Open pwalczysko opened 2 years ago
cc @sbesson @joshmoore cf. discussion this morning
@pwalczysko: can you try this in a subdirectory that has no parents ending in .zarr
? I wonder if logic is going up the stack to find a .z
file.
I was about to say part of the concern is redundant with #25 around defining precisely how much should be detected/discovered by the reader.
In this case, things are further complicated by the fact the output of a bioformats2raw
execution is nested within the output of a format bioformats2raw
execution. Sounds like the kind of edge case situations that we don't want to support (and the reader should probably fail accordingly)
[pwalczysko@idr2-slot3 idr-scratch]$ pwd
/uod/idr-scratch
[pwalczysko@idr2-slot3 idr-scratch]$ sudo mv test.zarr/petr-test/ .
[sudo] password for pwalczysko:
mv: cannot move ‘test.zarr/petr-test/’ to ‘./petr-test’: Permission denied
I think I am starting to understand - but see above, the options are limited to have a better situation, not sure why admin cannot move the folder.
Edit:
As indicated, managed to do the above move whilst being on idr0-slot3
. Not sure why it works there, but not on idr2-slot3``
Indeed, the move of the petr-test
out of the test.zarr
directory solved the issue
ls
multifile-Z1.ome.tiff multifile-Z2.ome.tiff multifile-Z4.ome.tiff
multifile-Z1.ome.zarr multifile-Z3.ome.tiff multifile-Z5.ome.tiff
(.venv3) bash-4.2$ omero import -f multifile-Z1.ome.zarr/
2022-04-07 15:24:49,350 230 [ main] INFO ome.formats.importer.ImportConfig - OMERO.blitz Version: 5.5.9-SNAPSHOT
2022-04-07 15:24:49,369 249 [ main] INFO ome.formats.importer.ImportConfig - Bioformats version: 6.9.1-SNAPSHOT revision: 17f2529242af94162bfaff168c25746f39ca04ad date: 7 April 2022
2022-04-07 15:24:49,414 294 [ main] INFO formats.importer.cli.CommandLineImporter - Log levels -- Bio-Formats: ERROR OMERO.importer: INFO
2022-04-07 15:24:49,787 667 [ main] INFO ome.formats.importer.ImportCandidates - Depth: 4 Metadata Level: MINIMUM
2022-04-07 15:24:50,298 1178 [ main] INFO ome.formats.importer.ImportCandidates - 6 file(s) parsed into 1 group(s) with 1 call(s) to setId in 502ms. (511ms total) [0 unknowns]
#======================================
# Group: /uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/OME/METADATA.ome.xml SPW: false Reader: loci.formats.in.ZarrReader
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/OME/METADATA.ome.xml
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/.zattrs
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/.zattrs
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/.zarray
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/0/0/3/0/0
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/0/0/2/0/0
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/0/0/4/0/0
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/0/0/1/0/0
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/0/0/0/0/0/0
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/0/.zgroup
/uod/idr-scratch/petr-test/multifile-Z1.ome.zarr/.zgroup
The removal of the petr-test
out of the test.zarr
also magically solved https://github.com/ome/ZarrReader/issues/25
This problem is resurfacing in https://github.com/ome/omero-cli-transfer/issues/41#issuecomment-1412166773 - I would claim that it shows that the issue here is actually caused by the naming of the enclosing folder, and it is less marginal (as shown by the fact that the omero-cli-transfer
is hitting it) than previously judged
cc @sbesson @dgault
Note: The problem described below is only occurring if the
petr-test
folder is nested into another .zarr folder, such as/uod/idr-scratch/test.zarr/petr-test/multifile-Z1.ome.zarr
. If this nesting is removed, the problem does not occurr, and the resulting import does consit of 1 image with 5 z planes as expected.Have an ome.zarr folder inside a subfolder called
petr-test
, e.g.In this case, the importer detects the zarr files to import as follows:
Now, have following 5 files inside the
petr-test
folder as well next to theome.zarr
folder (not inside the ome.zarr) such asThe importer is detecting the ome.tiff files as well, although pointed only and directly at the ome.zarr