Pylhc-submitter

- Fixing `job_submitter`: Do not transfer any output files, when the `output_destination` is given.

- Fixing `job_submitter`: Discovers more invalid URIs.

- Fixing `job_submitter`: type error in `print_stats`, when job-names are integers.

- General code cleanup/refactoring/documentation:
- Partly breaks backward compatibility, if individual methods of the `job_submitter`-functionality have been used.
- Does not affect any setups simply calling the `main()` function of `job_submitter.py` or calling the `job_submitter` as a module.
- Apart from some fixed imports, following the new structure, the `autosix` module has been untouched.


- New Feature of `job_submitter`:
- `output_destination` input parameter, which sets an output directory in which the folder-stucture
for the jobs will be replicated and the job's `job_output_dir` will be copied into "manually" at the end of the job,
instead of having the directory transferred back to the `working directory` by htcondor.

- Uses `concat` instead of `append` to stack the DataFrames.

This release adds some changes the `autosix` module:

- New Features:
- Added `sixdesk_directory` option, which allows the user to choose their own
sixdesk environment (default is PRO on AFS).
- Added `max_materialize` option, which allows the user to specify the amount of jobs that
materialize at once per SixDesk Workspace (i.e. one realization in the `replace_dict`).
This enables the user to send more jobs to HTCondor than are allowed within their user limit.
See the HTCondor API for details. This option requires writing rights in the `sixdesk_directory`.
- Allow `ENERGY` and `EMITTANCE` to be set via `replace_dict`, which are then
passed to the `sixdeskenv` (`GAMMA` is calculated from the `EMITTANCE` automatically).

- Changes:
- Big object-oriented restructuring of the Stages and increased use of Dataclasses.
- Some other small changes to improve readablity.
- Fixed the DA-Plot labels from `sigma [sigma]` to `DA [sigma]`