View in biofile finder

[will-moore]

With $ ome_zarr view my_image.zarr we can view a single zarr image (or plate etc) in the ome-ngff-validator by serving the image from a local python server.

But this PR allows you to browse a collection of images in your local filesystem with:

$ ome_zarr finder path/to/my_data_folder/

This command will traverse your local filesystem, looking for zarr images (only .zattrs for now - can add zarr v3 support as a follow-up) and collect them into a CSV. This CSV is then served by the localhost python server and opened in BioFile Finder, allowing you to browse your local zarr images which are also served in the same way (and see thumbnails, copy URL etc).

Then in Bio-File Finder, you can "Group By" > "Folders" and you'll get a "tree" of all your folders containing images.

[codecov]

Codecov Report

Attention: Patch coverage is 80.00000% with 26 lines in your changes missing coverage. Please review.

Project coverage is 85.48%. Comparing base (1e3e3e2) to head (786602d).
Report is 80 commits behind head on master.

Files with missing lines	Patch %	Lines
ome_zarr/utils.py	79.83%	24 Missing ⚠️
ome_zarr/cli.py	81.81%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #436      +/-   ##
==========================================
+ Coverage   85.45%   85.48%   +0.03%     
==========================================
  Files          13       13              
  Lines        1540     1654     +114     
==========================================
+ Hits         1316     1414      +98     
- Misses        224      240      +16     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[joshmoore]

Could definitely use some help under -h:

usage: ome_zarr view [-h] [--port PORT] path

positional arguments:
  path

options:
  -h, --help   show this help message and exit
  --port PORT

Should the CSVs support file:///?

Also, some failure on HCS data:

Traceback (most recent call last):
  File "/opt/homebrew/Caskroom/mambaforge/base/envs/ome-zarr-py/bin/ome_zarr", line 8, in <module>
    sys.exit(main())
             ^^^^^^
  File "/opt/ome-zarr-py/ome_zarr/cli.py", line 180, in main
    ns.func(ns)
  File "/opt/ome-zarr-py/ome_zarr/cli.py", line 37, in view
    zarr_view(args.path, args.port)
  File "/opt/ome-zarr-py/ome_zarr/utils.py", line 171, in view
    zarrs = list(walk(Path(input_path)))
            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/opt/ome-zarr-py/ome_zarr/utils.py", line 155, in walk
    yield from walk(p)
  File "/opt/ome-zarr-py/ome_zarr/utils.py", line 153, in walk
    yield from find_multiscales(p)
               ^^^^^^^^^^^^^^^^^^^
  File "/opt/ome-zarr-py/ome_zarr/utils.py", line 70, in find_multiscales
    print("well", wells[0])
                  ~~~~~^^^
TypeError: 'NoneType' object is not subscriptable

[will-moore]

@joshmoore - Thanks. I fixed the help and handling of plates with no wells.

Not sure what you mean by "Should the CSVs support file:///?"
I don't think BFF in the browser will be able to load anything via file:///?

[joshmoore]

Nothing I tried previously got to a BFF screen -- I was trying to setup my own CSV.

[will-moore]

Any errors?
Does this work?

$ mkdir test && cd test
$ ome_zarr download https://uk1s3.embassy.ebi.ac.uk/idr/zarr/v0.4/idr0062A/6001240.zarr
$ ome_zarr download https://uk1s3.embassy.ebi.ac.uk/idr/zarr/v0.4/idr0101A/13457539.zarr
$ cd ../
$ ome_zarr view test

$ cat test/biofile_finder.csv 
File Path,File Name,Folders
http://localhost:8000/test/13457539.zarr,13457539.zarr,.
http://localhost:8000/test/6001240.zarr,6001240.zarr,.

[will-moore]

@joshmoore If you want to create your own CSV then you can open it in BFF using their File picker (CSV doesn't need to be served by a server). But ome_zarr open my_dir does everything all for you (creates the csv and opens it in BFF).

[joshmoore]

Ok, with the caveat that the CSV above requires something to be serving those files on localhost (right?), the above works for me. I think though that'd I'd feel better about this if the current view failed for non-Zarrs and then said, please use the command ome_zarr finder or something to not be too much of a surprise.

[will-moore]

Phew - finally got codecov down to only -0.05% change, so we're green again!

Now, $ ome_zarr view ... will always open validator (even for a collection like ome_zarr view bioformats2raw/images.zarr.

Similarly, $ ome_zarr finder ... will always open BFF, even for a single image, or bf2raw.zarr or directory (unless no images found).

[joshmoore]

ome_zarr view /tmp/test (from above)

No .zattrs or zarr.json at http://localhost:8000/test: File not found

👍 for this PR, but we could consider improving it as things evolve, e.g., could view detect this before opening and suggest finder?

ome_zarr finder /tmp/test

👍

ome_zarr finder /tmp/test/13457539.zarr

👍 (possible to start with the thumbnail view?)

[will-moore]

@joshmoore That should address the last comments:

• Opens BFF with small thumbnails view
• Handle $ ome_zarr view dir with no images:

$ ome_zarr view data/boats.zarr/
No OME-Zarr images found in data/boats.zarr/. Try $ ome_zarr finder data/boats.zarr/