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

#Optimize to get rid of redundant code (3 separate copy pastes for 3 addresses)
#Update to use Head call to check for existence of page.  Cheaper than full get
#Maybe should just parse his main pages to find URL's of items....
#Add links to bottom of items referencing his various sites

def fetch_url(address)
  body = ""
  
  #Hit the URL and see if anything is there
  url = URI.parse(address)
  
  req = Net::HTTP::Head.new(url.path)
  res = Net::HTTP.start(url.host, url.port) {|http|
    http.request(req)
  }
    
  #puts res.http_version
  #puts res.message
  
  if(res.code == "200")
    #200 = HTTP OK
    puts res.code + " - " + address
  
    #Exists so go ahead and do a full get
    req = Net::HTTP::Get.new(url.path)
    res = Net::HTTP.start(url.host, url.port) {|http|
      http.request(req)
    }
    
    body = res.body

  end 

  return body  
  
end

def create_item(address, body, regexp, existing_filename, d, title)
      #puts address
      reg = Regexp.new('<hr width="98%" size="2" color="#336699" noshade>(.+)', Regexp::MULTILINE)
      regtitle = (body.scan(/(<TITLE>(.+)<\/TITLE>)/im)[0])[1]
      body2 = body[reg]
      
      if body2 == nil then
        puts "Body was nil"
        #puts address
        reg = Regexp.new('<HR width="98%" color=#336699 noShade SIZE=2>(.+)', Regexp::MULTILINE)
        regtitle = (body.scan(/(<TITLE>(.+)<\/TITLE>)/im)[0])[1]
        body = body[reg]
        puts "body action" 
      else
        body = body2
      end 
      
      body = body.gsub(/<hr width="98%" size="2" color="#336699" noshade>/, "")
      body = body.gsub(/<HR width="98%" color=#336699 noShade SIZE=2>/, "")

        #Replace EM with apostrophe (encoding issue I presume)
      body = body.gsub(/\x19/, "'")
      body = body.gsub('’', "'")
      
      #Get rid of weird MS specific content
      body = body.gsub(/<o:p>/i, '<p>')
      body = body.gsub(/<\/o:p>/i, '</p>')
      body = body.gsub(/style=".+?"/im, '')

      
      #get rid of any pesky non-ascii characters
      body = body.gsub(/[^\x20-\x7F]/, ' ') 
      #body = body.gsub(/<div class="storyhed">(.+?)<\/div>/, '<b>\1</b>')

      #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
      #existing_filename = filepath + 'ronpaultst_' + d.strftime('%m-%d-%y')
      does_file_exist = File.exist?(existing_filename)
      
      if(does_file_exist)
        #Fetch our existing file contents
        existing_file = File.new(existing_filename)
        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 #{existing_filename} does not match content from web, overwriting"
          existing_file = File.new(existing_filename, 'w')
          existing_file.puts body
          address = address + '#' + Time.now.strftime('%H%M%S')
        end
        
        existing_file.close()
      else
        existing_file = File.new(existing_filename, 'w')
        existing_file.puts body
      end # end file exist
      
      item =  '  <item>'
      item += "      <title>#{title}: #{regtitle}</title>"
      item += "      <link>#{address}</link>"
      item += '      <description>'
      item += "          <![CDATA[#{body}"
      item += '          <br><hr><br>'
        
      #adding in some related links (Home Page, House of Reps webpage
      #Campaign webpage and donation webpage) to the end of each item 
      item += '          <a href="http://www.ronpaul2008.com/">Ron Paul 2008 Webpage</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
      item += '          <a href="https://www.ronpaul2008.com/forms/contribute.cfm">Donate to the Ron Paul 2008 Campaign</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
      item += '          <a href="http://www.ronpaul.org/">Ron Paul Personal Webpage</a>&nbsp;&nbsp;|&nbsp;&nbsp;'
      item += '          <a href="http://www.house.gov/paul/">Ron Paul House of Representatives Webpage</a>'      
      item += '        ]]>'
      item += '      </description>'
      item += '      <dc:creator>Ron Paul</dc:creator>'
      item += "      <pubDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</pubDate>"
      item += "      <guid isPermaLink=\"false\">#{address}</guid>"
      
      item += '  </item>'

end 

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 Ron Paul RSS Feed</title>'
  f.puts '        <link>http://www.house.gov/paul/</link>'
  f.puts '        <description>Unofficial Ron Paul RSS feed containing his writings and site updates</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 ) 
    
    #We have 3 places to check
    #Texas Straight Talk
    address = "http://www.house.gov/paul/tst/tst#{d.strftime('%Y')}/tst#{d.strftime('%m%d%y')}.htm"

    #Speeches And Statements
    address2 = "http://www.house.gov/paul/congrec/congrec#{d.strftime('%Y')}/cr#{d.strftime('%m%d%y')}.htm"
    
    #Press Releases
    address3 = "http://www.house.gov/paul/press/press#{d.strftime('%Y')}/pr#{d.strftime('%m%d%y')}.htm"

