Unicode filenames on Windows

[Lukáš Lalinský]

On Ut, 2007-11-06 at 19:02 +0100, Scott Wheeler wrote:
> Lukáš Lalinský wrote:
> > Hi Scott,
> >
> > having support for Unicode filename API is a must for any more serious
> > usage on Windows, so I choose probably the only binary compatible way,
> > that makes sense, to do it -- duplicating constructors of all
> > TagLib::File subclasses to support also wchar_t* in addition to char*.
> 
> Here I'm not really happy with having the ifdef's all over the code. 
> Would it be possible to simply have one ifdef that's used for the 
> character type? I don't really care about changing the API for Windows 
> users sine TagLib hasn't officially supported Windows in the past. 
> (Plus, on Windows TagLib isn't used as a system API in the way that it 
> is on Unix.)

Well, the idea was to support both types, char* in the local encoding
and wchar_t* in utf-16. But since the conversion could be done by a
simple function without any information loss, I guess it's ok to
support only wchar_t*. And at least it will force lazy developers to
think about unicode if they still use ANSI encoded filenames. :)

Lukas