2016-03-30 14:00 GMT+02:00 Andrea Aime <[email protected]>:
> On Wed, Mar 30, 2016 at 12:58 PM, Alessandro Pasotti <[email protected]>
> wrote:
>
>> Hi All,
>>
>> I'm testing the official geoserver 2.8.2 on Linux, I've enabled WFS-T
>> transactional on all workspaces.
>> I'm doing some insert tests using topp: roads, state_boundaries and
>> cities as test layers for linestrings, polygons and points.
>>
>> Using QGIS as a client, this works:
>> - insert on cities
>>
>> This doesn't:
>> - insert on roads and state_boundaries
>>
>> First issue: geoserver returns <wfs:SUCCESS/> (no errors and no warnings
>> in the logs) but data are not inserted in the shapefiles.
>>
>> By comparing example XML I've tracked down the problem (at least for
>> linestrings) to the simple GML used by QGIS, compared with the multi
>> geometry GML used in the geoserver demo requests (which works):
>>
>> QGIS (silently fails):
>> <the_geom xmlns="http://www.openplans.org/topp";>
>> <gml:LineString srsName="EPSG:4326">
>> <gml:coordinates cs="," ts=" ">146.1911....
>>
>>
>> GeoServer Demo Requests (works):
>> <topp:the_geom>
>> <gml:MultiLineString srsName="
>> http://www.opengis.net/gml/srs/epsg.xml#4326";>
>> <gml:lineStringMember>
>> <gml:LineString>
>> <gml:coordinates decimal="." cs="," ts=" ">146.1911....
>>
>>
>>
> What geometry type is GeoServer declaring in DescribeFeatureType? If they
> do not match you'll
> get the result you are describing.
>
> As far as I know the original developer of the WFS-T subsystem said we
> cannot xml validate the
> request for some reason, but at the same time the parser is unfortunately
> schema driven,
> meaning that it will only look for the elements it expects to find,
> ignoring the others.
> So if it's looking for a attribute that's not mandatory, and does not find
> it because it's mis-encoded,
> the transaction will succeeded ignoring the unexpected attributes.
>
> We have had a number of reports due to this issue over the years, with a
> few variants (e.g., people encoding
> geometries in gml 2 in wfs 1.1 transactions), and I agree it would be best
> if the
> system validated the input xml against the expected schema and threw an
> error,
> but apparently it's not an easy thing to fix (meaning, not a spare time
> deal),
> and one that nobody has gathered funds to use working hours to fix so far.
>
> Cheers
> Andrea
>
>
Thank you Andrea,
I have clear picture now, and you are right, the problem is in that
DescribeFeatureType returns type="gml:MultiLineStringPropertyType" while
QGIS thinks that is a simple geometry.
I'll have a look into QGIS if we can do something to fix this.
--
Alessandro Pasotti
w3: www.itopen.it
------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785471&iu=/4140
_______________________________________________
Geoserver-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-users
