Unfiltering tests and cleaning up edge-cases

[krisfed]

Several minor changes (mostly resulting from going over all the tests that were filtered or commented out):

1. Change in behavior for fill value. Previously the fill value was silently cast to the dataset's datatype. This could sometimes result in unexpected/xBuggy behavior. E.g.:

>> zarrcreate("mygroup/myzarr", [1,2], Datatype="uint8", FillValue=-9)
>> d = zarrread("mygroup/myzarr")
 
d =
 
  1×2 uint8 row vector
 
   0   0   % not -9 because -9 cannot be represented with unsigned int

With this change, users will have to explicitly cast their fill value to the datasets datatype (which is more cumbersome, but avoids wrong expectations). A bit on the fence about this change, so let me know if you prefer the original approach.

The mismatch in the fill value datatype will now result in a new MATLAB:zarrcreate:invalidFillValueType error.
Also un-filtering and updating the corresponding tZarrCreate/invalidFillValue test.

2. Ensuring the dataset size cannot include zeros or be empty. Un-filtering and updatingtZarrCreate/invalidSizeInput test.

3. Ensuring the chunk size cannot include zeros or be empty. Un-filtering and updating tZarrCreate/invalidChunkSize test.

4. Making extractS3BucketNameAndPath into a static method. It is a helper method that does not use anything instance-specific - making it static will allow to unit-test it (see Add test to lock pattern verification for cloud-storage workflows #88 ).

5. Un-filtering and updating tZarrAttributes/verifyArrayAttributeInfo and tZarrAttributes/verifyAttrOverwrite tests (see Attribute transpose and type change behavior looks incorrect #34 - due to JSON limitations, the round-trip of attribute data is imperfect, so we should just capture the current behavior instead of keeping the tests filtered).

6. Adding tZarrCreate/createDefaultArray test - I don't think we were previously capturing the behavior of which default attributes zarrcreate generates.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 96.72%. Comparing base (1ad7173) to head (249c0e1).
Report is 6 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #105      +/-   ##
==========================================
+ Coverage   96.66%   96.72%   +0.06%     
==========================================
  Files           8        8              
  Lines         210      214       +4     
==========================================
+ Hits          203      207       +4     
  Misses          7        7              

☔ 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.

[jm9176]

Thanks for updating the tests.