GoldStandard table should qualify the source system associated with the accepted or rejected object

[paulalbert1]

The GoldStandard table is overly PubMed-centric. As we prepare to track from feedback from N other source systems (Scopus, WorldCat, etc.), our syntax shouldn't presume that every record is from PubMed.

Top level nodes

I suggest we change the data model for top level nodes from:

{
  "auditLog": [
    ],
  "knownpmids": [
    ],
  "rejectedpmids": [
  ],
  "uid": "jobuck"
}

to:

{
  "auditLog": [
    ],
  "PubMed": [
      "accepted": [
       ],
      "rejected": [
       ],
  "Scopus": [
      "accepted": [
       ],
      "rejected": [
       ],
  "uid": "jobuck"
}

There's probably no need to create Scopus or any other source system until we get feedback from this user.

Audit log

Here, I suggest we go from:

  "auditLog": [
    {
      "action": "ACCEPTED",
      "dateTime": "2019-04-26T16:07:57.179Z",
      "pmids": [
        10869379
      ],
      "uid": "jobuck",
      "userVerbose": "Jochen null Buck "
    },

to:

  "auditLog": [
    {
      "action": "ACCEPTED",
      "dateTime": "2019-04-26T16:07:57.179Z",
      "objectId": [
        10869379
      ],
      "source": "PubMed",
      "uid": "jobuck",
      "userVerbose": "Jochen null Buck "
    },