Examples of Morpheme

org.snu.ids.ha.ma.Morpheme

 형태소 정보를 가지는 Class 형태소, + 부가 class 정보 가짐 string : 형태소 정보 composed : 복합어 여부 tag : 형태소의 구분 품사, 어미, 어근, infoEnc 에 encoding한 상태로 저장하도록 수정함 [composed] 추가 색인어를 추출할 필요가 있는지를 구분하는 어휘 사전에 있는 어휘에서 추가로 분해하여 반환한다. 사랑방손님 -> 사랑방손님, 사랑방,손님 둥글게둥글게 -> 둥글게둥글게, 둥글게

@author therocks @since 2007. 6. 3

Examples of org.snu.ids.ha.ma.Morpheme

    Keyword ret = null;
    if( mc == null || mc.size() < 2 ) return null;


    int nnCnt = 0;
    for( int i = 0; i < mc.size(); i++ ) {
      Morpheme mp = mc.get(i);
      if( mp.isTagOf(POSTag.NN) ) {
        if( ret == null ) {
          ret = new Keyword(mp);
          ret.setComposed(true);
          nnCnt++;
        } else if( nnCnt == 0 ) {
          return null;
        } else {
          ret.setString(ret.getString() + mp.getString());
          nnCnt++;
        }
      } else if( ret != null && nnCnt > 1 ) {
        return ret;
      } else {

View Full Code Here

Examples of org.snu.ids.ha.ma.Morpheme

   */
  public String getMorpXmlStr()
  {
    StringBuffer sb = new StringBuffer();
    for( int i = 0, size = (eojeol == null ? 0 : eojeol.size()); i < size; i++ ) {
      Morpheme morp = eojeol.get(i);
      if( i > 0 ) sb.append("+");
      sb.append(Util.rplcXMLSpclChar(morp.getString()) + "/" + morp.getTag());
    }
    return sb.toString();
  }

View Full Code Here

Examples of org.snu.ids.ha.ma.Morpheme

    EnglishStemmer engStemmer = new EnglishStemmer();
    
    try {
      List<MExpression> meList = leaveJustBest(postProcess(analyze(string)));


      Morpheme mp = null;
      MCandidate mc = null;
      MExpression me = null;
      Keyword keyword = null;
      List<Morpheme> mpList = new ArrayList<Morpheme>();
      for( int i = 0, size = meList == null ? 0 : meList.size(); i < size; i++ ) {
        me = meList.get(i);
        mc = me.get(0);


        int jSize = mc.size();
        if( jSize == 1 ) {
          mp = mc.get(0);
          mp.setString(me.getExp());
          mpList.add(mp);
        } else {
          // 분할되지 않은 리스트 형태로 형태소를 넣어준다.
          for( int j = 0; j < jSize; j++ )
            mpList.add(mc.get(j));
        }


      }


      // 복합 UOM 확인
      for( int endIdx = mpList.size() - 1; endIdx > 0; endIdx-- ) {
        for( int startIdx = Math.max(endIdx - MAX_UOM_SIZE, 0); startIdx < endIdx; startIdx++ ) {
          String tempName = "";
          for( int i = startIdx; i <= endIdx; i++ ) {
            tempName += mpList.get(i).getString();
          }


          // 다수의 토큰으로 이루어진 UOM 확인
          if( UOMDic.contains(tempName) ) {
            for( ; startIdx < endIdx; endIdx-- ) {
              mpList.remove(startIdx + 1);
            }
            mp = mpList.get(startIdx);
            mp.setString(tempName);
            mp.setCharSet(CharSetType.COMBINED);
            mp.setTag(POSTag.NNM);
          }
          // 다수의 토큰으로 이루어진 화학식 확인
          else if( ChemFormulaDic.contains(tempName) ) {
            for( ; startIdx < endIdx; endIdx-- ) {
              mpList.remove(startIdx + 1);
            }
            mp = mpList.get(startIdx);
            mp.setString(tempName);
            mp.setCharSet(CharSetType.COMBINED);
            mp.setTag(POSTag.UN);
          }
          // 다수의 토큰으로 이루어진 명사 확인 ((주), Web2.0)류의 키워드
          else if( CompNounDic.contains(tempName) ) {
            for( ; startIdx < endIdx; endIdx-- ) {
              mpList.remove(startIdx + 1);
            }
            if( !JunkWordDic.contains(tempName) ) {
              mp = mpList.get(startIdx);
              mp.setString(tempName);
              mp.setCharSet(CharSetType.COMBINED);
              mp.setTag(POSTag.NNG);
              mp.setComposed(true);
            }
          }
        }
      }


