첫눈 님이 쓰신 글 :
:
https://anal.tistory.com/entry/Delphi에서-한글관련-작업의-사용방법
:
: 위 링크에 있는 소스코드입니다.
:
:
: const
:
: ChoSungTbl : String = 'ㄱㄲㄴㄷㄸㄹㅁㅂㅃㅅㅆㅇㅈㅉㅊㅋㅌㅍㅎ';
: JungSungTbl : String = 'ㅏㅐㅑㅒㅓㅔㅕㅖㅗㅘㅙㅚㅛㅜㅝㅞㅟㅠㅡㅢㅣ';
: JongSungTbl : String = ' ㄱㄲㄳㄴㄵㄶㄷㄹㄺㄻㄼㄽㄾㄿㅀㅁㅂㅄㅅㅆㅇㅈㅊㅋㅌㅍㅎ';
:
: EngChosung : String = 'r R s e E f a q Q t T d w W c z x v g ';
: EngJungsung : String = 'k o i O j p u P h hkhohly n njnpnlb m mll ';
: EngJongsung : String = ' r R rts swsge f frfafqftfxfvfga q qtt T d w c z x v g ';
:
: UniCodeHangeulBase = $AC00;
: UniCodeHangeulLast = $D79F;
:
:
: //한글관련
: function ConvEngToHan(src:String):String;
: function CutFirstHangul(src:String):String;
: function DecodingEngToHan(src:String):String;
: function DivChoJungJong(src:String):String;
: function CharKeyCheck(c : char):boolean;
: function MoumCheck(c : char):boolean;
: function HanDiv(const Han: PChar; Han3: PChar): Boolean;
: function HanDivPas(const Src: String;Check:integer): String;
: Function GetHanType(const Src: string): Boolean;
:
:
: ////////////////////////////////////////////////////////////////////////////////
: function MoumCheck(c : char):boolean;
: Begin
: Result := c in ['k','o','i','O','j','p','u','P','h','y','n','b','m','l'];
: end;
:
: function MarkCheck(c : char):boolean;
: Begin
: Result := c in ['@','#','$','%','^','&','*','(',')','_','-',' ','+','=','{','}','[',']','|','\','<','>','.',',','/','?',':',';','"',''''];
: end;
:
: function CharKeyCheck(c : char):boolean;
: Begin
: Result := c in ['A'..'Z','a'..'z'];
: end;
:
: function DivChoJungJong(src:String):String;
: var i, j, l : integer;
: Tmp, Res : String;
: Begin
: l := Length(src);
: If (l < 2) or (not CharKeyCheck(src[1])) Then Begin
: Result := Src[1];
: Exit;
: end;
: Tmp := src + ' ';
: {초성}
: i := Pos(Copy(Tmp,1,1),EngChosung);
: If i > 0 Then Begin
: res := Copy(tmp,1,1) + ' ';
: j := 2;
: End Else Begin
: Result := Tmp[1];
: Exit;
: End;
: {중성}
: If (not CharKeyCheck(tmp[j])) or (not MoumCheck(tmp[j]))Then Begin
: Result := Res;
: Exit;
: End Else Begin
: i := pos(Copy(tmp,j,2),EngJungsung);
: If i > 0 Then Begin
: res := res + Copy(tmp,j,2);
: j := j + 2;
: End Else Begin
: res := res + tmp[j] + ' ';
: j := j + 1;
: End;
: End;
: {종성}
: If (not (tmp[j] in [' ','A'..'Z','a'..'z'])) or (MoumCheck(tmp[j]))Then Begin
: Result := res + ' ';
: exit;
: End Else Begin
: i := pos(Copy(tmp,j,2),EngJongsung);
: If i > 0 Then Begin res := res + Copy(tmp,j,2);
: End Else Begin res := res + tmp[j] + ' ' ;
: End;
: End;
: Result := res;
: end;
:
: function DecodingEngToHan(src:String):String;
: var UniCode : Integer;
: ChoSung , JungSung , JongSung : Integer;
