0_rapbd_cleaner.rb

contents = File.read('raw_rapbd.txt')
EXCEPTIONS = [/^Urusan Pemerintahan :/, /^Organisasi :/, /^Jumlah/, /^$/, /^Surplus\/\(Defisit\)/, /^Pembiayaan Netto/]

contents.each_line do |l|
	l = l.strip
	unless /^\d\.\d\d/.match(l)
		if EXCEPTIONS.none? {|e| e.match(l)}
			print(" #{l}")
		else 
			print("\n#{l}")
		end
	else
		print("\n#{l}")
	end
end

1_rapbd_to_csv.rb

contents = File.read('raw_rapbd_2.txt')

puts "Urusan Pemerintahan|Organisasi|Kode Rekening|Uraian|Jumlah"
contents.each_line do |l|
	if l.strip.length == 0 then next end
	if (m = /^Urusan Pemerintahan : (.+)/.match(l)) 
		@urusan_pemerintah = m[1]
	elsif (m = /^Organisasi : (.+)/.match(l))
		@organisasi = m[1]
	else
		l.strip!
		if (m = /(\d\.\d{2} \d{3} \d{2} \d{3} (\d )?(\d )?(\d )?)(.+)/.match(l)) then
			kode_rekening = m[1].strip
			remaining = m[-1]
			if (m = /(.+?)([\.\d]+\d{3})$/.match(remaining)) then
				uraian = m[1].strip
				jumlah = m[2].strip
				puts "#{@urusan_pemerintah}|#{@organisasi}|#{kode_rekening}|#{uraian}|#{jumlah}"
			else
				puts "#{@urusan_pemerintah}|#{@organisasi}|#{kode_rekening}|#{remaining}|"
			end
		end
	end
end

2_rapbd_flattener.rb

def separate_with_dot(number)
  number.to_s.chars.to_a.reverse.each_slice(3).map(&:join).join(".").reverse
end

contents = File.read('raw_rapbd_3.txt')

puts "Urusan Pemerintahan|Organisasi|Kode Rekening|Uraian|Belanja Barang dan Jasa|Belanja Modal|Belanja Pegawai|Total"
first_line = true
@last_jumlah = @total = @total_child = 0
@belanja_barang_dan_jasa = @belanja_modal = @belanja_pegawai = 0

contents.each_line do |l|
	urusan_pemerintahan, organisasi, kode_rekening, uraian, jumlah = l.split("|")
	jumlah.gsub!(/\./, '')
	jumlah = jumlah.to_i

	if first_line then first_line = false
	else
		last_kode_rekening = kode_rekening[15..-1].strip

		case uraian
		when 'BELANJA BARANG DAN JASA'
			@belanja_barang_dan_jasa += jumlah
		when 'BELANJA PEGAWAI'
			@belanja_pegawai += jumlah
		when 'BELANJA MODAL'
			@belanja_modal += jumlah
		end

		if last_kode_rekening.length <= 1 and @prev_kode_rekening
			puts "#{@prev_urusan_pemerintahan}|#{@prev_organisasi}|#{@prev_kode_rekening}|#{@prev_uraian}|#{separate_with_dot(@belanja_barang_dan_jasa)}|#{separate_with_dot(@belanja_modal)}|#{separate_with_dot(@belanja_pegawai)}|#{separate_with_dot(@total)}"
			@total = @last_jumlah = @belanja_modal = @belanja_pegawai = @belanja_barang_dan_jasa = 0
		elsif last_kode_rekening.length == 3
			if @last_jumlah != 0
				if @last_jumlah != @total_child
					puts "Error! #{@last_jumlah} != #{@total_child}"
				end
				@last_jumlah = 0
			end
			@total += jumlah
			@last_jumlah = jumlah
			@total_child = 0
		elsif last_kode_rekening.length == 5
			@total_child += jumlah
		end

		if last_kode_rekening.length <= 1
			@prev_urusan_pemerintahan, @prev_organisasi, @prev_kode_rekening, @prev_uraian, @prev_jumlah = urusan_pemerintahan, organisasi, kode_rekening, uraian, jumlah
		end

		# Original Algorithm by Arief:
		#
		# if (last_kode_rekening.length() <= 1) {
		# 	print(kode_rekening + uraian + total);
		# 	total = 0;
		# }

