wfstream::get() function and multibyte encoding

[Paolo Carlini]

Joseph D. Wagner wrote:

>Ok, I think I've figured it out.  Documentation about this part of C++ is sparse.
>  
>
Yes, you are right: the problem is also that much of this is not 
mandated by the Standard
(see also the recent messages from Nathan and Benjamin)...

>So is there a codecvt or locale for UTF-16 or do I need to create one?
>  
>
Well, frankly, I can't find one: this is expected since all those 
encodings are really meant as
*internal* encodings (beware that the standard internal encoding on GNU 
system is 4
bytes). I don't know if you already have this very nice doc:

    http://www.cl.cam.ac.uk/~mgk25/unicode.html

Anyway, preparing a suited codecvt facet seems easy (it would just pack 
bytes of the internal
representation to an appropriate endianness): grep in our testsuite for 
'public std::codecvt' for
many examples...

Paolo.