Possible bug in sliced h5f%write() in the latest release (v4.10.2)?

[gtokic]

I've noticed that since updating h5fortran from v4.2.11 to v4.10.2 (the latest release) that the sliced write of a lower-rank array into a larger-rank dataspace raises an error.

I'm referring to the case when some of the indices in istart and iend in sliced write are equal

call h5f_p%write('dset_name',dset,istart=istart,iend=iend)

This occurs when I want to write an essentially a lower-rank array dset into a larger-rank dataspace, and I extend the rank of dset with size-1 dimensions to match that of the dataspace. For those dimensions i, istart(i) == iend(i) and h5fortran raises an error.

It seems this is due to line 84 in write.f90, where an error is raised if any(iend <= istart)

if(any(iend <= istart)) error stop 'ERROR:h5fortran:create: iend must be > istart'

If this condition is modified to allow for some iend == istart, as well as addressing lines 293--295 in utils.f90, the code executes without issues and the HDF5 file is written as expected.

Am I using the new API in a wrong way or is this a bug?