Provide for a cleaner API for querying NASA archive

[sashank27]

As per the discussions in #1351 and #1605, refactor the NASA exoplanet API to query for a specific planet (using query_planet()) as well as for host stellar system (using query_star()), using their API.
This prevents unnecessary hacking with the downloaded full table.

[keflavich]

Please add tests to replace those removed.

[pep8speaks]

Hello @sashank27! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2019-12-30 17:03:30 UTC

[codecov]

Codecov Report

Merging #1606 into master will decrease coverage by 0.07%.
The diff coverage is 32.25%.

@@            Coverage Diff             @@
##           master    #1606      +/-   ##
==========================================
- Coverage   62.91%   62.83%   -0.08%     
==========================================
  Files         185      185              
  Lines       14956    14976      +20     
==========================================
+ Hits         9409     9410       +1     
- Misses       5547     5566      +19

Impacted Files	Coverage Δ
...y/nasa_exoplanet_archive/nasa_exoplanet_archive.py	64.06% <32.25%> (-26.85%)	⬇️

---

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 56025c6...f7dadcf. Read the comment docs.

[keflavich]

lgtm! If you want to add a changelog entry, I'll merge this afterward, otherwise I'll merge next time I have access to command line (I'm writing this from mobile, so I can't easily add the changelog entry myself now)

[sashank27]

@keflavich what about the coverage? Also, do @bsipocz also need to review this?

[keflavich]

Coverage isn't an issue; we generally see reduction in coverage for any new features. We don't generally require multiple reviews.

[sashank27]

@keflavich @bsipocz thanks for the support!

[bmorris3]

Hi all – this PR is causing backwards compatibility failures because it removes the get_confirmed_planets_table function. Was that intentional?

[keflavich]

If I understand correctly, get_confirmed_planets_table retrieved the whole table? I think this should be deprecated, but should not have simply been removed.

Can we re-implement get_confirmed_planets_table as a wrapped of query_planets, but with a deprecation warning suggesting that you should use the newer syntax?