IfcCompositeCurve missing Derive: Dim attribute #74

pjanck · 2021-03-19T09:12:25Z

Looking at the code below, though: there is no DERIVE: Dim parameter specified that would return Segments[0].Dim or something similar. So it makes sense, that the checker returns false.

https://github.com/bSI-InfraRoom/IFC-Specification/blob/64f1fc0e8cdf8aba82d2b99a423ddefae2f5a2ba/IFC4x3_RC3.exp#L5105-L5119

Originally posted by @pjanck in #73 (comment)

The text was updated successfully, but these errors were encountered:

SergejMuhic · 2021-03-19T10:57:56Z

Can you post the checker message? IfcCurve has the derived Dim.

The IfcCurve.DIm calls IfcCurveDim, so I am not sure what the checker is complaining about.

larswik · 2021-03-19T11:13:11Z

Looks like it is connected to the "dim" warning.

Errors:

    <MVDMessage>
      <MessageType>ERROR</MessageType>
      <MVDType>SCHEMADEFINITION</MVDType>
      <MessageCode>IFCSHAPEREPRESENTATION.CORRECTITEMSFORTYPE</MessageCode>
      <Instance>#56=IFCSHAPEREPRESENTATION(#55,'Axis','Curve2D',(#50))</Instance>
      <InstanceLink></InstanceLink>
      <ObjectInstance>#24=IFCALIGNMENTHORIZONTAL('3$D$u$Lvj4IRsKVLePDIuA',$,$,$,$,#21,#57,2103.72056)</ObjectInstance>
      <ObjectInstanceLink></ObjectInstanceLink>
      <Messages>
        <MessageString> where rule 'IFCSHAPEREPRESENTATION.CORRECTITEMSFORTYPE' evaluates to FALSE</MessageString>
      </Messages>
    </MVDMessage>
    <MVDMessage>
      <MessageType>ERROR</MessageType>
      <MVDType>SCHEMADEFINITION</MVDType>
      <MessageCode>IFCSHAPEREPRESENTATION.CORRECTITEMSFORTYPE</MessageCode>
      <Instance>#107=IFCSHAPEREPRESENTATION(#55,'Axis','Curve3D',(#106))</Instance>
      <InstanceLink></InstanceLink>
      <ObjectInstance>#65=IFCALIGNMENTVERTICAL('1bR$_QwLnBJv2m8fn2FwZ0',$,'PR_Twin_Branch_section',$,$,#21,#108)</ObjectInstance>
      <ObjectInstanceLink></ObjectInstanceLink>
      <Messages>
        <MessageString> where rule 'IFCSHAPEREPRESENTATION.CORRECTITEMSFORTYPE' evaluates to FALSE</MessageString>
      </Messages>
    </MVDMessage>

Warnings (maybe one per curve segment?):

    <MVDMessage>
      <MessageType>WARNING</MessageType>
      <MVDType>SCHEMADEFINITION</MVDType>
      <MessageCode>SYNTAX_VALIDATION_22</MessageCode>
      <Instance>#42=IFCCURVESEGMENT(CONTINUOUS,#36,0.,741.371391339357,#41)</Instance>
      <InstanceLink></InstanceLink>
      <ObjectInstance>#24=IFCALIGNMENTHORIZONTAL('3$D$u$Lvj4IRsKVLePDIuA',$,$,$,$,#21,#57,2103.72056)</ObjectInstance>
      <ObjectInstanceLink></ObjectInstanceLink>
      <Messages>
        <MessageString> access to non-existing attribute 'dim'</MessageString>
      </Messages>
    </MVDMessage>
    <MVDMessage>
      <MessageType>WARNING</MessageType>
      <MVDType>SCHEMADEFINITION</MVDType>
      <MessageCode>SYNTAX_VALIDATION_22</MessageCode>
      <Instance>#78=IFCCURVESEGMENT(CONTINUOUS,#76,-63.4469995157475,346.277532651367,#77)</Instance>
      <InstanceLink></InstanceLink>
      <ObjectInstance>#65=IFCALIGNMENTVERTICAL('1bR$_QwLnBJv2m8fn2FwZ0',$,'PR_Twin_Branch_section',$,$,#21,#108)</ObjectInstance>
      <ObjectInstanceLink></ObjectInstanceLink>
      <Messages>
        <MessageString> access to non-existing attribute 'dim'</MessageString>
      </Messages>
    </MVDMessage>

SergejMuhic · 2021-03-19T11:53:23Z

This is not IfcCompositeCurve though. It is IfcCurveSegment. This was addressed though in c5a9282615009b56fb6b2e8eb6faf25f7241d2db. I guess Andreas did not use the latest exp file to insert the new WRs. I have not inserted the new WRs into ifcdoc as we were testing. I have to do that first and then regenerate the exp for him. Probably some time next week, unless it is super urgent.

pjanck · 2021-03-19T13:04:06Z

Right, I see. So just an update on the checker is necessary?

larswik · 2021-03-22T10:39:54Z

Please let me know when the checker has been updated.

larswik · 2021-03-30T08:01:03Z

I tested with new alignments and new checker and got this message. It happens for the IfcGradientCurve and I suspect that it has to do with IfcPolynomialCurve...