      // 키워드 추출
      for( int i = 0, size = mpList.size(); i < size; i++ ) {
        mp = mpList.get(i);
        mp.setString(mp.getString().toLowerCase());


        // stemming 및 키워드 추출
        if( (!onlyNoun || mp.isTagOf(POSTag.N) )  
            && !JunkWordDic.contains(mp.getString()) ) 
        {


          // do stemming english word
          if( mp.isTagOf(POSTag.UN) 
              && mp.getCharSet() == CharSetType.ENGLISH ) 
          {
            keyword = new Keyword(mp);
            engStemmer.setCurrent(keyword.getString().toLowerCase());
            engStemmer.stem();
            keyword.setString(engStemmer.getCurrent());
            ret.add(keyword);
          }
          // 사랑하 로 추출된 경우 명사 '사랑'을 색인어로 추출
          else if( mp.isTagOf(POSTag.V) ) {
            String temp = mp.getString();
            int tempLen = temp.length();
            char ch = temp.charAt(tempLen - 1);
            if( tempLen > 2 && (ch == '하' || ch == '되') 
                && VerbNounDic.contains(temp = temp.substring(0, tempLen - 1)))
            {
              keyword = new Keyword(mp);
              keyword.setString(temp);
              keyword.setTag(POSTag.NNG);
              ret.add(keyword);
            }
            // 일반 용언 처리
            else {
              keyword = new Keyword(mp);
              ret.add(keyword);
            }
          }
          // 이외 적합한 경우에 추가
          else if( !mp.isTagOf(POSTag.NP) || true ) {
            keyword = new Keyword(mp);
            ret.add(keyword);
          }
        }
      }
      
      Morpheme mp0 = null, mp1 = null, mp2 = null, mp3 = null;
      for(int i=0, size = mpList.size(), step = 0; i < size; i++) {
        mp0 = mpList.get(i);
        step = 0;
        
        // 복합 명사 추출 --------------
        // 두글자 복합 명사 추출
        if( i + 1 < size
            && mp0.isTagOf(POSTag.NN) 
            && (mp1 = mpList.get(i + 1)).isTagOf(POSTag.NN)
            && mp0.getIndex() + mp0.getString().length() == mp1.getIndex() ) 
        {
          // 세글자 복합 명사 추출
          if( i + 2 < size
              && (mp2 = mpList.get(i + 2)).isTagOf(POSTag.NN)
              && mp1.getIndex() + mp1.getString().length() == mp2.getIndex() )
          {
            // 네글자 복합명사 추출
            if( i + 3 < size 
                && (mp3 = mpList.get(i + 3)).isTagOf(POSTag.NN)
                && mp2.getIndex() + mp2.getString().length() == mp3.getIndex() )
            {
              keyword = new Keyword(mp0);
              keyword.setComposed(true);
              keyword.setString(mp0.getString() + mp1.getString() + mp2.getString() + mp3.getString());
              ret.add(keyword);
              step++;
            } else {
              keyword = new Keyword(mp0);
              keyword.setComposed(true);
              keyword.setString(mp0.getString() + mp1.getString() + mp2.getString());
              ret.add(keyword);
            }
            step++;
          } else {
            keyword = new Keyword(mp0);
            keyword.setComposed(true);
            keyword.setString(mp0.getString() + mp1.getString());
            ret.add(keyword);
          }
          step++;
        }
        i += step;

View Full Code Here

Examples of org.snu.ids.ha.ma.Morpheme

    // 2009-10-17 일단 넣어줌.
    if( stem.length() == 1 && !lastHg.hasJong() && lastHg.cho != 'ㅎ' ) {
      exp = stem;
      if( lastHg.jung == 'ㅏ' ) {
        mCandidateClone = mCandidate.copy();
        mCandidateClone.add(new Morpheme("아", POSTag.ECS));
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.addHavingCond(Condition.AH);
        mCandidateClone.setRealDicLen((byte) exp.length());
        ret.add(mCandidateClone);
      } else if( lastHg.jung == 'ㅓ' ) {
        mCandidateClone = mCandidate.copy();
        mCandidateClone.add(new Morpheme("어", POSTag.ECS));
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.addHavingCond(Condition.AH);
        mCandidateClone.setRealDicLen((byte) exp.length());
        ret.add(mCandidateClone);
      }
    }