: ChoSungPos, JungSungPos, JongSungPos: Integer;
: Tmp : String;
: Tmp2 : Array[0..2] of Char;
: Begin
: Result := '';
: Tmp := DivChoJungJong(Src);
: If Length(tmp) = 1 Then Begin
: ChoSungPos := Pos(Tmp+' ', EngChoSung );
: JungSungPos := Pos(Tmp+' ', EngJungSung);
: JongSungPos := Pos(Tmp+' ', EngJongSung);
: If ChoSungPos > 0 Then Result := Copy(ChoSungTbl ,ChoSungPos ,2)
: Else If JungSungPos > 0 Then Result := Copy(JungSungTbl,JungSungPos,2)
: Else If JongSungPos > 0 Then Result := Copy(JongSungTbl,JongSungPos,2)
: Else Result := Tmp;
: Exit;
: End Else If Length(Tmp) <> 6 Then Begin
: Result := Tmp;
: Exit;
: End;
: ChoSungPos := Pos(Copy(tmp, 1, 2), EngChoSung );
: JungSungPos := Pos(Copy(tmp, 3, 2), EngJungSung);
: JongSungPos := Pos(Copy(tmp, 5, 2), EngJongSung);
: //
: If (ChoSungPos And JungSungPos And JongSungPos) = 0 Then Exit;
: ChoSung := (ChoSungPos - 1) Div 2;
: JungSung := (JungSungPos - 1) Div 2;
: JongSung := (JongSungPos - 1) Div 2;
: UniCode := UniCodeHangeulBase +(ChoSung * 21 + JungSung) * 28 + JongSung;
: WideCharToMultiByte(CP_ACP, WC_COMPOSITECHECK, @UniCode, 1, @tmp2, 2, nil, nil);
: Result := tmp2[0]+tmp2[1];
: end;
:
: function CutFirstHangul(src:String):String;
: var l : integer;
: Begin
: l := Length(src);
: If l < 2 Then Result := src
: Else If (Length(DivChoJungJong(src)) <> 6 ) Then
: Result := src[1]
: Else If (l < 3) Then Result := src
: Else If (Length(DivChoJungJong(Copy(src,1,2))) = 6 ) And (not CharKeyCheck(src[3])) Then
: Result := Copy(src,1,2)
: Else If (l < 4) Then Result := src
: Else If (Length(DivChoJungJong(Copy(src,1,2))) = 6 ) And ((not CharKeyCheck(src[3])) or (Length(DivChoJungJong(Copy(src,3,2))) = 6)) Then
: Result := Copy(src,1,2)
: Else If (l < 5) Then Result := src
: Else If (Length(DivChoJungJong(Copy(src,1,3))) = 6 ) And ((not CharKeyCheck(src[4])) or (Length(DivChoJungJong(Copy(src,4,2))) = 6)) Then
: Result := Copy(src,1,3)
: Else If (l < 6) Then Result := src
: Else If (Length(DivChoJungJong(Copy(src,1,4))) = 6 ) And ((not CharKeyCheck(src[5])) Or (Length(DivChoJungJong(Copy(src,5,2))) = 6)) Then
: Result := Copy(src,1,4)
: Else Result := Copy(src,1,5);
: end;
:
: function ConvEngToHan(src:String):String;
: var s,r,t : String;
: Begin
: s := Src;
: r := '';
: If Trim(Src)<>'' Then Begin
: Repeat
: t := CutFirstHangul(s) ;
: r := r + DecodingEngToHan(t);
: Delete(s,1,Length(t));
: Until s = '';
: End;
: Result := r;
: end;
:
: function HanDiv(const Han: PChar; Han3: PChar): Boolean;
: var UniCode: Word;
: ChoSung, JungSung, JongSung: Integer;
: begin
: Result := False;
: MultiByteToWideChar(CP_ACP, MB_PRECOMPOSED, Han, 2, @UniCode, 1);
