[vtk-developers] Generic method for getting a sub-cell

[Will Schroeder]

I've seen this problem too (and fixed it in my versions of VTK). I
think in the initial implementation the developers were more
interested in knowing whether there was a hit, and getting that
information as quickly as possible (hence bailing out of the loop). In
recent times, users have started depending on VTK to provide precise
information....I think it's time to change this.

On Fri, Apr 2, 2010 at 10:33 AM, David Gobbi <david.gobbi at gmail.com> wrote:
> Speaking of subcells, I've identified a problem with the way that
> IntersectWithLine is implemented in vtkTriangleStrip et al.  The
> way IntersectWithLine is implemented in these classes is:
>
> for triangle in strip:
>  if triangle->IntersectWithLine(tol)
>     return
>
> In other words, the loop alway stops at first hit, but due to
> tolerance there might be more than one potential hit.  It
> could, instead, look for all potential hits and then use
> ParametricDistance() as a tiebreaker.
>
>  - David
>
>
> On Thu, Apr 1, 2010 at 2:38 PM, David Gobbi <david.gobbi at gmail.com> wrote:
>> Hi Will,
>>
>> Yes, all the efficient (npts, pts) API styles would also have to be
>> supported.  Ideally every GetCell method signature would need an
>> equivalent GetSubCell method.  The following classes would need to be
>> changed:
>>
>> vtkDataSet
>> vtkCell (the base class)
>> vtkGenericCell
>> vtkTriangleStrip et all
>>
>> You are correct that most of the cell classes would be unaffected.
>> Only vtkTriangleStrip and similar classes would have to be modified.
>>
>>  David
>>
>>
>> On Thu, Apr 1, 2010 at 2:17 PM, Will Schroeder
>> <will.schroeder at kitware.com> wrote:
>>> David in the long term I don't see why we couldn't easily add
>>> something like this to both the cell and dataset APIs. To minimize the
>>> initial effort (i.e., so you don't have to do it to all the datasets)
>>> I'd initially just add it to the affected cell types (e.g., triangle
>>> strip, polyline, polyvert, etc.) Also, if I understand correctly you
>>> are returning a new cell (or a pointer to an internal cell instance),
>>> have you looked at alternative APIs to return the information faster,
>>> and more compactly? I'm thinking of something akin to
>>> vtkCellArray::GetNextCell(npts,pts)...just curious.
>>>
>>>
>>> On Thu, Apr 1, 2010 at 1:40 PM, David Gobbi <david.gobbi at gmail.com> wrote:
>>>> Hi All,
>>>>
>>>> Would it be possible to add a generic method for getting a sub-cell from a cell?
>>>>
>>>> vtkCell *cell->GetSubCell(int subId);
>>>> void cell->GetSubCell(int subId, vtkGenericCell *subCell);
>>>>
>>>> Even better would be an easy way of iterating through the sub-cells in
>>>> a vtkDataSet directly:
>>>>
>>>> data->GetNumberOfSubCells(vtkIdType cellId);
>>>> data->GetSubCell(vtkIdType cellId, int subId, vtkGenericCell *subCell);
>>>>
>>>> There are several times when I've had to add special cases to my code
>>>> to handle triangle strips, and if there was a generic method for
>>>> getting sub-cells, my code could be a lot simpler.
>>>>
>>>>   David
>>>> _______________________________________________
>>>> 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
>>>
>>
>



-- 
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