VC 编程ANSI环境下读写Unicode文件和将CStdioFile类扩展,读取UNICODE文本文件

长平狐 发布于 2013/01/06 11:24
阅读 734
收藏 0

VC 编程ANSI环境下读写Unicode文件

没有注意到文件编码的不同会产生这么多的问题,在动手以前查询了很多资料,在本博客中收藏了不少先辈的成果,在这里一并表示致敬!
       关于ANSI和Unicode编码的原理在这里也不说了,主要讲下如何读写!
       首先确定你的工程是采用的是什么编码环境,默认是ANSI,不同的字符集读写文件的差别也比较大,我这里只在ANSI环境下做的,下一步在探索在Unicode环境下如何读写!(原先这个没搞懂,读了不少代码发现自己试验都是有误的)。
       在ANSI的字符集下,CString等都是单字节版本的,所以一定要注意。而多要读取的Unicode文件却是双字节的,这里就要转换了,当然在ANSI字符集下,还是用二进制的方式打开Unicode文件,自己判断是否是换行,在转化成ANSI编码。而在写Unicode的时候,先将所字符转化成Unicode编码再写入,而且在写文件之前一定要加上Unicode文件的标识。

下面是读
   CFile mFile(UnicodefilePath,CFile::modeRead);
   byte head[2];
   mFile.Read(head,2); 
   if((head[0]==0xff&&head[1]==0xfe)||(head[0]==0xfe&&head[1]==0xff)   )
   {
    //AfxMessageBox(_T("File is Unicode!"));
    isUnicode = true;
   }
if(isUnicode) mFile.Seek(2,CFile::begin); //0xfffe
wchar_t wch;
wchar_t wstr[300];

CString strvalue ;
hile(mFile.Read((char *)&wch,2)>0)
{
   if(wch==0x000D)  //by line
   {
    //chang to ansi
    int nLen = i;
    char *buf = new char[2*nLen];
    WideCharToMultiByte(CP_ACP, 0, wstr, nLen, buf, 2*nLen, NULL, NULL);
    buf[2*nLen-1] = 0;   //some assertion failed,这个比较重要,小问题可以折腾人啊
   strvalue = buf;
    mFile.Seek(2,CFile::current);  //跳过行开头符号
    i=0;
}
   else
   {
    wstr[i++] = wch;
   }
}

//下面是写
CStdioFile transFile;
transFile.Open(strUnicodeSavepath,CFile::modeCreate|CFile::modeWrite|CFile::typeBinary);
WORD wSignature = 0xFEFF;
transFile.Write(&wSignature, 2);  //Unicode的文件符号
   CHAR *pszAnsi = new TCHAR[strvalue.GetLength()+1];
   _tcscpy(pszAnsi, strvalue);
   WCHAR * szwBuffer = new WCHAR[strvalue.GetLength()+1];
   MultiByteToWideChar(CP_ACP, 0, pszAnsi, -1, szwBuffer, strvalue.GetLength()+1);
   //write to files
   transFile.Write(szwBuffer, lstrlenW(szwBuffer) * sizeof(WCHAR));


当然你可以把你的工程设置成Unicode的字符集,这个时候在Unicode的工程下读取ANSI文件又是一个烦人的事情,讲文件读到CString中的时候,每个单字节的ANSI被转成了双字节,需要自己来处理,后面我再探索探索再来记录。
 

本文来自CSDN博客,转载请标明出处:http://blog.csdn.net/Augusdi/archive/2009/10/15/4677528.aspx

============================================================================

 

将CStdioFile类扩展,读取UNICODE文本文件

 

最近因为要读取SQL脚本文件,用CStdioFile来读取脚本文件,却在程序调试时读取不了文件。
后来看了一下文本文件格式,竟然是UNICODE格式的,原来在导出SQL脚本的时候,选项默认的是UNICODE格式。为了同时支持ANSI和UNICODE两种格式,在Codeproject站点上找到了CStdioFileEx类的代码,但在调试运行过程中发现,在生成UNICODE版本的执行文件时,运行没有错误,但在非UNICODE版本中却出现错误,原来在代码中此部分没有考虑文件读到末尾的情况,经修改,CStdioFileEx类就可以正常使用了,在读取文本文件时,自动识别ANSI和UNICODE两种格式。

实现头文件如下:
#define nUNICODE_BOM      0xFEFF  // Unicode "byte order mark" which goes at start of file
#define sNEWLINE        _T("/r/n") // New line characters
#define sDEFAULT_UNICODE_FILLER_CHAR "#"   // Filler char used when no conversion from Unicode to local code page is possible

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);
 virtual void WriteString( LPCTSTR lpsz );
 bool    IsFileUnicodeText() { return m_bIsUnicodeText; }
 unsigned long GetCharCount();

