Update dtype usage in zarr/meta.py

[joshmoore]

Changes in zarr's annotations, numpy, and/or mypy have led to failures on mypy zarr/across a large number of PRs. This tries to take the most direct route to passing GitHub actions so that PRs can be unblocked.

Thanks to @rgommers and @seberg for jumping in.

TODO:

• Add unit tests and/or doctests in docstrings
• Add docstrings and API docs for any new/modified user-facing classes and functions
• New/modified features documented in docs/tutorial.rst
• Changes documented in docs/release.rst
• AppVeyor and Travis CI passes
• Test coverage is 100% (Coveralls passes)

[joshmoore]

Failures can be seen in https://github.com/zarr-developers/zarr-python/pull/699/checks?check_run_id=1904975788

arr/meta.py:82: error: Incompatible return value type (got "List[Union[Tuple[str, str], Tuple[str, str, Tuple[int, ...]]]]", expected "str")
zarr/meta.py:183: error: "generic" has no attribute "real"
zarr/meta.py:184: error: "generic" has no attribute "imag"
Found 3 errors in 1 file (checked 30 source files)

[rgommers]

Hmm, the dtype.descr change surprises me. Good chance it's on purpose, but I don't really want to comb through multiple NEPs - so let's just ask @seberg.

[seberg]

Do you know what changed? This has nothing to do with the dtype NEPs, but np.dtype(...).descr has never returned a string as far as I can tell. The only actual change I can think of (and find!) is that we might be returning unicode strings instead of bytes somewhere around here.

[joshmoore]

Thanks, both. I was assuming it was a numpy bump that led to this, but I'll need to try to find what the change was, unless @Carreau knows.

Another possibility is that mypy has gotten stricter, in which case I could update the annotation rather than the logic.

[seberg]

I would guess the typing is due to NumPy 1.20 now including typing. The .descr change I am confused about though, since I am not aware of a change.

[codecov]

Codecov Report

Merging #700 (79733be) into master (b0b161b) will increase coverage by 0.48%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #700      +/-   ##
==========================================
+ Coverage   99.45%   99.94%   +0.48%     
==========================================
  Files          28       28              
  Lines       10262    10263       +1     
==========================================
+ Hits        10206    10257      +51     
+ Misses         56        6      -50     

Impacted Files	Coverage Δ
zarr/meta.py	100.00% <100.00%> (ø)
zarr/util.py	100.00% <0.00%> (+15.06%)	⬆️

[joshmoore]

Asked for vetoes in gitter. Since all PRs are currently failing, I'm going to merge and start re-running the builds. If there are any issues with the changes here, I'll be happy to open a follow-up PR.