how to localize console and GUI apps in Windows

[Andrei]

On Thursday, 28 December 2017 at 18:45:39 UTC, H. S. Teoh wrote:
> On Thu, Dec 28, 2017 at 05:56:32PM +0000, Andrei via 
> Digitalmars-d-learn wrote:
> ...
> The string / wstring / dstring types in D are intended to be 
> Unicode strings.  If you need to use other encodings, you 
> really should be using ubyte[] or const(ubyte)[] or 
> immutable(ubyte)[], instead of string.

Thank you Teoh for advise and good example! I was looking towards 
writing something like that if no decision exists. Still this way 
of deliberate translations seems to be not the best. It supposes 
explicit workaround for every ahchoo in Russian and steady 
converting ubyte[] to string and back around. No formatting gems, 
no simple and elegant I/O statements or string/char comparisons. 
This may be endurable if you write an application where Russian 
is only one of rare options, and what if your whole environment 
is totally Russian?

Or some other nonASCII locale... Many other cultures have same 
mix of DOS/Window/Unix code pages. The decision to use only 
Unicode for strings in D language seems excellent just because of 
this, but the realization turns out to be delusive. Folks in such 
countries won’t appreciate a language which is elegant only for 
English-spoken intercommunications.

This problem is common for most programming languages and 
runtimes I know of. The only system which has decided the whole 
case is .NET I think.

The way proposed by zabruk70 below seems more appropriate though 
more particular too - I feel it suits only console type of 
applications. Alas, this type of application proved to be buggy 
too.

On Thursday, 28 December 2017 at 22:49:30 UTC, zabruk70 wrote:
> you can just set console CP to UTF-8:
>
> https://github.com/CyberShadow/ae/blob/master/sys/console.d

Yes! This seems to be the required, thank you very much! Though 
it is not suitable for GUI type of a Windows application.

Still some testing showed that this way conforms only console 
output. Simple read/write/compare script listed below works very 
well until the user enters something Russian. It then prints 
**empty** response and falls into indefinite loop printing the 
prompt and then immediately empty response without actually 
reading it.

But I think this is subject for ”Issues” part of this forum.

p.s. I’ve found that I may set “Consolas” font for a console 
window and then you can output properly localized UTF8 strings 
without any special code in D script managing code pages. Still 
this does not decide localized input problem: any localized input 
throws an exception “std.utf.UTFException... Invalid UTF-8 
sequence”.

The script:

import core.sys.windows.windows;
import std.stdio;
import std.string;

int main(string[] args)
{
     const UTF8CP = 65001;
     UINT oldCP, oldOutputCP;
     oldCP = GetConsoleCP();
     oldOutputCP = GetConsoleOutputCP();

     SetConsoleCP(UTF8CP);
     SetConsoleOutputCP(UTF8CP);

     writeln("hello world, привет всем!");

     bool quit = false;
     string response;
     while (!quit)
     {
         write("responde something: ");
         response=readln().strip();
         writefln("your response is \"%s\"", response);
         if (response == "quit")
         {
             writeln("good buy then!");
             quit = true;
         }
     }

     SetConsoleCP(oldCP);
     SetConsoleOutputCP(oldOutputCP);

     return 0;
}