On 2/23/07, Felix Nawothnig flexo@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 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
On 2/23/07, Felix Nawothnig flexo@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.
-
Felix Nawothnig -
James Hawkins