[vtk-developers] A possible bug and questions regarding the "polyhedral challenge" branch

[Takuya OSHIMA]

Thanks for the clarifications, shall try it out!

Takuya

From: Will Schroeder <will.schroeder at kitware.com>
Subject: Re: [vtk-developers] A possible bug and questions regarding the "polyhedral challenge" branch
Date: Thu, 1 Apr 2010 06:32:09 -0400

> Thanks for the feedback Takuya. Hua and I are aware of some of these
> issues (the work is actively being implemented). This will definitely
> help us produce better code in the end.
> 
> Based on what I understand of your questions and on the current state
> of development the answers to your questions are 1) we haven't fully
> addressed this yet, for many algorithms it doesn't matter but I
> believe that in the end we will have to enforce a consistent ordering
> (for rendering if nothing else); 2) global ids, inside the class a
> mapping to local ids is made; 3) FindCell() will work, further testing
> and possibly implementation will be required to make this so.
> 
> W
> 
> On Thu, Apr 1, 2010 at 5:59 AM, Takuya OSHIMA
> <oshima at eng.niigata-u.ac.jp> wrote:
>> Hi,
>>
>> I have a possible bug report and a couple of questions regarding the
>> "polyhedral challenge" fork of VTK (git://github.com/huayang/VTK.git).
>>
>> The possible bug is in vtkUnstructuredGrid::InsertNextCell() that is
>> used to insert a polyhedron. When a polyhedron is inserted for the
>> first time in a vtkUnstructuredGrid that already has non-polyhedral
>> cells, perhaps either FaceLocations should be padded until the current
>> position (as attached below), or offset to the current position should
>> be memorized and subtracted from cellId of
>> this->FaceLocations->GetValue(cellId) in
>> vtkUnstructuredGrid::GetFaces().
>>
>> The questions are as follows, which I hope to be clarified in the
>> Doxygen docs eventually.
>>
>> 1. When inserting a polyhedral cell, does the ordering of the points
>> in the faces array matter (clockwise/counterclockwise from inside of
>> the cell)?
>>
>> 2. Should the point labels in the faces array be numbered globally, or
>> locally?
>>
>> 3. Is vtkUnstructuredGrid::FindCell() supposed to work for polyhedral
>> cells?
>>
>> Thanks for all the efforts of the developers!
>>
>> Takuya
>>
>> diff --git a/Filtering/vtkUnstructuredGrid.cxx b/Filtering/vtkUnstructuredGrid.cxx
>> index a5c0deb..a6b917a 100644
>> --- a/Filtering/vtkUnstructuredGrid.cxx
>> +++ b/Filtering/vtkUnstructuredGrid.cxx
>> @@ -900,6 +900,11 @@ InsertNextCell(int type, vtkIdType npts, vtkIdType *pts,
>>     this->Faces->Allocate(this->Types->GetSize());
>>     this->FaceLocations = vtkIdTypeArray::New();
>>     this->FaceLocations->Allocate(this->Types->GetSize());
>> +    // FaceLocations must be padded until the current position
>> +    for(vtkIdType i = 0; i <= this->Types->GetMaxId(); i++)
>> +      {
>> +      this->FaceLocations->InsertNextValue(-1);
>> +      }
>>     }
>>
>>   // Okay the faces go in
>> @@ -932,6 +937,11 @@ vtkIdType vtkUnstructuredGrid::SetAllFacesAtOnce(vtkCellArray *faces)
>>     this->Faces->Allocate(this->Types->GetSize());
>>     this->FaceLocations = vtkIdTypeArray::New();
>>     this->FaceLocations->Allocate(this->Types->GetSize());
>> +    // FaceLocations must be padded until the current position
>> +    for(vtkIdType i = 0; i <= this->Types->GetMaxId(); i++)
>> +      {
>> +      this->FaceLocations->InsertNextValue(-1);
>> +      }
>>     }
>>
>> Takuya OSHIMA, Ph.D.
>> Faculty of Engineering, Niigata University
>> 8050 Ikarashi-Ninocho, Nishi-ku, Niigata, 950-2181, JAPAN
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtk-developers
>>
>>
> 
> 
> 
> -- 
> William J. Schroeder, PhD
> Kitware, Inc.
> 28 Corporate Drive
> Clifton Park, NY 12065
> will.schroeder at kitware.com
> http://www.kitware.com
> (518) 881-4902