Issues with runcard includes: stability, dependencies #687
Description
While integrating and extending Stefan's WIP MR #620 I noticed a couple of issues in the runcard handling, which I would say are issues in upstream.
Preliminary note: MR #520 adds three new parameters to the run_card.dat in gg_tt.mad. When rebuilding, this results in the following messages
treatcards run
Found unexpected entry in run_card: "exec_mode" with value "1 ".
The type was assigned to int.
The definition of that variable will be automatically added to fortran file
The value of that variable will be passed to the fortran code via fortran file
Found unexpected entry in run_card: "bridge_mode" with value "1 ".
The type was assigned to int.
The definition of that variable will be automatically added to fortran file
The value of that variable will be passed to the fortran code via fortran file
Found unexpected entry in run_card: "vector_size" with value "32 ".
The type was assigned to int.
The definition of that variable will be automatically added to fortran file
The value of that variable will be passed to the fortran code via fortran file
Now what happens here is that both run.inc and run_card.inc are modified
modified: ../../Source/run.inc
modified: ../../Source/run_card.inc
Issue 1, but this is about MR #520: as our run.inc and run_card.inc are in the repo, the new versions should be committed. I will fix this.
Issue 2 (or at least IMO this is an issue): while the changes to run_card.inc are minimal (just three lines with the new parameters are added), the changes to run.inc are massive. A lot of code that was previously lowercase is now uppercase. In addition, a lot of code that was previously longer than 72 characters is now fitting within 72 characters, with continuation characters added. And all these changes affect ALL parameters, not only the three newly added ones. I find this very annoying because it becomes impossible to understand what has changed. I may try to fix this by including run.inc that is already all uppercase? Not sure.
Issue 3. I modified the runcard.dat to remove the new parameters and rebuilt. The build fails, because the run.inc is NOT rebuilt. There is a missing dependency I think, only run_card.inc depends on run_card.dat, while also run.inc should depend on it.
Issue 4. I think that the "Found unexpected entry" are because in MR #520 one should also modify banner.py to make these parameters "default" parameters as opposed to new custom parameters. I will actually try to work around some of the previous problems this way. But some of them remain issues that should independently be fixed, probably...