Thanks. That code does need to be fixed, once we get the time.
Oliver Steinau wrote:
> I have a question concerning the CVTUTF.C file that is on the CD in the
> Unicode 3.0 book. There's a piece of code which I don't think is correct...
>
> Function ConvertUTF8toUTF16 contains the following piece of code (line 210):
> [ch is the result of the conversion so far]:
>
> if (ch <= kMaximumUCS2) {
> *target++ = ch;
> } else if (ch > kMaximumUCS4) {
> *target++ = kReplacementCharacter;
> } else {
> if (target + 1 >= targetEnd) {
> result = targetExhausted; break;
> };
> ch -= halfBase;
> *target++ = (ch >> halfShift) + kSurrogateHighStart;
> *target++ = (ch & halfMask) + kSurrogateLowStart;
> };
>
> with kMaximumUCS2 = 0xffff and kMaximumUCS4 = 0x7fffffff.
>
> Shouldn't the first comparison read "if (ch <= kMaximumUTF16)..."?
>
> In addition, function ConvertUTF8toUCS4 is a **COPY** of ConvertUTF8toUTF16,
> which sure isn't what it's intended to be. To correct this, would it be
> correct to just replace the above code with
>
> if (ch > kMaximumUCS4) {
> ch = kReplacementCharacter;
> }
> *target++ = ch;
>
> ? It would be great if someone could comment on this...
>
> Thanks a lot,
>
> /oliver
>
> ----------------------------------------------------------------------
> Name: winmail.dat
> winmail.dat Type: DAT File (application/x-unknown-content-type-dat_auto_file)
> Encoding: base64
