require 'fileutils'
require 'date/format'
require 'date'
require 'time'
require 'net/http'

def print_last_x_weekdays(filepath, copypath, x)
  d = Date.today()
  counter = 1
  badcounter = 0
  filename = 'index.rss'
  
  f = File.new(filepath + filename, "w")
  
  f.puts '<?xml version="1.0"?>'
  f.puts '<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">'
  f.puts '    <channel>'
  f.puts '        <title>Unofficial Don and Mike Show RSS Feed</title>'
  f.puts '        <link>http://www.donandmikewebsite.com</link>'
  f.puts '        <description>Unofficial Don and Mike Show rss feed containing daily recap and links to audio recap</description>'
  f.puts '        <language>en-us</language>'
  f.puts "        <pubDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</pubDate>"
  f.puts "        <lastBuildDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</lastBuildDate>"
  
  
  #Go through until we get a correct number of 
  #days or the number of bad hits exceeds limit
  while(counter <= x and badcounter < 5)
    
    #If the date is a weekday
    if(d.wday > 0 and d.wday < 6)
    
    address = "http://www.donandmikewebsite.com/archive/DAILY_LOG-archive-#{d.mday().to_s}-#{d.mon().to_s}-#{d.strftime('%Y')}.shtml"
    #address = "http://www.donandmikewebsite.com/Product/#{d.strftime('%m-%d-%y-%m%d%y')}.html"
      audio = "http://easylink.playstream.com/donandmikewebsite/daily-segment/#{d.strftime('%m-%d-%y')}.wax";
      
      #Hit the URL and see if anything is there
      url = URI.parse(address)
      req = Net::HTTP::Get.new(url.path)
      res = Net::HTTP.start(url.host, url.port) {|http|
        http.request(req)
      }
      
      #200 = HTTP OK
      puts res.code + " - " + address
      if(res.code=="200") 
        counter +=1
        body = res.body
        reg = Regexp.new('(<!-- START MAIN CONTENT HERE -->.+<!-- END MAIN CONTENT HERE -->)', Regexp::MULTILINE)
        body = body[reg]
       
        #body = body[/(<div style="padding:10px 20px 10px 20px">.+<\/div>)/]
        body = body.gsub(/<h1.+<\/h1>/, ' ')
        body = body.gsub(/<X-WYSIWYGv1.0>/, '')
        
        #Replace EM with apostrophe (encoding issue I presume)
        body = body.gsub(/\x19/, "'")
        #get rid of any pesky non-ascii characters
        body = body.gsub(/[^\x20-\x7F]/, ' ') 

        #Need to see if we have a file that corresponds to this day,
        #if we do, we need to compare it to the content we retrieved
        #to see if anything has changed.  If so, we need to update the 
        #link to be different so that feed readers will notice the change
        does_file_exist = File.exist?(filepath + d.strftime('%m-%d-%y'))
        
        if(does_file_exist)
          #Fetch our existing file contents
          existing_file = File.new(filepath + d.strftime('%m-%d-%y'))
          existing_contents = ""
          while (existing_file.gets != nil)
            existing_contents += $_
          end
          existing_contents = existing_contents.chomp
                    
          #Compare to what we fetched from the web
          #If different update our file on disk
          #along with the address variable (this is what the 
          #feedreaders will key off of to recognize a change    
          if(!body.eql?(existing_contents))
            puts "File #{filepath + d.strftime('%m-%d-%y')} does not match content from web, overwriting"
            existing_file = File.new(filepath + d.strftime('%m-%d-%y'), 'w')
            existing_file.puts body
            address = address + '#' + Time.now.strftime('%H%M%S')
          end
          
          existing_file.close()
        else
          existing_file = File.new(filepath + d.strftime('%m-%d-%y'), 'w')
          existing_file.puts body
        end
        
        f.puts '  <item>'
        f.puts "      <title>Show Recap for #{d.strftime('%A, %B %d, %Y')}</title>"
        f.puts "      <link>#{address}</link>"
        f.puts '      <description>'
        
        #uses regex on body variable to remove extraneous HTML content
        f.puts "          <![CDATA[#{body}" 
        f.puts '          <br><hr><br>'
        
        #adding in some show related links (Listen Live, Buy Swag, 
        #Eat at Hank's Look-Around Cafe) to the end of each item 
        f.puts '          <a href="http://tuner1.dc1.sonixtream.com/solon/media/tuner/Tuner?aff=1067wjfk&useSame=true&type=NN&activex=no">Listen Live 3-7 EST</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
        f.puts '          <a href="http://phobos.apple.com/WebObjects/MZStore.woa/wa/viewPodcast?id=114454618">Don and Mike Podcasts on iTunes</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
        f.puts '          <a href="http://www.logoradio.com/Department/Shop-Don-Mike.cfm">Buy Don and Mike Swag</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
        f.puts '          <a href="http://www.omearasrestaurant.com/home.htm">Eat at O\'Meara\'s</a>'
        f.puts '          ]]>'
        
        f.puts '      </description>'
        f.puts '      <dc:creator>Don and Mike</dc:creator>'
        f.puts "      <pubDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</pubDate>"
        f.puts "      <guid isPermaLink=\"false\">#{address}</guid>"
        
        
        #See if we have an audio feed for this day       
        audiourl = URI.parse(audio)
        audiores = Net::HTTP.start(audiourl.host, audiourl.port) {|http|
          http.request_head(audio)
        }     
        
        #Add RSS enclosure if we found an audio feed
        #if(audiores.code=="200")
        #  f.puts "      <enclosure url=\"#{audio}\" length=\"#{audiores['content-length']}\" type=\"#{audiores['content-type']}\" />"
        #end
        
        f.puts '  </item>'
        
      elsif
        badcounter+=1
      end          
    end
    
    #move back one day
    d = d-1
  end
  
  f.puts '    </channel>'
  f.puts '</rss>'
  
  f.close()
  
  #Copy our completed file to its final location
  FileUtils.copy(filepath + filename, copypath + filename)
end

#call our method with command line arguments
#ARGV[0]: file path to use for file construction
#ARGV[1]: file path to copy finished file to
#ARGV[2]: number of days back to fetch into file
print_last_x_weekdays(ARGV[0], ARGV[1], ARGV[2].to_i())