International UTF-32 strings output to console in Linux

Question

#include <stdio.h>
#include <iostream>
#include <locale>

int main()
{
    const wchar_t *str = L"\u041F\u043E\u0440\u044F\u0434\u043E\u043A";
    std::locale::global(std::locale(""));
    std::wcout << str << std::endl;
}

Here's a piece of code that outputs a russian phrase in UTF-32 wchar_t string as:

• The correct one: Порядок when run from UTF-8 gnome terminal in Ubuntu 11.10
• РџРѕСЂСЏРґРѕРє in Eclipse in the test run as above
• 45=B8D8:0B>@ in Eclipse in a real program (where I don't even know who does what and where, but I suppose someone does mess with locales)
• ??????? if I don't call locale
• str is shown as Details:0x400960 L"\320\237\320\276\321\200\321\217\320\264\320\276\320\272" in Eclipse Watch window
• is shown as ASCII only byte chars in Eclipse memory window (and there's no way to specify that this is UTF-32 string)

I believe this is a misconfiguration in either eclipse console or the program, because, for example, other people that just run my code in Eclipse they do see the correct output.

Could someone shed a light on this confusion? What is the correct way to setup all the pieces (OS, gcc, terminal, Eclipse, sources...) to output international symbols that are stored in UTF-32 wchar_t strings?

And as a side note, why should I still care about all this when we have UTF-32 and that should be enough to know what is inside...

Answer 1

It turned out to be that other code changed locale.