  <MVDMessage>
      <MessageType>ERROR</MessageType>
      <MVDType>SCHEMADEFINITION</MVDType>
      <MessageCode>IFCCOMPOSITECURVE.SAMEDIM</MessageCode>
      <Instance>#106=IFCGRADIENTCURVE((#78,#82,#87,#91,#96,#100,#105),UNKNOWN,#50,$)</Instance>
      <InstanceLink></InstanceLink>
      <ObjectInstance>#65=IFCALIGNMENTVERTICAL('1mggcWhRT1buFoX8TmC_xS',$,'PR_Twin_Branch_section',$,$,#21,#108)</ObjectInstance>
      <ObjectInstanceLink></ObjectInstanceLink>
      <Messages>
        <MessageString> where rule 'IFCCOMPOSITECURVE.SAMEDIM' evaluates to FALSE</MessageString>
      </Messages>
    </MVDMessage>
  </MVDMessageSet>

SergejMuhic · 2021-03-30T08:09:49Z

Do your segments have a dim of 2? I think I have to update the CurveDim for IfcPolynomialCurve then. It currently returns 3. Probably should return 2 if one of the coefficient lists is '$'.

larswik · 2021-03-30T08:43:04Z

CoefficientsZ = '$'

SergejMuhic · 2021-03-30T09:11:41Z

Thanks @larswik. Compiling this function is going to be interesting. :)

SergejMuhic · 2021-03-30T12:18:26Z

Here my proposal:

CurveDim function update for IfcPolynomialCurve:
IF ('IFCGEOMETRYRESOURCE.IFCPOLYNOMIALCURVE' IN TYPEOF(Curve))
IF (NOT EXISTS(Curve\IfcPolynomialCurve.CoefficientsX) OR NOT EXISTS(Curve\IfcPolynomialCurve.CoefficientsY) OR NOT EXISTS(Curve\IfcPolynomialCurve.CoefficientsZ))
THEN RETURN(2);
ELSE
THEN RETURN(3);
END_IF
END_IF;
Where rule for IfcPolynomialCurve:
ValidCoefficient : (EXISTS(CoefficientsX) AND EXISTS(CoefficientsY)) OR (EXISTS(CoefficientsX) AND EXISTS(CoefficientsZ)) OR (EXISTS(CoefficientsY) AND EXISTS(CoefficientsZ))

Looking forward to feedback.

larswik · 2021-03-30T12:50:05Z

Needs at least X/Y, X/Z or Y/Z and if one of X, Y or Z is missing then dim=2, else 3. Looks ok to me!

SergejMuhic · 2021-03-30T17:16:07Z

Probably should fix the where clause to:
ValidCoefficient : (EXISTS(CoefficientsX) AND EXISTS(CoefficientsY)) OR (EXISTS(CoefficientsX) AND EXISTS(CoefficientsZ)) OR (EXISTS(CoefficientsY) AND EXISTS(CoefficientsZ) OR (EXISTS(CoefficientsX) AND EXISTS(CoefficientsY) AND EXISTS(CoefficientsZ))

jmirtsch · 2021-03-30T22:27:24Z

As the polynomial has a placement, seems to me it should also validate if the placement is 2d or 3d.
It would be quite good if we could start establishing a unit test based approach to cross check the where rule (different scenarios of 2d and 3d configurations as well as valid and invalid).

SergejMuhic · 2021-03-31T09:55:24Z

Yep, that was my first thought also. But IfcAxis2PlacementLinear gets its Dim from its basis curve. Then we could have a problem. Probably a polynomial would not be placed on itself, but you never know. :) Seemed to be more explicit to do it based on coefficients because there is no other restriction in setting the Coefficients.

Possible scenario: Could be that I have a dataset where IfcPolynomialCurve.Placement is Dim 2 but all three Coefficients are set. It would not be forbidden, but how to interpret it?

SergejMuhic · 2021-04-01T13:32:53Z

Any objections?

pjanck · 2021-04-06T06:55:33Z

Discussion during the tech call - proposal:

IfcPolynomialCurve: Introduce a where rule that enforces SameDim for the number of coefficients and placement:

CorrectPlacementDim : (NOT EXISTS(CoefficientsX) OR NOT EXISTS(CoefficientsY) OR NOT EXISTS(CoefficientsZ)) OR (Position.Dim = 3);

Reasoning:

a curve that is missing one of the coefficients can still be placed in 3D (but may as well be placed in 2D),
a curve that has all coefficients, HAS to be placed 3D.

This works well with the ValidCoefficients where rule.

