Re: [1/2] comctl32: add WM_PAINT message sequence tests

[James Hawkins]

On 2/23/07, Felix Nawothnig <flexo(a)holycrap.org> wrote:
> Passed on XP SP2.
> ---
>   dlls/comctl32/tests/listview.c |  209
> +++++++++++++++++++++++++++++++++++++++-
>   dlls/comctl32/tests/msg.c      |    3 +
>   dlls/comctl32/tests/msg.h      |    3 +-
>   3 files changed, 211 insertions(+), 4 deletions(-)

This is wrong:

+static LRESULT WINAPI header_subclass_proc(HWND hwnd, UINT message,
WPARAM wParam, LPARAM lParam)
+{
+    struct subclass_info *info = (struct subclass_info
*)GetWindowLongA(hwnd, GWL_USERDATA);
+    static long defwndproc_counter = 0;
+    LRESULT ret;
+    struct message msg;
+
+    trace("header: %p, %04x, %08x, %08lx\n", hwnd, message, wParam, lParam);
+
+    msg.message = message;
+    msg.flags = sent|wparam|lparam|hdr;
+    if (defwndproc_counter) msg.flags |= defwinproc;
+    msg.wParam = wParam;
+    msg.lParam = lParam;
+    add_message(sequences, LISTVIEW_SEQ_INDEX, &msg);
+
+    defwndproc_counter++;
+    ret = CallWindowProcA(info->oldproc, hwnd, message, wParam, lParam);
+    defwndproc_counter--;
+    return ret;
+}

If you're subclassing the header control and checking messages from
the header control, then you need to add another message sequence,
HEADER_SEQ_INDEX.

--- a/dlls/comctl32/tests/msg.c
+++ b/dlls/comctl32/tests/msg.c
@@ -158,6 +158,9 @@ void ok_sequence_(struct msg_sequence **
             ok_(file, line) ((expected->flags & winevent_hook) ==
(actual->flags & winevent_hook),
                 "%s: the msg 0x%04x should have been sent by a
winevent hook\n",
                 context, expected->message);
+            ok_(file, line) ((expected->flags & hdr) == (actual->flags & hdr),
+                "%s: the msg 0x%04x should have been sent by the
header control\n",
+                context, expected->message);
             expected++;
             actual++;
         }

msg.c is the message sequence testing code abstracted...keyword is
abstraction.  Nothing in msg.c should know anything about specific
types of controls.

-- 
James Hawkins

[1/2] comctl32: add WM_PAINT message sequence tests

[Felix Nawothnig]

James Hawkins wrote:
> If you're subclassing the header control and checking messages from
> the header control, then you need to add another message sequence,
> HEADER_SEQ_INDEX.

Problem is I need to test whether the header msgs are sent between the 
WM_PAINT and WM_NOTIFY of the listview - right now the 
WM_PAINT/WM_NCPAINT/WM_ERASEBKGND _are_ sent - just not in between but 
after WM_PAINT processing (which causes the flickering).

I agree that the "hdr" flag is kinda hackish, any idea how to properly 
abstract away what I'm trying to do?

Felix

[1/2] comctl32: add WM_PAINT message sequence tests

[James Hawkins]

On 2/23/07, Felix Nawothnig <flexo(a)holycrap.org> wrote:
> James Hawkins wrote:
> > If you're subclassing the header control and checking messages from
> > the header control, then you need to add another message sequence,
> > HEADER_SEQ_INDEX.
>
> Problem is I need to test whether the header msgs are sent between the
> WM_PAINT and WM_NOTIFY of the listview - right now the
> WM_PAINT/WM_NCPAINT/WM_ERASEBKGND _are_ sent - just not in between but
> after WM_PAINT processing (which causes the flickering).
>
> I agree that the "hdr" flag is kinda hackish, any idea how to properly
> abstract away what I'm trying to do?

Patch is attached.  If you send it in to wine-patches, you need to put
both our names in the Changelog.

-- 
James Hawkins