I am concerned that this would undo one of the things we fixed in GEOT-2505: http://jira.codehaus.org/browse/GEOT-2505
This might break a bunch of use cases. I'll check with the users. On 05/10/10 04:35, Justin Deoliveira wrote: > That makes sense. That GMLSchema contains autogenerated types for all gml > schema types and gets customized in a need by need basis, as with the regular > geometry types. +1 here but I think you may want to wait for Ben or Rini to > weigh in before proceeding. > > 2c. > > -Justin > > On Mon, Oct 4, 2010 at 11:51 AM, Andrea > Aime<andrea.a...@geo-solutions.it<mailto:andrea.a...@geo-solutions.it>> > wrote: > On Mon, Oct 4, 2010 at 7:26 PM, Andrea Aime > <andrea.a...@geo-solutions.it<mailto:andrea.a...@geo-solutions.it>> wrote: >> Hi, >> I'm trying to use app-schema with a target schema that, unfortunately, >> uses a DirectPosition as >> the feature "geometry". I don't have any control over the schema so I >> have to try and make it work >> anyways. >> >> What I get overall is that my points are turned into their WKT >> representation when the >> GML encoding happens. Which is not what I want. >> >> Looking a bit more deeply I see that the attribute is correctly bound >> to DirectPosition: >> ComplexTypeImpl http://www.opengis.net/gml:DirectPositionType extends >> doubleList(axisLabels:NCNameList,srsDimension:positiveInteger,srsName:anyURI,uomLabels:NCNameList) >> but nowhere in the code there is an attempt to use converters to turn >> the JTS Point into a DirectPosition. >> >> Now, I tracked down the issue to XPath.convertValue: >> >> private Object convertValue(final AttributeDescriptor descriptor, >> final Object value) { >> final AttributeType type = descriptor.getType(); >> Class<?> binding = type.getBinding(); >> >> if (type instanceof ComplexType&& binding == Collection.class) { >> if (!(value instanceof Collection)&& isSimpleContentType(type)) >> { >> ArrayList<Property> list = new ArrayList<Property>(); >> if (value == null&& !descriptor.isNillable()) { >> return list; >> } >> list.add(buildSimpleContent(type, value)); >> return list; >> } >> } >> if (binding == String.class&& value instanceof Collection) { >> // if it's a single value in a collection, strip the square >> brackets >> String collectionString = value.toString(); >> return collectionString.substring(1, collectionString.length() - >> 1); >> } >> return FF.literal(value).evaluate(value, binding); >> } >> >> type is a complex type, but binding for some reason is exactly >> Collection.class, >> which results in the wrong code path being chosen (the right one, imho, would >> be the usage of the last line, FF.literal(value).evaluate(value, binding), >> which >> could be expressed also as Converters.convert(value, binding)). >> >> Now... does anybody know why the binding ends up being a collection instead >> of org.opengis.geometry.DirectPosition? >> I still haven't quite tracked down where the bindings are built... > > Ok, tracked it down. > The GMLSchema class contains the definitions of many gml foundation types, > for point the definition is: > > > public static final AttributeType POINTTYPE_TYPE = build_POINTTYPE_TYPE(); > > private static AttributeType build_POINTTYPE_TYPE() { > AttributeType builtType; > builtType = new AttributeTypeImpl( > new NameImpl("http://www.opengis.net/gml","PointType";), > com.vividsolutions.jts.geom.Point.class, false, > false, Collections.<Filter>emptyList(), > ABSTRACTGEOMETRICPRIMITIVETYPE_TYPE, null > ); > return builtType; > } > > Notice the direct binding to Point.class. > But for DirectPosition it is: > > public static final ComplexType DIRECTPOSITIONTYPE_TYPE = > build_DIRECTPOSITIONTYPE_TYPE(); > > private static ComplexType build_DIRECTPOSITIONTYPE_TYPE() { > ComplexType builtType; > List<PropertyDescriptor> schema = new > ArrayList<PropertyDescriptor>(); > schema.add( > new AttributeDescriptorImpl( > NCNAMELIST_TYPE, new > NameImpl("http://www.opengis.net/gml","axisLabels";), 0, 1, true, null > ) > ); > schema.add( > new AttributeDescriptorImpl( > XSSchema.POSITIVEINTEGER_TYPE, new > NameImpl("http://www.opengis.net/gml","srsDimension";), 0, 1, true, > null > ) > ); > schema.add( > new AttributeDescriptorImpl( > XSSchema.ANYURI_TYPE, new > NameImpl("http://www.opengis.net/gml","srsName";), 0, 1, true, null > ) > ); > schema.add( > new AttributeDescriptorImpl( > NCNAMELIST_TYPE, new > NameImpl("http://www.opengis.net/gml","uomLabels";), 0, 1, true, null > ) > ); > builtType = new ComplexTypeImpl( > new > NameImpl("http://www.opengis.net/gml","DirectPositionType";), schema, > false, > false, Collections.<Filter>emptyList(), DOUBLELIST_TYPE, null > ); > return builtType; > } > > That is, it's built as a complex type, and all complex types bind to > Collection.class (that's inside the > ComplexTypeImpl constructor). > > The change I'm tempted to perform is to make the direct position bound > to a opengis DirectPosition > instead. It would contain the ordinates and the crs, but not the other > attributes... > And then make a converter between Point and DirectPosition... the > DirectPositionTypeBinding > is actually already there and should work > > Thoughts? > > 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 > > ----------------------------------------------------- > > ------------------------------------------------------------------------------ > Virtualization is moving to the mainstream and overtaking non-virtualized > environment for deploying applications. Does it make network security > easier or more difficult to achieve? Read this whitepaper to separate the > two and get a better understanding. > http://p.sf.net/sfu/hp-phase2-d2d > _______________________________________________ > Geotools-devel mailing list > Geotools-devel@lists.sourceforge.net<mailto:Geotools-devel@lists.sourceforge.net> > https://lists.sourceforge.net/lists/listinfo/geotools-devel > > > > -- > Justin Deoliveira > OpenGeo - http://opengeo.org > Enterprise support for open source geospatial. > > -- Ben Caradoc-Davies <ben.caradoc-dav...@csiro.au> Software Engineering Team Leader CSIRO Earth Science and Resource Engineering Australian Resources Research Centre ------------------------------------------------------------------------------ Beautiful is writing same markup. Internet Explorer 9 supports standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. Spend less time writing and rewriting code and more time creating great experiences on the web. Be a part of the beta today. http://p.sf.net/sfu/beautyoftheweb _______________________________________________ Geotools-devel mailing list Geotools-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geotools-devel
