First release of fVTKHDF: A modern Fortran library for writing VTKHDF files

If you’re a Fortran’er, this may be of interest to you.

I’ve just published the first release of fVTKHDF, a high-level, object-oriented Fortran library for writing VTKHDF files.

It currently supports UnstructuredGrid (UG) and MultiBlockDataSet (MB) types of VTK datasets with temporal data – the itch I needed to scratch. The MB dataset is limited to a flat assembly of UG blocks (no hierarchical nesting of MB), so it’s essentially equivalent in this form to a PartitionedDataSetCollection.

The library is fully parallel, using parallel HDF5 (MPI-IO) for collective writes to partitioned datasets.

See the Github project page for more information and the Reference Manual.

This is glorious @nncarlson ! Thanks a lot!

In your readme you mention:

ParaView 5.13+ recommended for full support of the generated files.

I would recommend 6.0 (or even 6.1!) to get the last improvements that have been added to the format :slight_smile:

Any specific reason to not support PDC, it should be pretty similar to MB ?

@lgivord @Louis_Gombert

Any specific reason to not support PDC, it should be pretty similar to MB ?

In fact for my “flat” Assembly of leaf nodes it is identical; I only needed to assign a different value to the Type attribute. My sole reason for reverting to MB was because of this bug. It was fixed very recently, but I don’t think it made it into 9.6. My plan is to shift to PDC as soon as the fix is available in ParaView.

Correct

we should probably backport it in release in case we plan to do a 9.6.1, it’s an important bug fix imo

Agreed

Backport is here: https://gitlab.kitware.com/vtk/vtk/-/merge_requests/12972

Excellent. Thanks!