[Gvsig_desarrolladores] Possibly two small bugs

[S.O.]

I've moved my code to gvSIG 2.0 (trunk) and I've tested the patch: all seems 
OK. I was only wondering why letting the 'getBounds' method of the 
'Fpoint2D' class return an incorrect value (i.e. when x = -1.8 and y = -1.8 
the returned Rectangle has the origin in (-1,-1) and 0 width and height, 
thus not bounding the point); IMHO it should be safer surrounding the point 
with a minimum Rectangle (width = 1, height = 1) that really bounds it.

Debugging my code, I've also noticed that now the ConcreteMemoryDriver is 
not "tolerant" to rows that have a null associated to the geometry field 
(i.e. a row without a geometry on the map) raising a Null Pointer Ex in the 
'intersect' method of the 'SpatialQueryFeatureIterator' class; perhaps it 
should be safer to put a check there not allowing the call of 
'XRectangle2D' method when the rectangle bounding the feature 
(featureBounds) is null.

Thanks for your advice!
Stefano