Use the numpy dtype in get_type if the array is empty

[oruebel]

Motivation

Fix issue identified in oruebel/ndx-icephys-meta#57 with validating empty datasets. This is triggered in the icephys extension because it may use an empty DynamicTable (with just and id column) resulting in an empty columns dataset.

How to test the behavior?

See oruebel/ndx-icephys-meta#57 (comment)

Checklist

• Have you checked our Contributing document?
• Have you ensured the PR description clearly describes problem and the solution?
• Is your contribution compliant with our coding style ? This can be checked running flake8 from the source directory.
• Have you checked to ensure that there aren't other open Pull Requests for the same change?
• Have you included the relevant issue number using #XXX notation where XXX is the issue number ? By including "Fix #XXX" you allow GitHub to close the corresponding issue.

[codecov]

Codecov Report

Merging #318 into dev will decrease coverage by 0.02%.
The diff coverage is 33.33%.

@@            Coverage Diff             @@
##              dev     #318      +/-   ##
==========================================
- Coverage   72.74%   72.72%   -0.03%     
==========================================
  Files          33       33              
  Lines        6436     6438       +2     
  Branches     1409     1410       +1     
==========================================
  Hits         4682     4682              
- Misses       1327     1328       +1     
- Partials      427      428       +1     

Impacted Files	Coverage Δ
src/hdmf/validate/validator.py	70.90% <33.33%> (-0.41%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9da9162...ecf2a0a. Read the comment docs.

[oruebel]

@rly I'm not sure whether the changes that are on the dev branch fix this issue. If yes, then we should just close this PR and issue.

It looks like #319 is trying to fix the same issue, so I guess this PR can be closed.