#!/usr/bin/ruby -w

autoload :XMLTV, 'xmltv/xmltv'
require 'cgi'



module XMLTV

  class TvgidsGrabber < Grabber
    Cattrans = { 
      'amusement'        => 'Talk',
      'animatie'         => 'Animated',
      'comedy'           => 'Comedy',
      'documentaire'     => 'Documentary',
      'educatief'        => 'Educational',
      'erotiek'          => 'Adult',
      'film'             => 'Movies',
      'muziek'           => 'Art/Music',
      'informatief'      => 'Educational',
      'jeugd'            => 'Children',
      'kunst/cultuur'    => 'Arts/Culture',
      'misdaad'          => 'Crime/Mystery',
      'muziek'           => 'Music',
      'natuur'           => 'Science/Nature',
      'nieuws/actualiteiten' => 'News',
      'overige'          => 'Unknown',
      'religieus'        => 'Religion',
      'serie/soap'       => 'Drama',
      'sport'            => 'Sports',
      'theater'          => 'Arts/Culture',
      'wetenschap'       => 'Science/Nature'
    }
    Roletrans = {
      'regie'             => 'director',
      'acteurs'           => 'actor',
      'presentatie'       => 'presenter',
      'scenario'          => 'writer'
    }

    Titeltrans = {
      'titel aflevering' => 'sub-title',
      'jaar van premiere' => 'date',
      'aflevering' => 'episode-num'
    }

    def grab_detail(href)
      if href[0] == ?/
        href="#{base_url}#{href}"
      end
      STDERR.puts "#{Time.now}: #{href} #{@channelhash.size}" if XmltvOptions.verbose
      program = Hash.new
      details = fetch(href)
      desc = []
      details.at('//table#progDetail').search('//tr//p').each do |p|
        break if p['class'] == 'meerLinks'
        line = p.inner_text.strip
        desc << line unless line.empty?
      end
      program['desc'] = desc.join(' ').to_utf
      details.search('//div#progPropt//tr/th').each do |pg|
        content = pg.at('../td')
        if content['class'] == 'personen'
          rsl = content.at('div').search('.').find_all { |x| x.text? }.map{|x| x.to_s.strip.to_utf}.find_all{|x| ! x.empty?}
        else
          rsl = content.inner_text.strip.to_utf
        end
        program[pg.inner_text.strip.gsub(':','').downcase] = rsl 
      end
  #    PP.pp program, STDERR
      program
    end

    def channel_url(chan_id)
      "#{base_url}//zoeken/?periode=9&station=#{chan_id}"
    end

    def fetch_all_channels
      page = fetch(channel_url(1))
      channels = Hash.new
      page.search('//optgroup')[0..1].each do |og|
        og.search('/option').each do |g|
          channels[g['value']] = g.inner_text
        end
      end
      save_object(channels, channel_list)
      channels
    end

    def clean_cache(cache)
      count = 0
      cache.delete_if do |dt, en|
        rsl = (Date.dutch(en['datum']) < Vandaag)
  #      puts Date.dutch(en['datum']), Vandaag, rsl, '==='
        count += 1 if rsl
        rsl
      end
      count
    end

    def grab_channel(chan_id)
      url = channel_url(chan_id)
      page = fetch(url)
      @channelhash = load_cachefile(chan_id)
  #    get_file(chan_id)
      period = datum = nil
      fetched = 0
      begin
        found = remaining = page.at("//table.overzicht//tr//td/strong").inner_text.to_i
      rescue NoMethodError
        niks =  page.at("//div#resultaten").at("//td").inner_text
        STDERR.puts url, niks
        return
      end
  #    STDERR.puts("#{found} on site, #{@channelhash.size} in cache")
      page.search("//table.overzicht//tr").each do |pg|
        td = pg.at('td')
        next if td.nil? || pg['class'] == 'zoekstring'
        if td['class'] == 'bloktitel'
          period = td.at('h5').inner_text rescue period
          datum = td.at('h4').inner_text rescue datum
          next
        end
  #      puts "Period: #{period} "
        if (tijd = pg.at('/th').inner_text) =~ /\d\d:\d\d/
          det =  pg.at('/td//a')
          href = det['href']
          id = href[/ID=(\d+)/,1]
  #        puts @channelhash[id]
          remaining -= 1
          next if @channelhash[id]
          fetched += 1
          begin
            @channelhash[id] = program = grab_detail(href)
          rescue
            STDERR.puts href, pg, '====='
            raise
          end
          program['title'] = det.inner_text.strip.to_utf
          program['period'] = period
          program['datum'] = datum
          program['tijd'] = tijd
          program['progtip'] = '4/5' if pg['class'] == 'progTip'
        end
      end
      STDERR.puts "Something wrong remaining: #{remaining}" if remaining != 0
      save_object(@channelhash, cachefile(chan_id)) if fetched > 0
      found
    end

    def parse_times(str)
      rsl = nil
#      STDERR.puts str
      md = /(\d+)\s(\w+)\s(\d+),\s(\d+):(\d+)/.match(str)
      if md
        rsl = md.captures.map do |x|
            x =~ /\d/ ? x.to_i : Date::Maanden.index(x.downcase)
        end
        mdstop = /(\d+):(\d+)/.match(str[md.offset(0)[1]..-1])
        if mdstop
          rsl << mdstop.captures.map {|x| x.to_i}
        end
        rsl.flatten!
      end
      raise DateError.new(str) if rsl.index(nil)
      rsl
    end
    def transform(chan_id)
  #    get_file(chan_id)
  #    STDERR.print "#{chan_id} #{@channelhash.size}"
      progdata_array = Array.new
      @channelhash.each_pair do |id, entry|
        begin
          progdata = proghash(entry, chan_id)
          a=entry['datum en tijdstip']
          shift = entry['period'] == 'Nacht' ? Dag : 0
          dag, maand, jaar, startuur, startmin, stopuur, stopmin = parse_times(a)
          next if dag.nil?
          progdata['start'] = start = Time.local(jaar, maand, dag, startuur, startmin) + shift
          
          if stopuur
            stop = Time.local(jaar, maand, dag, stopuur, stopmin) + shift
            if start > stop && start.hour >= 21 && stop.hour <= 7
              stop += Dag
            end
            progdata['stop'] = stop
          end
          date_stats(chan_id, progdata['start'])
          if (b = entry['bijzonderheden'])
            b.downcase.split(',').each do |bijz|
              case bijz
                when /breedbeeld/
                  progdata['video']['aspect'] = '16:9'
                when /zwart/
                  progdata['video']['colour'] = 'no'
               when /teletekst/
                  progdata['subtitles']['type'] = 'teletext'
                when /stereo/
                  progdata['audio']['stereo'] = 'stereo'
              end
            end
          end
          %w{ regie acteurs scenario presentatie }.each do |role|
            if entry[role]
              progdata['credits'][Roletrans[role]] = entry[role]
            end
          end
          progdata['category'] = Cattrans[entry['genre'].downcase] if entry['genre']
          progdata['star-rating']['value'] = entry['progtip'] if entry['progtip']
          Titeltrans.each do |key|
            progdata[Titeltrans[key]] = entry[key] if entry[key]
          end
          progdata_array << progdata
        rescue DateError => exc
          STDERR.puts exc.class, exc.message
          PP.pp(entry, STDERR)
          next
        rescue StandardError => exc
          STDERR.puts exc, exc.message, exc.backtrace
          PP.pp(entry, STDERR)
          raise
        end
      end
      progdata_array
    end
  end
  
end
XMLTV::TvgidsGrabber.new.run