re_expand 0.0.1 → 0.0.2

data/lib/Generator.rb

@@ -17,217 +17,221 @@ require 'Scanner'
 require 'Node'
 require 'Asearch'
 
-class GenNode
-  def initialize(id, state=[], s="", substrings=[], accept=false)
-    @id = id
-    @s = s
-    @substrings = substrings
-    @accept = accept
-    @state = state
-  end
-
-  attr :id, true
-  attr :s, true
-  attr :substrings, true
-  attr :accept, true
-  attr :state, true
-end
-
-class Generator
-  def initialize(s = nil)
-    @s = (s ? [s] : [])
-    @matchedlist = []
-    @par = 0
-    @commands = []
-  end
-
-  def add(pat,command)
-    @s << pat
-    @commands << command
-  end
-
-  def delete
-    @s.pop
-    @commands.pop
+module ReExpand
+  class GenNode
+    def initialize(id, state=[], s="", substrings=[], accept=false)
+      @id = id
+      @s = s
+      @substrings = substrings
+      @accept = accept
+      @state = state
+    end
+    
+    attr :id, true
+    attr :s, true
+    attr :substrings, true
+    attr :accept, true
+    attr :state, true
   end
-
-  #
-  # ルールを解析して状態遷移機械を作成し、patにマッチするもののリストを返す
-  #
-  def generate(pat, app = nil)
-    res = [[],[],[]] # 曖昧度0,1,2のマッチ結果
-    patterns = pat.split.map { |p| p.downcase }
-
-    @asearch = Asearch.new(pat)
-    scanner = Scanner.new(@s.join('|'))
-
-    # HelpDataで指定した状態遷移機械全体を生成
-    # (少し時間がかかる)
-    (startnode, endnode) = regexp(scanner,true) # top level
-
-    #
-    # 状態遷移機械からDepth-Firstで文字列を生成する
-    # n個のノードを経由して生成される状態の集合をlists[n]に入れる
-    # 生成しながらマッチングも計算する
-    #
-    lists = []
-    listed = [{},{},{}]
-    #
-    # 初期状態
+  
+  class Generator
+    def initialize(s = nil)
+      @s = (s ? [s] : [])
+      @matchedlist = []
+      @par = 0
+      @commands = []
+    end
+    
+    def add(pat,command)
+      @s << pat
+      @commands << command
+    end
+    
+    def delete
+      @s.pop
+      @commands.pop
+    end
+    
     #
-    list = []
-    list[0] = GenNode.new(startnode.id, @asearch.initstate)
-    lists[0] = list
+    # ルールを解析して状態遷移機械を作成し、patにマッチするもののリストを返す
     #
-    (0..1000).each { |length|
-      break if app && app.inputPending
-      list = lists[length]
-      newlist = []
-      # puts "#{length} - #{list.length}"
-      list.each { |entry|
-        srcnode = Node.node(entry.id)
-        if list.length * srcnode.trans.length < 10000 then
-          srcnode.trans.each { |trans|
-            ss = entry.substrings.dup
-            srcnode.pars.each { |i|
-              ss[i-1] = ss[i-1].to_s + trans.arg
-            }
-            newstate = @asearch.state(entry.state, trans.str) # 新しいマッチング状態を計算してノードに保存
-            s = entry.s + trans.str
-            acceptno = trans.dest.accept
-            newlist << GenNode.new(trans.dest.id, newstate, s, ss, acceptno)
-            #
-            # この時点で、マッチしているかどうかをstateとacceptpatで判断できる
-            # マッチしてたら出力リストに加える
-            #
-            if acceptno then
-              maxambig = 2
-              (0..maxambig).each { |ambig|
-                if !listed[ambig][s] then
-                  if (newstate[ambig] & @asearch.acceptpat) != 0 then # マッチ
-                    maxambig = ambig if ambig < maxambig # 曖昧度0でマッチすれば曖昧度1の検索は不要
-                    listed[ambig][s] = true
-                    sslen = ss.length
-                    if sslen > 0 then
-                      # patstr = "(.*)\t" * (sslen-1) + "(.*)"
-                      patstr = (["(.*)"] * sslen).join("\t")
-                      /#{patstr}/ =~ ss.join("\t")
+    def generate(pat, app = nil)
+      res = [[],[],[]] # 曖昧度0,1,2のマッチ結果
+      patterns = pat.split.map { |p| p.downcase }
+      
+      @asearch = Asearch.new(pat)
+      scanner = Scanner.new(@s.join('|'))
+      
+      # HelpDataで指定した状態遷移機械全体を生成
+      # (少し時間がかかる)
+      (startnode, endnode) = regexp(scanner,true) # top level
