VC中的UTF8解码算法

为了节省网络带宽，很多网络应用程序都是使用了UTF8编码方式，比如MSN等，UTF8的编码原理其实很简单，这里给一段Windows下面的VC实例。

#define UNKOWN_CHAR       '?'
#define isByteOfOneByteChar(b)    (((b) & 0x80) == 0x00)
#define isHeadOfTwoBytesChar(b)    (((b) & 0xE0) == 0xC0)
#define isHeadOfThreeBytesChar(b)   (((b) & 0xF0) == 0xE0)
#define isComingByteOfMultBytesChar(b)  (((b) & 0xC0)==0x80)
#define isHeadOfUTF8Char(b) /
 isByteOfOneByteChar(b) || /
 isByteOfOneByteChar(b) || /
 isHeadOfThreeBytesChar(b)

Data Data::UTF8Decode(const Data &utf8hex)
{
 int L = utf8hex.length();
 unsigned short* unicode = new unsigned short[L * 2];
 int outi = 0;
 
 for(int i = 0; i < L - 1;) //1-
 {
  BYTE b1 =(BYTE)utf8hex.getChar(i);
  if(isByteOfOneByteChar(b1))
  {
   unicode[outi++] = b1;
   i++;
  }
  else if(isHeadOfTwoBytesChar(b1))
  {
   if(i + 2 > L)
   {
    unicode[outi++] = UNKOWN_CHAR;
    break;
   }
   BYTE b2 =(BYTE)utf8hex.getChar(i + 1);
   if(isComingByteOfMultBytesChar(b2))
   {
    unsigned short tmp = b1 & 0x1F;
    tmp = (tmp << 6) + (b2 & 0x3F);
    unicode[outi++] = tmp;
    i += 2;
   }
   else if(isHeadOfUTF8Char(b2))
   {
    unicode[outi++] = UNKOWN_CHAR;
    i += 1;
   }else
   {
    unicode[outi++] = UNKOWN_CHAR;
    i +=2;
   }
  }
  else if(isHeadOfThreeBytesChar(b1))
  {
   if(i + 2 > L)
   {
    unicode[outi++] = UNKOWN_CHAR;
    break;
   }
   else if(i + 3 > L)
   {
    BYTE b2 =(BYTE)utf8hex.getChar(i + 1);
    if(isByteOfOneByteChar(b2))
    {
     unicode[outi++] = UNKOWN_CHAR;
     i += 1;
     continue;
    }else
    {
     unicode[outi++] = UNKOWN_CHAR;
     break;
    }
   }
   
   BYTE b2 =(BYTE)utf8hex.getChar(i + 1);
   BYTE b3 =(BYTE)utf8hex.getChar(i + 2);
   if(isComingByteOfMultBytesChar(b2) && isComingByteOfMultBytesChar(b3))
   {
    unsigned short tmp = b1 & 0x0F;
    tmp = tmp << 12;
    unsigned short tmp2 = b2 & 0x3F;
    tmp2 = tmp2 << 6;
    unicode[outi++] = tmp + tmp2 + (b3 & 0x3F);
    i +=3;
   }
   else if(isHeadOfUTF8Char(b2))
   {
    unicode[outi++] = UNKOWN_CHAR;
    i += 1;
   }else if(isHeadOfUTF8Char(b3))
   {
    unicode[outi++] = UNKOWN_CHAR;
    i += 2;
   }else
   {
    unicode[outi++] = UNKOWN_CHAR;
    i +=3;
   }
  }else
  {
   unicode[outi++] = UNKOWN_CHAR;
   i += 1;
  }
 }
 unicode[outi] = 0;

 Data ret(outi * 4, true);
#ifdef WIN32
 int count = WideCharToMultiByte(0, 0, unicode, outi, ret.mBuf, outi * 4, NULL, NULL);
 if(count == 0)
  cpLog(LOG_ERR, "Get UTF byte, error number=%lu", GetLastError());
#else
#endif
 if(count > 0)
 {
  ret.mBuf[count] = '/0';
  ret.mLength = count;
 }
 delete [] unicode;

 return ret;
}