Adding array support for unflatten_dict

[galvana]

Closes #3493

Description Of Changes

This ticket updates the unflatten_dict function to be able to "unflatten" array values

Code Changes

• Updated the unflatten_dict function and updated the test cases

Pre-Merge Checklist

• All CI Pipelines Succeeded
• Issue Requirements are Met
• Update CHANGELOG.md

[cypress]

Passing run #2968 ↗︎

0	4	0	0	0
⚠️ You've recorded test results over your free plan limit. Upgrade your plan to view test results.

Details:

Merge bd3e2d7 into c7effb1...
Project: fides	Commit: c51e8c8570 ℹ️
Status: Passed	Duration: 00:48 💡
Started: Jun 28, 2023 10:24 PM	Ended: Jun 28, 2023 10:25 PM

_{This comment has been generated by cypress-bot as a result of this project's GitHub integration settings.}

[codecov]

Codecov Report

Patch coverage: 90.90% and project coverage change: +38.36 🎉

Comparison is base (c7effb1) 48.72% compared to head (035a1b6) 87.09%.

❗ Current head 035a1b6 differs from pull request most recent head bd3e2d7. Consider uploading reports for the commit bd3e2d7 to get more accurate results

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #3629       +/-   ##
===========================================
+ Coverage   48.72%   87.09%   +38.36%     
===========================================
  Files         310      310               
  Lines       19052    19042       -10     
  Branches     2439     2441        +2     
===========================================
+ Hits         9284    16585     +7301     
+ Misses       9604     2028     -7576     
- Partials      164      429      +265     

Impacted Files	Coverage Δ
src/fides/api/util/saas_util.py	87.64% <90.90%> (+54.50%)	⬆️

... and 207 files with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[adamsachs]

this looks good, tricky change! but great test coverage, the tests make clear the behavior of the change and confirms it's working on a number of different edge cases.

i still have trouble fully tracking all the flattened dict transformations that come thru with the masking engine codepath, so i just want to confirm that we're not going to hit an edge case here where an integer-like string, e.g. "0" or "1", is actually intended to be a dict key? rather than an indicator of the index of the element in the list as we're assuming/treating it now.