It seems that cPS parameter provided to bPainted doesn't have the expected data. Can you confirm this? If yes, any fix?
EMG
FiveTech Support Forums
FiveWin / Harbour / xBase community
Board index FiveWin for Harbour/xHarbour cPS in bPainted
Page: 1 2 (21 posts)
Enrico,
It should have it as see how it gets built:
It should have it as see how it gets built:
HB_FUNC( BEGINPAINT ) // ( hWnd, @cPS ) --> hDC
{
PAINTSTRUCT ps;
HDC hDC = BeginPaint( ( HWND ) hb_parnl( 1 ), &ps );
hb_storclen( ( char * ) &ps, sizeof( PAINTSTRUCT ), 2 );
hb_retnl( ( LONG ) hDC );
}Enrico,
There is a working example managing cPS at FWH\samples\formdes.prg
Antonio,
That's exactly what I was trying to do but unfortunately DrawGrid() makes no use of cPS parameter.
EMG
Antonio Linares wrote:Enrico,
There is a working example managing cPS at FWH\samples\formdes.prg
That's exactly what I was trying to do but unfortunately DrawGrid() makes no use of cPS parameter.
EMG
Enrico,
Have you tried to modify DrawGrid() this way ?
Have you tried to modify DrawGrid() this way ?
HB_FUNC( DRAWGRID ) // hWnd, hDC, @cPS, wGridX, wGridY
{
WORD wRow, wCol;
#ifdef _WIN64
HDC hDC = ( HDC ) hb_parnll( 2 );
HWND hWnd = ( HWND ) hb_parnll( 1 );
#else
HDC hDC = ( HDC ) hb_parnl( 2 );
HWND hWnd = ( HWND ) hb_parnl( 1 );
#endif
WORD wGridX = hb_parni( 4 );
WORD wGridY = hb_parni( 5 );
RECT rc;
WORD wWidth, wHeight;
PAINTSTRUCT ps;
memcpy( &ps, hb_parc( 3 ), sizeof( PAINTSTRUCT ) );
GetWindowRect( hWnd, &rc );
wWidth = rc.right - rc.left + 1;
wHeight = rc.bottom - rc.top + 1;
for( wRow = 0; wRow <= wHeight; wRow += wGridX )
for( wCol = 0; wCol <= wWidth; wCol += wGridY )
SetPixel( hDC, wCol, wRow, 0 );
}Antonio,
Yes, but ps.rcPaint.top and ps.rcPaint.left are always zero, unless I'm missing something.
EMG
Antonio Linares wrote:Enrico,
Have you tried to modify DrawGrid() this way ?
Yes, but ps.rcPaint.top and ps.rcPaint.left are always zero, unless I'm missing something.
EMG
Enrico,
ps.rcPaint.bottom and ps.rcPaint.right are non zero, so they seem valid.
Going to check top and left...
Enrico,
They are zero but this code seems to work fine:
They are zero but this code seems to work fine:
HB_FUNC( DRAWGRID ) // hWnd, hDC, @cPS, wGridX, wGridY
{
WORD wRow, wCol;
#ifdef _WIN64
HDC hDC = ( HDC ) hb_parnll( 2 );
HWND hWnd = ( HWND ) hb_parnll( 1 );
#else
HDC hDC = ( HDC ) hb_parnl( 2 );
HWND hWnd = ( HWND ) hb_parnl( 1 );
#endif
WORD wGridX = hb_parni( 4 );
WORD wGridY = hb_parni( 5 );
PAINTSTRUCT ps;
memcpy( &ps, hb_parc( 3 ), sizeof( PAINTSTRUCT ) );
for( wRow = 0; wRow <= ps.rcPaint.bottom; wRow += wGridX )
for( wCol = 0; wCol <= ps.rcPaint.right; wCol += wGridY )
SetPixel( hDC, wCol, wRow, 0 );
}Antonio,
Yes, but the benefit is greatly reduced if the area to be repainted always starts from 0, 0...
EMG
Antonio Linares wrote:Enrico,
They are zero but this code seems to work fine:
Yes, but the benefit is greatly reduced if the area to be repainted always starts from 0, 0...
EMG
That is what Windows is providing us 
Antonio,
I'm not so sure... how FWH provides cPS to bPainted codeblock?
EMG
Antonio Linares wrote:That is what Windows is providing us
I'm not so sure... how FWH provides cPS to bPainted codeblock?
EMG
Enrico,
cPS is obtained here:
and used from bPainted:
cPS is obtained here:
METHOD _BeginPaint() CLASS TWindow
local cPS
if ::nPaintCount == nil
::nPaintCount = 1
else
::nPaintCount++
endif
::hDC = BeginPaint( ::hWnd, @cPS )
::cPS = cPS
return niland used from bPainted:
if ValType( ::bPainted ) == "B"
uVal = Eval( ::bPainted, ::hDC, ::cPS, Self )
endifAntonio,
It looks correct, thank you. Any ideas on how to speed up DrawGrid() function?
EMG
Antonio Linares wrote:Enrico,
cPS is obtained here:
METHOD _BeginPaint() CLASS TWindow local cPS if ::nPaintCount == nil ::nPaintCount = 1 else ::nPaintCount++ endif ::hDC = BeginPaint( ::hWnd, @cPS ) ::cPS = cPS return nil
and used from bPainted:
if ValType( ::bPainted ) == "B" uVal = Eval( ::bPainted, ::hDC, ::cPS, Self ) endif
It looks correct, thank you. Any ideas on how to speed up DrawGrid() function?
EMG
Enrico,
Here it seems fast enough, I am using a Core 2 Duo
It could be speed up using a brush instead of calling SetPixel()
Antonio,
Interesting... any little sample?
EMG
Antonio Linares wrote:Enrico,
Here it seems fast enough, I am using a Core 2 Duo
It could be speed up using a brush instead of calling SetPixel()
Interesting... any little sample?
EMG
Page: 1 2 (21 posts)