Add scripts/managed_repo_symlinks.py

[will-moore]

For the importing-NGFF-Filesets-without-chunks workflow, we import a metadata-only copy of the image/plate, then need to view an image from each plate before creating symlinks from the Managed Repo to the directory that contains the image/plate.

E.g. see README at https://github.com/IDR/idr0125-way-cellpainting

This PR aims to replace the following manual steps described there...

• Need to view an image from each plate OR omero render set Image:ID rdef.yml for each plate, to avoid subsequent ResourceError after symlinking. See scripts/render_set_cmd.py

• Run scripts/get_import_paths.py to create imported_paths.txt with the managed repo path for each imported plate.

• Run scripts/symlink_cmd.py which consumes imported_paths.txt to output the commands to replace each plate in the managed repo with a symlink to the corresponding plate on a mounted s3 bucket. These can be saved in scripts/symlinks.bash. Need to run sudo chmod +x symlinks.bash then sudo -u omero-server -s and bash symlinks.bash.

...with a single script run:

E.g. For a Screen that contains Plates, where the plates are mounted at /path/to/plates/dir/ we can run:

Default repo is /data/OMERO/ManagedRepository
Script also supports the --dry-run and --report options.

$ sudo -u omero-server -s
$ source /opt/omero/server/venv3/bin/activate
$ python idr-utils/scripts/managed_repo_symlinks Screen:123 /path/to/plates/dir/ --repo /path/to/ManagedRepository --report

[will-moore]

As discussed in IDR meeting yesterday: don't expect to use the omero import workflow for NGFF data since this will be too slow.

[will-moore]

Reopened PR for now: Using this branch to update symlinks following upload to BioStudies: see IDR/idr-metadata#659 (comment)