HLR for 10-20 times faster

Hi guys,
The HLR algoritm (not Poly) is really working for 10-20 times faster when we put the sewed shape in.

for the sample:

...
Handle_HLRBRep_Algo myHlAlgo = new HLRBRep_Algo();
BRepBuilderAPI_Sewing mySewing(1.0e-06,true,false,false,false);

for(aListIterator.Initialize(aList);aListIterator.More();aListIterator.Next())
{
...
mySewing.Add(myShape);
}

mySewing.Perform();
myHlAlgo->Add(mySewing.SewedShape(),0);
myHlAlgo->Update();
myHlAlgo->Hide();
...

In this jpg-image without optimization I'm need wait for 10 seconds while building HLR, but now only 0.91 of second.

Attachments: 
JuryS's picture

Maybe this optimization can be used when we are produce the boolean operations ? I think that the boolean operations like fuse, cut or common may work fastly with next?

Fotis Sioutis's picture

Having a topologically and geometrically correct shape is a must for all occ operations (specially for boolean) ! Sewing is only one fix that should be applied among the others !

Fotis

JuryS's picture

I'm make some tests around this. And this only option how to disable compute internal intersections. There is more easy way with HLR Hider. Only to check status of line and thats all. Sewed shape can't be fastly with boolean operation. Only in some case when some shapes build with compound and don't sewing before...

heXus's picture

Hello, Yuriy Sinithin! I'm saw your picture. And I have some questions to you. Please, contact with me by mail:
denis_andreevich (a) mail dot ru