Added a tool to investicate statistics of csa kifu files
shogi-server/trunk/utils/csa-filter.rb (diff) shogi-server/trunk/utils/statistics.rb (diff)| @@ -28,6 +28,7 @@ | ||
| 28 | 28 | require 'time' |
| 29 | 29 | require 'pathname' |
| 30 | 30 | require 'getoptlong' |
| 31 | +require 'nkf' | |
| 31 | 32 | |
| 32 | 33 | class CsaFileReader |
| 33 | 34 | WIN_MARK = "win" |
| @@ -35,9 +36,11 @@ | ||
| 35 | 36 | DRAW_MARK = "draw" |
| 36 | 37 | |
| 37 | 38 | attr_reader :file_name |
| 39 | + attr_reader :str | |
| 38 | 40 | attr_reader :black_name, :white_name |
| 39 | 41 | attr_reader :black_id, :white_id |
| 40 | 42 | attr_reader :winner, :loser |
| 43 | + attr_reader :state | |
| 41 | 44 | attr_reader :start_time, :end_time |
| 42 | 45 | |
| 43 | 46 | def initialize(file_name) |
| @@ -46,13 +49,14 @@ | ||
| 46 | 49 | end |
| 47 | 50 | |
| 48 | 51 | def grep |
| 49 | - str = File.open(@file_name).read | |
| 52 | + @str = File.open(@file_name, "r:Shift_JIS:EUC-JP").read | |
| 50 | 53 | |
| 51 | - if /^N\+(.*)$/ =~ str then @black_name = $1.strip end | |
| 52 | - if /^N\-(.*)$/ =~ str then @white_name = $1.strip end | |
| 53 | - if /^'summary:(.*)$/ =~ str | |
| 54 | - state, p1, p2 = $1.split(":").map {|a| a.strip} | |
| 55 | - return if state == "abnormal" | |
| 54 | + | |
| 55 | + if /^N\+(.*)$/ =~ @str then @black_name = $1.strip end | |
| 56 | + if /^N\-(.*)$/ =~ @str then @white_name = $1.strip end | |
| 57 | + if /^'summary:(.*)$/ =~ @str | |
| 58 | + @state, p1, p2 = $1.split(":").map {|a| a.strip} | |
| 59 | + return if @state == "abnormal" | |
| 56 | 60 | p1_name, p1_mark = p1.split(" ") |
| 57 | 61 | p2_name, p2_mark = p2.split(" ") |
| 58 | 62 | if p1_name == @black_name |
| @@ -65,13 +69,13 @@ | ||
| 65 | 69 | raise "Never reach!: #{black} #{white} #{p3} #{p2}" |
| 66 | 70 | end |
| 67 | 71 | end |
| 68 | - if /^\$START_TIME:(.*)$/ =~ str | |
| 72 | + if /^\$START_TIME:(.*)$/ =~ @str | |
| 69 | 73 | @start_time = Time.parse($1.strip) |
| 70 | 74 | end |
| 71 | - if /^'\$END_TIME:(.*)$/ =~ str | |
| 75 | + if /^'\$END_TIME:(.*)$/ =~ @str | |
| 72 | 76 | @end_time = Time.parse($1.strip) |
| 73 | 77 | end |
| 74 | - if /^'rating:(.*)$/ =~ str | |
| 78 | + if /^'rating:(.*)$/ =~ @str | |
| 75 | 79 | black_id, white_id = $1.split(":").map {|a| a.strip} |
| 76 | 80 | @black_id = identify_id(black_id) |
| 77 | 81 | @white_id = identify_id(white_id) |
| @@ -90,6 +94,14 @@ | ||
| 90 | 94 | end |
| 91 | 95 | end |
| 92 | 96 | |
| 97 | + def movetimes | |
| 98 | + ret = [] | |
| 99 | + @str.gsub(%r!^T(\d+)!) do |match| | |
| 100 | + ret << $1.to_i | |
| 101 | + end | |
| 102 | + return ret | |
| 103 | + end | |
| 104 | + | |
| 93 | 105 | def to_s |
| 94 | 106 | return "Summary: #{@file_name}\n" + |
| 95 | 107 | "BlackName #{@black_name}, WhiteName #{@white_name}\n" + |
| @@ -0,0 +1,205 @@ | ||
| 1 | +#!/usr/bin/ruby1.9.1 | |
| 2 | +# This program filters CSA files. For example, if you want only CSA files | |
| 3 | +# played by GPS vs Bonanza, | |
| 4 | +# $ ./csa-filter.rb --players gps-l,bonanza some_dir | |
| 5 | +# you will see such files under the some_dir directory. | |
| 6 | +# | |
