Microsoft MVP성태의 닷넷 이야기
글쓴 사람
정성태 (techsharer at outlook.com)
홈페이지
첨부 파일

(시리즈 글이 7개 있습니다.)
.NET Framework: 248. 닷넷에서 지원되는 문자열 인코딩 이름 목록
; https://www.sysnet.pe.kr/2/0/1147

.NET Framework: 368. Encoding 타입의 대체(fallback) 메카니즘
; https://www.sysnet.pe.kr/2/0/1446

.NET Framework: 373. C# 문자열의 인코딩이란?
; https://www.sysnet.pe.kr/2/0/1461

.NET Framework: 701. 한글이 포함된 바이트 배열을 나눈 경우 한글이 깨지지 않도록 다시 조합하는 방법
; https://www.sysnet.pe.kr/2/0/11378

.NET Framework: 702. 한글이 포함된 바이트 배열을 나눈 경우 한글이 깨지지 않도록 다시 조합하는 방법(두 번째 이야기)
; https://www.sysnet.pe.kr/2/0/11381

.NET Framework: 867. C# - Encoding.Default 값을 바꿀 수 있을까요?
; https://www.sysnet.pe.kr/2/0/12037

닷넷: 2190. C# - 닷넷 코어/5+에서 달라지는 System.Text.Encoding 지원
; https://www.sysnet.pe.kr/2/0/13506




한글이 포함된 바이트 배열을 나눈 경우 한글이 깨지지 않도록 다시 조합하는 방법(두 번째 이야기)

지난 글에서 Decoder 타입의 사용법을 알아봤는데요.

한글이 포함된 바이트 배열을 나눈 경우 한글이 깨지지 않도록 다시 조합하는 방법
; https://www.sysnet.pe.kr/2/0/11378

그런데, 실제 통신 환경에서 Decoder 타입을 유지하는 것은 왠지 오버헤드 같습니다. 그보다, 한글 조합이 안 된 상태의 바이트 배열을 다음번 조합으로 미루는 방법을 사용하는 것도 괜찮을 것 같습니다.

Decoder 타입을 보면, GetChars 메서드의 호출로 미완성의 char가 있는 경우 내부의 HasState 속성 값이 true로 바뀌면서 bits 필드로 상태 값을 유지하는데 모두 internal 접근자를 갖고 있습니다. 따라서 Reflection을 이용해 다음과 같이 처리하는 것도 가능합니다.

static void Test()
{
    int bits = 0;

    {
        Decoder utf8Decoder = Encoding.UTF8.GetDecoder();
        SendBuffer(utf8Decoder, buf1);
        bits = GetBits(utf8Decoder);
    }

    {
        Decoder utf8Decoder = Encoding.UTF8.GetDecoder();
        SetBits(utf8Decoder, bits);

        SendBuffer(utf8Decoder, buf2);
    }
}

private static void SetBits(Decoder utf8Decoder, int bits)
{
    FieldInfo fi = GetFieldInfo(utf8Decoder);
    fi.SetValue(utf8Decoder, bits);
}

private static FieldInfo GetFieldInfo(Decoder decoder)
{
    return decoder.GetType().GetField("bits", System.Reflection.BindingFlags.NonPublic | System.Reflection.BindingFlags.Instance);
}

private static int GetBits(Decoder utf8Decoder)
{
    FieldInfo fi = GetFieldInfo(utf8Decoder);
    return (int)fi.GetValue(utf8Decoder);
}

보는 바와 같이 decoder 인스턴스를 유지하지 않아도 bits 값을 보존하는 것만으로 디코딩을 정상적으로 수행할 수 있습니다.




그런데, 왠지 public 접근자가 아닌 것이 좀 걸립니다. 그런 경우라면 그냥 Decoder 인스턴스를 직렬화해서 통신 간에 들고 있으면 됩니다.

byte[] buf = null;
{
    Decoder utf8Decoder = Encoding.UTF8.GetDecoder();
    SendBuffer(utf8Decoder, buf1);

    MemoryStream ms = new MemoryStream();
    BinaryFormatter formatter = new BinaryFormatter();
    formatter.Serialize(ms, utf8Decoder);

    buf = ms.ToArray();
}

{
    BinaryFormatter formatter = new BinaryFormatter();
    Decoder utf8Decoder = formatter.Deserialize(new MemoryStream(buf)) as Decoder;

    SendBuffer(utf8Decoder, buf2);
}

유지해야 할 값이 직접적인 Decoder 인스턴스에서 byte 배열로 바뀌었습니다.




그 외에, 또 다른 방법으로는 디코딩 시에 아직 char로 변환되지 못한 바이트 배열(UTF8의 경우 최대 6개의 바이트)이 있는 경우 다음번 디코딩 과정에 참여하도록 만드는 것입니다. 그러니까 대충 다음과 같은 과정을 거치는 것입니다.

byte[] remains = null;

{
    remains = SendBuffer(buf1);
}

{
    List<byte> buf = new List<byte>();

    if (remains != null)
    {
        buf.AddRange(remains);
    }

    buf.AddRange(buf2);

    SendBuffer(buf.ToArray());
}

이 글을 쓰기 위해 급조한 코드라 별로 마음에 들지 않아 이것은 첨부 파일로 대신합니다. 나중에 실제로 쓰게 되면 그때나 다듬어야 겠습니다. ^^; 끝!