		# if (last_kode_rekening.length() == 3) {\

		# 	if (last_jumlah != 0) {
		# 		if (last_jumlah != total_child)
		# 			gak sama

		# 		last_jumlah = 0;
		# 	}

		# 	total += jumlah;
		# 	last_jumlah = jumlah;
		# 	total_child = 0;
		# }

		# if (last_kode_rekening.length() == 5) {
		# 	total_child += jumlah;
		# }
	end
end
if @prev_kode_rekening
	puts "#{@prev_urusan_pemerintahan}|#{@prev_organisasi}|#{@prev_kode_rekening}|#{@prev_uraian}|#{separate_with_dot(@belanja_barang_dan_jasa)}|#{separate_with_dot(@belanja_modal)}|#{separate_with_dot(@belanja_pegawai)}|#{separate_with_dot(@total)}"
end

3_rapbd_differ.rb

def separate_with_comma(number)
  res = number.abs.to_s.chars.to_a.reverse.each_slice(3).map(&:join).join(",").reverse
  if number < 0 then return "-#{res}"
  else return res end
end

dprd_version = File.read('join_apbd.csv')

kode_kegiatan_to_pagu = Hash.new
nama_kegiatan_to_pagu = Hash.new
kode_nama_kegiatan_to_pagu = Hash.new

# kode_kegiatan_to_line = Hash.new

# dprd_version.each_line do |l|
# 	_, kode_skpd, nama_skpd, komisi, kode_kegiatan, nama_kegiatan, pagu, tambah, kurang, _, _, _, hasil_pembahasan = /(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|/.match(l).to_a
# 	unless kode_kegiatan.strip.empty?
# 		k = [kode_skpd, nama_skpd, komisi, kode_kegiatan].join("_")
# 		if kode_kegiatan_to_line.key?(k) 
# 			if kode_kegiatan_to_line[k]['name'] == nama_kegiatan then print("* ") end
# 			puts "Duplicate keys: #{kode_skpd}|#{nama_skpd}|#{komisi}|#{kode_kegiatan} -> #{kode_kegiatan_to_line[k]['name']} (#{kode_kegiatan_to_line[k]['jumlah']}) vs #{nama_kegiatan} (#{hasil_pembahasan})"
# 		else	
# 			kode_kegiatan_to_line[k] = Hash.new
# 			kode_kegiatan_to_line[k]['name'] = nama_kegiatan
# 			kode_kegiatan_to_line[k]['jumlah'] = hasil_pembahasan
# 		end
# 	end
# end

dprd_version.each_line do |l|
	_, kode_skpd, nama_skpd, komisi, kode_kegiatan, nama_kegiatan, pagu, tambah, kurang, _, _, _, hasil_pembahasan = /(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|/.match(l).to_a
	unless (kode_kegiatan.strip.empty? or nama_kegiatan.strip.empty?)
		if kode_nama_kegiatan_to_pagu.key?([kode_skpd, kode_kegiatan, nama_kegiatan].join("_")) then puts "Duplicate kode & nama: #{[kode_skpd, kode_kegiatan, nama_kegiatan].join("_")} -> #{pagu} vs #{kode_nama_kegiatan_to_pagu[[kode_skpd, kode_kegiatan, nama_kegiatan].join("_")]}" end
		kode_nama_kegiatan_to_pagu[[kode_skpd, kode_kegiatan, nama_kegiatan].join("_")] = pagu
	end
	unless kode_kegiatan.strip.empty?
		if kode_kegiatan_to_pagu.key?(kode_kegiatan) then puts "Duplicate kode: #{kode_kegiatan}" end
		kode_kegiatan_to_pagu[kode_kegiatan] = pagu
	end
	unless nama_kegiatan.strip.empty?
		if nama_kegiatan_to_pagu.key?(nama_kegiatan) then puts "Duplicate nama: #{nama_kegiatan}" end
		nama_kegiatan_to_pagu[nama_kegiatan] = pagu	
	end
end