| 7 | +# Author:: Daigo Moriwaki <daigo at debian dot org> | |
| 8 | +# Copyright:: Copyright (C) 2006-2008 Daigo Moriwaki <daigo at debian dot org> | |
| 9 | +# | |
| 10 | +# $Id$ | |
| 11 | +# | |
| 12 | +#-- | |
| 13 | +# This program is free software; you can redistribute it and/or modify | |
| 14 | +# it under the terms of the GNU General Public License as published by | |
| 15 | +# the Free Software Foundation; either version 2 of the License, or | |
| 16 | +# (at your option) any later version. | |
| 17 | +# | |
| 18 | +# This program is distributed in the hope that it will be useful, | |
| 19 | +# but WITHOUT ANY WARRANTY; without even the implied warranty of | |
| 20 | +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
| 21 | +# GNU General Public License for more details. | |
| 22 | +# | |
| 23 | +# You should have received a copy of the GNU General Public License | |
| 24 | +# along with this program; if not, write to the Free Software | |
| 25 | +# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA | |
| 26 | +#++ | |
| 27 | + | |
| 28 | +$:.unshift File.dirname(__FILE__) | |
| 29 | +require 'csa-filter' | |
| 30 | +require 'set' | |
| 31 | + | |
| 32 | +class Monthly | |
| 33 | + def initialize | |
| 34 | + @games = Hash.new {|hash,key| hash[key] = 0} | |
| 35 | + @players = Hash.new {|hash,key| hash[key] = Set.new} | |
| 36 | + end | |
| 37 | + | |
| 38 | + def add(csa) | |
| 39 | + st = csa.start_time | |
| 40 | + month = st.strftime("%Y%m") | |
| 41 | + | |
| 42 | + @games[month] += 1 | |
| 43 | + | |
| 44 | + [csa.black_id, csa.white_id].each do |player| | |
| 45 | + @players[month].add(player) | |
| 46 | + end | |
| 47 | + end | |
| 48 | + | |
| 49 | + def print | |
| 50 | + puts "YYYYMM\t#games\t#players" | |
| 51 | + @games.sort {|a,b| a[0] <=> b[0]}.each do |key,value| | |
| 52 | + puts "%s\t% 6d\t% 2d" % [key, value, @players[key].size] | |
| 53 | + end | |
| 54 | + end | |
| 55 | +end | |
| 56 | + | |
| 57 | +class Values | |
| 58 | + def initialize | |
| 59 | + @v = [] | |
| 60 | + end | |
| 61 | + | |
| 62 | + def add(value) | |
| 63 | + case value | |
| 64 | + when Array | |
| 65 | + @v.concat value | |
| 66 | + else | |
| 67 | + @v << value | |
| 68 | + end | |
| 69 | + end | |
| 70 | + | |
| 71 | + def print(file) | |
| 72 | + total = @v.inject(0){|sum, item| sum+item} | |
| 73 | + avg = 1.0*total/@v.size | |
| 74 | + puts "avg: %f sec (size: %d)" % [avg, @v.size] | |
| 75 | + | |
| 76 | + File.open(file, "w") do |f| | |
| 77 | + @v.each {|v| f.puts v} | |
| 78 | + end | |
| 79 | + end | |
| 80 | +end | |
| 81 | + | |
| 82 | +class State | |
| 83 | + def initialize | |
| 84 | + @hash = Hash.new {|hash,key| hash[key] = 0} | |
| 85 | + end | |
| 86 | + | |
| 87 | + def add(value) | |
| 88 | + if value.nil? || value.empty? | |
| 89 | + value = "error" | |
| 90 | + end | |
| 91 | + @hash[value] += 1 | |
| 92 | + end | |
| 93 | + | |
| 94 | + def print | |
| 95 | + puts "status\t#games" | |
| 96 | + @hash.sort {|a,b| b[1] <=> a[1]}.each do |key, value| | |
| 97 | + puts "%s\t% 6d" % [key, value] | |
| 98 | + end | |
| 99 | + end | |
| 100 | +end | |
| 101 | + | |
| 102 | +$monthly = Monthly.new | |
