[bulk data] Import Single CSV cannot import cardinality n tiles by legacy ID

[jacobtylerwalls]

With a CSV like this that provides a legacy ID and multiple tiles for a node...

ResourceID,Concept List n
4,condition
4,credit line

... the single CSV Import in the Bulk Data Manager fails with:

2023-10-20 20:16:33,480 arches.app.etl_modules.save ERROR    duplicate key value violates unique constraint "resource_instances_legacyid_key"
DETAIL:  Key (legacyid)=(4) already exists.
CONTEXT:  SQL statement "INSERT INTO resource_instances(resourceinstanceid, graphid, legacyid, createdtime)
                                VALUES (instance_id, graph_id, legacy_id, now())"
PL/pgSQL function __arches_staging_to_tile(uuid) line 39 at SQL statement

Internal Server Error: /etl-manager

Ensure you use a fresh ResourceID each test iteration, because loading a second time with the same one would be expected to fail.

[jacobtylerwalls]

Just following up from the planning meeting:

• Yes, this affects the Bulk Data Manager (web tool)
• Cardinality n imports do work with a UUID
• This ticket is just noting they don't work with a legacy ID.

Here is a successfully imported resource with two tiles (using a UUID):