This allowed me to remove its custom GetAttr method.

There is no guarantee that tp_name is unique, and it would be bad
if someone fooled the type checking by simply creating a new type
with the same name as an existing type.

The casts are unecessary for Python 2.5 and higher.

This reduces the number of exported symbols.  Wrapped enum types are
still exported in order to allow them to be type-checked elsewhere.

This includes the use of PyVTKMethodDescriptor, which fixes the issue
where overloads of static and non-static methods did not work.

Now that the python wrappers use type objects for VTK classes, calling
type(obj) returns the class type, rather than 'vtkobject'.

This is the first step towards Python 3 compatibility. Wrapped VTK classes
used to use the old-style python metaclass mechanism, and they worked like
the old-style classes from python 1 and python 2. Now each wrapped VTK
class is a new-style class with its own PyTypeObject.

These were required up to Python 2.4, which is no longer supported.

This allows a lot of obsolete code to be removed.

Ben Boeckel authored 9 years ago and Kitware Robot committed 9 years ago

2166c067 FiltersStatisticsGnuR: fix R_NO_REMAP juggling

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Brad King <brad.king@kitware.com>
Merge-request: !437

Joachim Pouderoux authored 9 years ago and Kitware Robot committed 9 years ago

bf1cd7e2 Change iterator from postincrement to preincrement
5c97e44a Move variable declaration inside loop to limit its scope
5e1fc1bf Fix indentation and restore empty line
761a643e Add field data to vtkUniformGrid to save GDALDataset NoDataValue

Acked-by: Kitware Robot <kwrobot@kitware.com>
Merge-request: !344

Fix uninitialized memory reported by valgrind.

23dff2bd introduced a bug where it didn't copy the point data correctly
for all the layers. Fixed that.

See merge request !438

23dff2bd introduced a bug where it didn't copy the point data correctly
for all the layers. Fixed that.

Berk Geveci authored 9 years ago and Kitware Robot committed 9 years ago

f4126e3a Fixed bug with loading 2D or ASCII datasets in parallel.

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Utkarsh Ayachit <utkarsh.ayachit@kitware.com>
Merge-request: !434

David Lonie authored 9 years ago and Kitware Robot committed 9 years ago

1118440f Make ScalarBarRepresentation orientation set-able.

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Utkarsh Ayachit <utkarsh.ayachit@kitware.com>
Merge-request: !423

Had to move the code that broadcasted settings to ranks > 1
because the settings are needed to make a decision about
2D / ASCII parallel code path.

Berk Geveci authored 9 years ago and Kitware Robot committed 9 years ago

d91ecf98 BUG: Was using the wrong extent in computation.
005e70d3 Fixed warnings.
5f15c59d Cleaned up error handling in plot3d reader.

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Utkarsh Ayachit <utkarsh.ayachit@kitware.com>
Merge-request: !433

- Minor error handling cleanup.
- Made sure that reading error on rank 0 causes all
ranks to report error.

Berk Geveci authored 9 years ago and Kitware Robot committed 9 years ago

557ec619 Error checking: check status of collective call.
d9555377 Fix typo in call to superclass method.

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Berk Geveci <berk.geveci@kitware.com>
Merge-request: !432

Ensure we check status for all collective calls correctly.

Copy/paste error!

Utkarsh Ayachit authored 9 years ago and Kitware Robot committed 9 years ago

70cb8dde ENH: Move test that discovers inconsistent structured grids in an AMR grid.

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Berk Geveci <berk.geveci@kitware.com>
Merge-request: !365

Utkarsh Ayachit authored 9 years ago and Kitware Robot committed 9 years ago

07869098 Win64 fixes for vtkMultiBlockPLOT3DReader.
9c1667f6 Do not override DATA_NUMBER_OF_GHOST_LEVELS() unnecesarily.
0027efa6 Do not rexecute based on ghosts for serial execution
1c0cfb9a Improved ghost array generation.
44c7f0b8 Cleaning python testing macros.
695b4817 Adding a basic test to test vtkMPIMultiBlockPLOT3DReader.
bc18ac79 Add vtkMPIMultiBlockPLOT3DReader -- MPI-IO enabled reader for Plot3D.
d3dee7d6 Make vtkMultiBlockPLOT3DReader factory overridable.
...

Acked-by: Kitware Robot <kwrobot@kitware.com>
Reviewed-by: Berk Geveci <berk.geveci@kitware.com>
Merge-request: !428

vtkMultiBlockPLOT3DReader now handles reading large files on Win64
machines. We need to use 64-bit API on Windows since sizeof(long) is 4
bytes on Windows.

If an algorithm produces more ghost levels than asked for, it
should be able to set DATA_NUMBER_OF_GHOST_LEVELS() and let the
pipeline know. This was not possible because the executive was
overwriting the value.

- Added the option to generate only cell values.
- Bypass if there are no ghosts
- Do not replace existing ghost arrays. Add to them.

When attemping to add tests using vtk_add_test_python_mpi(), issues were
spotted in the vtkTestingMacros code. This commit addresses those
issues.

VTK_PYTHON_EXE is no longer globally defined in CMakeLists.txt.
vtk_add_test_python_mpi() and vtk_add_test_python() assume a good
default is none is set (this allows for applications like ParaView to
use these macros while overriding the executable to run).

We use generator expressions to point to the default
vtkpython/pvtkpython executables which works better when MPI
(mpiexec/mpirun) is being used to launch the test executable.

Also fixed vtkTestingMacros to properly locate vtkpython executable when
adding MPI tests.

Adding new vtkMPIMultiBlockPLOT3DReader which uses MPI-IO to read plot3D
files in parallel. vtkMPIMultiBlockPLOT3DReader overrides
vtkMultiBlockPLOT3DReader using vtkObjectFactory mechanisms. It only
supports reading binary 3D-Plot3D files.

Make vtkMultiBlockPLOT3DReader use vtkObjectFactoryNewMacro so that it
can be overridden using vtkObjectFactory.

Updating vtkMultiBlockPLOT3DReader to enable parallel reading support.
vtkMultiBlockPLOT3DReader can now read subextents in parallel on
different ranks using POSIX I/O.

In preparation to add parallel support to Plot3D reader.