: if (UniCode < UniCodeHangeulBase) or
: (UniCode > UniCodeHangeulLast) then Exit;
: UniCode := UniCode - UniCodeHangeulBase;
: ChoSung := UniCode div (21 * 28);
: UniCode := UniCode mod (21 * 28);
: JungSung := UniCode div 28;
: UniCode := UniCode mod 28;
: JongSung := UniCode;
: StrLCopy(Han3, Pchar(ChoSungTbl ) + ChoSung * 2, 2);
: StrLCopy(Han3 + 2, Pchar(JungSungTbl) + JungSung * 2, 2);
: StrLCopy(Han3 + 4, Pchar(JongSungTbl) + JongSung * 2, 2);
: Result := True;
: end;
:
: function HanDivPas(const Src: String;Check:integer): String;
: var Buff: array[0..6] of Char;
: B,C:String ;
: i :Integer;
: begin
: B:='';
: C:='';
: For i:=1 To Length(Src) Do Begin
: B:=Copy(Src,((i-1)*2)+1,2);
: if Length(Trim(B))=2 then begin
: if HanDiv(PChar(B), Buff) then begin
: Buff[6]:=#0;
: If Check=1 Then C :=C + Copy(String(Buff),1,2)
: Else If Check=2 Then C :=C + Copy(String(Buff),3,2)
: Else If Check=3 Then C :=C + Copy(String(Buff),5,2);
: end;
: end;
: End;
: HanDivPas:=C;
: end;
:
: Function GetHanType(const Src: string): Boolean;
: var Len, Hi, Lo: Integer;
: begin
: Result := False;
: if Length(Src) < 2 then Exit;
: Len := Length(Src);
: Hi := Ord(Src[Len - 1]);
: Lo := Ord(Src[Len]);
: //if ($A1 > Lo) or ($FE < Lo) then Exit;
: if ($B0 <= Hi) and ($C8 >= Hi) then Result := True //한글
: Else if ($CA <= Hi) and ($FD >= Hi) then Result := False //한자
: Else Result := False; //기타
: end;
:
:
: 이 코드를 DelphiXE2에서 사용하려니 에러가 납니다.
: 에러메세지는
: Incompatible types: 'PAnsiChar' and 'PWideChar'
: 이렇게 나오는데요.
:
: 수정 부탁드립니다.
:
: 공부가 잛아 죄송합니다.
: 공부 열심히 하겠습니다.
답변:
델파이로 구현되어 있는 위의 코드는 권장하고 싶지 않네요.
1.
한글 오토마타 알고리즘이야 뻔한건데, 코드가 너무 장황하게 작성되어 있고.
2.
카피가 너무 빈번하게 일어나므로 런타임 효율성이 매우 떨어지고.
C++17 테크닉 이용해서 구현하면...
#include < iostream >
#include < vector >
#include < string >
#include < unordered_map >
using namespace std;
const static vector< int > ivc[] = {
{ 12593, 12594, 12596, 12599, 12600, 12601, 12609, 12610, 12611, 12613, 12614, 12615, 12616, 12617, 12618, 12619, 12620, 12621, 12622 },
{ 12623, 12624,12625,12626,12627,12628,12629,12630,12631,12634,12633,12634,12635,12636,12637,12638,12639,12640,12641,12642,12643},
{ 12593,12594,12595,12596, 12597,12598,12599,12601,12602,12603,12604,12605,12606,12607,12608,12609,12610,12612,12613,12614,12615,12616,12618,12619,12620,12621,12622 } };
const static unordered_map< int, int > map = {
{0x740066, 12605}, {0x780066, 12606}, {0x760066, 12607}, {0x760066, 12608}, {0x740072, 12595}, {0x740071, 12612}, {0x6B0068, 12632},
{0x6F0068, 12633}, {0x6C0068, 12634}, {0x6A006E, 12637}, {0x70006E, 12638}, {0x6C006E, 12639},
