ifsbench.files for data pipeline

[johannesbulin]

Is your feature request related to a problem? Please describe.

The data handling in ifsbench is currently limited to extracting an archive and moving some files around. In the future we may need a more extensive pipeline like approach to support things like

• different namelist/file modifications, depending on the application that is run.
• options at runtime that modify some namelist entries to enable/disable additional debugging or GPU capabilities.

Describe the solution you'd like

Implement a DataHandler class that performs a given operation in a working directory. It should be possible to execute several DataHandler sequentially (in a pipeline manner), for example

1. Extract an archive with raw input data
2. Move some files around to adapt to a given benchmark.
3. Update namelist entries.

To do this, implement the DataHandler base class and specialisations for

• archive extraction
• renaming/copying
• namelist modifications

Describe alternatives you've considered

No response

Additional context

No response

Organisation

ECMWF