Statistical Units Vector - geometry test for GM_MultiSurface is too strict

[ulbrichtd]

Hi,

currently the validation of GML in the data theme Statistical Unit (Vector) fails when "gml:Surface" is used to characterize the area, i.e. the test "su-vector-as.b.2: Area Statistical Units" fails with the message "Check that the reference geometry of an area statistical units is a GM_MultiSurface.".

According to the specification for Statistical Unit ( https://inspire.ec.europa.eu/documents/Data_Specifications/INSPIRE_DataSpecification_SU_v3.0.pdf ) the UML class diagram for "Vector package" (figure 7 in section 5.3.1.2.3) suggests that GM_Object is allowed to describe the geometry.

I think the validation test is to strict and should allow gml:Surface.

Link to the data file: https://fbinter.stadt-berlin.de/fb/atom/SU/SU_BLK.zip
Link to the test report: SU_BLK_2021_06_29_Test.zip

Best regards and many thanks
Damian

[fabiovinci]

Dear @ulbrichtd,

we will investigate the check performed by the validator since your encoding should be ok.

As you mentioned the dataType of the geometry element is GM_Object, there is a specific constraint for the spatial object type VectorStatisticalUnit which states: "Vector statistical units with a reference geometry instance of GM_MultiSurface must be instances of the specialised class AreaStatisticalUnit.", but in your case, a "gml:Surface" geometry should be valid for the spatial object type AreaStatisticalUnit.

[fabiovinci]

Dear @ulbrichtd,

I had a deep look at the ATS and the related constraints defined by the IR 1089 and its amendment Commission Regulation (EU) No 1253/2013 defines a constraint about the geometry type of the AreaStatisticalUnit feature type.
See below:

The test in the validator reflects this constraint, although it is probably not well formulated since it is a constraint but makes use of the verb "should", which is, in general, used for recommendations.

Moreover, this constraint is not present in either the TG or the UML.

I think we can move the discussion to the TG repository, opening an issue for the inclusion of the constraint in the TG and asking feedback about its implementation, i.e. if it can be considered as a recommendation.

Could you please open the relevant issue in the TG repository?

[fabiovinci]

According to the conclusions of the related TG issue, the related ATS has been modified and the test will be relaxed in the Validator.

[arantzaetxebarria]

Dear @ulbrichtd

We have done the changes in the ETS in order to relax the test, now you can test it in staging environment. Let us know if all is correct, thanks in advance.

Regards,

Arantza

[ulbrichtd]

Thank you for the fix. The validator in the staging area accepts now files with "su-vector:geometry/gml:Surface" and "su-vector:geometry/gml:MultiSurface".

Best regards,
Damian

[arantzaetxebarria]

Dear @ulbrichtd

The issue will remain open until the next release 2022.2 when the changes will be put into the production environment.

Regards