What do you mean exactly with 'crash'? Do OCC raise an exception catched by pythonOCC (like Standard_ConstructionError for instance) or is it a segfault?
If the error is not reproductible (ie. according to the wires passed to the BRepOffsetAPI_MakeOffset class, the Perform() method passes or fails), it may come from an issue in the algorithm used by OCC de perform the operation. In that case, it's really difficult to identify the problem, unless diving into the C++ code. Thomas 2010/4/18 Dave Cowden <dave.cow...@gmail.com> > Hi, Thomas, thanks! > > That definitely makes my code much more readable: but unfortunately the > crash still occurs at that same place. I also switched to using Topo in a > couple of other places ( where the face you see passed into that method is > created ), but that did not help either. > > How do you go about finding what's going on when you encounter these? > > On Sun, Apr 18, 2010 at 12:30 AM, Thomas Paviot <tpav...@gmail.com> wrote: > >> 2010/4/18 Dave Cowden <dave.cow...@gmail.com> >> >> Hi, all: >>> >>> I'm struggling with a couple of problems that I presume to be null >>> pointers that are crashing OCC. I have a feeling this may be related to >>> memory management. >>> >>> For example, I've tracked down one crash to one line: >>> BRepBuilderAPI_MakeOffset.Perform(). I load it with several wires, each of >>> which I can display successfully immediately before invoking perform. Yet, >>> Perform() crashes. See example below. I do not believe any object scoping >>> issues are at play that I can see: all of the objects in question seem to be >>> in scope when the crash happens. >>> >>> I am also fairly certain that its something I'm doing: all of the >>> 'simple' test cases I create work fine. >>> >>> So my question is: are there any tricks/techniques I can use to isolate >>> the problem when a crash occurs? Right now i'm using old school print >>> statements to get to what method call dies, but I still cannot see what >>> exactly is going wrong. Help is appreciated, this is completely kicking my >>> butt :( >>> >>> >>> #offset a face, returning the offset shape >>> def _offsetFace(self,face,offset ): >>> brt = BRepTools.BRepTools(); >>> ow = brt.OuterWire(face); >>> bo = BRepOffsetAPI.BRepOffsetAPI_MakeOffset(); >>> >>> bo.AddWire(ow); >>> >>> #now get the other wires >>> te = TopExp.TopExp_Explorer(); >>> te.Init(face,TopAbs.TopAbs_WIRE); >>> while te.More(): >>> w = ts.Wire(te.Current()); >>> TestDisplay.display.showShape(w); # this line succesfully shows the >>> wires on screen >>> if not w.IsSame(ow): >>> bo.AddWire(w); >>> te.Next(); >>> te.ReInit(); >>> print "about to offset..."; >>> >>> bo.Perform(offset,0.00001); #this line crashes hard, but only >>> sometimes. >>> print "done offsetting.."; >>> >>> if not bo.IsDone(): >>> raise Exception, "Offset Was Not Successful."; >>> else: >>> return bo.Shape(); >>> >>> >> Hi Dave, >> >> If you want to traverse topology, I suggest you use the Topo class >> available from the High-Level Topology subpackage (the 'Topo' name for this >> class is maybe not the best one). This is a wrapper over the TopExp class >> which fixes issues related to TopExp (with the use of the __hash__ method): >> >> from OCC.Utils.Topology import Topo >> >> #offset a face, returning the offset shape >> def _offsetFace(self,face,offset ): >> brt = BRepTools.BRepTools(); >> ow = brt.OuterWire(face); >> bo = BRepOffsetAPI.BRepOffsetAPI_MakeOffset(); >> >> bo.AddWire(ow); >> >> #now get the other wires >> for w in Topo(face).wires(): >> bo.AddWire(w); >> print "about to offset..."; >> >> bo.Perform(offset,0.00001); >> print "done offsetting.."; >> >> if not bo.IsDone(): >> raise Exception, "Offset Was Not Successful."; >> else: >> return bo.Shape(); >> >> Hope this helps, >> >> Regards, >> >> Thomas >> >> >> >> _______________________________________________ >> Pythonocc-users mailing list >> Pythonocc-users@gna.org >> https://mail.gna.org/listinfo/pythonocc-users >> >> > > _______________________________________________ > Pythonocc-users mailing list > Pythonocc-users@gna.org > https://mail.gna.org/listinfo/pythonocc-users > >
_______________________________________________ Pythonocc-users mailing list Pythonocc-users@gna.org https://mail.gna.org/listinfo/pythonocc-users