The dimension is then handled within IfcCurveDim:

IF ('IFC4X3_RC3.IFCPOLYNOMIALCURVE' IN TYPEOF(Curve))
    IF ( NOT EXISTS(Curve\IfcPolynomialCurve.CoefficientsZ) AND Curve\IfcPolynomialCurve.Position.Dim = 2 )
        THEN RETURN(2);
    END_IF;
    RETURN(3);
END_IF;

This rule assumes that as soon as we have a z-dimension, the curve is 3D. This argumentation is based on the way conics handle their dimension.

AlexBrad1eyCT · 2021-04-12T09:12:38Z

Is this Resolved, by This Commit, if so can we close the issue??

SergejMuhic · 2021-04-13T09:13:41Z

By itself, this rule does not enforce the same Dim of placement and curve. It states that if one of the coefficients is missing Placement.Dim does not have to be 3.

To actually enforce the dims to be the same, we would have to look at it from a different direction. We have two valid Position.Dim values. If we are enforcing those, we need to cover that one case where Position.Dim is 2. In my view, a valid definition looks like this: either Position.Dim equals 2 and CoefficientsZ is empty or Position.Dim is 3. So, my proposal would be:

((Position.DIm=2) AND (NOT EXISTS(CoefficientsZ))) OR (Position.Dim=3)

Or fix text of where rule.

pjanck · 2021-04-13T09:16:12Z

Correction for my post above:

Introduce a where rule that enforces SameDim for the number of coefficients and placement:

Should read:

Introduce a where rule that enforces CorrectPlacementDim for the number of coefficients and placement:

pjanck · 2021-04-13T09:28:24Z

Proposed documentation for the where rule:

A curve that is missing one of the coefficients can be positioned in 2D or 3D; otherwise, its position has to be in 3D.

Possible namings:

CorrectPositionDim
Enforce3dPositionFor3dCurve
ValidPositionDim

SergejMuhic · 2021-04-22T17:55:35Z

Closing. See commit bSI-InfraRoom/IFC-Specification@c6c3aa8

pjanck assigned SergejMuhic Mar 19, 2021

pjanck added the BUG A bug in code or checking actions label Mar 19, 2021

pjanck mentioned this issue Mar 19, 2021

IfcCompositeCurve::Dim attribute missing bSI-InfraRoom/IFC-Specification#12

Closed

SergejMuhic assigned jmirtsch Mar 30, 2021

SergejMuhic closed this as completed Apr 22, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

IfcCompositeCurve missing Derive: Dim attribute #74

IfcCompositeCurve missing Derive: Dim attribute #74

pjanck commented Mar 19, 2021

SergejMuhic commented Mar 19, 2021 •

edited

Loading

larswik commented Mar 19, 2021

SergejMuhic commented Mar 19, 2021

pjanck commented Mar 19, 2021

larswik commented Mar 22, 2021

larswik commented Mar 30, 2021 •

edited

Loading

SergejMuhic commented Mar 30, 2021

larswik commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021 •

edited

Loading

larswik commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021

jmirtsch commented Mar 30, 2021

SergejMuhic commented Mar 31, 2021 •

edited

Loading

SergejMuhic commented Apr 1, 2021

pjanck commented Apr 6, 2021

AlexBrad1eyCT commented Apr 12, 2021

SergejMuhic commented Apr 13, 2021

pjanck commented Apr 13, 2021

pjanck commented Apr 13, 2021

SergejMuhic commented Apr 22, 2021 •

edited

Loading

IfcCompositeCurve missing Derive: Dim attribute #74

IfcCompositeCurve missing Derive: Dim attribute #74

Comments

pjanck commented Mar 19, 2021

SergejMuhic commented Mar 19, 2021 • edited Loading

larswik commented Mar 19, 2021

SergejMuhic commented Mar 19, 2021

pjanck commented Mar 19, 2021

larswik commented Mar 22, 2021

larswik commented Mar 30, 2021 • edited Loading

SergejMuhic commented Mar 30, 2021

larswik commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021 • edited Loading

larswik commented Mar 30, 2021

SergejMuhic commented Mar 30, 2021

jmirtsch commented Mar 30, 2021

SergejMuhic commented Mar 31, 2021 • edited Loading

SergejMuhic commented Apr 1, 2021

pjanck commented Apr 6, 2021

AlexBrad1eyCT commented Apr 12, 2021

SergejMuhic commented Apr 13, 2021

pjanck commented Apr 13, 2021

pjanck commented Apr 13, 2021

SergejMuhic commented Apr 22, 2021 • edited Loading

SergejMuhic commented Mar 19, 2021 •

edited

Loading

larswik commented Mar 30, 2021 •

edited

Loading

SergejMuhic commented Mar 30, 2021 •

edited

Loading

SergejMuhic commented Mar 31, 2021 •

edited

Loading

SergejMuhic commented Apr 22, 2021 •

edited

Loading