Hi all,
I have a quick update on the CS-FileSystem work.
I integrated the official AI-for-science dataset list into the filesystem
as a metadata registry. The filesystem now exposes the list in two ways:
registry.json
datasets/
alphafold_db/
metadata.json
dandi/
metadata.json
allen_institute/
metadata.json
...
ATLAS remains the fully working reference dataset connector, and the
generic TSV/CSV path can expose additional metadata datasets with name=path
arguments. This means the official dataset list is now browseable through
the filesystem, while larger or restricted datasets can start with
metadata/manifests before adding raw data access where practical.
Also, I've updated the README file and kept the sample dataset workflow
working for demos.
Validation:
cargo fmt
cargo test
All tests pass on my local machine.
Thanks,
Jayanth Vennamreddy
