Centralize/generalize declaration of convention attributes

[emiliom]

This is a demonstration step for centralizing where coordinate and variable attributes are declared. Moving in this direction will make it easier and more robust to maintain adherence to the convention.

Not ready for merging. Though it completes all tests, this PR is intended as a demonstration for discussion.

Sorry for the clutter, but it also contains changes from PR #492, which is not merged yet; those are all in echopype/echodata/echodata.py alone.

Partially addresses #373 (adds range_bin attributes, but not a description in the docs).

[codecov-commenter]

Codecov Report

Merging #493 (6449630) into dev (d996e28) will decrease coverage by 7.00%.
The diff coverage is 60.00%.

@@            Coverage Diff             @@
##              dev     #493      +/-   ##
==========================================
- Coverage   77.13%   70.12%   -7.01%     
==========================================
  Files          39       23      -16     
  Lines        3415     2638     -777     
==========================================
- Hits         2634     1850     -784     
- Misses        781      788       +7     

Flag	Coverage Δ
unittests	70.12% <60.00%> (-7.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
echopype/convert/set_groups_base.py	84.28% <ø> (-3.04%)	⬇️
echopype/echodata/convention/attrs.py	0.00% <0.00%> (ø)
echopype/echodata/echodata.py	61.23% <0.00%> (-27.32%)	⬇️
echopype/convert/set_groups_azfp.py	98.21% <100.00%> (+0.03%)	⬆️
echopype/convert/set_groups_ek60.py	91.26% <100.00%> (+0.08%)	⬆️
echopype/convert/set_groups_ek80.py	94.32% <100.00%> (-1.39%)	⬇️
echopype/__init__.py	0.00% <0.00%> (-100.00%)	⬇️
echopype/echodata/__init__.py	0.00% <0.00%> (-100.00%)	⬇️
echopype/echodata/convention/__init__.py	0.00% <0.00%> (-100.00%)	⬇️
echopype/echodata/api.py	16.66% <0.00%> (-66.67%)	⬇️
... and 24 more

---

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 d996e28...6449630. Read the comment docs.

[emiliom]

TODO: Change from using the attribute definitions in attrs.py to reading it dynamically from a yaml file.

[emiliom]

I resolved the merge conflict we saw today. But: @leewujung & @lsetiawan : a test failed with this PR. When I first created the PR 19 days ago, all tests were successful. The error occurs with test_convert_ek80_cw_bb_in_single_file, and involves a failure to find a raw file:

E           FileNotFoundError: There is no file named Summer2018--D20180905-T033113.raw

I don't see how that could be caused by the changes in this PR, so my guess is that it has to do with something else altogether. Has there been a change in the availability of EK80 test files in the last ~24 hours (eg, related to #498)? I'm hoping this is a transient hiccup. Could either of you take a quick look to diagnose if it looks like it's an external problem that will be addressed elsewhere?

If the problem persists into tomorrow and it's hard to diagnose, let's skip this PR for 0.5.5.

[leewujung]

Oops, that's my fault, as part of the work for #498. It's a tiny _ vs - that took me half an hour to debug out... and I forgot to rename the file in the CI_test_data folder. It should run now.

[leewujung]

Also, @emiliom there seems to be some unused import thing found by the pre-commit.ci bot. Could you check on those?

[emiliom]

Oops, that's my fault, as part of the work for #498. It's a tiny _ vs - that took me half an hour to debug out... and I forgot to rename the file in the CI_test_data folder. It should run now.

Thanks! Good for me, bad for you

Also, @emiliom there seems to be some unused import thing found by the pre-commit.ci bot. Could you check on those?

Nice catch. I'll admit that I had ignored the pre-commit failures altogether. I've fixed it.

[leewujung]

Ok this is ready to merge now. Thanks @emiliom!