I have a small application that uses a DrawingArea
to draw a simple map using PyGObject
and GTK3
.
I load a Pixbuf
using
from gi.repository import Gtk, GdkPixbuf
pixbuf = GdkPixbuf.Pixbuf.new_from_file_at_size("logo.png", 25, 25)
and then try to draw it in the DrawingArea
's draw event signal
def draw(self, widget, context):
window = widget.get_window()
ctx = window.cairo_create()
ctx.set_source_pixbuf(pixbuf, 0, 0)
but I get the error message
"AttributeError: 'cairo.Context' object has no attribute 'set_source_pixbuf'"
If I'm reading the Gtk2 to Gtk3 migration guide correctly, this should work.
What am I doing wrong?
The new draw signal uses a callback that already passes the cairo context as a parameter, you don't need to do stuff like window = widget.get_window()
like you did in PyGtk to get the cairo context while attending the expose-event signal. In PYGObject is simpler:
import cairo
class Foo(object):
def __init__(self):
(...)
self.image = cairo.ImageSurface.create_from_png('logo.png')
(...)
def draw(self, widget, context):
if self.image is not None:
context.set_source_surface(self.image, 0.0, 0.0)
context.paint()
else:
print('Invalid image')
return False
That is if you don't need the PixBuf, but if you need it for something else you have several options:
- To have both objects in memory. If both are loaded from a PNG there should not be much problems other than the waste of memory.
- Convert GdkPixbuf to PIL Image, then PIL Image to data array and then create a Cairo ImageSurface from that data array using create_for_data(). Yak :S I don't know better, sorry :S
- Use Gdk.cairo_set_source_pixbuf() proposed by hock. This seems to be the correct way to draw a Pixbuf in a ImageSurface, but it is totally unpythonic (and that's why I hate this Introspection stuff, all looks like C, like a bad C port).
If you choose the awful second option here is how:
import Image
import array
from gi.repository import Gtk, GdkPixbuf
width = 25
height = 25
pixbuf = GdkPixbuf.Pixbuf.new_from_file_at_size('logo.png', width, height)
pil_image = Image.fromstring('RGBA', (width, height), pixbuf.get_pixels())
byte_array = array.array('B', pil_image.tostring())
cairo_surface = cairo.ImageSurface.create_for_data(byte_array, cairo.FORMAT_ARGB32, width, height, width * 4)
Note that create_for_data() is not yet available for Python3, only for Python2.
Check also my answer on how to use a double buffer in PyGObject if this is what you're trying to achieve: Drawing in PyGobject (python3)
Kind regards
The following seems to do the job:
def draw(self, widget, context):
Gdk.cairo_set_source_pixbuf(context, self.pixbuf, 0, 0)
context.paint()
One question still remains: Is this the preferred way of doing things?