# pemda_version = File.read('raw_rapbd_4.txt')
# puts("Urusan Pemerintahan|Organisasi|Kode Rekening|Uraian|Jumlah|Pagu_DPRD|Flag")
# first_line = true
# pemda_version.each_line do |l|
# 	if first_line then 
# 		first_line = false 
# 	else
# 		_, urusan_pemerintahan, organisasi, kode_rekening, uraian, jumlah = /(.*)\|(.*)\|(.*)\|(.*)\|(.*)/.match(l).to_a
# 		# 1.01 001 01 121 -> 1.01.01.001.121
# 		translatted_kode_rekening = kode_rekening[0..3] + "." + kode_rekening[9..10] + "." + kode_rekening[5..7] + "." + kode_rekening[12..-1]
# 		pagu = nil
# 		pagu ||= kode_nama_kegiatan_to_pagu[[translatted_kode_rekening, uraian].join("_")]
# 		pagu ||= kode_kegiatan_to_pagu[translatted_kode_rekening]
# 		# pagu ||= nama_kegiatan_to_pagu[uraian]

# 		jumlah.gsub!(/\./, ',')
# 		if pagu
# 			if pagu == jumlah
# 				flag = 0
# 			else
# 				flag = separate_with_comma(jumlah.gsub(/\,/, '').to_i - pagu.gsub(/\,/, '').to_i)
# 			end
# 		else
# 			flag = "Doesn't exist in DPRD"
# 		end
# 		print "#{urusan_pemerintahan}|#{organisasi}|#{kode_rekening}|#{uraian}|#{jumlah}|#{pagu}|#{flag}\n"
# 	end
# end

4_rapbd_differ_reverse.rb

# This code needs refactoring

require 'fuzzy_match'
require 'set'
require 'amatch'

def separate_with_comma(number)
  if number.empty? then return 0 end
  
  number = number.to_i
  res = number.abs.to_s.chars.to_a.reverse.each_slice(3).map(&:join).join(",").reverse
  if number < 0 then return "-#{res}"
  else return res end
end
 
pemda_version = File.read('mata.anggaran.csv')
 
kode_rekening_program_to_total = Hash.new
kode_rekening_to_total = Hash.new
program_to_total = Hash.new

first_line = true
all_programs = Set.new

pemda_version.each_line do |l|
  if first_line 
    first_line = false
  else
    _, urusan_pemerintahan, organisasi, kategori, kode_rekening, program, _, _, _, total = /(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)/.match(l).to_a
    total = total.to_s.gsub(/\,/, "").gsub(/\./, "")

    # puts kode_rekening, program
 
    # if kode_rekening_to_total.key? kode_rekening then puts "Duplicate kode_rekening: #{kode_rekening}" end
    # if kode_rekening_program_to_total.key? [kode_rekening, program].join("_") then puts "Duplicate kode_rekening_program: #{kode_rekening}_#{program}" end
    # if program_to_total.key? program then puts "Duplicate program: #{program}" end

    kode_rekening_program_to_total[[kode_rekening, program.downcase].join("_")] ||= []
    kode_rekening_to_total[kode_rekening] ||= []
    program_to_total[program.downcase] ||= []

    kode_rekening_program_to_total[[kode_rekening, program.downcase].join("_")] << total
    kode_rekening_to_total[kode_rekening] << total
    program_to_total[program.downcase] << total

    all_programs.add(program.downcase)
  end
end
# fm = FuzzyMatch.new(all_programs.to_a)

dprd_version = File.read('join_apbd.csv')
first_line = true

puts "no|kode_skpd|nama_skpd|komisi|kode_kegiatan|nama_kegiatan|pagu|tambah|kurang|hasil_pembahasan|hasil_pemprov_version|delta = hasil_pembahasan - hasil_pemprov_version|flag|kegiatan_versi_pemprov_yang_mirip|% kemiripan|hasil_pemprov_version_yang_mirip|delta = hasil_pemprov_version_yang_mirip - hasil_pemprov_version"

number = 1
dprd_version.each_line do |l|
  if first_line 
    first_line = false
  else
    _, kode_skpd, nama_skpd, komisi, kode_kegiatan, nama_kegiatan, pagu, tambah, kurang, hasil_pembahasan = /(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|(.*)\|/.match(l).to_a
    # 1.01.01.001.121 -> 1.01 001 01 121
    # puts kode_kegiatan
    unless kode_kegiatan.empty?
      translatted_kode_kegiatan = kode_kegiatan[0..3] + " " + kode_kegiatan[8..10] + " " + kode_kegiatan[5..6] + " " + kode_kegiatan[12..-1]
    else
      translatted_kode_kegiatan = ""
    end
      
