ruby wrapper for japanese dependency parser cabocha 0.5

cabo_wrapper.rb

#!/usr/bin/ruby -Kw
$KCODE="UTF-8"
require "CaboCha"
require "kconv"

class DepAnalyzeByCabocha
  @cabo
  def initialize(args=[])
    @cabo = CaboCha::Parser.new(args)
  end
  
  def analyze(str)
    if str.split(//).length > 500
      return []
    else
      return analyzeSub(str)
    end
  end
  
  def analyzeSub(str)
    [email protected](str.toeuc)
    phrases=Array.new
    
    begin#解析誤りへの対策
      size =  tree.size()
    rescue
      return []
    end
    
    cid = 0
    phrase=nil
    (0..size-1).each{|i|
      token = tree.token(i)
      
      if (token.hasChunk())
        if phrase != nil
          phrases.push(phrase)
        end
        phrase=Phrase.new(token.chunk,phrases,cid)
        cid = cid + 1
      end
      phrase.pushToken(token)
    }
    
    if phrase!=nil
      phrases.push(phrase)
    end
    phrases.each{|x|
      x.setDep
    }    
    phrases
  end
  
  class Phrase
    @chunk
    @tokens
    @head
    @dependents
    @sentence
    @id
    def initialize(chunk,phrases,id)
      @chunk=chunk
      @tokens=Array.new
      @dependents=Array.new
      @sentence=phrases
      @id=id
    end
    attr_accessor :dependents,:head,:tokens,:chunk,:id,:sentence
    
    def pushToken(token)
      @tokens.push(token)
    end
    
    def setDep
      @head=@sentence[@chunk.link()]
      @head.setDependent(@id)
    end
    
    def getHead
      [@chunk.rel,@head]
    end
    
    def setDependent(id)
      @dependents.push(@sentence[id])
    end
    
    def ne?()
      @tokens.each{|x|
        if x.ne!="O"
          return x.ne
        end
      }
      nil
    end
    
    def num?()
      @tokens.each{|x|
        if x.pos.toutf8=~/名詞-数/
          return true
        end
      }
      nil
    end
    
    def getSurface
      @tokens.map{|x|
        x.surface()
      }.join("")
    end
  end
end