Revise test_reforms.py so that reform info is in a JSON file

[martinholmer]

This pull request refactors the test_reforms.py code so that information about all of the 55 tested reforms is in one JSON file named reforms.json. These changes are based on the suggestion made by @hdoupe in an recent discussion of how to approach the reform tests.

These changes permit TaxBrain to specify an HTML link to the more readable reforms.json file, which contains no Python test code. There is no change in test logic and there is no change in tax-calculating logic or test results.

This pull request will require a change in one HTML link in the "Transparency and Replicability" section of the "About TaxBrain" page.

@MattHJensen @Amy-Xu @hdoupe @brittainhard

[codecov-io]

Codecov Report

Merging #1576 into master will not change coverage.
The diff coverage is n/a.

@@          Coverage Diff           @@
##           master   #1576   +/-   ##
======================================
  Coverage     100%    100%           
======================================
  Files          37      37           
  Lines        2728    2728           
======================================
  Hits         2728    2728

---

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 8976879...d6a8f72. Read the comment docs.

[MattHJensen]

This is definitely an improvement. I could see this leading to a system where we write the results into reforms_actual and copy them to reforms_expect.json

[martinholmer]

@MattHJensen said about pull request #1576:

This is definitely an improvement. I could see this leading to a system where we write the results into reforms_actual and copy them to reforms_expect.json.

I didn't see how to do that because the 55 tests are not necessarily run in order (because of the use of py.test -n4 option). This makes writing an actual file impossible. Or, if not impossible, more work than it is worth.

[MattHJensen]

@martinholmer said:

I didn't see how to do that because the 55 tests are not necessarily run in order (because of the use of py.test -n4 option). This makes writing an actual file impossible. Or, if not impossible, more work than it is worth.

I see, thanks for the explanation, @martinholmer

[hdoupe]

Nice work @martinholmer. This looks good

[martinholmer]

Pull request #1576 has been open for seven days and discussion of it seems to be resolved, so it is being merged.