audit: M2C should alert when it finds something on disk that isn't where it's expected to be

[jmartin-sul]

e.g., if MoabToCatalog finds a moab on disk16, and that moab is listed in the catalog as being on disk05, we should send a honeybadger alert, and we should probably also update the status for the CompleteMoab. my gut feeling is that we should not silently update the catalog to point to the new location: if things are in this state, it indicates, for example that someone manually moved a moab, but didn't update the catalog properly or clean things up completely. we should be alerted to this situation to make sure we've actually cleaned things up properly.

spawned by #1177

[ndushay]

IMO, these things should all happen:

• the M2C audit code should trap this error case
  • IFF we want to autofix
    • error handling code ensures Moab is Not at the location indicated by catalog
    • error handling code ensures Moab at the current location matches the catalog in all ways (e.g. version, full validation) except for disk location
    • error handling code updates the disk location in the catalog if the above 2 checks are true (o.w. cannot auto-fix)
    • error handling code reports to honeybadger that the update has been made
    • error handling code reports to preservation_audit WF that update has been made
  • if no autofix
    • pres_cat status for the object should NOT be 'ok'
    • preservation_audit workflow should be used to indicate error (with useful message)
    • alert sent to honeybadger

[julianmorley]

Related to #1139, I think.

[jmartin-sul]

yup, agreed w/ @julianmorley that those are related