Created
November 8, 2011 09:40
-
-
Save nebuta/1347366 to your computer and use it in GitHub Desktop.
Test algorithm with smaller dictionaries
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| require 'rubygems' | |
| require 'hpricot' | |
| $asciilist = (0x20..0x7e).to_a | [0x09,0x0a,0x0c,0x0d] | |
| def parse(lines) | |
| arr = Array.new(65536) | |
| start = 0 | |
| lines.each{|line| | |
| arr[start,256]=line.chomp.split("\t").map{|e| e.to_f} | |
| start += 256 | |
| } | |
| return arr | |
| end | |
| def isAscii?(b) | |
| $asciilist.include? b | |
| end | |
| def test_judge(out,vector,bytes) | |
| len = bytes.length | |
| score = Hash.new | |
| score[:utf8]=0 | |
| score[:eucjp]=0 | |
| score[:iso]=0 | |
| score[:shiftjis]=0 | |
| i = 0 | |
| ch_count = 0 | |
| while i < len | |
| i += 1 | |
| next if isAscii? bytes[i-1] | |
| while i<len and ch_count <= 100 | |
| break if isAscii? bytes[i] | |
| score.each_key{|key| | |
| score[key] += vector[key][(bytes[i-1])*256+bytes[i]] | |
| } | |
| ch_count += 1 | |
| i += 1 | |
| if ch_count>=100 | |
| sorted = score.to_a.sort{|a,b| a[1] <=> b[1]}.reverse | |
| out.puts score.to_a.sort{|a,b| a[0].to_s <=> b[0].to_s }.map{|a| "%.1f"%a[1]}.join("\t") | |
| return sorted[0][0].to_s, ch_count, i | |
| end | |
| end | |
| end | |
| sorted = score.to_a.sort{|a,b| a[1] <=> b[1]}.reverse | |
| out.puts score.to_a.sort{|a,b| a[0].to_s <=> b[0].to_s }.map{|a| "%.1f"%a[1]}.join("\t") | |
| return sorted[0][0].to_s, ch_count, i | |
| end | |
| def judge(res,answer) | |
| res = | |
| case answer | |
| when /shift[_-]jis/i | |
| res == "shiftjis" | |
| when /euc[_-]jp/i | |
| res == "eucjp" | |
| when /ISO-2022-JP/i | |
| res == "iso" | |
| when /UTF-8/i | |
| res == "utf8" | |
| else | |
| "N/A" | |
| end | |
| if res == true | |
| return "OK" | |
| elsif res == false | |
| return "NG" | |
| else | |
| return "N/A" | |
| end | |
| end | |
| def main | |
| coarse = [1,2,4,8,16,32,64,128] | |
| encode = [:utf8,:eucjp,:iso,:shiftjis] | |
| outs = Hash.new | |
| coarse.each{|c| | |
| f= open("result_c#{c}.txt","w") | |
| outs[c] = f | |
| } | |
| vs = Hash.new | |
| coarse.each{|c| | |
| v = Hash.new | |
| encode.each{|k| | |
| filename = "sub_#{k.to_s}_c#{c}.txt" | |
| puts "Parsing: #{filename}" | |
| v[k] = parse(IO.readlines(filename)) | |
| } | |
| vs[c] = v | |
| } | |
| Dir::chdir("web") | |
| Dir::glob("web*.html").each{|file| | |
| begin | |
| doc = Hpricot(open(file)) | |
| (doc/'script').remove | |
| text = (doc/:body).inner_text | |
| (doc/:head).inner_html =~ /charset=['"]?(.+?)['"]/ | |
| $stderr.puts File.basename(file) | |
| if $1 then | |
| answer = $1 | |
| open(file+".txt","w"){|out| | |
| out.print text | |
| } | |
| coarse.each{|c| | |
| res, ch, index = test_judge(outs[c],vs[c],text.unpack('C*')) | |
| outs[c].puts File.basename(file) | |
| outs[c].puts [res,answer,judge(res,answer),ch,index].join("\t") | |
| outs[c].puts | |
| } | |
| else | |
| puts "skipped (no encoding info in HTML)" | |
| end | |
| rescue => e | |
| puts "Error: " + File.basename(file) | |
| p e | |
| end | |
| } | |
| outs.each{|k,f| | |
| f.close | |
| } | |
| end | |
| main |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment