Restart with diags start from non-zero step

[mulingLHY]

When I restart simulation with beam from_file, the new diags always start from step 0 openpmd_000000.h5.
I wonder if there is any parameter to set inital step of diags so that I can put the new diags together with the previous diags files with continuous step.

[AlexanderSinn]

I opened a PR to add an input parameter for this: #1333

[ax3l]

Just to understand, is this a new simulation (from zero) or a true checkpoint-restart of HiPACE++?

For the former, I agree the new parameter makes sense. For the latter, I think we should restore the output step during restart and keep counting.

[mulingLHY]

Just to understand, is this a new simulation (from zero) or a true checkpoint-restart of HiPACE++?

For the former, I agree the new parameter makes sense. For the latter, I think we should restore the output step during restart and keep counting.

What I want is the latter one.
Additionally, I think we should implement a set of parameters to control simulation restart behavior.

[AlexanderSinn]

Usually people restart hipace simulations to slightly tweak some settings or add/remove a particle beam or laser.  So I think it is better to have separate input parameters for each component of the simulation, like now, that can be combined arbitrarily up to a full restart with identical settings. 

[mulingLHY]

Usually people restart hipace simulations to slightly tweak some settings or add/remove a particle beam or laser.  So I think it is better to have separate input parameters for each component of the simulation, like now, that can be combined arbitrarily up to a full restart with identical settings.

That can be done by import beam from_file, but sometimes the simulation was stopped by accident, and I think such a parameter is helpful.