NaverSearcher/NaverSearcher/StringParser.cs

using System;
using System.Collections.Generic;
using System.Text;

namespace NaverSearcher
{
    class StringParser
    {
        ///< 초성 테이블
        private static readonly char[] wcHead =
        {
            'ㄱ', 'ㄲ', 'ㄴ', 'ㄷ',
            'ㄸ', 'ㄹ', 'ㅁ', 'ㅂ',
            'ㅃ', 'ㅅ', 'ㅆ', 'ㅇ',
            'ㅈ', 'ㅉ', 'ㅊ', 'ㅋ',
            'ㅌ', 'ㅍ', 'ㅎ'
        };

        ///< 중성 테이블
        private static readonly char[] wcMid =
        {
            'ㅏ', 'ㅐ', 'ㅑ', 'ㅒ',
            'ㅓ', 'ㅔ', 'ㅕ', 'ㅖ',
            'ㅗ', 'ㅘ', 'ㅙ', 'ㅚ',
            'ㅛ', 'ㅜ', 'ㅝ', 'ㅞ',
            'ㅟ', 'ㅠ', 'ㅡ', 'ㅢ', 'ㅣ'
        };

        ///< 중성 합성 테이블
        private static readonly char[,] wcMidMix =
        {
            {'ㅗ', 'ㅏ', 'ㅘ'},
            {'ㅗ', 'ㅐ', 'ㅙ'},
            { 'ㅗ', 'ㅑ', 'ㅚ'},
            { 'ㅜ', 'ㅓ', 'ㅝ'},
            { 'ㅜ', 'ㅔ', 'ㅞ'},
            { 'ㅜ', 'ㅣ', 'ㅟ'},
            { 'ㅡ', 'ㅣ', 'ㅢ'},
        };

        ///< 종성 테이블
        private static readonly char[] wcTail =
        {
            ' ', 'ㄱ', 'ㄲ', 'ㄳ',
            'ㄴ', 'ㄵ', 'ㄶ', 'ㄷ',
            'ㄹ', 'ㄺ', 'ㄻ', 'ㄼ',
            'ㄽ', 'ㄾ', 'ㄿ', 'ㅀ',
            'ㅁ', 'ㅂ', 'ㅄ', 'ㅅ',
            'ㅆ', 'ㅇ', 'ㅈ', 'ㅊ',
            'ㅋ', 'ㅌ', 'ㅍ', 'ㅎ'
        };

        ///< 중성 합성 테이블
        private static readonly char[,] wcTailMix =
        {
            { 'ㄱ', 'ㅅ', 'ㄳ'},
            { 'ㄴ', 'ㅈ', 'ㄵ'},
            { 'ㄴ', 'ㅎ', 'ㄶ'},
            { 'ㄹ', 'ㄱ', 'ㄺ'},
            { 'ㄹ', 'ㅁ', 'ㄻ'},
            { 'ㄹ', 'ㅂ', 'ㄼ'},
            { 'ㄹ', 'ㅅ', 'ㄽ'},
            { 'ㄹ', 'ㅌ', 'ㄾ'},
            { 'ㄹ', 'ㅍ', 'ㄿ'},
            { 'ㄹ', 'ㅎ', 'ㅀ'},
            { 'ㅂ', 'ㅅ', 'ㅄ'},
        };

        // c++ -> C#
        private static bool IsMidMix(char wc, ref char pFirst, ref char pSeconde)
        {
            for (int i = 0; i < wcMidMix.GetLength(0); i++)
            {
                if (wcMidMix[i, 2] == wc)
                {
                    pFirst = wcMidMix[i, 0];
                    pSeconde = wcMidMix[i, 1];
                    return true;
                }
            }

            return false;
        }

        // C++ -> C#
        private static bool IsTailMix(char wc, ref char pFirst, ref char pSeconde)
        {
            for (int i = 0; i < wcTailMix.GetLength(0); i++)
            {
                if (wcTailMix[i, 2] == wc)
                {
                    pFirst = wcTailMix[i, 0];
                    pSeconde = wcTailMix[i, 1];
                    return true;
                }
            }

            return false;
        }

        // C++ -> C#
        private static bool IsFusionTailMix(char wcFirst, char wcSeconde, ref char pResult)
        {
            int dIndex = -1;

            for (int i = 0; i < wcTailMix.GetLength(0); i++)
            {
                if (wcTailMix[i, 0] == wcFirst && wcTailMix[i, 1] == wcSeconde)
                {
                    dIndex = i;
                    break;
                }
            }

            if (dIndex == -1)
                return false;

            pResult = wcTailMix[dIndex, 2];
            return true;
        }

        // C++ -> C#
        private static int Get_T_Index(char ch)
        {
            for (int i = 0; i < wcTail.GetLength(0); i++)
            {
                if (wcTail[i] == ch)
                    return i;
            }

            return -1;
        }

        // C++ -> C#
        private static int Get_M_Index(char ch)
        {
            for (int i = 0; i < wcMid.GetLength(0); i++)
            {
                if (wcMid[i] == ch)
                    return i;
            }

            return -1;
        }

        // C++ -> C#
        public static List<string> CharacterMakeHistory(string wstr)
        {
            List<string> m_buffer = new List<string>();

            string strMakeString = string.Empty;
            string strIngInput = string.Empty;

            char wcFirst = '\0';
            char wcSeconde = '\0';

            int dPos = 0;
            char[] str = wstr.ToCharArray();

            while (dPos < str.Length)
            {
                // KR 조합형
                if (str[dPos] >= 'ㄱ' && str[dPos] <= 'ㅣ')
                {
                    if (IsMidMix(str[dPos], ref wcFirst, ref wcSeconde))
                    {
                        strIngInput = strMakeString;
                        strIngInput += String.Format("{0}", wcFirst);

