[UICAL-268, UICAL-228, UICAL-269, UICAL-220, UICAL-270] Convert to react-final-form-arrays, stripes-types, react 18

[ncovercash]

Jira UICAL-268 UICAL-228 UICAL-269 UICAL-220 UICAL-270

Purpose

This clean up a lot of technical debt with the previous way of handling form state. This should lead to a significantly improved experience, for both users and developers of the module. While in here, I also added some tests for TimeField (UICAL-228) and MCLRowFormatter (UICAL-220)

Approach

Before, a lot of the state management within react-final-form was not done optimally; this PR moves a lot of this logic into react-final-form and react-final-form-arrays' state management.

Other notes

• A lot of the file difference here is from the removal of the old self-hosted types.
• I also performed upgrades of most dependencies, as applicable, and purged some unused ones.
• Some work will very likely need to be done to ensure this does not break platform-complete w.r.t. @types/react resolutions, etc.

[github-actions]

Jest Unit Test Statistics

    1 files  ±  0    70 suites   - 1   3m 55s ⏱️ +12s
261 tests  - 70  260 ✔️  - 70  1 💤 ±0  0 ❌ ±0 
282 runs   - 76  281 ✔️  - 76  1 💤 ±0  0 ❌ ±0 

Results for commit 746eea3. ± Comparison against base commit 0d6ad40.

This pull request removes 141 and adds 71 tests. Note that renamed tests count towards both.

 A closed row with days filled in is valid ‑  A closed row with days filled in is valid
 A closed row with nothing filled in is invalid ‑  A closed row with nothing filled in is invalid
 An open row with nothing filled in is invalid ‑  An open row with nothing filled in is invalid
 An open row with only days filled in is invalid ‑  An open row with only days filled in is invalid
 Bad ref inner closure rows are properly reported ‑  Bad ref inner closure rows are properly reported
 Bad ref inner opening row dates are properly reported ‑  Bad ref inner opening row dates are properly reported
 Conflicts are reported ‑  Conflicts are reported
 Date order is checked ‑  Date order is checked
 Date-time order is checked ‑  Date-time order is checked
 Empty errors are reported alone ‑  Empty errors are reported alone
…

 A non-filled entry does not split ‑  A non-filled entry does not split
 Closed exception missing any date value is not filled ‑  Closed exception missing any date value is not filled
 Closed rows are not considered ‑  Closed rows are not considered
 Default method works as expected ‑  Default method works as expected
 Error message is correct text ‑  Error message is correct text
 Exception with no rows nor name is considered filled ‑  Exception with no rows nor name is considered filled
 Exceptions convert properly ‑  Exceptions convert properly
 Exceptions with all date/time values is filled ‑  Exceptions with all date/time values is filled
 Issues are reported ‑  Issues are reported
 MCLRowFormatter renders ‑  MCLRowFormatter renders
…

♻️ This comment has been updated with latest results.

[ncovercash]

Also before y'all freak out at the number of line changes here: a lot of deletions are from the removal of the in-house stripes-components typings. Also, the rewriting of most all validation/calendar form tests made a pretty heavy impact, too.

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
2 Code Smells

98.7% Coverage
0.0% Duplication

[zburke]

Gotta be honest: there is no practical way to review this PR. Big refactors like "leverage stripes-types" or "reformat for new lint rules" etc that make for really noisy diffs but don't actually change the logic much should be handled in their own PRs. As is, it's hard to separate these noisy-but-likely-safe changes from the must-be-closely-reviewed changes related to react-final-form-arrays.

I think a good way to proceed could be

1. bump to react v18 and stripes-types
2. react-final-form-arrays
3. tests for TimeField, MCLRowFormatter (or maybe do this along with 2.)

I'm not gonna block it if you think the tests are sufficiently comprehensive, but I can't approve it either. Sorry.

[ncovercash]

Gotta be honest: there is no practical way to review this PR. Big refactors like "leverage stripes-types" or "reformat for new lint rules" etc that make for really noisy diffs but don't actually change the logic much should be handled in their own PRs. As is, it's hard to separate these noisy-but-likely-safe changes from the must-be-closely-reviewed changes related to react-final-form-arrays.

I think a good way to proceed could be

1. bump to react v18 and stripes-types
2. react-final-form-arrays
3. tests for TimeField, MCLRowFormatter (or maybe do this along with 2.)

I'm not gonna block it if you think the tests are sufficiently comprehensive, but I can't approve it either. Sorry.

Totally understood! It's a really noisy mess, no doubt about it. I'll try and separate out stripes-types vs. final form bits

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
8 Code Smells

98.7% Coverage
0.0% Duplication