Re: std::cout << Glib::ustring(utf8_literal) throws exception

[Sohail Somani <sohail taggedtype net>]

Murray Cumming wrote:
> On Sun, 2008-07-13 at 00:48 -0700, Sohail Somani wrote:
> > Murray Cumming wrote:
> > > On Sat, 2008-07-12 at 19:15 -0700, Sohail Somani wrote:
> > > > Yes, you are right. Copying the ustring to a std::string gets rid of the 
> > > > error but still doesn't work right because it is interpreted as ASCII in 
> > > > the console.
> > > std::cout does not understand UTF-8. That's why glibmm tries to do a
> > > conversion (with the Glib::convert_*() functions. That's documented by
> > > the way:
> > > http://www.gtkmm.org/docs/glibmm-2.4/docs/reference/html/classGlib_1_1ustring.html#_details
> > >
> > > But I don't know why the conversion is failing in your case. Maybe it's
> > > a problem with the glib installation on windows.
> > Isn't the whole point of UTF-8 that std::cout doesn't need to understand it?
> >
> > Shouldn't it be the console (or whatever is reading the output) who 
> > tries to understand it?
>
> No, I don't think that's how it works, though I guess that would be
> nice. I'm quite sure that something in C++ standard streams assumes that
> one character is one byte.

Probably.

> It's a pity that C++ _still_ doesn't have UTF-8 support in its standard,
> or apparently any sign that it will get into the standard any time soon.

Well, C++ is getting two new types. char16_t and char32_t (or something 
like that) whose intended purpose is to support utf-16 and utf-32.

But yes, I need to figure out why the conversion is failing in the first 
place rather than trying these hacks (ustring -> std::string).

--
Sohail Somani
http://uint32t.blogspot.com