GetCellsOverlappingRectangle should return the inclusive cell on Right and Bottom sides

[gazle]

class CollisionGrid
...
public CollisionGrid(...)
{
...
	WidthInPixels = Columns * CellWidth;
	HeightInPixels = Rows * CellHeight;
...
}
...
public IEnumerable<CollisionGridCell> GetCellsOverlappingRectangle(RectangleF rectangle)
{
	var sx = (int) (rectangle.Left/CellWidth);
	var sy = (int) (rectangle.Top/CellHeight);
	v̶a̶r̶ ̶e̶x̶ ̶=̶ ̶(̶i̶n̶t̶)̶ ̶(̶r̶e̶c̶t̶a̶n̶g̶l̶e̶.̶R̶i̶g̶h̶t̶/̶C̶e̶l̶l̶W̶i̶d̶t̶h̶ ̶+̶ ̶1̶)̶;̶
	var ex = Columns - (int) (WidthInPixels - rectangle.Right) / CellWidth - 1;
	v̶a̶r̶ ̶e̶y̶ ̶=̶ ̶(̶i̶n̶t̶)̶ ̶(̶r̶e̶c̶t̶a̶n̶g̶l̶e̶.̶B̶o̶t̶t̶o̶m̶/̶C̶e̶l̶l̶H̶e̶i̶g̶h̶t̶ ̶+̶ ̶1̶)̶;̶
	var ey = Rows - (int) (HeightInPixels - rectangle.Bottom) / CellHeight - 1;

	for (var y = sy; y <= ey; y++)
		for (var x = sx; x <= ex; x++)
			yield return GetCellAtIndex(x, y);
}

When a rectangle CellWidth by CellHeight is tile aligned, it would be better for the right and bottom sides to resolve into the same cell.

It may break a lot of code but this is a better way going forward IMO. Collision detection along those sides works better and you no longer have to do the kludge of making Width and Height "slightly" smaller to make it fit in, which I see in so many youtube tutorials.