[mapserver-users] Geojson output from WFS is missing ID

Tue May 10 04:20:14 PDT 2016

Le mardi 10 mai 2016 13:08:07, Rahkonen Jukka (MML) a écrit :
> Hi,
> 
> Geoserver breaking GML 3.2.1 schema is even documented in
> http://docs.geoserver.org/2.8.3/user/data/app-schema/supported-gml-version
> s.html, section GML 3.2.1 geometries require gml:id, including an
> application schema example about how to make it GML 3.2.1 compliant.
> 
> So Mapserver WFS 2.0 does not return a FeatureCollection with members with
> GetFeatureByID but only a feature. The gml:id "states.1" is in the very
> beginning of the GML and it is the gml:id of the feature. In the geometry
> there is gml:id "states.1.1" and it means roughly "first geometry of the
> feature states.1". However, with GeoJSON as an outputformat the geometry
> is wrapped into a FeatureCollection probably because GeoJSON required it
> to be so. But would it be wrong to wrap also the GML 3.2.1 output of
> GetFeatureByID into a FeatureCollection?

GML output and OGR output in MapServer take very different roads.

For GML, MapServer is right to return the result of GetFeatureById as a single 
feature since this is mandated by the WFS 2.0 spec (an horrible exception that 
was painful to implement AFAIR since normally GetFeaturebyId should be just a 
normal GetFeature once the stored query is resolved) : 

"""
11.3.5
GetFeatureById response
Executing a GetFeature operation containing the GetFeatureById stored query 
(see 7.9.3.6) shall generate a response composed of a single feature encoded 
according to the values of the outputFormat parameter (see 7.6.3.7). The 
response shall only contain the feature XML without any of the normal 
containing elements generated as a result of executing a GetFeature operation 
(see 11.3.3.4, 11.3.3.5, 11.3.3.6).
"""

For other output formats, I'm not sure what the WFS spec mandates, but anyway 
the OGR GeoJSON driver only exports features in a GeoJSON FeatureCollection.

> 
> -Jukka-
> 
> -----Alkuperäinen viesti-----
> Even Rouault wrote:
> 
> Le mardi 10 mai 2016 12:32:17, Rahkonen Jukka (MML) a écrit :
> > Hi,
> > 
> > Do you mean that the output from GeoServer WFS 2.0 is not valid
> > because it does not seem to have gml:id for the collection, nor for
> > the geometries, but it does have gml:id for the members?
> > 
> > <wfs:FeatureCollection numberMatched="49" numberReturned="49"
> > timeStamp="2016-05-10T10:23:29.813Z"
> > xsi:schemaLocation="http://www.opengis.net/wfs/2.0
> > http://localhost:8080/geoserver/schemas/wfs/2.0/wfs.xsd
> > http://www.openplans.org/topp
> > http://localhost:8080/geoserver/wfs?service=WFS&version=2.0.0&request=
> > Desc ribeFeatureType&typeName=topp%3Astates
> > http://www.opengis.net/gml/3.2
> > http://localhost:8080/geoserver/schemas/gml/3.2.1/gml.xsd"><wfs:member
> > 
> > ><to pp:states gml:id="states.1"><topp:the_geom><gml:MultiSurface
> > 
> > srsName="urn:ogc:def:crs:EPSG::4326"
> > srsDimension="2"><gml:surfaceMember><gml:Polygon><gml:exterior><gml:Li
> > near
> > Ring><gml:posList>
> 
> Yes, this will not validate the gml 3.2.1 schema. See for example
> http://www.datypic.com/sc/niem21/e-gml32_MultiSurface.html gml:id being
> compulsory on geometries is one of the main difference between gml 3.2 and
> 3.1
> 
> > -Jukka Rahkonen-
> > 
> > Even Rouault wrote:
> > > Le mardi 10 mai 2016 12:08:07, Rahkonen Jukka (MML) a écrit :
> > >> Hi,
> > >> 
> > >> WFS output from Mapserver 7.0 in GML contain gml:id
> > >> <ms:msGeometry><gml:Polygon gml:id="states.1"
> > >> srsName="urn:ogc:def:crs:EPSG::4326">
> > > 
> > > It looks like you must use WFS 2.0 and/or GML 3.2 output. In that
> > > case the GML
> > 
> > 3.2 schemas require each geometry to have a gml:id. So this is a
> > geometry id, not a feature id ( in case of geometry collections,
> > you'll have one id for the collection and as many ids as members) To
> > address that technical requirement, MapServer generates it in a
> > sequential way for the output of each request. Not completely sure it
> > matches the uniqueness intended but at least it validates schemas. So
> > this particular one is not a good candidate to be exported to other
> > formats, but I see Thomas has just replied for the feature id.
> > 
> > >> However, with GeoJSON as outputformat the id is missing:
> > >> {
> > >> "type": "FeatureCollection",
> > >> "crs": { "type": "name", "properties": { "name":
> > >> "urn:ogc:def:crs:OGC:1.3:CRS84" } }, "features": [ { "type":
> > >> "Feature", "properties": { "PK_UID": 1, "STATE_NAME": "Illinois",
> > >> 
> > >> 
> > >> GeoJSON specification says in
> > >> http://geojson.org/geojson-spec.html#feature-objects: "If a feature
> > >> has a commonly used identifier, that identifier should be included
> > >> as a member of the feature object with the name "id"."
> > >> 
> > >> GeoJSON from Geoserver WFS does contain the id:
> > >> 
> > >> {"type":"FeatureCollection","totalFeatures":49,"features":[{"type":
> > >> "Fe ature ","id":"states.1","geometry":{"type":"MultiPolygon"
> > >> 
> > >> Should this be classified as a bug or a candidate for a featere
> > >> request?
> > >> 
> > >> -Jukka Rahkonen-
> > 
> > --
> > Spatialys - Geospatial professional services http://www.spatialys.com
> 
> --
> Spatialys - Geospatial professional services http://www.spatialys.com
> _______________________________________________
> mapserver-users mailing list
> mapserver-users at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/mapserver-users

-- 
Spatialys - Geospatial professional services
http://www.spatialys.com