Hello Andrea
On Tue, Jul 25, 2017 at 12:14 AM, Andrea Aime <[email protected]>
wrote:
> HI Torben,
> thanks a lot for the update. There is only one thing giving me pause as it
> read it:
>
> "If a style group refers to an invalid layer, it should just drop the
> layer from the output, to match with the current treatment of layer groups.
> It may be valuable to extend this implementation to the treatment of
> external styles as well (Perhaps as a configurable option if this is
> necessary to remain compliant with the SLD / WMS spec)."
>
> Is it really working that way, if someone manages (probably due to a bug)
> to create a pending reference to
> a missing layer or style, then you get no exceptions?
>
Looking back at that, I got that info from an earlier mail; not quite sure
where it came from. Turns out, the logic is a bit more complicated (and
completely opposite to) than that:
- An entry in the LAYERS list referencing a remote layer will try and
find the appropriate WFS source; if this fails it will be treated as a
regular layer
- An entry in the LAYERS list with an invalid name will throw a
ServiceException
- A reference to a null layer within a layer group will throw a NPE
- A reference to a nonexistant layer in a layergroup will not throw an
immediate exception, but may fail further down the line.
So, yes, you're right, this does and should actually fail with an
exception. I'll update the proposal accordingly; it will throw exceptions
if a layer is missing (thanks for the catch there).
I'd prefer an exception being though (fail early), but if you have
> compelling arguments not to go that way, then
> at least a log at SEVERE level?
> Silent failure would make it hard to figure out.
>
> One more little detail, mostly just a reminder, style treatment in
> GeoServer is pretty lax (validation being optional),
> so be prepared to find also styles with no named layer reference at all,
> or invalid UserLayer structures.
> Right, what happens if someone actually put in a style a _valid_
> UserLayer, with a locally defined set of features?
>
> I've looked at this functionality, and I am not certain what GeoServer
does right now, although I think there is logic to handle that. It will
definitely be something to be aware of when implementing this
functionality.
Torben
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Geoserver-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