    # puts translatted_kode_kegiatan, nama_kegiatan
    flag = "Not found in Pemprov version based on kode & nama kegiatan"

    total = kode_rekening_program_to_total[[translatted_kode_kegiatan, nama_kegiatan.downcase].join("_")]
    if total then flag = "Perfect match" end

    if total.nil? then 
      total = kode_rekening_to_total[translatted_kode_kegiatan]
      if total then flag = "Matched by kode kegiatan" end
    end

    if total.nil? then 
      total = program_to_total[nama_kegiatan.downcase]
      if total then flag = "Matched by nama kegiatan" end
    end

    pagu = separate_with_comma(pagu.gsub(/\,/, "").gsub(/\./, ""))
    tambah = separate_with_comma(tambah.gsub(/\,/, "").gsub(/\./, ""))
    kurang = separate_with_comma(kurang.gsub(/\,/, "").gsub(/\./, ""))
    original_hasil_pembahasan = hasil_pembahasan.gsub(/\,/, "").gsub(/\./, "").to_i
    hasil_pembahasan = separate_with_comma(hasil_pembahasan.gsub(/\,/, "").gsub(/\./, ""))

    recommended_nama_kegiatan = nil
    recommended_total = nil
    recommended_score = nil
    
    if total.nil?
      # recommended_nama_kegiatan = fm.find(nama_kegiatan)
      recommended_nama_kegiatan = nil
      recommended_score = -1
      all_programs.each do |p|
        score = p.levenshtein_similar(nama_kegiatan.downcase)
        if recommended_score < score
          recommended_score = score
          recommended_nama_kegiatan = p
        end
      end

      if program_to_total[recommended_nama_kegiatan].length == 1
        recommended_delta = separate_with_comma((original_hasil_pembahasan - program_to_total[recommended_nama_kegiatan][0].to_i).to_s)
      end

      if program_to_total[recommended_nama_kegiatan].length > 3
        recommended_total = "Too many variants"
      else
        t = program_to_total[recommended_nama_kegiatan].map{|e| separate_with_comma(e)}
        recommended_total = t.join("/")
      end
    end

    if total.nil?
      puts "#{number}|#{kode_skpd}|#{nama_skpd}|#{komisi}|#{kode_kegiatan}|#{nama_kegiatan}|#{pagu}|#{tambah}|#{kurang}|#{hasil_pembahasan}||#{hasil_pembahasan}|#{flag}|#{recommended_nama_kegiatan}|#{recommended_score}|#{recommended_total}|#{recommended_delta}"
    else
      if total.length == 1 
        delta = separate_with_comma((original_hasil_pembahasan - total[0].to_i).to_s)
      end
      if total.length > 3
        total = "Too many variants"
      else
        total = total.map{|e| separate_with_comma(e)}
        total = total.join("/")
      end
      puts "#{number}|#{kode_skpd}|#{nama_skpd}|#{komisi}|#{kode_kegiatan}|#{nama_kegiatan}|#{pagu}|#{tambah}|#{kurang}|#{hasil_pembahasan}|#{total}|#{delta}|#{flag}|#{recommended_nama_kegiatan}|#{recommended_score}|#{recommended_total}|#{recommended_delta}"
    end

    number += 1
  end
end

file raw_rapbd.txt, raw_rapbd_2.txt dapat dari mana ya mas. pengen nyobain

raw_rapbd_2.txt dari output 0_rapbd_cleaner.rb. raw_rapbd_3.txt dari output 1_rapbd_to_csv.rb.
raw_rapbd.txt dari copy paste PDF di sini www.jakarta.go.id/v2/news/2015/03/rapbd-pemprov-dki-dan-dprd#.VPLDDCwxs7w -> https://drive.google.com/file/d/0B9EJd8nA4QE8R0s1ellncWtPcmc/view?usp=sharing

terimakasih mas.. tak cobai melihat dengan cara yang berbeda