Prepare zrisher PR #1179 for merging into master branch

[martinholmer]

This pull request resolves a few conflicts with the master branch that have arisen over the course of the two months since #1179 was submitted by @zrisher. Many thanks, @zrisher, for the very useful idea and all your implementation work. See the conversation at #1179 for more details on the rationale for these changes.

@MattHJensen @feenberg @Amy-Xu @andersonfrailey @GoFroggyRun @codykallen

[codecov-io]

Codecov Report

Merging #1285 into master will decrease coverage by 0.03%.
The diff coverage is 94.11%.

@@            Coverage Diff             @@
##           master    #1285      +/-   ##
==========================================
- Coverage   99.68%   99.65%   -0.04%     
==========================================
  Files          38       38              
  Lines        2857     2867      +10     
==========================================
+ Hits         2848     2857       +9     
- Misses          9       10       +1

Impacted Files	Coverage Δ
taxcalc/records.py	98.73% <94.11%> (-0.39%)	⬇️

---

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 0807386...37dc3f7. Read the comment docs.

[martinholmer]

Note that the issues raised by @zrisher in #1179 on Feb 20th will be resolved after #1285 is merged.

[martinholmer]

Note that the codecov problem with #1185 is inescapable given that we have not figured out how to test the reading of .csv and .json data from a Python egg. The basic idea of #1179 is to put Records class variable metadata into a separate JSON file. And this file will be read from the egg when the Records class is used in a conda package. As a result, this pull request increases the number of untested statements from 9 to 10.

[MattHJensen]

+1 here. Thanks @martinholmer and @zrisher. Merging now so that we can have these changes in 0.8.1, which I am releasing today.

[MattHJensen]

@zrisher, by merging now I don't mean that you shouldn't feel free to look over @martinholmer's additions and suggest any follow ons in a separate PR.