{0x72, 12593}, {0x52, 12594}, {0x73, 12596}, {0x65, 12599}, {0x66, 12601}, {0x61, 12609}, {0x71, 12610}, {0x45, 12600},
{0x51, 12611}, {0x74, 12613}, {0x54, 12614}, {0x64, 12615}, {0x77, 12616}, {0x57, 12617}, {0x63, 12618}, {0x7A, 12619},
{0x78, 12620}, {0x76, 12621}, {0x67, 12622}, {0x6B, 12623}, {0x6F, 12624}, {0x4F, 12626}, {0x69, 12625}, {0x6A, 12627},
{0x70, 12628}, {0x50, 12630}, {0x75, 12629}, {0x68, 12631}, {0x79, 12635}, {0x6E, 12636}, {0x62, 12640}, {0x6C, 12643},
{0x6D, 12641}, {0x6C006D, 12642}, {0x770073, 12597}, {0x670073, 12598}, {0x720066, 12602}, {0x610066, 12603}, {0x710066, 12604},
};
wstring ConvEng2Kor(const wstring&& in)
{
if (!in.size()) return L"";
static auto cnv = [&](size_t i, int n, int& l, bool c1 = false)
{
if (i >= in.size())
return 0;
static auto map2i = [](const int& _) {
if (auto it = map.find(_); it == map.end()) return 0;
else return it->second;
};
auto tc = 0;
return ([&](const auto& _) {
_ && !c1 && i + 1 < in.size() ?
(tc = map2i(*(int*)in.data()), l = 2) : 0;
!tc ? (tc = map2i(in[i]), l = 1) : 0;
}(n), [](const auto& v, auto e,int tc = 0) {
if (auto _ = ::find(v.begin(), v.end(), e); _ == v.end()) return -1;
else return ::distance(v.begin(), _);
}(ivc[n], tc) != -1 ? tc : 0);
};
wstring ret;
for (int i = 0, c1, c2, c3, l1, l2; i < (int)in.size();) {
if (!(c1 = cnv(i, 0, l1))) {
if (!(c2 = cnv(i, 1, l1))) {
ret += in[i++];
continue;
}
ret += c2, i += l1;
continue;
}
if (cnv(i + 1, 0, l2)) {
if (cnv(i + 2, 1, l2)) {
ret += c1, i += l1;
continue;
}
if ((c3 = cnv(i, 2, l2), l2 == 2)) {
ret += c3, i += l2;
continue;
}
else {
ret += c1, i += l1;
continue;
}
}
if ([&ret, c1](const auto _) { return _ ? (ret += c1, 1) : 0; }(!(c2 = cnv(++i, 1, l1)))
) continue;
static auto cvt = [](const std::vector< int > & v, int e) {
if (auto _ = ::find(v.begin(), v.end(), e); _ != v.end()) return ::distance(v.begin(), _);
return -1;
};
static auto cv2 = [&]() { return cvt(ivc[1], c2) * 0x1C; };
static auto cvi = [&]() { return (c1 ? 0xAC00 + cvt(ivc[0], c1) * 0x24C : 0); };
if (!(c3 = cnv(i += l1, 2, l1))) {
ret += [&](int _) { return (c2 ? _ += cv2(): 0, _); }(cvi());
continue;
}
if (l1 == 2) {
if (cnv(i + 2, 0, l2)) {
ret += [&](auto _) { return (c2 ? _ += cv2(): 0, c3 ? _ += cvt(ivc[2], c3) : 0, i += l1, _); }(cvi());
continue;
}
ret += [&](auto _) { return (c2 ? _ += cv2(): 0, c3 = cnv(i, 2, l1, true), c3 ? _ += cvt(ivc[2], c3) + 1 : 0, i += l1, _); }(cvi());
continue;
}
else {
if (cnv(i + 1, 1, l2)) {
ret += [&](auto _) { return (c2 ? _ += cv2(): 0, _); }(cvi());
continue;
}
ret += [&](auto _) { return (c2 ? _ += cv2(): 0, c3 ? _ += cvt(ivc[2], c3) + 1 : 0, i += l1, _); }(cvi());
continue;
}
}
return ret;
}
