- Nov 12, 2019
-
-
Allison Vacanti authored
See the vtkCellArray.h class docs for details.
-
- Nov 11, 2019
-
-
Allison Vacanti authored
The SplineDrivenImageSlicer remote module needs to accept MR#2 before it will compile with the new vtkCellArray API. Once merged, the url changes here should be reverted.
-
-
3643774c libLAS: find Boost when finding libLAS Acked-by:
Kitware Robot <kwrobot@kitware.com> Acked-by:
Ben Boeckel <ben.boeckel@kitware.com> Merge-request: !6176
-
ec9bb829 Expose VTK-m ImageConnectivity Filter Acked-by:
Sebastien Jourdain <sebastien.jourdain@kitware.com> Merge-request: !6019
-
Ben Boeckel authoredIf building with Boost 1.70 and newer, Boost uses imported targets. We need to find Boost so that libLAS' usage of the imported targets works as well. (cherry picked from commit 6b8e3db8)
-
d4721aa8 Fix warning in new GDALookup code for 32-bit IdType builds. 4284edfb Minor typo/markup fixes in DataArrayRange implementation. Acked-by:
-
Allison Vacanti authored
-
Allison Vacanti authored
-
Robert Maynard authored
-
- Nov 10, 2019
-
-
979dd604 Replaced vtkAtomicTypes.h by using std::atomic Acked-by:
-
- Nov 09, 2019
-
-
0ba48c4e Fix unneeded "extern" declaration in vtkWrapJava 9e408e90 Change macro to function to avoid export issues 1880a022 Add export macros for the WrappingTools API e0238253 Wrapping tools no longer need vtkType.h Acked-by:
-
b3fe3a26 Fixed index values for multicomponent array add33ea1 Fixed iterating over components and inadvertent cast to int 6228530d Stable value lookup for vtkGenericDataArray Acked-by:
Allison Vacanti <allison.vacanti@kitware.com> Merge-request: !6169
-
- Nov 08, 2019
-
-
David Gobbi authored
This "extern" declaration is legacy and hasn't been necessary for at least ten years.
-
David Gobbi authored
Exporting a function is easier than exporting a variable.
-
beb6307e Clean up some of the range documentation. a3429edc Fix unused variable warnings. Acked-by:
Robert Maynard <robert.maynard@kitware.com> Merge-request: !6172
-
David Gobbi authored
This library was originally forced to be static to ensure that the wrapping executables (vtkWrapPython, vtkWrapHierarchy, etc) were statically linked. The idea was that since the wrapping tools are used prior to installation, static linking was more robust, since otherwise they would have to rely on a baked-in RPATH. However, these worries were unfounded, and it is simpler and more consistent for BUILD_SHARED to control whether WrappingTools is static.
-
Allison Vacanti authored
Clang is warning about temporary lifetime extension. The examples have been reworked to avoid using this feature.
-
David Gobbi authored
The inclusion of this header file was removed Jan 2019 (1a51e69f)
-
Allison Vacanti authored
Compilers are starting to warn when an object is instantiated and only used to call static methods, such as the dispatchers. This patch cleans up some of these warnings and updates documentation. The TestScaledSOADataArrayDispatchers test has been removed, as it's redundant wrt the TestArrayDispatchers test (it doesn't test ScaledSOA functionality, just dispatching).
-
Andreas Buykx authored
-
Andreas Buykx authored
Added test to TestArrayLookup to verify lookup behavior for multicomponent array.
-
7cd9f55d Revert "Do not force WrappingTools library to be static" Acked-by:
-
ce1fcb19 FindPostgreSQL: sync with CMake's FindPostgreSQL Acked-by:
Brad King <brad.king@kitware.com> Merge-request: !6168
-
-
121a5f16 FindHDF5: clear any `HDF5_FOUND` status before starting Acked-by:
Utkarsh Ayachit <utkarsh.ayachit@kitware.com> Acked-by:
-
121a5f16 FindHDF5: clear any `HDF5_FOUND` status before starting Acked-by:
-
David Gobbi authored
This reverts commit ff91938b. The library can't be made "shared" until the export specifiers are in place for all the functions.
-
Ben Boeckel authored
-
Andreas Buykx authored
The vtkGenericDataArrayLookupHelper is reimplemented using std::unordered_map which provides a stable implementation (LookupValue always returns the first index, or a sorted vtkIdList) and performs faster lookup in most cases.
-
Ben Boeckel authored
-
- Nov 07, 2019
-
-
fd2d5f95 vtkGenerateGlobalIds: make ids geometry independent 239ab414 vtkGenerateGlobalIds: make results reproducible Acked-by:
Cory Quammen <cory.quammen@kitware.com>
Merge-request: !6167 -
ff91938b Do not force WrappingTools library to be static Acked-by:
-
Utkarsh Ayachit authored
vtkGenerateGlobalIds now does not depend on the dataset geometry to assign ids. That resulted in ids changing in time varying datasets where the geometry changed over time but topology remained consistent. The new implementation consistently assigns global cell and point ids preferring to name ids on lower ranks and before those on higher ranks. It needs a few more passes than before, but now the ids don't change over time for time varying datasets where only the geometry changes over time. Addresses paraview/paraview#19435
-
Utkarsh Ayachit authored
vtkGenerateGlobalIds was using `diy::kdtree_sampling` which uses `rand()` which may return different kdtree splits on repeated runs. Fixed by using `diy::kdtree` instead which generates a histogram of the entire block which will produce reproducible results.
-
ab396fbc Support parameter forwarding in ArrayDispatch. Acked-by:
-
07b9c96c Fix depth peeling Acked-by:
Joachim Pouderoux <joachim.pouderoux@kitware.com> Acked-by:
-
David Gobbi authored
This library was originally forced to be static to ensure that the wrapping executables (vtkWrapPython, vtkWrapHierarchy, etc) were statically linked. The idea was that since the wrapping tools are used prior to installation, static linking was more robust, since otherwise they would have to rely on a baked-in RPATH. However, these worries were unfounded, nd it is simpler and more consistent for BUILD_SHARED to control whether WrappingTools is static.
-
Allison Vacanti authored
-
662f285a Merge branch 'upstream-netcdf' into fix-netcdf-preprocessor-define c61168c2 netcdf 2019-11-07 (c4458542) bfd7c849 netcdf: update for _WIN32 definition usage Acked-by:
-
