You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Edits to k-buffering and new features in pp_ser.py (#245)
This PR contains two minor changes to the k-buffering feature used by `pp_ser.py` and implemented in `utils_ppser_kbuff.f90` (increase in maximum number of buffers, more verbose error message in case of a problem) and adds two new features to `pp_ser.py`.
For the latter, the changes are located in `pp_ser.py` and `utils_ppser.f90` only and entail:
- When initializing `pp_ser.py` with reference data one can specify another directory where the data is located. If this is not available, new data will be serialized into the same directory as the reference data which can lead to conflicts or corrupt data.
- Up to now, the directive syntax for specifying a savepoint was `!$ser savepoint gugus` where gugus was the name of the savepoint. This syntax does not allow the dynamic generation of savepoint names (e.g. from runtime information such as timestep or a unique ID). A new command line option to `pp_ser.py` named `-s` or `--sp-as-var` allows to use `!$ser savepoint "gugus"` or `!$ser savepoint my_string` where `my_string` is a variable.
0 commit comments