, 31 Oct 2007 08:32:00 +0000

@mislav you are replacing doc = Hpricot.XML(response.body.to_s) by a bloc using a helper. I like the idea of using a helper, but why not doing something like:

def parse_xml_response
Hpricot.XML(response.body.to_s)
end

then use it as:

it "should have the high average as values of the first graph" do
render "/averages/index.xml.builder"
doc = parse_xml_response
(doc/:graph/:value).first.inner_html.should == "74.5"
end

Didn’t try it but looks better than a bloc to me. What do you think? Did you have a specific reason to use a bloc?

Good point about inner_text!

By: Matt Aimonetti

Matt Aimonetti — Wed, 31 Oct 2007 08:32:00 +0000

stupid textle!

Here is the snippet of code which got messed up:

it "should have the high average as values of the first graph" do
render "/averages/index.xml.builder"
doc = parse_xml_response
(doc/:graph/:value).first.inner_html.should == "74.5"
end

@joshknowles

it "should render the months in the series" do
render "/averages/index.xml.builder"
response.should have_tag("value", ‘February’)
response.should have_tag("value", ‘November’)
doc = Hpricot.XML(response.body.to_s)
(doc/:value).first.inner_html.should == ‘February’
(doc/:value)[1].inner_html.should == ‘November’
end

works well but I can’t find a way to check a node’s attribute:

it "should set the xid attributes for the series" do
render "/averages/index.xml.builder"
doc = Hpricot.XML(response.body.to_s)
(doc/:value).first["xid"].should == ‘0′
(doc/:value).last["xid"].should == ‘1′
end

-Matt

By: Matt Aimonetti

Matt Aimonetti — Wed, 31 Oct 2007 08:32:00 +0000

after chatting offline with Josh, I found a solution:

it "should set the xid attributes for the series" do
render "/averages/index.xml.builder"
response.should have_tag("value[xid=0]:first-child")
response.should have_tag("value[xid=1]:last-child")
# same as:
doc = Hpricot.XML(response.body.to_s)
(doc/:value).first["xid"].should == ‘0′
(doc/:value).last["xid"].should == ‘1′
end

have_tag wraps assert_select, check out all the cool options you get with assert_select:

http://cheat.errtheblog.com/s/assert_select/

Thanks Josh!

By: josh

josh — Wed, 31 Oct 2007 08:32:00 +0000

Just FYI, your xml feed refreshes several times a day. All articles are marked as new. NetNewsWire tells me I have tons of new stuff to read, but it’s a lie!

By: Matt Aimonetti

Matt Aimonetti — Wed, 31 Oct 2007 08:32:00 +0000

@josh make sure you are using the feedburner feed: http://feeds.feedburner.com/railsontherun for some reason the atom feed provided by mephisto resets itself.