[vtkusers] matrix output VERY hard to read

[David Gobbi]

Yes, that's true, I guess it really is that easy.

Still, I wish that VTK did have a vtkPrintFormat object because it
would make writing PrintSelf methods a lot less tedious.

For example, there are tons of lines of code in VTK like this
(probably many with little errors and inconsistencies here and there):

 os << "Origin: " << this->Origin[0] << " " << this->Origin[1] << " "
<< this->Origin[2] << "\n";

These could have been simply like this:

 format->PrintAttribute(os, "Origin", this->Origin, 3);

Or even a macro:

 vtkPrintAttributeMacro(Origin,3);

Unfortunately all those "<<" operators have already been typed... but
it's not too late to switch to a better system.

   David


On Wed, Oct 21, 2009 at 5:08 PM, Francois Bertel
<francois.bertel at kitware.com> wrote:
> You don't need all these things if you're working in C++.
> You just have to change the flag on your ostream:
>
> Example with cout:
>
> vtkObject *a;
>
> cout.setf(ios::scientific,ios::floatfield); // ios::fixed or ios::scientific
> cout.precision(5);
>
> a->Print(cout);
>
> ref: http://www.cplusplus.com/reference/iostream/ostream/
> ref: http://www.cplusplus.com/reference/iostream/ios_base/fmtflags/
>
>
> On Wed, Oct 21, 2009 at 7:03 PM, David Doria <daviddoria+vtk at gmail.com> wrote:
>> On Wed, Oct 21, 2009 at 6:52 PM, David Gobbi <david.gobbi at gmail.com> wrote:
>>> I agree with Jeff that a method for pretty-printing would be handy.  A
>>> generic way of doing this is to create a "formatting" object that can
>>> be sent to PrintSelf:
>>>
>>> void PrintSelf(ostream& os, vtkIndent indent, vtkPrintFormat *format);
>>>
>>> The "format" object could provide information about precision, column
>>> widths, and "level of detail" i.e. whether to print all attributes or
>>> just the important ones i.e. the data values.
>>>
>>> This would be an optional method, i.e. if any object that didn't
>>> define the new form of PrintSelf would fall back to the old PrintSelf.
>>>  The reason to modify PrintSelf, instead of adding new methods, is
>>> because it makes it easy to convert relevant object (matrix,
>>> transform, array) over to the new system.
>>>
>>> There could be two ways of invoking the pretty-print methods:
>>>
>>> vtkPrintFormat::Print(ostream &stream, vtkObject *object);
>>>
>>> vtkObject::Print(ostream &stream, vtkPrintFormat *format);
>>>
>>> Overall, this would be fairly easy to implement and would be very
>>> general-purpose.  In fact it is general enough to even encompass such
>>> things as serialization of VTK objects.
>>>
>>>    David
>>> ___
>>
>> David Gobbi,
>>
>> That sounds like a very reasonable idea to me. Are you or Jeff
>> interested in setting up the framework? Once it's setup, I'd be happy
>> to try to write a few of the Print functions. We should make a list of
>> things that could really use the new idea: vtkMatrix4x4,
>> vtk*Transform, vtk*Array, ?
>>
>> This "level of detail" is also a nice feature, almost all the classes
>> could eventually be updated to have a few different levels of detail.
>>
>> Thanks,
>>
>> David
>> _______________________________________________
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>>
>> Please keep messages on-topic and check the VTK FAQ at: http://www.vtk.org/Wiki/VTK_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtkusers
>>
>
>
>
> --
> François Bertel, PhD  | Kitware Inc. Suite 204
> 1 (518) 371 3971 x113 | 28 Corporate Drive
>                      | Clifton Park NY 12065, USA
> _______________________________________________
> Powered by www.kitware.com
>
> Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
>
> Please keep messages on-topic and check the VTK FAQ at: http://www.vtk.org/Wiki/VTK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.vtk.org/mailman/listinfo/vtkusers
>