search

ome / omero-ms-zarr

Experimental micro-service for implementing the Zarr spec over HTTP
https://www.openmicroscopy.org/omero
GNU General Public License v2.0
17 stars 12 forks source link

Export masks from omero-ms-zarr #52

Open joshmoore opened 4 years ago

joshmoore commented 4 years ago

Once the specification is decided (#49), the proposed URLs should all generate mask arrays on the fly.

mtbc commented 4 years ago

For image/123.zarr/0/ we have the directory listing controlled by omero.ms.zarr.folder.layout so we know, alongside the .zarray, if to list flattened chunks or the top level of the nested chunk directories.

I would suggest that image/123.zarr/roi/45/ would redirect to roi/45.zarr/ and roi/45.zarr/mask/67/ (also shape/67/) would redirect to mask/67/ which would have a .zarray. However, that path format does not quite match with omero-cli-zarr at present.

Interpreting a plural as a kind of joining operation, for image/123.zarr/rois/masks/ we could have a .zarray there too, but which? Should labeled or 6D be another configuration property or should there be two different URLs?

joshmoore commented 4 years ago

would redirect to roi/45.zarr/

I don't disagree that this is a potential URL we would want to support, but I'd err on the side of simplicity initially since we will likely want to change the general layout anyway.

However, that path format does not quite match with omero-cli-zarr at present.

Remembering from your implementation, I'd think that that would be another, well-known. regex

Should labeled or 6D be another configuration property or should there be two different URLs?

I don't think we know yet. Again, I think we try for the smallest number of initial options since we may need to update these files.