separate read/display png functions

[okyeron]

While looking at the png functions today I wanted to revisit the idea of separating read and display functions for pngs.

Am I on the right track with something like the following?

void screen_read_png(const char *filename){
	image = cairo_image_surface_create_from_png (filename);
	if(cairo_surface_status (image)) { 
		fprintf(stderr, "display_png: %s\n", cairo_status_to_string (cairo_surface_status(image))); 
		return; 
	} 
}
		
void screen_show_png(double x, double y){
	int img_w, img_h;
	img_w = cairo_image_surface_get_width (image);
	img_h = cairo_image_surface_get_height (image);

	cairo_set_source_surface (cr, image, x, y);
	cairo_rectangle (cr, x, y, img_w, img_h);
	cairo_fill (cr);
	cairo_surface_destroy (image);
}

void screen_display_png(const char *filename, double x, double y){
	screen_read_png(filename);
	screen_show_png(x, y);
}

[artfwo]

@okyeron the code above will destroy the image after you first display it. it would be better to return an image object from screen_read_png (with a finalizer that will call cairo_surface_destroy in C). screen_show_png can then accept the image object as first parameter, so you can load and display images separately.