From: Alexandre Julliard <julliard@winehq.org>
Subject: Re: [PATCH] kernel32/tests: Georgian and Hindi now get a thread codepage of CP_UTF8.
Message-Id: <87pnhe5qmh.fsf@wine>
Date: Tue, 26 Nov 2019 21:28:38 +0100
In-Reply-To: <E1iXo9M-0002E3-5l@amboise> (Francois Gouget's message of "Thu, 21 Nov 2019 16:13:40 +0100 (CET)")
References: <E1iXo9M-0002E3-5l@amboise>

Francois Gouget <fgouget@free.fr> writes:

> @@ -1151,9 +1151,9 @@ static void test_threadcp(void)
>          if (lcids[i].threadcp)
>              ok(cpi.CodePage == lcids[i].threadcp, "wrong codepage %u for lcid %04x, should be %u\n",
>                 cpi.CodePage, lcids[i].lcid, lcids[i].threadcp);
> -        else
> -            ok(cpi.CodePage == acp, "wrong codepage %u for lcid %04x, should be %u\n",
> -               cpi.CodePage, lcids[i].lcid, acp);
> +        else todo_wine
> +             ok(broken(cpi.CodePage == acp) || cpi.CodePage == CP_UTF8, "wrong codepage %u for lcid %04x, should be CP_UTF8\n",
> +               cpi.CodePage, lcids[i].lcid);
>  
>          /* WideCharToMultiByte - CP_THREAD_ACP */
>          num = WideCharToMultiByte(CP_THREAD_ACP, 0, foobarW, -1, NULL, 0, NULL, NULL);

It would be interesting to test if the WideCharToMultiByte conversion is
actually using utf-8.

-- 
Alexandre Julliard
julliard@winehq.org