    #puts address
    #puts address2
    #puts address3
            
    body = fetch_url(address)
    
    if(body.length > 0) 
      counter +=1

      existing_filename = filepath + 'ronpaultst_' + d.strftime('%m-%d-%y')
      regexp = '<hr width="98%" size="2" color="#336699" noshade>(.+)'

      item = create_item(address, body, regexp, existing_filename, d, "Texas Straight Talk")
      
      f.puts item
    end  # end res.code test

    body = fetch_url(address2)

    if(body.length > 0) 
      counter +=1

      existing_filename = filepath + 'ronpaulcongrec_' + d.strftime('%m-%d-%y')
      regexp = '<hr width="98%" size="2" color="#336699" noshade>(.+)'

      item = create_item(address2, body, regexp, existing_filename, d, "Speeches And Statements")
      
      f.puts item
    end  # end res.code test

    body = fetch_url(address3)

    if(body.length > 0) 
      counter +=1

      existing_filename = filepath + 'ronpaulpress_' + d.strftime('%m-%d-%y')
      regexp = '<hr width="98%" size="2" color="#336699" noshade>(.+)'

      item = create_item(address3, body, regexp, existing_filename, d, "Press Releases")
      
      f.puts item
    end  # end res.code test


    #Hit the URL and see if anything is there
    #url = URI.parse(address2)
    #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 + " - " + address2
=begin    
    #if(res.code=="200") 
    body = fetch_url(address2)
    
    if(body.length > 0) 

      counter +=1
      body = res.body
      #puts body
      reg = Regexp.new('<hr width="98%" size="2" color="#336699" noshade>(.+)', Regexp::MULTILINE)
      regtitle = (body.scan(/(<title>(.+)<\/title>)/)[0])[1]
      body = body[reg]

      body = body.gsub(/<hr width="98%" size="2" color="#336699" noshade>/, "")

        #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]/, ' ') 
      #body = body.gsub(/<div class="storyhed">(.+?)<\/div>/, '<b>\1</b>')

      #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
      existing_filename = filepath + 'ronpaulcongrec_' + d.strftime('%m-%d-%y')
      does_file_exist = File.exist?(existing_filename)
      
      if(does_file_exist)
        #Fetch our existing file contents
        existing_file = File.new(existing_filename)
        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 #{existing_filename} does not match content from web, overwriting"
          existing_file = File.new(existing_filename, 'w')
          existing_file.puts body
          address = address + '#' + Time.now.strftime('%H%M%S')
        end
        
        existing_file.close()
      else
        existing_file = File.new(existing_filename, 'w')
        existing_file.puts body
      end # end file exist
      
      item =  '  <item>'
      item += "      <title>Speeches And Statements: #{regtitle}</title>"
      item += "      <link>#{address}</link>"
      item += '      <description>'
      item += "          <![CDATA[#{body}]]>"
      
      item += '      </description>'
      item += '      <dc:creator>Ron Paul</dc:creator>'
      item += "      <pubDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</pubDate>"
      item += "      <guid isPermaLink=\"false\">#{address}</guid>"
      
      item += '  </item>'
      
      f.puts item
    elsif
      badcounter+=1
    end  # end res.code test

#Hit the URL and see if anything is there
    #url = URI.parse(address3)
    #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 + " - " + address3
    
    #if(res.code=="200") 
    body = fetch_url(address3)
    
    if(body.length > 0) 

      counter +=1
      body = res.body
      #puts body
      reg = Regexp.new('<hr width="98%" size="2" color="#336699" noshade>(.+)', Regexp::MULTILINE)
      regtitle = Regexp.new('(<HEAD><TITLE>(.+)<\/TITLE><\/HEAD>)', Regexp::MULTILINE)
      regtitle = (body.scan(regtitle)[0])[1]
      body = body[reg]

      body = body.gsub(/<hr width="98%" size="2" color="#336699" noshade>/, "")

        #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]/, ' ') 
      #body = body.gsub(/<div class="storyhed">(.+?)<\/div>/, '<b>\1</b>')

      #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
      existing_filename = filepath + 'ronpaulpress_' + d.strftime('%m-%d-%y')
      does_file_exist = File.exist?(existing_filename)
      
      if(does_file_exist)
        #Fetch our existing file contents
        existing_file = File.new(existing_filename)
        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 #{existing_filename} does not match content from web, overwriting"
          existing_file = File.new(existing_filename, 'w')
          existing_file.puts body
          address = address + '#' + Time.now.strftime('%H%M%S')
        end
        
        existing_file.close()
      else
        existing_file = File.new(existing_filename, 'w')
        existing_file.puts body
      end # end file exist
      
      item =  '  <item>'
      item += "      <title>Press Releases: #{regtitle}</title>"
      item += "      <link>#{address}</link>"
      item += '      <description>'
      item += "          <![CDATA[#{body}]]>"
      
      item += '      </description>'
      item += '      <dc:creator>Ron Paul</dc:creator>'
      item += "      <pubDate>#{d.strftime('%a, %d %b %Y 23:00:00 CST')}</pubDate>"
      item += "      <guid isPermaLink=\"false\">#{address}</guid>"
      
      item += '  </item>'
      
      f.puts item
    elsif
      badcounter+=1
    end  # end res.code test
=end    
    #move back one day
    d = d-1
end #end while loop
  
  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()) 
#puts fetch_url("http://www.house.gov/paul/tst/tst2007/tst041607.htm")
#print_last_x_weekdays("c:\\test\\", "c:\\test\\test2\\", 1)