* Early pushout of restore handling for core/container

* Adjust after rebase

* Move logging definition, we should only do this if we go ahead with the restore

* First round

* More paths

* Add async_restore_backup to base class

* Block restore of new backup files

* manager tests

* Add websocket test

* Add testing to main

* Add coverage for missing backup file

* Catch FileNotFoundError instead

* Patch Path.read_text instead

* Remove HA_RESTORE from keep

* Use secure paths

* Fix restart test

* extend coverage

* Mock argv

* Adjustments