Standardize deletion mechanics across ~13 BAOs

[colemanw]

Overview

Refactors various BAO code to use new deleteRecord function and deprecate ad-hoc del() functions. Gets the test in #22205 to pass.

Note: This fixes some intra-alpha regressions with ManagedEntities. Following #21989 and #22068, it becomes mandatory to support the hook_post for APIv4-based cleanup of managed-entities. (This is true regardless of whether downstream's specific record is inserted via APIv4. It only matters if the entity declares support for APIv4 management.)

Before

Less standardization of delete functions and the hooks they call.

After

More standardization.

Technical Details

Annotating a BAO::del() function as @deprecated tells APIv4 not to use it. It will call deleteRecord instead.

Entities failing the test in #22205:

• CaseType
• Contact (false-positive, test needs fixing)
• ContactType
• FinancialAccount
• FinancialType
• LocationType
• MembershipStatus
• MembershipType
• MessageTemplate
• OptionGroup
• PaymentProcessor
• PaymentProcessorType
• RelationshipType
• WordReplacement

[civibot]

(Standard links)

• If this is your first pull-request for CiviCRM, please browse CONTRIBUTING.md for information about the development and testing processes.
• If you are reviewing this pull-request, you may wish to consult the test sites and the Review Standards (long template, short template).

[colemanw]

@totten I'm not sure how to interpret the test output. Retest this please.

[colemanw]

@totten this now fixes all 13 noncompliant BAOs, but Jenkins seems to be choking on something and I think it's unrelated to this PR. "Fatal: At least one command failed abnormally [ mixin]"

[colemanw]

Retest this please

[colemanw]

@totten this is passing now. We could merge it and then retest #22205

[totten]

this is passing now. We could merge it and then retest #22205

Yeah, sounds good. This looks like it was a significant effort - requiring some cleanup in ~13 entities!

I had some concern/hesitation about the size of the changes, and I want to leave a few notes on that. At a high level, moving logic from the create/delete methods to the hook-listeners poses a risk for contrib hook-listeners (because a contrib listener could find its relative placement changed in the flow of code). However, I think we should go ahead and merge. There are two major mitigating factors here:

1. These specific entities do not currently emit pre/post hooks - that's why they've been targeted for changes. No hooks fired in the past -- so no existing contrib-listeners. It's hard to be absolutely confident, but at a high level it should be fine.

2. The change is not just an idle refactoring - it's needed to fix intra-alpha regressions with handling of managed entities. And we need to move to beta ASAP. It's better to get the fix we have (and which now passes the main test-suite) so that we can freeze (and then get more r-run exposure in the beta period). The alternatives would either lock-in the regression or reduce the length of the beta period.

So, yeah, let's merge - and branch for 5.45.beta soon.