require "rexml/document"
require 'test/unit'
require 'uri'

# Parse an atom:title, atom:summary, atom:content, or atom:rights element and
# return its contents as text; i.e., unencode that which is encoded, and
# return the rest as is.
#
# This processing model is described in sections 3.1.1 and 4.1.3 of
# the Atom Syndication Format.  Most of the bits relevant to the subject
# at hand can be found at: <http://tinyurl.com/945r5>.

if not REXML::Element.public_instance_methods.include? "getBaseURI"
  class REXML::Element
    def getBaseURI
      if not attribute('xml:base')
        return parent.getBaseURI
      elsif parent
        return URI.join(parent.getBaseURI, attribute('xml:base').value).to_s
      else
        return attribute('xml:base').value || ''
      end
    end 
  end
end


# TestText provides unit test cases for the text function described above.
# 
# One parse method is defined to take care of the small bit of administrivia
# necessary.  The remainder are the actual test cases, and all begin with the
# string "test_".  Each invokes the parse method with a given feed, and makes
# an assertion on what the expected results are.

class TestText < Test::Unit::TestCase

  @@base = "http://example.com/blog/index.atom"

  # The parse method takes care of the following:
  # * parsing the feed as XML
  # * finding the first link element
  # * extracting the href attribute from it
  # * resolving the href

  def parse feed, base=@@base
    doc = REXML::Document.new feed
    ns = {"atom" => "http://www.w3.org/2005/Atom"}
    link = REXML::XPath.first doc, "//atom:link", ns
    doc.add_attribute('xml:base', base)
    base = link.getBaseURI.to_s
    return URI.join(base, link.attribute('href').value).to_s
  end

  def test_no_xml_base_abs
    assert_equal("http://example.org/archives/2005/08/123.html",parse(<<-END))
      <feed xmlns="http://www.w3.org/2005/Atom">
        <entry>
          <link href="http://example.org/archives/2005/08/123.html"/>
        </entry>
      </feed>
    END
  end

  def test_no_xml_base_rel
    assert_equal("http://example.com/archives/2005/08/123.html",parse(<<-END))
      <feed xmlns="http://www.w3.org/2005/Atom">
        <entry>
          <link href="/archives/2005/08/123.html"/>
        </entry>
      </feed>
    END
  end

  def test_xml_base_on_feed_and_entry
    assert_equal("http://example.com/archives/2005/08/123.html",parse(<<-END))
      <feed xmlns="http://www.w3.org/2005/Atom" xml:base="/archives/">
        <entry xml:base="2005/08/">
          <link href="123.html"/>
        </entry>
      </feed>
    END
  end

  def test_xml_base_overridden_on_link
    assert_equal("http://example.com/123.html",parse(<<-END))
      <feed xmlns="http://www.w3.org/2005/Atom" xml:base="/archives/">
        <entry xml:base="2005/08/">
          <link href="123.html" xml:base="http://example.com/"/>
        </entry>
      </feed>
    END
  end

end