Re: mlang: tests, converstion test was falsely failing on windows xp for codepage 1361 (Korean, Johab) because of lack of roman characters

[Alexandre Julliard]

Aric Stewart  writes:
> @@ -259,6 +259,15 @@ static BOOL check_convertible(IMultiLanguage2 *iML2, UINT from, UINT to)
>  static WCHAR strW[] = {'a','b','c',0};
>
> - if (from == 1361 || to == 1361)
> - {
> -
>    /*
> -
>     * This is Korean (Johab).  This codepages does not have latin
> -
>     * characters so these checks fail causing confusions later.
> -
>     */
> -
>    return TRUE;
> - }

There should be a way to detect this with a generic check, instead of
having to hardcode specific codepages.