peter koch wrote:
>
Such an answer asks for more! Immediate questions are: what is it that
has serious deficiencies? C++ or Unicode? And what are those
deficiencies? And if the OP is to look up wchar_t and related
functions/classes, what does it mean that C++ is ignorant of Unicode?
Your posting looks and feels to me like FUD, but knowing you from other
posts, my belief is that I must have misunderstood something.
C++ has char's which serve double duty as the native character type
and also as the minimum addressable unit of storage (this itself
has its own problems). It also has the concept of making up multibyte
characters (larger character encoding formed of variable numbers of
consecutive chars). The standard also provides for "wide" chars
made wchar_t types. Unfortunately a number of interfaces only have
char based arguments and no wchar_t versions. This includes file
names and the program arguments. This is a deficiency. It assumes
that there is a single known mapping form wchar_t to multibyte to
get around the problem.
C++ is ignorant of Unicode in the same way that it is ignorant of
ASCII. It has no concept of character encoding. While ASCII
is common it's not mandated. The multibyte encoding might
be UTF-8 Unicode, but it might be EUC or some one of the older
Chineese/Japanese encodings.
>