첨부 파일은 위의 예제 순서에 따라 decoder_sample1, decoder_sample2, decoder_sample3 폴더로 나뉩니다.




[이 글에 대해서 여러분들과 의견을 공유하고 싶습니다. 틀리거나 미흡한 부분 또는 의문 사항이 있으시면 언제든 댓글 남겨주십시오.]







[최초 등록일: ]
[최종 수정일: 12/27/2022]

Creative Commons License
이 저작물은 크리에이티브 커먼즈 코리아 저작자표시-비영리-변경금지 2.0 대한민국 라이센스에 따라 이용하실 수 있습니다.
by SeongTae Jeong, mailto:techsharer at outlook.com

비밀번호

댓글 작성자
 




1  2  3  4  5  6  [7]  8  9  10  11  12  13  14  15  ...
NoWriterDateCnt.TitleFile(s)
13804정성태11/7/20247003Linux: 101. eBPF 함수의 인자를 다루는 방법
13803정성태11/7/20246609닷넷: 2309. C# - .NET Core에서 바뀐 DateTime.Ticks의 정밀도
13802정성태11/6/20247031Windows: 269. GetSystemTimeAsFileTime과 GetSystemTimePreciseAsFileTime의 차이점파일 다운로드1
13801정성태11/5/20246852Linux: 100. eBPF의 2가지 방식 - libbcc와 libbpf(CO-RE)
13800정성태11/3/20247829닷넷: 2308. C# - ICU 라이브러리를 활용한 문자열의 대소문자 변환 [2]파일 다운로드1
13799정성태11/2/20245612개발 환경 구성: 732. 모바일 웹 브라우저에서 유니코드 문자가 표시되지 않는 경우
13798정성태11/2/20247265개발 환경 구성: 731. 유니코드 - 출력 예시 및 폰트 찾기
13797정성태11/1/20247244C/C++: 185. C++ - 문자열의 대소문자를 변환하는 transform + std::tolower/toupper 방식의 문제점파일 다운로드1
13796정성태10/31/20246759C/C++: 184. C++ - ICU dll을 이용하는 예제 코드 (Windows)파일 다운로드1
13795정성태10/31/20245956Windows: 268. Windows - 리눅스 환경처럼 공백으로 끝나는 프롬프트 만들기
13794정성태10/30/20246032닷넷: 2307. C# - 윈도우에서 한글(및 유니코드)을 포함한 콘솔 프로그램을 컴파일 및 실행하는 방법
13793정성태10/28/20245994C/C++: 183. C++ - 윈도우에서 한글(및 유니코드)을 포함한 콘솔 프로그램을 컴파일 및 실행하는 방법
13792정성태10/27/20245395Linux: 99. Linux - 프로세스의 실행 파일 경로 확인
13791정성태10/27/20245857Windows: 267. Win32 API의 A(ANSI) 버전은 DBCS를 사용할까요?파일 다운로드1
13790정성태10/27/20245352Linux: 98. Ubuntu 22.04 - 리눅스 커널 빌드 및 업그레이드
13789정성태10/27/20246319Linux: 97. menuconfig에 CONFIG_DEBUG_INFO_BTF, CONFIG_DEBUG_INFO_BTF_MODULES 옵션이 없는 경우
13788정성태10/26/20245109Linux: 96. eBPF (bpf2go) - fentry, fexit를 이용한 트레이스
13787정성태10/26/20246470개발 환경 구성: 730. github - Linux 커널 repo를 윈도우 환경에서 git clone하는 방법 [1]
13786정성태10/26/20246257Windows: 266. Windows - 대소문자 구분이 가능한 파일 시스템
13785정성태10/23/20245601C/C++: 182. 윈도우가 운영하는 2개의 Code Page파일 다운로드1
13784정성태10/23/20246195Linux: 95. eBPF - kprobe를 이용한 트레이스
13783정성태10/23/20245719Linux: 94. eBPF - vmlinux.h 헤더 포함하는 방법 (bpf2go에서 사용)
13782정성태10/23/20245416Linux: 93. Ubuntu 22.04 - 커널 이미지로부터 커널 함수 역어셈블
13781정성태10/22/20245961오류 유형: 930. WSL + eBPF: modprobe: FATAL: Module kheaders not found in directory
13780정성태10/22/20247019Linux: 92. WSL 2 - 커널 이미지로부터 커널 함수 역어셈블
13779정성태10/22/20245598개발 환경 구성: 729. WSL 2 - Mariner VM 커널 이미지 업데이트 방법
1  2  3  4  5  6  [7]  8  9  10  11  12  13  14  15  ...