VC2005解决UNICODE字符集下CStdioFile的Writestring无法写入中文

在VC2005环境下， 以下代码无法实现使用CStdioFile向文本文件中写入中文（用notepad.exe查看不到写入的中文） CStdioFile file; file.Open(…); file.WriteString(_T("abc你好"));//只能写入abc 在VC2005中的解决办法： 使用setlocale语句设定区域 #include <locale>//头文件 CStdioFile file; file.Open(…); char* old_locale = _strdup( setlocale(LC_CTYPE,NULL) ); setlocale( LC_CTYPE, "chs" );//设定 file.WriteString(_T("abc你好"));//正常写入 setlocale( LC_CTYPE, old_locale ); free( old_locale );//还原区域设定 简化处理可以仅使用语句setlocale( LC_CTYPE, "chs" )。 setlocale: 函数原形为：char *setlocale( int category, const char *locale ); 头文件：<locale.h> 所支持的操作系统为:ANSI, Win 95, Win NT 对于简体中文可以使用如下设置：setlocale( LC_ALL, "chs" ); 为什么一定要调用setlocale呢？ 因为在C/C++语言标准中定义了其运行时的字符集环境为"C"，也就是ASCII字符集的一个子集，那么mbstowcs在工作时会将cstr中所包含的字符串看作是ASCII编码的字符，而不认为是一个包含有chs编码的字符串，所以他会将每一个中文拆成2个ASCII编码进行转换，这样得到的结果就是会形成4个wchar_t的字符组成的串，那么如何才能够让mbstowcs正常工作呢？在调用mbstowcs进行转换之间必须明确的告诉mbstowcs目前cstr串中包含的是chs编码的字符串，通过setlocale( LC_ALL, "chs" )函数调用来完成，需要注意的是这个函数会改变整个应用程序的字符集编码方式，必须要通过重新调用setlocale( LC_ALL, "C" )函数来还原，这样就可以保证mbstowcs在转换时将cstr中的串看作是中文串，并且转换成为2个wchar_t字符，而不是4个。 本地化设置需要具备三个条件： a. 语言代码 (Language Code) b. 国家代码 (Country Code) c. 编码(Encoding) 本地名字可以用下面这些部分来构造： 语言代码_国家代码.编码 比如（zh_CN.UTF-8, en_US等） locale的别名表见 /usr/lib/X11/locale/locale.alias(以Debian GNU/Linux为例) setlocale语言字符串参考 另外还有一种方法就是重新写CStdioFile的派生类CStdioFileEx（网上有）。 //好像C++中没有类能够读些Unicode格式的文本文件，所以我写了下面这个类。用法很简单，大家尝试几下就明白了。 #pragma once class CStdioFileEx: public CStdioFile { public: CStdioFileEx(); CStdioFileEx( LPCTSTR lpszFileName, UINT nOpenFlags ); virtual BOOL Open( LPCTSTR lpszFileName, UINT nOpenFlags, CFileException* pError = NULL ); virtual BOOL ReadString(CString& rString); BOOL ReadWideString(CStringW& rString); BOOL ReadAnsiString(CStringA& rString); virtual void WriteString(LPCTSTR lpsz); void WriteWideString(LPCWSTR lpsz); void WriteAnsiString(LPCSTR lpsz); bool IsUnicodeFormat() {return m_bIsUnicodeText;} unsigned long GetCharCount(); // Additional flag to allow Unicode text format writing enum {modeWriteUnicode = 0x100000}; static bool IsFileUnicode(const CString& sFilePath); protected: UINT PreprocessFlags(const CString& sFilePath, UINT& nOpenFlags); bool m_bIsUnicodeText; }; //。cpp文件 #include "stdafx.h" #include "StdioFileEx.h" //在UCS 编码中有一个叫做"ZERO WIDTH NO-BREAK SPACE"的字符，它的编码是FEFF。而FFFE在UCS中是不存在的字符， //所以不应该出现在实际传输中。UCS规范建议我们在传输字节流前，先传输字符"ZERO WIDTH NO-BREAK SPACE"。这样 //如果接收者收到FEFF，就表明这个字节流是Big-Endian的；如果收到FFFE，就表明这个字节流是Little-Endian的。 //因此字符"ZERO WIDTH NO-BREAK SPACE"又被称作BOM。 //UTF-8不需要BOM来表明字节顺序，但可以用BOM来表明编码方式。字符"ZERO WIDTH NO-BREAK SPACE"的UTF-8编码是 //EF BB BF。所以如果接收者收到以EF BB BF开头的字节流，就知道这是UTF-8编码了。 //Windows就是使用BOM来标记文本文件的编码方式的。 //有些老的浏览器和文本编辑器不支持BOM。 #define UNICODE_BOM 0xFEFF//Unicode "byte order mark" which goes at start of file CStdioFileEx::CStdioFileEx(): CStdioFile() { m_bIsUnicodeText = false; } CStdioFileEx::CStdioFileEx(LPCTSTR lpszFileName,UINT nOpenFlags) :CStdioFile(lpszFileName, PreprocessFlags(lpszFileName, nOpenFlags)) { } BOOL CStdioFileEx::Open(LPCTSTR lpszFileName,UINT nOpenFlags,CFileException* pError /*=NULL*/) { PreprocessFlags(lpszFileName, nOpenFlags); return CStdioFile::Open(lpszFileName, nOpenFlags, pError); } BOOL CStdioFileEx::ReadString(CString& rString) { #ifdef _UNICODE return ReadWideString(rString); #else return ReadAnsiString(rString); #endif } BOOL CStdioFileEx::ReadWideString(CStringW& rString) { _ASSERTE(m_pStream); rString = L""; // empty string without deallocating if(m_bIsUnicodeText) { // If at position 0, discard byte-order mark before reading if(GetPosition() == 0) { wchar_t bom; Read(&bom, sizeof(wchar_t)); } const int nMaxSize = 128; LPWSTR lpsz = rString.GetBuffer(nMaxSize); LPWSTR lpszResult; int nLen = 0; for (;;) { lpszResult = fgetws(lpsz, nMaxSize+1, m_pStream); rString.ReleaseBuffer(); // handle error/eof case if (lpszResult == NULL && !feof(m_pStream)) { Afx_clearerr_s(m_pStream); AfxThrowFileException(CFileException::genericException, _doserrno, m_strFileName); } // if string is read completely or EOF if (lpszResult == NULL || (nLen = (int)lstrlenW(lpsz)) < nMaxSize || lpsz[nLen-1] == '\n') break; nLen = rString.GetLength(); lpsz = rString.GetBuffer(nMaxSize + nLen) + nLen; } //remove crlf if exist. nLen = rString.GetLength(); if (nLen > 1 && rString.Mid(nLen-2) == L"\r\n") { rString.GetBufferSetLength(nLen-2); } return rString.GetLength() > 0; } else { CStringA ansiString; BOOL bRetval = ReadAnsiString(ansiString); //setlocale(LC_ALL, "chs_chn.936");//no need rString = ansiString; return bRetval; } } BOOL CStdioFileEx::ReadAnsiString(CStringA& rString) { _ASSERTE(m_pStream); rString = ""; // empty string without deallocating if(!m_bIsUnicodeText) { const int nMaxSize = 128; LPSTR lpsz = rString.GetBuffer(nMaxSize); LPSTR lpszResult; int nLen = 0; for (;;) { lpszResult = fgets(lpsz, nMaxSize+1, m_pStream); rString.ReleaseBuffer(); // handle error/eof case if (lpszResult == NULL && !feof(m_pStream)) { Afx_clearerr_s(m_pStream); AfxThrowFileException(CFileException::genericException, _doserrno, m_strFileName); } // if string is read completely or EOF if (lpszResult == NULL || (nLen = (int)lstrlenA(lpsz)) < nMaxSize || lpsz[nLen-1] == '\n') break; nLen = rString.GetLength(); lpsz = rString.GetBuffer(nMaxSize + nLen) + nLen; } //remove crlf if exist. nLen = rString.GetLength(); if (nLen > 1 && rString.Mid(nLen-2) == "\r\n") { rString.GetBufferSetLength(nLen-2); } return rString.GetLength() > 0; } else { CStringW wideString; BOOL bRetval = ReadWideString(wideString); //setlocale(LC_ALL, "chs_chn.936");//no need rString = wideString; return bRetval; } } // Purpose: Writes string to file either in Unicode or multibyte, depending on whether the caller specified the // CStdioFileEx::modeWriteUnicode