+      
+      #
+      # 状態遷移機械からDepth-Firstで文字列を生成する
+      # n個のノードを経由して生成される状態の集合をlists[n]に入れる
+      # 生成しながらマッチングも計算する
+      #
+      lists = []
+      listed = [{},{},{}]
+      #
+      # 初期状態
+      #
+      list = []
+      list[0] = GenNode.new(startnode.id, @asearch.initstate)
+      lists[0] = list
+      #
+      (0..1000).each { |length|
+        break if app && app.inputPending
+        list = lists[length]
+        newlist = []
+        # puts "#{length} - #{list.length}"
+        list.each { |entry|
+          srcnode = Node.node(entry.id)
+          if list.length * srcnode.trans.length < 10000 then
+            srcnode.trans.each { |trans|
+              ss = entry.substrings.dup
+              srcnode.pars.each { |i|
+                ss[i-1] = ss[i-1].to_s + trans.arg
+              }
+              newstate = @asearch.state(entry.state, trans.str) # 新しいマッチング状態を計算してノードに保存
+              s = entry.s + trans.str
+              acceptno = trans.dest.accept
+              newlist << GenNode.new(trans.dest.id, newstate, s, ss, acceptno)
+              #
+              # この時点で、マッチしているかどうかをstateとacceptpatで判断できる
+              # マッチしてたら出力リストに加える
+              #
+              if acceptno then
+                maxambig = 2
+                (0..maxambig).each { |ambig|
+                  if !listed[ambig][s] then
+                    if (newstate[ambig] & @asearch.acceptpat) != 0 then # マッチ
+                      maxambig = ambig if ambig < maxambig # 曖昧度0でマッチすれば曖昧度1の検索は不要
+                      listed[ambig][s] = true
+                      sslen = ss.length
+                      if sslen > 0 then
+                        # patstr = "(.*)\t" * (sslen-1) + "(.*)"
+                        patstr = (["(.*)"] * sslen).join("\t")
+                        /#{patstr}/ =~ ss.join("\t")
+                      end
+                      ##                    next if $1 == $2
+                      # 'set date #{$2}' のような記述の$変数にsubstringの値を代入
+                      File.open("/tmp/log","a"){ |f|
+                        f.puts "#{s}-----" + eval('%('+@commands[acceptno]+')')
+                      }
+                      res[ambig] << [s, eval('%('+@commands[acceptno]+')')]
                     end
-##                    next if $1 == $2
-                    # 'set date #{$2}' のような記述の$変数にsubstringの値を代入
-                File.open("/tmp/log","a"){ |f|
-                  f.puts "#{s}-----" + eval('%('+@commands[acceptno]+')')
-                }
-                    res[ambig] << [s, eval('%('+@commands[acceptno]+')')]
                   end
-                end
-              }
-            end
-          }
-        end
+                }
+              end
+            }
+          end
+        }
+        break if newlist.length == 0
+        lists << newlist
+        break if res[0].length > 100
       }
-      break if newlist.length == 0
-      lists << newlist
-      break if res[0].length > 100
-    }
-    [res[0], res[1], res[2]]
-  end
-
-  #
-  # 正規表現をパースして状態遷移機械を作る
-  #
-  private
-  #            n1     n2
-  #        +-->□.....□--+
-  # start /                \  end
-  #     □ --->□.....□---> □
-  #       \                /
-  #        +-->□.....□--+
-  #
-  def regexp(s,toplevel=false) # regcat { '|' regcat }
-    startnode = Node.new
-    endnode = Node.new
-    if toplevel then
-      @pars = []
-      @parno = 0
-      @ruleid = 0
+      [res[0], res[1], res[2]]
     end
-    startnode.pars = @pars
-    endnode.pars = @pars
-    (n1, n2) = regcat(s)
-    startnode.addTrans('',n1)
-    if toplevel then
-      n2.accept = @ruleid
-    end
-    n2.addTrans('',endnode)
-    while s.gettoken == '|' && s.nexttoken != '' do
+    
+    #
+    # 正規表現をパースして状態遷移機械を作る
+    #
+    private
