fix irregular HTML by replacing not closed '<' with '&lt;'

fix_irregular_html.rb

module SanitizeHelper
  def fix_irregular_html(html)
    regexp = /<([^<>]*)(<|$)/
    #we need to do this multiple time as regex are overlapping
    while (fixed_html = html.gsub(regexp, "&lt;\\1\\2")) && fixed_html != html
      html = fixed_html
    end
    fixed_html
  end
end
 
  # Test
describe SanitizeHelper do
  it "should fix irregular html" do
    replaces = {
    "Foo"       => "Foo",

    "Foo <3"    => "Foo &lt;3",
    "Foo 3>"    => "Foo 3>",
    "Foo <3>"   => "Foo <3>",

    "Foo <3<4"  => "Foo &lt;3&lt;4",
    "Foo 3><4"  => "Foo 3>&lt;4",
    "Foo <3><4" => "Foo <3>&lt;4",

    "Foo <3<4>"  => "Foo &lt;3<4>",
    "Foo <3<<4>"  => "Foo &lt;3&lt;<4>",
    "Foo 3><4>"  => "Foo 3><4>",
    "Foo <3><4>" => "Foo <3><4>",

    "Foo <3<4<5"  => "Foo &lt;3&lt;4&lt;5",
    "Foo 3><4<5"  => "Foo 3>&lt;4&lt;5",
    "Foo <3><4<5" => "Foo <3>&lt;4&lt;5",

    "Foo <3<4><5"  => "Foo &lt;3<4>&lt;5",
    "Foo 3><4><5"  => "Foo 3><4>&lt;5",
    "Foo <3><4><5" => "Foo <3><4>&lt;5",

    "Foo <3<4<5<6"  => "Foo &lt;3&lt;4&lt;5&lt;6",
    "Foo 3><4<5<6"  => "Foo 3>&lt;4&lt;5&lt;6",
    "Foo <3><4<5<6" => "Foo <3>&lt;4&lt;5&lt;6",

    "Foo <3<4><5<6"  => "Foo &lt;3<4>&lt;5&lt;6",
    "Foo 3><4><5<6"  => "Foo 3><4>&lt;5&lt;6",
    "Foo <3><4><5<6" => "Foo <3><4>&lt;5&lt;6",
    
    "Foo <3<4<5<6<7"  => "Foo &lt;3&lt;4&lt;5&lt;6&lt;7",
    "Foo <3<4<5<6<7<8"  => "Foo &lt;3&lt;4&lt;5&lt;6&lt;7&lt;8",
    "Foo <3<4<5<6<7<8<9"  => "Foo &lt;3&lt;4&lt;5&lt;6&lt;7&lt;8&lt;9",
 
    }.each do |is, should|
      fix_irregular_html(is).should == should
    end

  end
end

See discussion here:
http://stackoverflow.com/questions/3268387/nokogiri-parsing-irregular

Thanks a lot !