    // 겹모음 'ㄶ'의 경우 'ㅎ'을 빼먹고 사용하는 경우가 많으므로 이를 처리해줌
    if( lastCh == '찮' || lastCh == '잖' ) {
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㄴ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      mCandidateClone.decreaseNumOfPrfrdCond();
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.setExp(exp);
      ret.add(mCandidateClone);
    }




    // 과거형 붙여주기
    if( lastCh == '하' ) {
      // 했 -> 하였
      mCandidateClone = mCandidate.copy();
      exp = preStem + "했";
      mCandidateClone.add(new Morpheme("였", POSTag.EPT));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 해 -> 하여
      mCandidateClone = mCandidate.copy();
      exp = preStem + "해";
      mCandidateClone.add(new Morpheme("여", POSTag.ECS));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);
      
      // 종결형
      mCandidateClone = mCandidate.copy();
      exp = preStem + "해";
      mCandidateClone.add(new Morpheme("여", POSTag.EFN));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 형용사는 하지 -> 치 로 줄여질 수 있다.
      if( mCandidate.isTagOf(POSTag.VA | POSTag.VXA) ) {
        mCandidateClone = mCandidate.copy();
        exp = preStem + "치";
        mCandidateClone.add(new Morpheme("지", POSTag.ECS));
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.setRealDicLen((byte) exp.length());
        ret.add(mCandidateClone);
      }
    }
    // '이'로 끝나는 말
    else if( !lastHg.hasJong() && lastHg.jung == 'ㅣ' ) {
      // ㅣ -> ㅣ었->ㅕㅆ
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, 'ㅕ', 'ㅆ');
      mCandidateClone.add(new Morpheme("었", POSTag.EPT));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // ㅣ -> ㅣ어->ㅕ
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, 'ㅕ', ' ');
      mCandidateClone.add(new Morpheme("어", POSTag.ECS));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.MOEUM | Condition.EUMSEONG | Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // ㅆ, ㅏㅆ, ㅐㅆ, ㅕㅆ  결합에 의한 어간 출력
    else if( !lastHg.hasJong() && MO_SET1.contains(lastHg.jung) ) {
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㅆ');
      mCandidateClone.add(new Morpheme("었", POSTag.EPT));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // '르'불규칙
    else if( lastCh == '르' ) {
      // 았
      mCandidateClone = mCandidate.copy();
      mCandidateClone.clearHavingCondition();
      if( preLastCh == '따' ) {
        exp = preStem + "랐";
        mCandidateClone.add(new Morpheme("았", POSTag.EPT));
        mCandidateClone.addHavingCond(Condition.EUT);
      } else if( preLastCh == '푸' ) {
        exp = stem + "렀";
        mCandidateClone.add(new Morpheme("었", POSTag.EPT));
        mCandidateClone.addHavingCond(Condition.EUT);
      } else {
        mo = getMoeum(lastHg, preLastHg);
        exp = stem.substring(0, stemLen - 2)
        + Hangul.combine(preLastHg.cho, preLastHg.jung, 'ㄹ')
        + Hangul.combine(lastHg.cho, mo, 'ㅆ');
        if( mo == 'ㅏ' ) {
          mCandidateClone.add(new Morpheme("았", POSTag.EPT));
        } else {
          mCandidateClone.add(new Morpheme("었", POSTag.EPT));
        }
        mCandidateClone.addHavingCond(Condition.EUT);
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setCandDicLen((byte)exp.length());
      
      ret.add(mCandidateClone);


      // 아
      mCandidateClone = mCandidate.copy();
      mCandidateClone.clearHavingCondition();
      if( preLastCh == '따' ) {
        exp = preStem + "라";
        mCandidateClone.add(new Morpheme("아", POSTag.ECS));
        mCandidateClone.addHavingCond(Condition.AH);
      } else if( preLastCh == '푸' ) {
        exp = stem + "러";
        mCandidateClone.add(new Morpheme("어", POSTag.ECS));
        mCandidateClone.addHavingCond(Condition.AH);
      } else {
        mo = getMoeum(lastHg, preLastHg);
        exp = stem.substring(0, stemLen - 2)
        + Hangul.combine(preLastHg.cho, preLastHg.jung, 'ㄹ')
        + Hangul.combine(lastHg.cho, mo, ' ');
        if( mo == 'ㅏ' ) {
          mCandidateClone.add(new Morpheme("아", POSTag.ECS));
          mCandidateClone.addHavingCond(Condition.AH);
        } else {
          mCandidateClone.add(new Morpheme("어", POSTag.ECS));
          mCandidateClone.addHavingCond(Condition.AH);
        }
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