+    #            n1     n2
+    #        +-->□.....□--+
+    # start /                \  end
+    #     □ --->□.....□---> □
+    #       \                /
+    #        +-->□.....□--+
+    #
+    def regexp(s,toplevel=false) # regcat { '|' regcat }
+      startnode = Node.new
+      endnode = Node.new
       if toplevel then
         @pars = []
         @parno = 0
-        @ruleid += 1
+        @ruleid = 0
       end
+      startnode.pars = @pars
+      endnode.pars = @pars
       (n1, n2) = regcat(s)
       startnode.addTrans('',n1)
       if toplevel then
         n2.accept = @ruleid
       end
       n2.addTrans('',endnode)
-    end
-    s.ungettoken
-    return [startnode, endnode]
-  end
-
-  def regcat(s) # regfactor { regfactor }
-    (startnode, endnode) = regfactor(s)
-    while s.gettoken !~ /^[\)\]\|]$/ && s.nexttoken != '' do
+      while s.gettoken == '|' && s.nexttoken != '' do
+        if toplevel then
+          @pars = []
+          @parno = 0
+          @ruleid += 1
+        end
+        (n1, n2) = regcat(s)
+        startnode.addTrans('',n1)
+        if toplevel then
+          n2.accept = @ruleid
+        end
+        n2.addTrans('',endnode)
+      end
       s.ungettoken
-      (n1, n2) = regfactor(s)
-      endnode.addTrans('',n1)
-      endnode = n2
+      return [startnode, endnode]
     end
-    s.ungettoken
-    return [startnode, endnode]
-  end
-
-  def regfactor(s) # regterm [ '?' | '+' | '*' ]
-    (startnode, endnode) = regterm(s)
-    t = s.gettoken
-    if t =~ /^[\?]$/ then
-      startnode.addTrans('',endnode)
-    elsif t =~ /^[\+]$/ then
-      endnode.addTrans('',startnode)
-    elsif t =~ /^[\*]$/ then
-      startnode.addTrans('',endnode)
-      endnode.addTrans('',startnode)
-    else
+    
+    def regcat(s) # regfactor { regfactor }
+      (startnode, endnode) = regfactor(s)
+      while s.gettoken !~ /^[\)\]\|]$/ && s.nexttoken != '' do
+        s.ungettoken
+        (n1, n2) = regfactor(s)
+        endnode.addTrans('',n1)
+        endnode = n2
+      end
       s.ungettoken
+      return [startnode, endnode]
     end
-    return [startnode,endnode]
-  end
-
-  def regterm(s) # '(' regexp ')' | token
-    t = s.gettoken
-    if t == '(' then
-      @parno += 1
-      @pars.push(@parno)
-      (n1, n2) = regexp(s)
-      n1.pars = @pars.dup
+    
+    def regfactor(s) # regterm [ '?' | '+' | '*' ]
+      (startnode, endnode) = regterm(s)
       t = s.gettoken
-      if t == ')' then
-        @pars.pop
-        n2.pars = @pars.dup
-        return [n1, n2]
+      if t =~ /^[\?]$/ then
+        startnode.addTrans('',endnode)
+      elsif t =~ /^[\+]$/ then
+        endnode.addTrans('',startnode)
+      elsif t =~ /^[\*]$/ then
+        startnode.addTrans('',endnode)
+        endnode.addTrans('',startnode)
       else
-        puts 'missing )'
-        exit
+        s.ungettoken
+      end
+      return [startnode,endnode]
+    end
+    
+    def regterm(s) # '(' regexp ')' | token
+      t = s.gettoken
+      if t == '(' then
+        @parno += 1
+        @pars.push(@parno)
+        (n1, n2) = regexp(s)
+        n1.pars = @pars.dup
+        t = s.gettoken
+        if t == ')' then
+          @pars.pop
+          n2.pars = @pars.dup
+          return [n1, n2]
+        else
+          puts 'missing )'
+          exit
+        end
+      else
+        startnode = Node.new
+        startnode.pars = @pars.dup
+        endnode = Node.new
+        endnode.pars = @pars.dup
+        startnode.addTrans(t,endnode)
+        return [startnode, endnode]
       end
-    else
-      startnode = Node.new
-      startnode.pars = @pars.dup
-      endnode = Node.new
-      endnode.pars = @pars.dup
-      startnode.addTrans(t,endnode)
-      return [startnode, endnode]
     end
   end
+  
 end
+