 // Additional flag to allow Unicode text writing
 static const UINT modeWriteUnicode;

 // static utility functions

 // --------------------------------------------------------------------------------------------
 //
 // CStdioFileEx::GetUnicodeStringFromMultiByteString()
 //
 // --------------------------------------------------------------------------------------------
 // Returns:    bool
 // Parameters: char *  szMultiByteString  (IN) Multi-byte input string
 //     wchar_t*  szUnicodeString  (OUT) Unicode output string
 //     short   nUnicodeBufferSize (IN) Size of Unicode output buffer
 //     UINT   nCodePage    (IN) Code page used to perform conversion
 //                  Default = -1 (Get local code page).
 //
 // Purpose:  Gets a Unicode string from a MultiByte string.
 // Notes:  None.
 // Exceptions: None.
 //
 static bool  GetUnicodeStringFromMultiByteString(char * szMultiByteString,wchar_t* szUnicodeString,
                   short nUnicodeBufferSize,UINT nCodePage=-1);

 // --------------------------------------------------------------------------------------------
 //
 // CStdioFileEx::GetMultiByteStringFromUnicodeString()
 //
 // --------------------------------------------------------------------------------------------
 // Returns:    BOOL
 // Parameters: wchar_t * szUnicodeString   (IN) Unicode input string
 //     char*   szMultiByteString   (OUT) Multibyte output string
 //     short   nMultiByteBufferSize  (IN) Multibyte buffer size
 //     UINT   nCodePage     (IN) Code page used to perform conversion
 //                   Default = -1 (Get local code page).
 //
 // Purpose:  Gets a MultiByte string from a Unicode string.
 // Notes:  .
 // Exceptions: None.
 //
 static BOOL   GetMultiByteStringFromUnicodeString(wchar_t * szUnicodeString,char* szMultiByteString,
                   short nMultiByteBufferSize,UINT nCodePage=-1);


 // --------------------------------------------------------------------------------------------
 //
 // CStdioFileEx::IsFileUnicode()
 //
 // --------------------------------------------------------------------------------------------
 // Returns:    bool
 // Parameters: const CString& sFilePath
 //
 // Purpose:  Determines whether a file is Unicode by reading the first character and detecting
 //     whether it's the Unicode byte marker.
 // Notes:  None.
 // Exceptions: None.
 //
 static bool IsFileUnicode(const CString& sFilePath);


protected:
 UINT ProcessFlags(const CString& sFilePath, UINT& nOpenFlags);

 bool  m_bIsUnicodeText;
 UINT  m_nFlags;
};

 

实现文件如下:
/*static*/ const UINT CStdioFileEx::modeWriteUnicode = 0x20000; // Add this flag to write in Unicode

CStdioFileEx::CStdioFileEx(): CStdioFile()
{
 m_bIsUnicodeText = false;
}

CStdioFileEx::CStdioFileEx(LPCTSTR lpszFileName,UINT nOpenFlags)
 :CStdioFile(lpszFileName, ProcessFlags(lpszFileName, nOpenFlags))
{
}

BOOL CStdioFileEx::Open(LPCTSTR lpszFileName,UINT nOpenFlags,CFileException* pError /*=NULL*/)
{
 // Process any Unicode stuff
 ProcessFlags(lpszFileName, nOpenFlags);

 return CStdioFile::Open(lpszFileName, nOpenFlags, pError);
}

BOOL CStdioFileEx::ReadString(CString& rString)
{
 const int nMAX_LINE_CHARS = 4096;
 BOOL   bReadData;
 LPTSTR  lpsz;
 int   nLen = 0; //, nMultiByteBufferLength = 0, nChars = 0;
 CString  sTemp;
 wchar_t*  pszUnicodeString = NULL;
 char *  pszMultiByteString= NULL;

 // If at position 0, discard byte-order mark before reading
 if (!m_pStream || (GetPosition() == 0 && m_bIsUnicodeText))
 {
  wchar_t cDummy;
//  Read(&cDummy, sizeof(_TCHAR));
  Read(&cDummy, sizeof(wchar_t));
 }

// If compiled for Unicode
#ifdef _UNICODE
 // Do standard stuff -- both ANSI and Unicode cases seem to work OK
 bReadData = CStdioFile::ReadString(rString);
#else

 if (!m_bIsUnicodeText)
 {
  // Do standard stuff -- read ANSI in ANSI
  bReadData = CStdioFile::ReadString(rString);
 }
 else
 {
  pszUnicodeString = new wchar_t[nMAX_LINE_CHARS];
  pszMultiByteString= new char[nMAX_LINE_CHARS]; 