                        m_buffer.Add(strIngInput);
                    }
                    else if (IsTailMix(str[dPos], ref wcFirst, ref wcSeconde))
                    {

                        strIngInput = strMakeString;
                        strIngInput += String.Format("{0}", wcFirst);

                        m_buffer.Add(strIngInput);
                    }

                    strIngInput = strMakeString;
                    strIngInput += String.Format("{0}", wcFirst);

                    m_buffer.Add(strIngInput);
                }
                // KR 완성형
                else if (str[dPos] >= '가' && str[dPos] <= '힣')
                {
                    // 초성
                    int dHeadPos = Math.Abs(str[dPos] - 0xAC00) / (21 * 28);

                    char wcTemp = '\0';

                    // 앞 글자에 초성 합성이 가능할 경우 체크하여 합성
                    if ((strIngInput.Length >= 1))
                    {
                        wcTemp = strIngInput[strIngInput.Length - 1];

                        // 이전 글자 종성과 현재 글자 초성으로 혼합된 종성을 얻는다.
                        char wcMixTail = ' ';

                        // 합성에 성공한 경우
                        if (IsFusionTailMix(wcTail[Math.Abs(wcTemp - 0xAC00) % 28], wcHead[dHeadPos], ref wcMixTail))
                        {
                            // 이전 글자 종성 제거
                            wcTemp = (char)(wcTemp - Get_T_Index(wcTail[(Math.Abs(wcTemp - 0xAC00)) % 28]));

                            // 이전 글자 새로운 종성 삽입
                            wcTemp = (char)(wcTemp + Get_T_Index(wcMixTail));
                        }
                        // 이전 글자에 종성이 없고 현재 글자의 초성이 정상일 경우
                        else if (Get_T_Index(wcTail[Math.Abs(wcTemp - 0xAC00) % 28]) == 0 && Get_T_Index(wcHead[dHeadPos]) != -1)
                        {
                            wcTemp = (char)(wcTemp + Get_T_Index(wcHead[dHeadPos]));
                        }
                        else
                        {
                            wcTemp = '\0';
                        }
                    }

                    strIngInput = strMakeString;

                    if (wcTemp != '\0')
                    {
                        strIngInput = strIngInput.Substring(0, strIngInput.Length - 1);
                        strIngInput += String.Format("{0}", wcTemp);
                    }
                    else
                    {
                        strIngInput += String.Format("{0}", wcHead[dHeadPos]);
                    }

                    m_buffer.Add(strIngInput);

                    // 중성
                    int dMidPos = Math.Abs(str[dPos] - 0xAC00) % (21 * 28) / 28;

                    if (IsMidMix(wcMid[dMidPos], ref wcFirst, ref wcSeconde))
                    {
                        int szMindex2 = Get_M_Index(wcFirst);
                        char TszM_0 = (char)(0xAC00 + (dHeadPos * 588) + (szMindex2 * 28));

                        strIngInput = strMakeString;
                        strIngInput += String.Format("{0}", TszM_0);

                        m_buffer.Add(strIngInput);
                    }

                    char TszM = (char)(0xAC00 + (dHeadPos * 588) + (dMidPos * 28));

                    strIngInput = strMakeString;
                    strIngInput += String.Format("{0}", TszM);

                    m_buffer.Add(strIngInput);

                    // 종성
                    int dTailPos = Math.Abs(str[dPos] - 0xAC00) % 28;

                    if (wcTail[dTailPos] != ' ')
                    {
                        if (IsTailMix(wcTail[dTailPos], ref wcFirst, ref wcSeconde))
                        {
                            int szTindex2 = Get_T_Index(wcFirst);

                            char TszT_0 = (char)(0xAC00 + (dHeadPos * 588) + (dMidPos * 28) + szTindex2);

                            strIngInput = strMakeString;
                            strIngInput += String.Format("{0}", TszT_0);

                            m_buffer.Add(strIngInput);
                        }

                        char TszT = (char)(0xAC00 + (dHeadPos * 588) + (dMidPos * 28) + dTailPos);

                        strIngInput = strMakeString;
                        strIngInput += String.Format("{0}", TszT);

                        m_buffer.Add(strIngInput);
                    }
                }
                else
                {
                    strIngInput = strMakeString;
                    strIngInput += String.Format("{0}", str[dPos]);

                    m_buffer.Add(strIngInput);
                }

                strMakeString = strIngInput;

                dPos++;
            }

            return m_buffer;
        }
    }
}