data/lib/Node.rb

@@ -8,56 +8,58 @@
 #       ----------> □
 #
 
-class Trans
-  def initialize(pat,dest)
-    # pat にマッチしたら dest に遷移
-    @pat = pat
-    @dest = dest
-  end
-
-  attr_reader :pat, :dest
-
-  def str
-    @pat.split(/\t/)[0].to_s
-  end
-
-  def arg
-    if @pat =~ /^(.*)\t(.*)$/ then
-      return $2
-    else
-      return @pat
+module ReExpand
+  class Trans
+    def initialize(pat,dest)
+      # pat にマッチしたら dest に遷移
+      @pat = pat
+      @dest = dest
+    end
+    
+    attr_reader :pat, :dest
+    
+    def str
+      @pat.split(/\t/)[0].to_s
+    end
+    
+    def arg
+      if @pat =~ /^(.*)\t(.*)$/ then
+        return $2
+      else
+        return @pat
+      end
     end
   end
-end
-
-class Node
-  @@id = 1
-  @@nodes = {}
-
-  def initialize
-    @accept = nil
-    @trans = []
-    @id = @@id
-    @@nodes[@id] = self
-    @@id += 1
-    @pars = []
-  end
-
-  attr_reader :id
-  attr_reader :trans
-  attr :accept, true
-  attr :pars, true
-
-  def addTrans(pat,dest)
-    t = Trans.new(pat,dest)
-    @trans << t
-  end
-
-  def Node.node(id) # ノードidからノードを取得
-    @@nodes[id.to_i]
-  end
-
-  def Node.nodes
-    @@nodes.values
+  
+  class Node
+    @@id = 1
+    @@nodes = {}
+    
+    def initialize
+      @accept = nil
+      @trans = []
+      @id = @@id
+      @@nodes[@id] = self
+      @@id += 1
+      @pars = []
+    end
+    
+    attr_reader :id
+    attr_reader :trans
+    attr :accept, true
+    attr :pars, true
+    
+    def addTrans(pat,dest)
+      t = Trans.new(pat,dest)
+      @trans << t
+    end
+    
+    def Node.node(id) # ノードidからノードを取得
+      @@nodes[id.to_i]
+    end
+    
+    def Node.nodes
+      @@nodes.values
+    end
   end
 end

data/lib/Scanner.rb

@@ -5,55 +5,57 @@
 
 $KCODE = 'utf8'
 
-class Scanner
-  def initialize(s)
-    @s = s
-    @a = s.split(//)
-    @p = 0
-    @t = ''
-    @u = ''
-  end
-
-  def gettoken
-    if @u != '' then
-      @t = @u
-      @u = ''
-      return @t
-    end
-    if @p >= @a.length then
+module ReExpand
+  class Scanner
+    def initialize(s)
+      @s = s
+      @a = s.split(//)
+      @p = 0
       @t = ''
-      return ''
+      @u = ''
     end
-    @t = @a[@p]
-    if @t =~ /^[\(\|\)\*\?\[\]]$/ then
-      @p += 1
-      return @t
-    elsif @t == '\\' then
-      @p += 1
+    
+    def gettoken
+      if @u != '' then
+        @t = @u
+        @u = ''
+        return @t
+      end
+      if @p >= @a.length then
+        @t = ''
+        return ''
+      end
       @t = @a[@p]
-      @t = "\n" if @t == 'n'
-      @t = "\t" if @t == 't'
-      @p += 1
-      return @t
-    else
-      @p += 1
-      while @p < @a.length && @a[@p] !~ /^[\(\|\)\*\+\?\[\]\\]$/ do
-        @t += @a[@p]
+      if @t =~ /^[\(\|\)\*\?\[\]]$/ then
+        @p += 1
+        return @t
+      elsif @t == '\\' then
         @p += 1
+        @t = @a[@p]
+        @t = "\n" if @t == 'n'
+        @t = "\t" if @t == 't'
+        @p += 1
+        return @t
+      else
+        @p += 1
+        while @p < @a.length && @a[@p] !~ /^[\(\|\)\*\+\?\[\]\\]$/ do
+          @t += @a[@p]
+          @p += 1
+        end
+        return @t
       end
-      return @t
     end
-  end
-
-  def ungettoken
-    if @u == '' then
-      @u = @t
-    else
-      puts "Can't ungettoken(#{token})"
+    
+    def ungettoken
+      if @u == '' then
+        @u = @t
+      else
+        puts "Can't ungettoken(#{token})"
+      end
+    end
+    
+    def nexttoken
+      return @t
     end
-  end
-
-  def nexttoken
-    return @t
   end
 end

data/lib/re_expand.rb

@@ -3,7 +3,7 @@ $:.unshift(File.dirname(__FILE__)) unless
   $:.include?(File.dirname(__FILE__)) || $:.include?(File.expand_path(File.dirname(__FILE__)))
 
 module ReExpand
-  VERSION = '0.0.1'
+  VERSION = '0.0.2'
 end
 
 require 'Generator'
@@ -14,7 +14,7 @@ require 'Scanner'
 class String
   # ExpandHelp用のライブラリを利用
   def expand(filterpat=' ')
-    g = Generator.new
+    g = ReExpand::Generator.new
     g.add(self,'')
     strings = []
     if filterpat.class == String then

