Is there a way to make all character sequences UNICODE by default? For instance, now I have to say:
std::wstring wstr(L"rofl");
instead, I'd like to say
std::wstring wstr("rofl");
Thanks!
Visual C++ 8.0
+2
A:
No, there is no such thing. Many platforms intended for compiliation with either Unicode or ASCII (but not both) provide a macros to allow easy selection of the desired character type (often _ is used, e.g. _("text")). You can use this if your issue is compiling for both options. Otherwise, it's only an additional character.
Also, bear in mind the exact encoding of wchar_t may vary from platform to platform. Sometimes it is UTF-16, sometimes it is UTF-32. C++1x will add explicit encoding-specified strings u8"string" u"string" and U"string". (I could be wrong about the prefixes, but there will be such strings).
coppro
2009-01-06 16:46:56
A:
I was surprised by this: Doesn't wstring have a wstring(const char *) constructor??
Roddy
2009-01-06 16:47:49
No, it has a wstring(const wchar_t*) constructor. It also has a constructor from a pair of iterators, but this is inappropriate in the case of a string literal.
coppro
2009-01-06 16:53:20
Nope. It has basic_string<T>(const T*), where T = wchar_t.
Rob Kennedy
2009-01-06 17:16:35
Perhaps; the standard doesn't require it, but it doesn't forbid it either.
MSalters
2009-01-09 14:08:27
A:
As an aside, CString supports cross-width construction (eg: CString("text") will work for Unicode builds, and CString(L"text") for MBCS builds), but still does a conversion in both cases. That might accomplish what you want, at the cost of performance.
I'd suggest getting used to using _T("text"), as that's generally what you want.
Nick
2009-01-06 18:39:32