Function oneUTF32toUTF8

engine/source/string/unicode.cc:496–541

-----------------------------------------------------------------------------

//-----------------------------------------------------------------------------
const U32 oneUTF32toUTF8(const UTF32 codepoint, UTF8 *threeByteCodeunitBuf)
{
   PROFILE_START(oneUTF32toUTF8);
   U32 bytecount = 0;
   UTF8 *buf;
   U32 working = codepoint;
   buf = threeByteCodeunitBuf;

   //-----------------
   if(isSurrogateRange(working))  // found an illegal codepoint!
      working = kReplacementChar;
      //return oneUTF32toUTF8(kReplacementChar, threeByteCodeunitBuf);
   
   if(isAboveBMP(working))        // these are legal, we just dont want to deal with them.
      working = kReplacementChar;
      //return oneUTF32toUTF8(kReplacementChar, threeByteCodeunitBuf);

   //-----------------
   if( working < (1 << 7))        // codeable in 7 bits
      bytecount = 1;
   else if( working < (1 << 11))  // codeable in 11 bits
      bytecount = 2;
   else if( working < (1 << 16))  // codeable in 16 bits
      bytecount = 3;

   AssertISV( bytecount > 0, "Error converting to UTF-8 in oneUTF32toUTF8(). isAboveBMP() should have caught this!");

   //-----------------
   U8  mask = byteMask8LUT[0];            // 0011 1111
   U8  marker = ( ~mask << 1);            // 1000 0000
   
   // Process the low order bytes, shifting the codepoint down 6 each pass.
   for( int i = bytecount-1; i > 0; i--)
   {
      threeByteCodeunitBuf[i] = marker | (working & mask); 
      working >>= 6;
   }

   // Process the 1st byte. filter based on the # of expected bytes.
   mask = byteMask8LUT[bytecount];
   marker = ( ~mask << 1 );
   threeByteCodeunitBuf[0] = marker | (working & mask);
   
   PROFILE_END();
   return bytecount;
}

//-----------------------------------------------------------------------------
const U32 dStrlen(const UTF16 *unistring)