ProcessCSV.rb

# encoding: UTF-8
require "fileutils"
###random sampling

class ProcessCSV
	def initialize (orig)
		#本当はコンストラクタで禁則文字，文字コードの設定もしたい．
		@origFileName = orig
	end

	def sampling(num,samp)
		sampFileName = samp
		remainFileName = "remain_#{sampFileName}"
		File.open(sampFileName,"w").close
		File.open(remainFileName,"w").close

		##元データの行数をカウントする部分
		orig = File.open(@origFileName,"r")
		numOrig = 0
		orig.each_line do |line|
			numOrig += 1 
		end
		orig.close

		begin
			orig = File.open(@origFileName,"r")
			sampled = File.open(sampFileName,"a")
			remain = File.open(remainFileName,"a")
			#元データの行数と同じ長さの配列を生成してシャッフル
			i = 0
			indicies = [*1..numOrig].shuffle

			orig.each_line do |line|
				if i == 0		#header行の検出
					sampled.write(line)
					remain.write(line)
				elsif indicies[i] <= num
				#入力値より小さい値が割り振られた行はsampledへ
					sampled.write(line)
				else
				#さもなくばremainへ
					remain.write(line)
				end
				i += 1
			end
			remain.close
			sampled.close
			orig.close
		rescue => ex
			print ex.message,"\n"
		end
	end

	def grepSeparate(word,samp)
		pattern = Regexp.new(word)

		sampFileName = samp
		remainFileName = "remain_#{sampFileName}"
		File.open(sampFileName,"w").close
		File.open(remainFileName,"w").close

		begin
			headerFlag = 1
			orig = File.open(@origFileName,"r")
			sampled = File.open(sampFileName,"a")
			remain = File.open(remainFileName,"a")
			orig.each_line do |line|
				if headerFlag == 1
				#header行の検出
					sampled.write(line)
					remain.write(line)
					headerFlag = 0
				elsif pattern =~ line
				#マッチしたらsampledへ
					sampled.write(line)
				else
				#さもなくばremainへ
					remain.write(line)
				end
			end
			remain.close
			sampled.close
			orig.close
		rescue => ex
			print ex.message,"\n"
		end

	end

end

begin
	##type commands you wanna execute
	file = ProcessCSV.new("ChangeLog")
	file.sampling(60,"samp.csv")
	# file.grepSeparate("Scott","greped.csv")
rescue => ex
	puts ex.message
end