Semantically, logical OR is what is desired here.
I suspect the use of bitwise was a performance hack for old hardware.
Logical OR can short circuit, giving the compiler an optimisation opportunity.

The source was searched with "grep -l [^ -~]" to find characters outside
of the printable ASCII range 0x20 to 0x7e.  The worst offender is the
en-dash character, which is visually indistinguishable from the minus
character in most code editors.  Several tabs were also found, which
is surprising since they should be blocked by the commit hooks.

Kunda authored 8 years ago and Cory Quammen committed 8 years ago

doxy and misc. typos pt2

Kunda authored 8 years ago and Cory Quammen committed 8 years ago

fixes some doxy typos
fixes misc. comment typos

This ensures that the table is up-to-date before it is applied to
the image.  Note that is change is needed for the correct behavior
of vtkImageResliceMapper.  Also note that the vtkImageMapToColors
class correctly calls Build() on the table.

- Made a few APIs' params const
- Changed many usages of the evil GetBounds() to expect a const
  return value, even though it does not (yet?) return const
- Slightly changed return values of vtkTextActor3D::GetBounds()
  to return its own ivar instead of its ImageActor's

The class vtkImageResample declared a member called OutputSpacing, even
though that member was already present in the superclass vtkImageReslice.
In order to ensure that the SetOutputSpacing() method (in vtkImageReslice)
can be used with vtkImageResample, it is necessary that vtkImageResample
does not redeclare this member.

Note that vtkImageResample uses a spacing of [0,0,0] as a signal that the
spacing is "undefined", whereas vtkImageReslice does not (it uses a default
spacing of [1,1,1] and a flag called ComputeOutputSpacing to indicate
whether the filter should use these values or compute new ones).  This
doesn't cause a bug since the OutputSpacing member only used in
RequestInformation, and each class has its own RequestInformation
implementation.

shut up warnings on mun

passing size_t as vtkIdType requires explicit cast

The GetIncrements() method calls ComputeIncrements(), and if you do not
supply your own array, then the vtkImageData::Increments ivar is modified
and returned.  This means that the signature of ComputeIncrements() that
returns a pointer is not thread safe.

This fixes performance warnings for cppcheck.

The class was writing to a memory location from
multiple threads bro. Also fix thread issues in
the TestAtomic and TestConditionVariable tests.

There were no C++ tests that directly tested this class, and this test
also covers the recently-discovered bug for TranformInputSamplingOff().

A simple omission in the code caused TransformInputSamplingOff() to produce
incorrect results.  The default is TransformInputSamplingOn(), and it seems
that no tests exercised the class with it off.

There were no C++ tests that directly tested this class, and this test
also covers the recently-discovered bug for TranformInputSamplingOff().

A simple omission in the code caused TransformInputSamplingOff() to produce
incorrect results.  The default is TransformInputSamplingOn(), and it seems
that no tests exercised the class with it off.

I used perl regex to find any groups that spanned a protected: or
private: access specifier.

Kitware Robot authored 8 years ago and David Gobbi committed 8 years ago

This commit reindents the code with the following utility:
Utilities/Maintenance/vtk-reindent-code.py

This utility changes the positions of the braces so that they are no
longer indented relative to the code block they reside in.

The bash command line used was the following:

for d in Charts Common Deprecated Domains Examples Filters GUISupport \
         Geovis IO Imaging Infovis Interaction Parallel Rendering \
         Testing Views Web Wrapping; do
  for e in cxx cxx.in txx txx.in hxx hxx.in h h.in c c.in; do
    find "${d}" -name "*.${e}" -exec \
      python Utilities/Maintenance/vtk_reindent_code.py {} +
  done
done

generated via:
cd $VTKSRC
perl Utilities/Doxygen/doc_header2doxygen.pl --to ~/tmp/vtkheaders .
cp -r ~/tmp/vtkheaders/* .

This is the vtkImageData equivalent of vtkPolyDataConnectivityFilter.
It performs connectivity checks on points with values within a specific
scalar range (by default, it checks all points with values > 0).  Its
output is a label image, where each connected region is assigned an
integer label value.  Unlabeled points are assigned a value of zero.

The current dependency relationship in vtk is unclear and misleading.
This MR tries to rewrite them based on header files inclusion of headers
and source files in each module. Corner cases are considered and
modules are sorted in alphabetical order to facilitate future reference.
See details in my gitlab python based script project. In future we can continue
using this script to clean the VTK Dependency easily from
time to time.

This commit fixes #16842 for vtkImageReslice.  If GenerateStencilOutput
is On, then stencil->InsertNextExtent() is called when vtkImageReslice's
execute method completes a row, but if a row has been split among more
than one piece, the stencil will be corrupted.  It's only when SplitMode
is "Block" that this is likely to occur.  The fix used here is to detect
when rows might be split, and to adjust the split mode so that it won't
occur if GenerateStencilOutput is On.

For C++11 (and for some pre-C++11 compilers via extensions) a template
instantiation can be declared "extern" so that the compiler will avoid
generating more than one copy of the code for that template.

This commit is an attempt to simplify the use of extern template in VTK.

This is a fix for:
#14310

Windows applications that run for a long time report that rendered
objects do not change. This is because the modified time on a Windows
system is 32 bits. This causes overflows that defeat the modified time
mechanism. This patch defines a new type, vtkMTimeType that is 64
unsigned integer regardless of the architecture.

A mechanism to provide backward compatibility is introduced. The
preprocessor define "VTK_HAS_MTIME_TYPE" can be used in applications
that must build against VTK versions that use the "unsigned long" type
for MTime's.

Methodology used to find MTime occurences:
1) Identify files as follows:
   git grep "unsigned long" | grep ime | cut -d":" -f1,1 | sort | uniq
2) Hand edit each of those files replacing "unsigned long" with
   "vtkMTimeType" where appropriate.
3) Temporarily change typedef for vtkMTimeType to "double" to detect
   missing conversions

Added special case for RGB to RGB and RGBA to RGBA,
where the loops are rewritten, transforming pointer arithmetic to indexing in order
to allow auto-vectorization by clang.
For the RGBA to RGBA case there was a noticeable gain of 3.2% in performance.

some dashboard warnings

Use the efficient "setpixels" method for setting out-of-bounds regions
to the background color.  Also, use a switch() to unroll the inner loop
for common pixels sizes, this provides a performance boost of a few %.

The clip extent provides the intersection of the output extent and the
input extent, taking the kernel size into account.  If it is empty, it
should always be specified the same way: the lower bound should be the
same as the output's lower bound, and the upper bound should be one unit
less than the lower bound (so that the computed size in any dimension is
zero).

For this path, the resulting speed increase can be 30% or more.

The vtkImagePointDataIterator provides a more efficient and sometimes
more convenient way to iterate through image data.  It also takes care
of progress reporting.

It doesn't seem to have got everything, but it's a start.

Some resulting whitespace is suboptimal, but one day
we can fix that with clang-format. :)

Just ran this command:
run-clang-tidy.py -checks=-*,modernize-use-override -fix

(I actually had to build my own clang to make clang-tidy
output "VTK_OVERRIDE" instead of "override".)