> I assume this is not true of the return from OGR_F_GetGeometryRef, which
> needs to be cloned or it can become invalid.
Yes, the geometry of a feature belongs to the feature. If you destroy the
feature, the geometry becomes invalid. So you need to clone it to get a
standalone version.
--
Ge
On Wed, Feb 26, 2014 at 3:02 PM, Even Rouault
wrote:
> Tim,
>
> > When I extract a feature from a layer, what issues arise with the
> lifetime
> > of the feature and the layer? If I keep the layer alive, will the feature
> > continue to be valid?
>
> Yes, feature objects (at least at C++ objects,
Tim,
> When I extract a feature from a layer, what issues arise with the lifetime
> of the feature and the layer? If I keep the layer alive, will the feature
> continue to be valid?
Yes, feature objects (at least at C++ objects, not speaking about bindings)
retrieved by GetNextFeature() and GetF
When I extract a feature from a layer, what issues arise with the lifetime
of the feature and the layer? If I keep the layer alive, will the feature
continue to be valid?
If I clone a feature, is it now standalone or does it still depend on the
layer/datasource in anyway?
I see the note about des
