IfcAnnotation with IfcLinearSpanPlacement

[seb-esser]

I've created a quick sample testing IfcAnnotation together with IfcLinearSpanPlacement. Not sure if I am violating any existing agreement but it seems compliant with the schema.

General pull request

Fixes nothing but could be a solid base for further discussions regarding semantics along LRS

Changes and improvements proposed by this pull request

• showcases a possible way to add semantics to a given alignment without violating the separation between geometry and semantics

@pjanck @larswik @czapplitec @jmirtsch @SergejMuhic @NigelPMPeters

IFC4.3 Unit test

Unit test

Positioning an IfcAnnotation with IfcLinearSpanPlacement, which references an IfcAlignmentCurve

Review requested

• Please carefully check the intent and discuss if my approach is suitable. If there are any concerns, I am looking forward to discuss any improvements (especially how to document that this might be a "wrong" approach)

NOTE: we (@pjanck and @seb-esser) will try to prettify the supporting files (dwgs, pngs, ...) within the next days. However, don't hesitate to check the IFC file already

[pjanck]

A good example. Even if anything changes in the way alignments or linear referencing is done, we should still be able to convey the contents of this UT.

[larswik]

I was wondering if it is correct to use IfcAlignment => IfcRelAggregates => IfcAnnotation. An alternative would be to use e.g. spatial containment (the alignment is being referenced by the span placement anyway). I think that advice from more experienced people than me would be good. This is an interesting (and generic) question that maybe should be standardized. For superelevation and width it was suggested to use spatial containment to indicate which (lateral) part of the road is being affected.
However keeping that discussion separate, I think that this pull request could be merged (by someone with the proper rights).

[seb-esser]

should we raise @larswik 's comment in the bSI forum? What is the workflow for such a discussion? @pjanck ?

[seb-esser]

I've published the source code for this sample file: GeometryGymIFCExamples

[seb-esser]

Can anybody guide me on how to update the scenario PRed here? I assume, my lack of understanding is also related to #52 . I am happy to update from RC1 to RC2 but it seems to cause major changes in my staged model.

[pjanck]

2 entities are missing in the file (although referenced by other entities).

[SergejMuhic]

2 entities are missing in the file (although referenced by other entities).

Would it not be helpful to comment directly on the commit? That way we can see what your comments are referring to.

[seb-esser]

You are missing the segments and I would use IfcShapeRepresentation.RepresentationIdentifier='Axis2/3D'.

• RepresentationIdentifier resolved by 1a4f8ee
• missing segments is still open due to issues stated in the GeometryGymIfc toolkit repo

[pjanck]

2 entities are missing in the file (although referenced by other entities).

Would it not be helpful to comment directly on the commit? That way we can see what your comments are referring to.

I believe you can see the two lines addressed by my comments directly below the comment in question. Although it is hard to point to a missing line = nullptr :)

[pjanck]

Great work!

[larswik]

Is the intention to place the annotation from 25 to 110 (i.e. the deprecated IfcLinearSpanPlacement replaced by IfcLinearPlacement using a (linear) relative placement? Or is it that the annotation starts at 110 + 25? I heard some discussions before about using IfcCurveSegment as a replacement for IfcLinearSpanPlacement. Good if this could be settled. Also, I think that the readme should be updated since it still mentions RC1 and IfcLinearSpanPlacement.

[pjanck]

Superseded by bSI-InfraRoom/IFC-infra-unit-test#9