On 2007-09-03 12:36, Barry wrote:
Quote:
Alf P. Steinbach wrote:
Quote:
>* Barry:
Quote:
>>Hi group,
>>>
>>As I found out
>>Boost has utf8_codecvt_facet
>>
>Super. I didn't know that.
>>
>>
Quote:
>>So is there any library completely support for all UTF in this way
>>
>The Dinkumware implementation of the standard library reportedly has
>fairly complete support for Unicode.
>>
>>
Quote:
>>or any other way to deal with Unicode more elegantly?
>>
>Depends how much of Unicode functionality you need. E.g., UCS2 is easy.
>>
>
well, I'm student
so language selection is not critical for me,
When come to Unicode issue, I choose Java or C#, I really appreciate
their native support for Unicode. Together with string and IO.
>
I just downloaded ICU yesterday.
with a quick view, I did't see how ICU works with C++ io streams using
codecvt.
If fully Unicode support needs different String and IO implementation.
that's really painful. Then we have to modify a lot of code to bridge
std::xstring and std::ioxxx with the new String and IO supporting Unicode.
There are two ways (that I know of, I have not made much use of Unicode)
to handle Unicode in C++, either use the wide verions (wstring, wcout,
etc.) for UTF-16 or UTF32 (depends on the platform) or use the normal
strings ect. for UTF-8. If you use normal strings just remember that
size() does not represent the number of characters (or whatever it's
called) but rather the number of bytes.
--
Erik Wikström
