[Paraview-developers] ParaView and rendering composite datasets

[Paul Edwards]

Utkarsh,

I've only just updated and noticed that some of my custom representations
are not working now.  I have been subclassing vtkPVGeometryFilter and
implementing RequestData as follows:

8<---------------------------------------------------------------------
int vtkMyGeometryFilter::RequestData(vtkInformation* request,
                                     vtkInformationVector** inputVector,
                                     vtkInformationVector* outputVector)
{
  if (!this->Superclass::RequestData(request, inputVector, outputVector))
    {
    return 0;
    }

  vtkPolyData *output = vtkPolyData::GetData(outputVector, 0);

  vtkPolyData* clone = vtkPolyData::New();
  clone->ShallowCopy(output);

  // now pass the clone through a filter

  vtkMyFilter* filter = vtkMyFilter::New();
  filter->SetInput(clone)
  filter->Update
  output->ShallowCopy(filter->GetOutputDataObject(0));
  filter->Delete();
  clone->Delete();

  return 1;
}
--------------------------------------------------------------------->8

This now fails as output is not necessarily polydata.  What's the best
solution here?

Thanks,
Paul

On 31 August 2010 21:57, Utkarsh Ayachit <utkarsh.ayachit at kitware.com>wrote:

> Folks,
>
> Since the introduction of composite datasets into ParaView, there was
> one oddity when it came to rendering: we appended all the blocks into
> a single polydata and then rendered it. It had several advantages -
> esp. since it kept the rendering code simple esp when dealing with
> delivery and M-to-N data movement when render server was involved and
> with surface selection. However, it was a major bottleneck as well
> when dealing with really large number of blocks. The appending of data
> was slow which resulted in delays.
>
> I've just pushed a change that resolves this issue
> (http://paraview.org/Bug/view.php?id=11133). We no longer append
> blocks together before rendering. Most of the issues like
> data-delivery, selection etc. have been addressed. But if you notice
> any other quirks, please feel free to raise them so that we can
> address them as soon as possible.
>
> Utkarsh
> _______________________________________________
> Paraview-developers mailing list
> Paraview-developers at paraview.org
> http://public.kitware.com/mailman/listinfo/paraview-developers
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/paraview-developers/attachments/20100921/dbfb3b7a/attachment.htm>