Hi,
I'm looking into what appears to be a 2.0.x regression vs GetFeature
and DescribeFeatureType behavior.
Basically it seems GetFeature can deal with names without a namespace
without issues, but DFT cannot
do the same:
This one works:
http://demo.opengeo.org/geoserver/wfs?request=GetFeature&typeName=states&version=1.0.0&service=wfs
This one breaks:
http://demo.opengeo.org/geoserver/wfs?request=DescribeFeatureType&typeName=states&version=1.0.0&service=wfs
Whilst using topp:states they both work.
Now, typeName is parsed by a QNameKvpParser, which used to have a bit
of code around line 80
to inject the default namespace into the qname that was commented out
in this commit:
r11502 | jdeolive | 2009-03-01 03:12:58 +0100 (dom, 01 mar 2009) | 1 line
GEOS-2662, ported wfs to new catalog + config
Now, uncommenting that bit of code would fix my issue but would
probably create others, as the catalog
behavior is to look in the default namespace first, but then fall back
into whatever first layer is named
"states" in case the default namespace does not contain it, right?
So what about having TypeNameKVPParser override QNameKvpParser parsing
and make it use
the catalog itself to lookup a non qualified feature type name?
I guess that would make DFT and GetFeature behaviour consistent and
backwards compatible at the
same time?
Cheers
Andrea
-----------------------------------------------------
Ing. Andrea Aime
Senior Software Engineer
GeoSolutions S.A.S.
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
phone: +39 0584962313
fax: +39 0584962313
http://www.geo-solutions.it
http://geo-solutions.blogspot.com/
http://www.linkedin.com/in/andreaaime
http://twitter.com/geowolf
-----------------------------------------------------