  // Read as Unicode, convert to ANSI

  if(fgetws(pszUnicodeString, nMAX_LINE_CHARS, m_pStream)==NULL)
  { 
   bReadData=FALSE;
  }
  else
  {
   bReadData=TRUE;
   if (GetMultiByteStringFromUnicodeString(pszUnicodeString, pszMultiByteString, nMAX_LINE_CHARS))
   {
    rString = (CString)pszMultiByteString;
   }

   if (pszUnicodeString)
   {
    delete pszUnicodeString;
   }

   if (pszMultiByteString)
   {
    delete pszMultiByteString;
   }
  }
 }
#endif

 // Then remove end-of-line character if in Unicode text mode
 if (bReadData)
 {
  // Copied from FileTxt.cpp but adapted to Unicode and then adapted for end-of-line being just '/r'.
  nLen = rString.GetLength();
  if (nLen > 1 && rString.Mid(nLen-2) == sNEWLINE)
  {
   rString.GetBufferSetLength(nLen-2);
  }
  else
  {
   lpsz = rString.GetBuffer(0);
   if (nLen != 0 && (lpsz[nLen-1] == _T('/r') || lpsz[nLen-1] == _T('/n')))
   {
    rString.GetBufferSetLength(nLen-1);
   }
  }
 }

 return bReadData;
}

// --------------------------------------------------------------------------------------------
//
// CStdioFileEx::WriteString()
//
// --------------------------------------------------------------------------------------------
// Returns:    void
// Parameters: LPCTSTR lpsz
//
// 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.
// Notes:  If writing in Unicode we need to:
//      a) Write the Byte-order-mark at the beginning of the file
//      b) Write all strings in byte-mode
//     - If we were compiled in Unicode, we need to convert Unicode to multibyte if
//      we want to write in multibyte
//     - If we were compiled in multi-byte, we need to convert multibyte to Unicode if
//      we want to write in Unicode.
// Exceptions: None.
//
void CStdioFileEx::WriteString(LPCTSTR lpsz)
{
 // If writing Unicode and at the start of the file, need to write byte mark
 if (m_nFlags & CStdioFileEx::modeWriteUnicode)
 {
  // If at position 0, write byte-order mark before writing anything else
  if (!m_pStream || GetPosition() == 0)
  {
   wchar_t cBOM = (wchar_t)nUNICODE_BOM;
   CFile::Write(&cBOM, sizeof(wchar_t));
  }
 }

// If compiled in Unicode...
#ifdef _UNICODE

 // If writing Unicode, no conversion needed
 if (m_nFlags & CStdioFileEx::modeWriteUnicode)
 {
  // Write in byte mode
  CFile::Write(lpsz, lstrlen(lpsz) * sizeof(wchar_t));
 }
 // Else if we don't want to write Unicode, need to convert
 else
 {
  int  nChars = lstrlen(lpsz) + 1;    // Why plus 1? Because yes
  int  nBufferSize = nChars * sizeof(char);
  wchar_t* pszUnicodeString = new wchar_t[nChars];
  char * pszMultiByteString= new char[nChars]; 

  // Copy string to Unicode buffer
  lstrcpy(pszUnicodeString, lpsz);

  // Get multibyte string
  if (GetMultiByteStringFromUnicodeString(pszUnicodeString, pszMultiByteString, nBufferSize, GetACP()))
  {
   // Do standard write
   CFile::Write((const void*)pszMultiByteString, lstrlen(lpsz));
  }

  if (pszUnicodeString && pszMultiByteString)
  {
   delete [] pszUnicodeString;
   delete [] pszMultiByteString;
  }
 }
// Else if *not* compiled in Unicode
#else
 // If writing Unicode, need to convert
 if (m_nFlags & CStdioFileEx::modeWriteUnicode)
 {
  int  nChars = lstrlen(lpsz) + 1;  // Why plus 1? Because yes
  int  nBufferSize = nChars * sizeof(wchar_t);
  wchar_t* pszUnicodeString = new wchar_t[nChars];
  char * pszMultiByteString= new char[nChars];

  // Copy string to multibyte buffer
  lstrcpy(pszMultiByteString, lpsz);

  if (GetUnicodeStringFromMultiByteString(pszMultiByteString, pszUnicodeString, nBufferSize, GetACP()))
  {
   // Write in byte mode
   CFile::Write(pszUnicodeString, lstrlen(lpsz) * sizeof(wchar_t));
  }
  else
  {
   ASSERT(false);
  }

  if (pszUnicodeString && pszMultiByteString)
  {
   delete [] pszUnicodeString;
   delete [] pszMultiByteString;
  }
 }
 // Else if we don't want to write Unicode, no conversion needed
 else
 {
  // Do standard stuff
  CStdioFile::WriteString(lpsz);
 }

#endif
}

UINT CStdioFileEx::ProcessFlags(const CString& sFilePath, UINT& nOpenFlags)
{
 m_bIsUnicodeText = false;

