Python binding for curve_network_to_surface

[wattrg]

Hi,

Firstly thanks for maintaining this great software!

I have a question about the python bindings for curve_networks_to_surface. I'm attempting to run the example at examples/python_internal/example_gordon_surface.py. I'm on a new version of pythonocc and tigl (pythonocc-core 7.6.2 and tigl 3.5.0rc1, both from conda), so I had to make some modifications to the pythonocc imports. When I run the code I get the following error message. (I've added some print statements at the top to convince myself that the brep files are being read and converted to curvers correctly)

Guide shapes: <class 'TopoDS_Shape'; Type:Compound>
Profile shapes: <class 'TopoDS_Shape'; Type:Compound>
Profile curves: [<class 'Geom_Curve'>, <class 'Geom_Curve'>, <class 'Geom_Curve'>, <class 'Geom_Curve'>, <class 'Geom_Curve'>]
Guide curves: [<class 'Geom_Curve'>, <class 'Geom_Curve'>, <class 'Geom_Curve'>]
Traceback (most recent call last):
  File "/home/rob/Documents/00_code/sandbox/gordon_surface/gordon_surface.py", line 60, in <module>
    main()
  File "/home/rob/Documents/00_code/sandbox/gordon_surface/gordon_surface.py", line 56, in main
    surface = curve_network_to_surface(profile_curves, guide_curves, 1.e-4)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/rob/.opt/miniconda3/envs/tigl/lib/python3.12/site-packages/tigl3/geometry.py", line 3049, in curve_network_to_surface
    return _geometry.curve_network_to_surface(profiles, guides, spatialTol)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
TypeError: Wrong number or type of arguments for overloaded function 'curve_network_to_surface'.
  Possible C/C++ prototypes are:
    tigl::curveNetworkToSurface(std::vector< opencascade::handle< Geom_Curve >,std::allocator< opencascade::handle< Geom_Curve > > > const &,std::vector< opencascade::handle< Geom_Curve >,std::allocator< opencascade::handle< Geom_Curve > > > const &,double)
    tigl::curveNetworkToSurface(std::vector< opencascade::handle< Geom_Curve >,std::allocator< opencascade::handle< Geom_Curve > > > const &,std::vector< opencascade::handle< Geom_Curve >,std::allocator< opencascade::handle< Geom_Curve > > > const &)

So my question is if anyone has used these bindings recently, or if maybe they are a bit stale, or if I am doing something wrong? I can see the standalone library, is using that preferable to the version within tigl itself? I was just hoping to use the tigl version because I already have it as a dependency.

Thanks for your help!

[joergbrech]

Hi @wattrg, thanks for reaching out! I just verified that the example works on my machine. Also, I didn't have to modify any of the import statements.

• You are talking about this example, right?
• Can you send me the output of conda list?
• Did you install pythonocc-core from our channel? This conda package is build with a specific OCCT patch that we use.

---

EDIT: The bindings are not stale. They are actively maintained by us. That being said, they are currently not well tested and it is possible that some examples become outdated every once and again and that a few C++ features of TiGL are not properly exposed to Python.

[wattrg]

Ahh I see where I went wrong. I followed a link from an issue on another github repo, which took me to an old example. The current one works properly for me.

I had been using pythonocc-core from conda-forge, not realising you had your own patched version, that is very helpful to know.

Thanks very much for your help!