Merging authority records does not change the attached bibliographic record(s) name

[Ditty Mulder]

Reporting in 2.0 beta3

Mergen: 'Brouwer, Eef' (linking 1 bib.record) and 'Brouwer, E.' (linking 1 bib.record)
Step 1: Manage authorities
Step 2: Search: 'Brouwer'
Step 3: Select 'Brouwer, Eef' (Record ID 50)and Mark for merge: MASTER
Step 4: Select 'Brouwer, E.' (Record ID 51) and Mark for merge: TARGET
Step 5: MERGE: Record merge is complete
Result: The authority list shows: 'Brouwer, Eef (Record ID 50) linking 2 records
The bib.record with the original 'Brouwer, E.' is still 'Brouwer, E.' BUT with record ID 50

Example:
Bib. record !:
=008 070101s\\\\\\\\\\\\\\\\\\\\\\\\\\\\eng\d
=100 \\$aBrouwer, Eef$0(CONS)50
=245 \\$aNieuwsbericht
=901 \\$a34$b$c34$tbiblio

Bib.record 2
=008 070101s\\\\\\\\\\\\\\\\\\\\\\\\\\\\eng\d
=100 \\$aBrouwer, E.$0(CONS)50
=245 \\$aPensioen
=901 \\$a35$b$c35$tbiblio

[Dan Scott (denials)]

Quick suspicion is that the ingest.reingest.force_on_same_marc flag - which is now respected by authority.record_entry - defeats the 'harmless update to target_record to trigger auto-update' in authority.propagate_changes() (which simply sets DELETED = FALSE, ergo the MARC will be the same).

[Dan Scott (denials)]

Of course, that ignores the forced setting of ingest.reingest.force_on_same_marc to TRUE earlier in that function. Hrm.

[Dan Scott (denials)]

Okay. I think it's just confusion between source_record and target_record. Testing...

[Dan Scott (denials)]

Committed a fix in r19033.