PTV-682 - temp fix to Pangenome viewer

[briehl]

Peripherally related to other copy/sharing problems, I'd say this is more a problem of policy. Basically, in an original narrative, the object name and current workspace name are passed to the pangenome viewer, so it looks up the object by the "ws_name_a/obj_name_a" formulation.

In a copied narrative, due to some deep flaws in the architecture (being addressed here: #1214 ), it gets passed the current workspace name still, but now the UPA of the object from the original workspace, so it tries to build a reference as "ws_name_b/ws_id_a/obj_id_a/obj_ver_a" which is wrong.

This PR fixes a few minor issues in other modules, then "fixes" the problem in the pangenome viewer by having it interpret the object identifier passed to it. If it looks like an UPA (has a '/' character), it gets treated as such, otherwise, it builds the reference with the workspace name and object name.

The Real Fix will come later from the other copy/sharing fixes, but this'll be enough to deal with the ticket as is.

[coveralls]

Coverage increased (+0.0002%) to 12.457% when pulling 9f670e3 on briehl:PTV-682-pangenome-widget into 2935f95 on kbase:develop.