rhinoscript command isBox(strObject) for VB.net

Replies to This Discussion

Permalink Reply by Vicente Soler on August 17, 2014 at 4:47pm

Weird box I got here:

Permalink Reply by Vicente Soler on August 17, 2014 at 5:36pm

Come on, lighten up.

Permalink Reply by Arend on August 17, 2014 at 5:44pm

This might work. Check for 6 faces, get the bounding box aligned to the first face, and check if the volume is equal to the volume of the boundingbox.


bool IsBrep (Brep x)
{
if (!(x.Faces.Count == 6))
{
return false;
}

// check for faces
Brep face = x.Faces[0].DuplicateFace(false);
if (!(face.Vertices.Count == 4))
{
return false;
}

// check volume
Plane pln = new Plane(face.Vertices[0].Location, face.Vertices[1].Location, face.Vertices[2].Location);
Box b = new Box(pln, x);

// 3d degree root of the tolerance - since we're doing volumes
double tolerance = Math.Pow(RhinoDocument.ModelAbsoluteTolerance, (1.0 / 3.0));
return RhinoMath.EpsilonEquals(b.Volume, x.GetVolume(), tolerance);
}


Permalink Reply by Vicente Soler on August 17, 2014 at 6:24pm

Yep, that's basically what Rhinoscript's IsBox(obj) function is doing:

https://github.com/mcneel/rhinoscript/blob/master/IsBox.rvb

It also checks for planar surfaces but I don't think is necessary (maybe to skip when possible calculating the brep volume?). It also gets the bbox orientation by creating a surface frame rather than with 3 corner points (I don't think it matters).

The IsBox function uses the default tolerance. Why are you square rooting the tolerance? Shouldn't be larger rather than smaller? (to the power of 3).

 

Permalink Reply by Vicente Soler on August 18, 2014 at 1:49am

cube rooting i mean

Permalink Reply by Heiko Bosse on August 18, 2014 at 2:19am

ah, that's a nice link. like a dictonary rhinoscribt to vbscribt

Permalink Reply by Arend on August 18, 2014 at 3:26am

I think checking for planar surfaces can be a good thing - calculating volume can become quite expensive, when it's not really a cube, checking planarity should be cheap.

Hmm, the math on the tolerance is still tricky, I agree. But cubing it is also not a solution right? If there's a tolerance of 0.01^3 = 0.0001, while any deviation on one axis would give quite a big difference.

Any ideas?

Permalink Reply by Vicente Soler on August 18, 2014 at 4:26am

Maybe cube root the volumes and compare them to the normal tolerance.

Permalink Reply by Heiko Bosse on August 18, 2014 at 2:13am

thank you guys.

i tried to do it by my own, but failed because of my bad vb.net knowledge...