| 103 | +$gametime = Values.new | |
| 104 | +$movetime = Values.new | |
| 105 | +$moves = Values.new | |
| 106 | +$states = State.new | |
| 107 | + | |
| 108 | +def do_file(file) | |
| 109 | + $OPT_REPEAT -= 1 if $OPT_REPEAT > 0 | |
| 110 | + csa = CsaFileReader.new(file) | |
| 111 | + | |
| 112 | + # Want to see complete games | |
| 113 | + $states.add csa.state | |
| 114 | + return unless csa.state == "toryo" | |
| 115 | + | |
| 116 | + # See games between 2008/03 to 2009/07 | |
| 117 | + return if csa.start_time < Time.parse("2008/03/01") || | |
| 118 | + csa.start_time >= Time.parse("2009/08/01") | |
| 119 | + | |
| 120 | + # Process monthly | |
| 121 | + $monthly.add(csa) | |
| 122 | + | |
| 123 | + # Process gametime | |
| 124 | + duration = (csa.end_time - csa.start_time).to_i | |
| 125 | + if duration > 2200 | |
| 126 | + $stderr.puts "Too long game: #{file}" | |
| 127 | + return | |
| 128 | + end | |
| 129 | + $gametime.add duration.to_i | |
| 130 | + | |
| 131 | + # Process movetime | |
| 132 | + values = csa.movetimes | |
| 133 | + $movetime.add values | |
| 134 | + | |
| 135 | + #Process moves | |
| 136 | + $moves.add values.size | |
| 137 | + | |
| 138 | +rescue => ex | |
| 139 | + $stderr.puts "ERROR: %s" % [file] | |
| 140 | + throw ex | |
| 141 | +end | |
| 142 | + | |
| 143 | +if $0 == __FILE__ | |
| 144 | + def usage | |
| 145 | + puts "Usage: #{$0} [OPTIONS] dir [...]" | |
| 146 | + puts "Options:" | |
| 147 | + exit 1 | |
| 148 | + end | |
| 149 | + | |
| 150 | + usage if ARGV.empty? | |
| 151 | + | |
| 152 | + parser = GetoptLong.new( | |
| 153 | + ['--repeat', '-n', GetoptLong::REQUIRED_ARGUMENT] | |
| 154 | + ) | |
| 155 | + begin | |
| 156 | + parser.each_option do |name, arg| | |
| 157 | + eval "$OPT_#{name.sub(/^--/, '').gsub(/-/, '_').upcase} = '#{arg}'" | |
| 158 | + end | |
| 159 | + rescue | |
| 160 | + usage | |
| 161 | + end | |
| 162 | + | |
| 163 | + $OPT_REPEAT = $OPT_REPEAT.to_i | |
| 164 | + if $OPT_REPEAT == 0 | |
| 165 | + $OPT_REPEAT = -1 | |
| 166 | + end | |
| 167 | + | |
| 168 | + while (cmd = ARGV.shift) | |
| 169 | + | |
| 170 | + if FileTest.directory?(cmd) | |
| 171 | + Dir.glob(File.join(cmd, "**", "*.csa")).each do |file| | |
| 172 | + break if $OPT_REPEAT == 0 | |
| 173 | + do_file(file) | |
| 174 | + end | |
| 175 | + elsif FileTest.file?(cmd) | |
| 176 | + break if $OPT_REPEAT == 0 | |
| 177 | + do_file(cmd) | |
| 178 | + else | |
| 179 | + throw "Unknown file or directory: #{cmd}" | |
| 180 | + end | |
| 181 | + | |
| 182 | + puts "States" | |
| 183 | + puts "------" | |
| 184 | + $states.print | |
| 185 | + puts | |
| 186 | + puts "=== Toryo Games ===" | |
| 187 | + puts | |
| 188 | + puts "Montly" | |
| 189 | + puts "------" | |
| 190 | + $monthly.print | |
| 191 | + puts | |
| 192 | + puts "Play Time" | |
| 193 | + puts "---------" | |
| 194 | + $gametime.print("gametime.dat") | |
| 195 | + puts | |
| 196 | + puts "Move Time" | |
| 197 | + puts "---------" | |
| 198 | + $movetime.print("movetime.dat") | |
| 199 | + puts | |
| 200 | + puts "Moves" | |
| 201 | + puts "-----" | |
| 202 | + $moves.print("moves.dat") | |
| 203 | + end | |
| 204 | +end | |
| 205 | + |
Fork