Fix scalar chunking

[dstansby]

Fixes #364. I noticed there wasn't a changelog, so I added one, I hope that's okay? I think it's important to communicate bug fixes like this.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 85.48%. Comparing base (1d113cd) to head (1927444).
Report is 41 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #365      +/-   ##
==========================================
+ Coverage   85.42%   85.48%   +0.05%     
==========================================
  Files          13       13              
  Lines        1537     1536       -1     
==========================================
  Hits         1313     1313              
+ Misses        224      223       -1     

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

[will-moore]

Thanks for the contribution and the changelog.
Could you add a link from the changelog to this issue? In other repos do this (e.g. https://github.com/ome/omero-web/blob/master/CHANGELOG.md) which also allows you to be more concise in the changelog if you wish?

The change looks good and makes sense. The only question I have is whether anyone would consider this a breaking API change (and therefore require a major version bump) or if this is really just a bug fix? cc @joshmoore

[will-moore]

Thanks for the update. LGTM.

Previously, doing write_image(....storage_options={"chunks": 32}) would have resulted in chunks of e.g (sizeZ, sizeY, 32) but this feature wasn't documented or have tests and the behaviour was different from zarr-python.
So, I think we could consider a change in this behaviour to be a bug-fix rather than a breaking API change, but happy to hear other opinions @joshmoore @sbesson ?

[joshmoore]

Please see https://github.com/ome/ome-zarr-py/blob/master/CHANGELOG.md

[dstansby]

👍 changelog in the right place now.