Modify the Archiver to read the ADS from each file and add them as independent Nodes with the full ADS name as the name of the file.
During restore the ADS files are restored using the ADS name and that automatically attaches them as ADS to the main file.

In the case where the main file gets restored after the ADS, it could overwrite and remove the ADS. Hence we try creating the file without the O_CREATE option first and if the file does not exist, then add the option and try again.

ADS are added as top level files in the repository. However, they must not be counted in the finished and total count, but their sizes must be counted.

The ADS is essentially a part of the main file. Hence during filtering, we need to derive the main file name from the ADS name for filtering to work.

Move SaveDir back to archiver and use hooks in the os specific archivers. Rename _unix to _notwin.