flag. Override of base class function. void CStdioFileEx::WriteString(LPCTSTR lpsz) { #ifdef _UNICODE WriteWideString(lpsz); #else WriteAnsiString(lpsz); #endif } void CStdioFileEx::WriteWideString(LPCWSTR lpsz) { ASSERT(lpsz != NULL); if (lpsz == NULL) { AfxThrowInvalidArgException(); } if(m_bIsUnicodeText) { ASSERT(m_pStream != NULL); // If writing Unicode and at the start of the file, need to write byte mark if(GetPosition() == 0) { wchar_t cBOM = (wchar_t)UNICODE_BOM; CFile::Write(&cBOM, sizeof(wchar_t)); } if (fputws(lpsz, m_pStream) == _TEOF) AfxThrowFileException(CFileException::diskFull, _doserrno, m_strFileName); } else { USES_CONVERSION; WriteAnsiString(CW2A(lpsz)); } } void CStdioFileEx::WriteAnsiString(LPCSTR lpsz) { ASSERT(lpsz != NULL); if (lpsz == NULL) { AfxThrowInvalidArgException(); } if(!m_bIsUnicodeText) { ASSERT(m_pStream != NULL); if (fputs(lpsz, m_pStream) == _TEOF) AfxThrowFileException(CFileException::diskFull, _doserrno, m_strFileName); } else { USES_CONVERSION; WriteWideString(CA2W(lpsz)); } } UINT CStdioFileEx::PreprocessFlags(const CString& sFilePath, UINT& nOpenFlags) { m_bIsUnicodeText = false; // If we have writeUnicode we must have write or writeRead as well if (nOpenFlags & CStdioFileEx::modeWriteUnicode) { ASSERT(nOpenFlags & CFile::modeWrite || nOpenFlags & CFile::modeReadWrite); m_bIsUnicodeText = true; } // If reading in text mode and not creating... else if (nOpenFlags & CFile::typeText && !(nOpenFlags & CFile::modeCreate) && !(nOpenFlags & CFile::modeWrite )) { m_bIsUnicodeText = IsFileUnicode(sFilePath); } //如果要读写Unicode格式的文本文件, 必须切换到typeBinary方式, 因为这会影响fputws/fgetws的工作方式(具体情况参考MSDN)。 if (m_bIsUnicodeText) { nOpenFlags &= ~(CFile::typeText); nOpenFlags |= CFile::typeBinary; } return nOpenFlags; } // Purpose: Determines whether a file is Unicode by reading the first character and detecting // whether it's the Unicode byte marker. bool CStdioFileEx::IsFileUnicode(const CString& sFilePath) { CFile file; wchar_t cFirstChar; CFileException exFile; bool bIsUnicode = false; // Open file in binary mode and read first character if (file.Open(sFilePath, CFile::typeBinary | CFile::modeRead, &exFile)) { // If byte is Unicode byte-order marker, let's say it's Unicode if (file.Read(&cFirstChar, sizeof(wchar_t)) > 0 && cFirstChar == (wchar_t)UNICODE_BOM) { bIsUnicode = true; } file.Close(); } else { // Handle error here if you like } return bIsUnicode; } unsigned long CStdioFileEx::GetCharCount() { int nCharSize; unsigned long nByteCount, nCharCount = 0; if (m_pStream) { // Get size of chars in file nCharSize = m_bIsUnicodeText ? sizeof(wchar_t): sizeof(char); // If Unicode, remove byte order mark from count nByteCount = (unsigned long)GetLength(); if (m_bIsUnicodeText) { nByteCount = nByteCount - sizeof(wchar_t); } // Calc chars nCharCount = (nByteCount / nCharSize); } return nCharCount; }