How can I stretch draw an area of a gdk pixbuf to a cairo surface with a user defined opacity?
I am trying to write a cross platform interface for working with bitmaps and want to add alpha blending to my cairo stretch draw method. What I have right now works just fine, but I am unable to think up a way of to mold alpha blending into a series of cairo/gdk apis which work somewhat like Microsoft's [AlphaBlend][1] function.
What I have so far is this:
procedure TGtkBitmap.Draw(const Source: TRect; Canvas: TCanvas;
const Dest: TRect; Alpha: Byte = $FF);
var
D: PGdkDrawable;
C: Pcairo_t;
M: cairo_matrix_t;
begin
if FBuffer = nil then
Exit;
if (WidthOf(Source) < 1) or (WidthOf(Dest) < 1) then
Exit;
if (HeightOf(Source) < 1) or (HeightOf(Dest) < 1) then
Exit;
D := TGtk2DeviceContext(Canvas.Handle).Drawable;
C := gdk_cairo_create(D);
gdk_cairo_set_source_pixbuf(C, FBuffer, 0, 0);
cairo_matrix_init_identity(@M);
cairo_matrix_translate(@M, Source.Left, Source.Top);
cairo_matrix_scale(@M, WidthOf(Source) / WidthOf(Dest),
HeightOf(Source) / HeightOf(Dest));
cairo_matrix_translate(@M, -Dest.Left, -Dest.Top);
cairo_pattern_set_matrix(cairo_get_source(C), @M);
cairo_rectangle(C, Dest.Left, Dest.Top, WidthOf(Dest),
HeightOf(Dest));
// what cairo functions can I combine here to vary
// the opacity of the pattern fill using Alpha argument?
cairo_fill(C);
cairo_destroy(C);
end;Everything works fine, but I am unsure how to get alpha blending with a pix buff pattern working. I can imagine one way which involves creating a second cairo surface, drawing the entire pixbuf to it with a user defined opacity, then creating pattern for the first surface using the new surface, which all gets a little messy and probably is much slower than something I'd be happy with.
Here is a [video recording][2] of what I have working so far. I'd like to know from someone familiar with cairo, what can I insert into my routine above to set the alpha level of the pixbuf source pattern?
[1]:
http://msdn.microsoft.com/en-us/library/windows/desktop/dd183351%28v=vs.85%29.aspx [2]: http://youtu.be/VJJIwA1hi7c
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20120824/ce9a3f01/attachment.html>