CEP2
| Status | Active |
| Author | David Ketcheson <dketch@gmail.com> |
| Created | November, 2014 |
| Updated | March 19, 2015 |
| Discussion | Clawpack #51 |
| Implementation |
The current output file names do not specify the file format. For instance, any PyClaw output using PETSc is named *.ptc (except for VTK files). Also, something inherited from Clawpack that goes against convention is that the file number is in the extension. Finally, it would be nice to have a filename that suggests the file came from Clawpack. This proposal would completely get rid of the fort prefix.
The following output file names are proposed:
- ASCII: claw_q_####.txt
- (Fortran) binary: claw_q_####.fort
- (PyClaw) PETSc binary: claw_q_####.petsc
- VTK: claw_q_####.vtk (same as current)
- HDF5: claw_q_####.h5
- NetCDF 4: claw_q_####.nc
- (PyClaw) metadata pickle: claw####.pickle (for both PyClaw and PetClaw output)
All aux files would have the same names as the q files, but with the q in the prefix replaced by aux.
Users will be strongly discouraged from writing output in multiple formats to the same directory. Then we can automatically detect what format to read in based on the names of files in the directory.
- Names for other types of output such as gauges and derived quantities (p and f)
- Do we want to include "claw" in the prefix or not?
- Can PETSc HDF5 output be read in with the PyClaw routines?
It will be slightly difficult to read (with the new code) files that were created with the old code.