    }
    // 'ㅡ' 결합에 의한 어간 출력
    else if( !lastHg.hasJong() && lastHg.jung == 'ㅡ' ) {
      // 양성으로 한번 결합
      mo = getMoeum(lastHg, preLastHg);
      mCandidateClone = mCandidate.copy();
      mCandidateClone.clearHavingCondition();
      exp = preStem + Hangul.combine(lastHg.cho, mo, 'ㅆ');
      if( mo == 'ㅏ' ) {
        mCandidateClone.add(new Morpheme("았", POSTag.EPT));
        mCandidateClone.addHavingCond(Condition.EUT);
      } else {
        mCandidateClone.add(new Morpheme("었", POSTag.EPT));
        mCandidateClone.addHavingCond(Condition.EUT);
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // ㅓ, ㅏ
      mCandidateClone = mCandidate.copy();
      mCandidateClone.clearHavingCondition();
      exp = preStem + Hangul.combine(lastHg.cho, mo, ' ');
      if( mo == 'ㅏ' ) {
        mCandidateClone.add(new Morpheme("아", POSTag.ECS));
        mCandidateClone.addHavingCond(Condition.AH);
      } else {
        mCandidateClone.add(new Morpheme("어", POSTag.ECS));
        mCandidateClone.addHavingCond(Condition.AH);
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // ㅜ, ㅗ결합에 의한 어간 출력
    else if( !lastHg.hasJong() && MO_SET2.contains(lastHg.jung) ) {
      // 었, 았
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, getMoeum(lastHg, preLastHg), 'ㅆ');
      if( lastHg.jung == 'ㅜ' ) {
        mCandidateClone.add(new Morpheme("었", POSTag.EPT));
      } else {
        mCandidateClone.add(new Morpheme("았", POSTag.EPT));
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 어, 아
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, getMoeum(lastHg, preLastHg), ' ');
      if( lastHg.jung == 'ㅜ' ) {
        mCandidateClone.add(new Morpheme("어", POSTag.ECS));
      } else {
        mCandidateClone.add(new Morpheme("아", POSTag.ECS));
      }
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // 겹모음 처리
    else if( !lastHg.hasJong() && lastHg.jung != 'ㅚ' ) {
      // 'ㅓ' 결합
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, 'ㅙ', ' ');
      mCandidateClone.add(new Morpheme("어", POSTag.ECS));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // '었' 결합
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, 'ㅙ', 'ㅆ');
      mCandidateClone.add(new Morpheme("었", POSTag.EPT));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }




    // ㅂ 불규칙
    // ㅂ불규칙 활용하는 어간의 마지막 어절
    // '뵙뽑씹업입잡접좁집' 들은 활용 안함~
    if( "갑겁겹곱굽깁깝껍꼽납눕답덥돕둡땁떱랍럽렵롭립맙맵밉볍섭쉽습엽줍쭙춥탑".indexOf(lastCh) > -1 ) {


      // ㅂ탈락된 음절 생성
      char bChar = Hangul.combine(lastHg.cho, lastHg.jung, ' ');


      // 럽은 '러운' 뿐만 아니라 짧게 '런' 등으로도 활용됨
      if( lastCh == '럽' ) {
        mCandidateClone = mCandidate.copy();
        exp = preStem + '런';
        mCandidateClone.add(new Morpheme("ㄴ", POSTag.ETD));
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.decreaseNumOfPrfrdCond();
        mCandidateClone.setRealDicLen((byte)exp.length());
        ret.add(mCandidateClone);
      }


