Ap identification

[SFJohnson24]

This PR updates logic for identifying AP class which used to seek out the parent dataset then ID based on that. This PR still uses that logic, if it fails, it attempts to use the AP suffix--testing it in the case of both standard and custom domains attached to AP--.

[pendingintent]

AP-- will not have an associated domain other than APRELSUB. MH will not have a relationship to APMH. All AP domains are only to record details for a person associated with a USUBJID. There is no parent/child relationship with AP--. The suffix only indicates the type of data collected in the dataset. In this case, this is medical history for the associated person. The APID in APMH will join to APRELSUB.APID and the APRELSUB.USUBJID will indicate to which subject the person is associated.

[SFJohnson24]

I removed the logic for AP that I preserved, that was built on a false premise.
I also updated ap_suffix for APRELSUB/APRELSPEC as it would return "" for ap_suffix. In returning this, the the suffix is used to ID the dataset as RELSUB/RELSPEC and thus relationship. It would not impact wildcard replacement as there are not wildcard variables in either dataset from what I can tell from library

I updated tests as the ap_suffix had several failing tests where there was no metadata on the dataset--meaning a dataset was submitted but completely empty. It no longer returns "" and instead would error out given the total lack of metadata. I would expect this given the blank data which would force a rule skip for that dataset and rule