Change logging strategy and add new logging for adaptivity metrics

[IshaanDesai]

This PR adds the following:

• A new logging framework with a dedicated Logger class. In this class, a logger is generated and logging options are provided for logging via one one rank, any rank of choice, or all ranks. Different types of logging are possible. Currently the options are info and error.
• Breaking change: changes the configuration parameter config_file_name to precice_config_file_name to differentiate it from the Micro Manager JSON configuration file.
• Add optional parameter output_cpu_time to adaptivity_settings which will output CPU time of the adaptivity computation on every rank. The data is written to preCICE with the intention of being exported. All the micro simulations (coupling vertices) on one rank will have the same scalar value of the adaptivity computation CPU time.
• Breaking change: Change the name of the diagnostics output micro_sim_time to solve_cpu_time.

Checklist:

• I made sure that the CI passed before I ask for a review.
• I added a summary of the changes (compared to the last release) in the CHANGELOG.md.
• I will remember to squash-and-merge, providing a useful summary of the changes of this PR.

[Fujikawas]

Would be good to put the "number of active simulations" into a format which is more machine-readable than the current text style.

[uekerman]

Could precice/precice#2015 be an alternative technical solution?

[IshaanDesai]

Could precice/precice#2015 be an alternative technical solution?

Potentially yes. I have implemented something very close to this, with additionally handling logging in parallel. Would it be possible to tell preCICE to log somethings only on one rank, and other things on all ranks? Also, such logging API would need to be added to the Python bindings.

[uekerman]

Would it be possible to tell preCICE to log somethings only on one rank, and other things on all ranks?

Yes, this is part of the runtime configuration for the existing preCICE logging already now.