[PATCH] Preliminary fix for codecvt_members_unicode_wchar_t

[Paolo Carlini]

Ulrich Weigand wrote:

>However, the test case specifies just "UCS4".  Does this always
>imply big-endian?
>
Now it occurs to me that, in the test:

    unicode_codecvt::state_type state01("UCS-2BE", "UCS4", 0xfeff, 0);

therefore, we convert to a /Big Endian/ representation. May be that by 
default it is assumed that the endianness of the source is the same of 
the destination.

Ciao, Paolo.