Final Curve Zero length Segment Continuous (not Discontinuous)

[jmirtsch]

ifc-gherkin-rules/features/ALS015_Alignment-representation-zero-length-final-segment.feature

Line 22 in e049f83

Scenario: Validating that the final alignment geometry segment is discontinuous.

This suggests that the final segment should have a transition code that is discontinuous.

However in the documentation https://standards.buildingsmart.org/IFC/RELEASE/IFC4_3/HTML/lexical/IfcSegment.htm
it refers to the transition code attribute relates to the previous segment. So shouldn't a discontinuous segment be expected at the first segment where there is no prior segment? A zero length segment should have the same location as the end of the previous segment, so this should be a continuous transition code? https://standards.buildingsmart.org/IFC/RELEASE/IFC4_3/HTML/lexical/IfcTransitionCode.htm

Thanks for clarification on this.

[aothms]

Thanks for bring this to our attention. It's really these things that improve the quality of the spec and the platform :)

1. Ok, so first of all, I created IfcSegment.Transition description doesn't describe attribute and mentioned non-existant name: "TangentialContinuity" IFC4.3.x-development#897 because I find the attribute description not satisfactory.

2. The IFC4 definition of the attribute is compared to next segment. So that's most likely where confusion comes from: https://standards.buildingsmart.org/IFC/RELEASE/IFC4/ADD2_TC1/HTML/link/ifccompositecurvesegment.htm

The state of transition (i.e., geometric continuity from the last point of this segment to the first point of the next segment) in a composite curve.

3. For N segments that are only (N-1) transitions so either the first or the last is to be marked as discontinuous.

My preference: we keep gherkin-rules as and and revert the attribute description to the one from IFC4.

[jmirtsch]

Thanks for the reply. When you refer to other documentation (ie Segments on https://ifc43-docs.standards.buildingsmart.org/IFC/RELEASE/IFC4x3/HTML/lexical/IfcCompositeCurve.htm ) it's clear the transition code does refer to the following segment. It seems a flaw in the curve segment documentation for 4.3. I am in favor of the proposal to change the documentation and retain the gherkin rule.