Bryan Cole a écrit :
> Hi,
>
Hi Bryan,
> I've started experimenting with TopExp_Explorer in order extract
> sub-shapes from Shapes. When I try to store the results of the iteration
> in a python list, all but the last extracted shapes become invalid. It
> seems that TopExp_Explorer is re-using some part of the data structure
> it outputs as it iterates, so keeping references to the shapes it
> generates doesn't work.
>
> Here's an example:
>
> from OCC import TopExp, BRepPrimAPI, TopAbs, TopoDS
>
> box = BRepPrimAPI.BRepPrimAPI_MakeBox(10., 20., 30.)
> ex = TopExp.TopExp_Explorer(box.Shape(), TopAbs.TopAbs_EDGE)
>
> results = []
>
> while ex.More():
> edge = TopoDS.TopoDS().Edge(ex.Current())
> print "is null?", bool(edge.IsNull())
> results.append(edge)
> ex.Next()
>
> for edge in results:
> print "null now?", bool(edge.IsNull())
>
> Inside the while loop, none of the shapes are Null. Once the iteration
> is complete, all the edges become Null.
>
At the end of the iteration, e.g. when ex.More() is False, the
TopExp_Explorer instance is cleared (with the Clear() method). Then all
the references to subshapes are lost and your list contains NULL elements.
> What's the correct way to store a references to sub-Shapes? I guess I
> need a shallow copy of the shape-object returned by Explorer.Current()
> but I can't figure how to achieve this.
>
You just have to ReInit() the TopExp_Explorer instance at the end of the
loop.
Here is the corrected code:
from OCC import TopExp, BRepPrimAPI, TopAbs, TopoDS
box = BRepPrimAPI.BRepPrimAPI_MakeBox(10., 20., 30.)
ex = TopExp.TopExp_Explorer(box.Shape(), TopAbs.TopAbs_EDGE)
results = []
while ex.More():
edge = TopoDS.TopoDS().Edge(ex.Current())
print "is null?", bool(edge.IsNull())
results.append(edge)
ex.Next()
ex.ReInit() #to avoid loosing reference to objects found in the loop
for edge in results:
print "null now?", bool(edge.IsNull())
> cheers,
>
Cheers,
> Bryan
>
Thomas
_______________________________________________
Pythonocc-users mailing list
Pythonocc-users@gna.org
https://mail.gna.org/listinfo/pythonocc-users
