"Damjan Jovanovic" <[EMAIL PROTECTED]> writes:
> + switch (dstDib.dsBm.bmBitsPixel)
> + {
> + case 1:
> + widthBytes = (width + 7) / 8;
> + srcOffsetBytes = (xSrc + 7) / 8;
> + dstOffsetBytes = (xDst + 7) / 8;
> + break;
> + case 4:
> + widthBytes = (width + 1) / 2;
> + srcOffsetBytes = (xSrc + 1) / 2;
> + dstOffsetBytes = (xDst + 1) / 2;
> + break;
You can't simply round size up, you have to handle partial bytes
properly.
> + case 24:
> + widthBytes = width * 2;
> + srcOffsetBytes = xSrc * 2;
> + dstOffsetBytes = xDst * 2;
> + break;
I don't think you can fix 24 bits in two bytes...
A bigger problem is that you also need to handle the clip region of
the destination DC, and only copy the pixels that aren't clipped out.
--
Alexandre Julliard
[EMAIL PROTECTED]