data/test/test_generator.rb

@@ -1,36 +1,32 @@
 # -*- coding: utf-8 -*-
 
-require 'Generator'
-
-class GeneratorTest < Test::Unit::TestCase
-  def setup
-  end
-  
-  def teardown
-  end
-  
-  def test_clock
-    g = Generator.new
-    g.add '(時計|時間|時刻)を(0|1|2|3|4|5|6|7|8|9|10|11|12)時に(セットする|設定する|あわせる)', 'set time #{$2}:00'
-    res = g.generate(' 10 ')
-    assert res[0].member?(['時刻を10時に設定する','set time 10:00'])
-    assert res[0].member?(['時計を10時にセットする','set time 10:00'])
-    assert !res[0].member?(['時計を8時にセットする','set time 8:00'])
-  end
-
-  def test_substring
-    g = Generator.new
-    g.add '(a)bcd(e(fg)h)i(jk)', '#{$1}/#{$2}/#{$3}/#{$4}'
-    res = g.generate(' b ')
-    assert res[0].member?(['abcdefghijk', 'a/efgh/fg/jk'])
-    assert !res[0].member?(['abcdefghijk', 'a/fgh/fg/jk'])
-    g.add '(ab|cd)efg(hij|klm)n', '#{$1}/#{$2}'
-    res = g.generate(' ef ')
-    assert res[0].member?(['abefghijn', 'ab/hij'])
-    assert res[0].member?(['cdefgklmn', 'cd/klm'])
+module ReExpand
+  class GeneratorTest < Test::Unit::TestCase
+    def setup
+    end
+    
+    def teardown
+    end
+    
+    def test_clock
+      g = Generator.new
+      g.add '(時計|時間|時刻)を(0|1|2|3|4|5|6|7|8|9|10|11|12)時に(セットする|設定する|あわせる)', 'set time #{$2}:00'
+      res = g.generate(' 10 ')
+      assert res[0].member?(['時刻を10時に設定する','set time 10:00'])
+      assert res[0].member?(['時計を10時にセットする','set time 10:00'])
+      assert !res[0].member?(['時計を8時にセットする','set time 8:00'])
+    end
+    
+    def test_substring
+      g = Generator.new
+      g.add '(a)bcd(e(fg)h)i(jk)', '#{$1}/#{$2}/#{$3}/#{$4}'
+      res = g.generate(' b ')
+      assert res[0].member?(['abcdefghijk', 'a/efgh/fg/jk'])
+      assert !res[0].member?(['abcdefghijk', 'a/fgh/fg/jk'])
+      g.add '(ab|cd)efg(hij|klm)n', '#{$1}/#{$2}'
+      res = g.generate(' ef ')
+      assert res[0].member?(['abefghijn', 'ab/hij'])
+      assert res[0].member?(['cdefgklmn', 'cd/klm'])
+    end
   end
 end
-
-
-
-

data/test/test_scanner.rb

@@ -1,27 +1,28 @@
 # -*- coding: utf-8 -*-
 
-require 'Scanner'
-
-class ScannerTest < Test::Unit::TestCase
-  def setup
-  end
-  
-  def teardown
-  end
-
-  def test_1
-    s = Scanner.new('abc*def')
-    assert_equal s.gettoken, 'abc'
-    assert_equal s.gettoken, '*'
-    assert_equal s.gettoken, 'def'
-  end
-
-  def test_2
-    s = Scanner.new("(a\tb|c\td)")
-    assert_equal s.gettoken, "("
-    assert_equal s.gettoken, "a\tb"
-    assert_equal s.gettoken, '|'
-    assert_equal s.gettoken, "c\td"
-    assert_equal s.gettoken, ')'
+module ReExpand
+  class ScannerTest < Test::Unit::TestCase
+    def setup
+    end
+    
+    def teardown
+    end
+    
+    def test_1
+      s = Scanner.new('abc*def')
+      assert_equal s.gettoken, 'abc'
+      assert_equal s.gettoken, '*'
+      assert_equal s.gettoken, 'def'
+    end
+    
+    def test_2
+      s = Scanner.new("(a\tb|c\td)")
+      assert_equal s.gettoken, "("
+      assert_equal s.gettoken, "a\tb"
+      assert_equal s.gettoken, '|'
+      assert_equal s.gettoken, "c\td"
+      assert_equal s.gettoken, ')'
+    end
   end
 end
+

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: re_expand
 version: !ruby/object:Gem::Version 
-  hash: 29
+  hash: 27
   prerelease: 
   segments: 
   - 0
   - 0
-  - 1
-  version: 0.0.1
+  - 2
+  version: 0.0.2
 platform: ruby
 authors: 
 - Toshiyuki Masui