reading unicode characters from standard input.

getuchar()

char32_t getuchar()
{
   int inp = getchar();

   if (inp == EOF)
      return 0;

   if (!(inp & 0x80))
      return (char32_t)inp;

   int unicode_bytes = 0;

   for (int i = 7; i > 0; --i)
   {
      if (~inp & (1 << i))
         break;
      unicode_bytes++;
   }

   unsigned char tmp[4] = {0, 0, 0, 0};

   tmp[unicode_bytes - 1] = inp;

   char32_t rval;

   for (int i = unicode_bytes -2; i >= 0; --i)
   {
      tmp[i] = getchar();
   }

   rval = (tmp[0] & 0x3f);
   for (int i = 1; i < unicode_bytes; ++i)
   {
      if (i == unicode_bytes-1)
      {
         rval |= (0x1f & tmp[i]) << (6 * i);
      }
      else 
         rval |= (0x3f & tmp[i]) << (6 * i);

   }
   return rval;
}