[10/23] user32/tests: Fix input tests compilation with __WINESRC__ defined.

[Saulius Krasuckas]

* On Wed, 23 Oct 2013, Dmitry Timoshkov wrote:

> --- a/dlls/user32/tests/input.c
> +++ b/dlls/user32/tests/input.c
> @@ -929,7 +929,7 @@ static void test_Input_blackbox(void)
>       * key state set by SendInput(). */
>      empty_message_queue();
>  
> -    prevWndProc = SetWindowLongPtr(window, GWLP_WNDPROC, (LONG_PTR) WndProc2);
> +    prevWndProc = SetWindowLongPtrA(window, GWLP_WNDPROC, (LONG_PTR) WndProc2);
>      ok(prevWndProc != 0 || (prevWndProc == 0 && GetLastError() == 0),
>         "error: %d\n", (int) GetLastError());
>  

* On Fri, 8 Jun 2007, Saulius Krasuckas wrote:

> --- a/dlls/user32/tests/input.c
> +++ b/dlls/user32/tests/input.c
> @@ -496,7 +496,7 @@ static void test_Input_blackbox(void)
>       * key state set by SendInput(). */
>      empty_message_queue();
>  
> -    prevWndProc = SetWindowLongPtr(window, GWLP_WNDPROC, (LONG_PTR) WndProc2);
> +    prevWndProc = SetWindowLongPtrA(window, GWLP_WNDPROC, (LONG_PTR) WndProc2);
>      ok(prevWndProc != 0 || (prevWndProc == 0 && GetLastError() == 0),
>         "error: %d\n", (int) GetLastError());
>  

It's nice to see the change is being offered by it's previous opponent six 
years later.  Since Dmitry was strongly against it, questions go to him:

* On Sat, 9 Jun 2007, Dmitry Timoshkov wrote:

> "Saulius Krasuckas" <saulius2 at ar.fi.lt> wrote:
> 
> > Hello.  I see some tests call [GS]etWindowLongPtr(..., GWLP_WNDPROC, 
> > ...). I think this is somewhat confusing.  API offers both A- an W- 
> > versions of this function, while at my first glance I cannot firmly 
> > decide, what used name does refer to.
> > 
> > Shouldn't occurences be renamed to ...WindowLongPtrA() here (or even 
> > all occurrences {by not requiring second parameter == GWLP_WNDPROC}) ?
> > 
> > Same question stands for [GS]etWindowLong().
> 
> The code is correct as it is now, there is no need to change anything.

Dmitry, what has changed since?
(now that every test subdir is being migrated to compile with __WINESRC__)
  
> > Did that restrict us from using single naming scheme?
> 
> Personally I don't see any point in that. The tests are not forced to 
> use A/W API suffixes, that's done on purpose.

Is the purpose gone now?

S.