 // If we have writeUnicode we must have write or writeRead as well
#ifdef _DEBUG
 if (nOpenFlags & CStdioFileEx::modeWriteUnicode)
 {
  ASSERT(nOpenFlags & CFile::modeWrite || nOpenFlags & CFile::modeReadWrite);
 }
#endif

 // If reading in text mode and not creating...
 if (nOpenFlags & CFile::typeText && !(m_nFlags & CFile::modeCreate) && !(m_nFlags & CFile::modeWrite ))
 {
  m_bIsUnicodeText = IsFileUnicode(sFilePath);

  // If it's Unicode, switch to binary mode
  if (m_bIsUnicodeText)
  {
   nOpenFlags ^= CFile::typeText;
   nOpenFlags |= CFile::typeBinary;
  }
 }

 m_nFlags = nOpenFlags;

 return nOpenFlags;
}

// --------------------------------------------------------------------------------------------
//
// CStdioFileEx::IsFileUnicode()
//
// --------------------------------------------------------------------------------------------
// Returns:    bool
// Parameters: const CString& sFilePath
//
// Purpose:  Determines whether a file is Unicode by reading the first character and detecting
//     whether it's the Unicode byte marker.
// Notes:  None.
// Exceptions: None.
//
/*static*/ bool CStdioFileEx::IsFileUnicode(const CString& sFilePath)
{
 CFile    file;
 bool    bIsUnicode = false;
 wchar_t   cFirstChar;
 CFileException exFile;

 // 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)nUNICODE_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;
}

// --------------------------------------------------------------------------------------------
//
// CStdioFileEx::GetUnicodeStringFromMultiByteString()
//
// --------------------------------------------------------------------------------------------
// Returns:    bool
// Parameters: char *  szMultiByteString  (IN) Multi-byte input string
//     wchar_t*  szUnicodeString  (OUT) Unicode outputstring
//     short   nUnicodeBufferSize (IN) Size of Unicode output buffer
//     UINT   nCodePage    (IN) Code page used to perform conversion
//                  Default = -1 (Get local code page).
//
// Purpose:  Gets a Unicode string from a MultiByte string.
// Notes:  None.
// Exceptions: None.
//
bool CStdioFileEx::GetUnicodeStringFromMultiByteString(char * szMultiByteString, wchar_t* szUnicodeString, short nUnicodeBufferSize, UINT nCodePage)
{
 bool  bOK = true;
 int  nReturn = 0;
 CString sErrorMsg;
 
 if (szUnicodeString && szMultiByteString)
 {
  // If no code page specified, take default for system
  if (nCodePage == -1)
  {
   nCodePage = GetACP();
  }

  try
  {
   nReturn = MultiByteToWideChar(nCodePage,MB_PRECOMPOSED,szMultiByteString,-1,szUnicodeString,nUnicodeBufferSize);

   if (nReturn == 0)
   {
    bOK = false;
   }
  }
  catch(...)
  {
   bOK = false;
  }
 }
 else
 {
  bOK = false;
 }

 ASSERT(bOK);
 return bOK;
}

// --------------------------------------------------------------------------------------------
//
// CStdioFileEx::GetMultiByteStringFromUnicodeString()
//
// --------------------------------------------------------------------------------------------
// Returns:    BOOL
// Parameters: wchar_t * szUnicodeString   (IN) Unicode input string
//     char*   szMultiByteString   (OUT) Multibyte output string
//     short   nMultiByteBufferSize  (IN) Multibyte buffer size
//     UINT   nCodePage     (IN) Code page used to perform conversion
//                   Default = -1 (Get local code page).
//
// Purpose:  Gets a MultiByte string from a Unicode string
// Notes:  None.
// Exceptions: None.
//
BOOL CStdioFileEx::GetMultiByteStringFromUnicodeString(wchar_t * szUnicodeString, char* szMultiByteString,
                   short nMultiByteBufferSize, UINT nCodePage)
{
 BOOL   bUsedDefChar = FALSE;
 BOOL   bGotIt = FALSE;

 if (szUnicodeString && szMultiByteString)
 {
  // If no code page specified, take default for system
  if (nCodePage == -1)
  {
   nCodePage = GetACP();
  }

  try
  {
   bGotIt = WideCharToMultiByte(nCodePage, WC_COMPOSITECHECK | WC_SEPCHARS,
       szUnicodeString,-1, szMultiByteString, nMultiByteBufferSize, sDEFAULT_UNICODE_FILLER_CHAR, &bUsedDefChar);
  }
  catch(...)
  {
   TRACE(_T("Controlled exception in WideCharToMultiByte!/n"));
  }
 }

 return bGotIt;
}

 

本文来自CSDN博客,转载请标明出处:http://blog.csdn.net/Augusdi/archive/2009/10/15/4677520.aspx


原文链接:http://blog.csdn.net/sunboy_2050/article/details/5019900
加载中
返回顶部
顶部