You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Integrate @Carreau ’s zarr-python implementation (https://github.com/Carreau/zarr-python/tree/spec-v3) into the zarr_implementations repository, generating both .zr and .zr3 files, again for all currently implemented codecs. This work may need to remain open as a PR if tests are failing as work on the spec-v3 branch continues. The goal is to have all V2 and V3 files readable by this branch.
The text was updated successfully, but these errors were encountered:
I did rebase that branch, fixing merge conflicts and then proposed some small changes in Carreau/zarr-python#1.
My updates were mainly modifying the list_dirs method of BaseV3Store to match the description in the spec and adding the metadata_key_suffix entry to zarr.json.
Using zarr.open does not yet seem to work for opening zr3 files. @Carreau, is there another way of opening these currently or should I hold off a bit on trying that?
Also, is there a plan to have hierarchy methods matching the API from zarrita and xtensor-zarr? e.g. create_hierarchy/get_hierarchy functions? (Such hierarchy methods are not described in the v3 spec so I wasn't sure)
Integrate @Carreau ’s zarr-python implementation (https://github.com/Carreau/zarr-python/tree/spec-v3) into the zarr_implementations repository, generating both .zr and .zr3 files, again for all currently implemented codecs. This work may need to remain open as a PR if tests are failing as work on the spec-v3 branch continues. The goal is to have all V2 and V3 files readable by this branch.
The text was updated successfully, but these errors were encountered: