%# begin user customizations
%
%# mod atom-root directory
% ROOTDIR = '/home/rubys/tmp/atom/test-pub'
%
%# add html5lib to your search path
%# svn checkout http://html5lib.googlecode.com/svn/trunk/ html5lib
% $:.unshift '/home/rubys/svn/html5lib/ruby/lib'
%
%# end user customizations
%
% require "rexml/document.rb"; include REXML
%
% # atom content model
% def text element
% return nil if element.attribute("src")
% type = element.attribute("type")
% case type && type.value
% when "xhtml"
% throw "missing xhtml:div" if element.elements[1].name != "div"
% element.elements[1].to_a.to_s.strip
% when "html"
% element.text.strip
% when "text", nil
% element.to_a.to_s.strip
% when /^text\//i, /\+xml$/i, /\/xml$/i
% element.to_a.to_s.strip
% else
% require 'base64'
% Base64.decode64(element.text.gsub(/\s/,''))
% end
% end
%
% # html5 sanitization
% require 'html5/html5parser'; require 'html5/sanitizer'; include HTML5
% def clean element
% HTMLParser.parseFragment(text(element), :tokenizer=>HTMLSanitizer, :encoding=>'utf-8')
% end
%
% # atom namespace
% ns = {'atom' => 'http://www.w3.org/2005/Atom'}
%
%
% Dir["#{ROOTDIR}/**/*"].sort.reverse.each_with_index do |file, index|
% break if index>12
% next if %w(lock. collection).include? File.basename(file)
% next if File.directory? file
% doc=Document.new(open(file)) rescue next
%
<%= clean(XPath.first(doc,'//atom:title',ns)) %>
<%= clean(XPath.first(doc,'//atom:summary',ns)) %>
%
% end
%