Ccx2paraview

- Added PyPi publishing workflow
- Turned into installable package
- Made the development version installable by pip
- Added alias scripts for converting to formats

- Improvements in von Mises and principal stress calculations
- Added description in readme: how to convert to a vtkhdf-file
- Added Exception message indicating missing vtk-dependency
- Streamline import of Converter (i.e. from ccx2paraview import Converter)

- Fixed syntax warning invalid escape sequence
- Fixed broken pvd-file (missing '.' in '.vtu')

Big thanks to soylentOrange, fsimonis and rsmith-nl for working on this release.

- Process FRD files and write output increment by increment.
- Open FRD only once to speed up the conversion.
- Tests: format time delta to mm:ss.
- Fixed steps processing in modal results conversion.
- Fixed bug in PVD file writer.

Fixed FRD reader's open file mode. File read in binary mode took a 25% more time.
Caught PermissionError in cache cleaner.
Merged code into a single file - for user's simplicity.
Updated README.

Since this release binary files will not be prepared, because no one binary release tool (pyinstaller, nuitka, cython3) can handle external VTK Python library correctly. So, please, use source code for your needs instead of binaries.



Full refactoring of .vtk/.vtu/.pvd writers - now they use native Python VTK library. Legacy .vtk format is fully supported - previously there was a problem with component names.

Convert .frd into both formats at once:
ccx2paraview.Converter(file_name, ['vtk', 'vtu']).run()

Von Mises and principal components are written as separate scalar fields. Tensor fields are exported as tensors with 6 components. It makes possible to apply TensorGlyph filter in Paraview.

Thanks to Nuitka - a new binary generation tool - ccx2paraview binaries are lightweight in comparison with the one generated by PyInstaller.

Perfect test results!

+ Fixed os.path.relpath bug in Windows.
+ Windows version was rebuilt with numpy.

Fixed bug with time steps and PVD writing.