      // 워, 와
      mCandidateClone = mCandidate.copy();
      if( lastHg.jung == 'ㅗ') {
        mo = 'ㅘ';
        mCandidateClone.add(new Morpheme("아", POSTag.ECS));
      } else {
        mo = 'ㅝ';
        mCandidateClone.add(new Morpheme("어", POSTag.ECS));
      }
      exp = preStem + bChar + Hangul.combine('ㅇ', mo, ' ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.AH);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 웠, 왔
      mCandidateClone = mCandidate.copy();
      mCandidateClone.clearHavingCondition();
      if( lastHg.jung == 'ㅗ') {
        mo = 'ㅘ';
        mCandidateClone.add(new Morpheme("았", POSTag.EPT));
      } else {
        mo = 'ㅝ';
        mCandidateClone.add(new Morpheme("었", POSTag.EPT));
      }
      exp = preStem + bChar + Hangul.combine('ㅇ', mo, 'ㅆ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.EUT);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 우
      mCandidateClone = mCandidate.copy();
      exp = preStem + bChar + '우';
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // ㄴ, ㄹ, ㅁ 에 의한 활용
      mCandidateClone = mCandidate.copy();
      exp = preStem + bChar + '운';
      mCandidateClone.add(new Morpheme("ㄴ", POSTag.ETD));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      mCandidateClone = mCandidate.copy();
      mCandidateClone.add(new Morpheme("ㄹ", POSTag.ETD));
      exp = preStem + bChar + '울';
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      mCandidateClone = mCandidate.copy();
      mCandidateClone.add(new Morpheme("ㅁ", POSTag.ETN));
      exp = preStem + bChar + '움';
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // 'ㅅ' 뷸규칙
    else if( "젓짓긋낫붓잇".indexOf(lastCh) > -1 )
    {
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, ' ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.MINUS_SIOT);
      mCandidateClone.decreaseNumOfPrfrdCond();
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // 'ㄷ' 뷸규칙
    else if( lastHg.jong == 'ㄷ' ) {
      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㄹ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.addHavingCond(Condition.MINUS_SIOT);
      mCandidateClone.decreaseNumOfPrfrdCond();
      mCandidateClone.setCandDicLen((byte)exp.length());
      ret.add(mCandidateClone);
    }
    // 그외 처리
    else if( !lastHg.hasJong() || lastHg.jong == 'ㄹ'
      // ㅎ 불규칙 처리
      || lastCh == '맣' || lastCh == '갛' || lastCh == '랗'
      )
    {
      // ㄴ, ㄹ, ㅁ, ㅂ 에 의한 활용
      mCandidateClone = mCandidate.copy();
      mCandidateClone.add(new Morpheme("ㄴ", POSTag.ETD));
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㄴ');
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      mCandidateClone = mCandidate.copy();
      exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㄹ');
      mCandidateClone.add(new Morpheme("ㄹ", POSTag.ETD));
      mCandidateClone.setExp(exp);
      mCandidateClone.setAutoExtd(true);
      mCandidateClone.clearHavingCondition();
      mCandidateClone.initHavingCond(exp);
      mCandidateClone.setRealDicLen((byte)exp.length());
      ret.add(mCandidateClone);


      // 'ㄹ' 불규칙
      if( lastHg.jong == 'ㄹ' ) {
        mCandidateClone = mCandidate.copy();
        exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㄻ');
        mCandidateClone.add(new Morpheme("ㅁ", POSTag.ETN));
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.setRealDicLen((byte)exp.length());
        ret.add(mCandidateClone);


        // ㄹ탈락 현상 처리
        mCandidateClone = mCandidate.copy();
        exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, ' ');
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.addHavingCond(Condition.MINUS_LIEUL);
        mCandidateClone.decreaseNumOfPrfrdCond();
        mCandidateClone.setCandDicLen((byte)exp.length());
        ret.add(mCandidateClone);
      }
      // 'ㅎ' 불규칙
      else if( lastHg.jong == 'ㅎ' ) {
        mCandidateClone = mCandidate.copy();
        exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, ' ');
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);
        mCandidateClone.addHavingCond(Condition.MINUS_HIEUT);
        mCandidateClone.decreaseNumOfPrfrdCond();
        mCandidateClone.setCandDicLen((byte)exp.length());
        ret.add(mCandidateClone);
      } else {
        mCandidateClone = mCandidate.copy();
        mCandidateClone.add(new Morpheme("ㅁ", POSTag.ETN));
        exp = preStem + Hangul.combine(lastHg.cho, lastHg.jung, 'ㅁ');
        mCandidateClone.setExp(exp);
        mCandidateClone.setAutoExtd(true);
        mCandidateClone.clearHavingCondition();
        mCandidateClone.initHavingCond(exp);

View Full Code Here

TOP

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.