[Bug]: Bulk Upload created a new changemaker when it should have used the existing

[jmergy]

Component or URL

Bulk Upload

Steps to reproduce

I confirmed the changemaker we were adding data to existed in PDC. It did not have a tax_id associated, so I added it via a changemaker field values batch. The name of the changemaker was different than the one that was in PDC already.

Expected behavior

Expected behavior would be that bulk upload would see an existing changemaker and use that vs. creating a new one. I had thought if the tax_id was present and the same, that would be the way bulk upload finds the existing.

Actual behavior

I ran bulk upload and the data came in but PDC created a new changemaker and associated the bulk upload proposal data with the new changemaker and not the old.

Additional context & screenshots

No response

[jmergy]

@hminsky2002 @kfogel again here might just be we need to understand exactly what we need to do with any data via bulk upload it we want it to land on an existing changemaker or create new one(s)

[slifty]

@jmergy based on current implementation this is expected because we wanted to be careful about the fact that multiple organizations can share a Tax ID despite having different names (e.g. "Foo Foundation Pennsylvania" vs "Foo Foundation Virginia").

We may find that the right solution here should be to introduce a pdc_changemaker_id column which allows an uploader to make the intention explicit.

[jmergy]

@slifty totally makes sense. Again, I think this might be more of a edge case.

I would be happy with something like the pdc_changemaker_id method. If we want to direct them, we do that.

I think we could spend a lot of time trying to solve for a dynamic way but still be thwarted with exceptions that would break it.

[kfogel]

@slifty Is the pdc_changemaker_id solution how we want to go? (I see this is scheduled as Todo already, so maybe the design is ready, or close enough to ready that we felt confident scheduling it.)

[slifty]

@kfogel I do think we want an improvement to prevent accidental changemaker creation; that said for that it might be cleaner to expand the task to accept a flag such as "halt on changemaker creation" which the bulk uploader could then honor and "error" if the uploaded file would have unexpectedly resulted in new changemakers.

There are two benefits to this second approach:

1. People don't have to look up changemaker IDs as part of their bulk upload flow
2. We avoid another "special column" in the bulk uploader (not a huge deal, but every non-basefield column we add is another weird thing the user has to know about).

---

We did mark this as to do last week and I do think this is close to actionable, but at the time we might have been thinking of it as a purely documentation issue.

Maybe we move it to under discussion for today and then next week we can determine if we want to take it on immediately based on all priorities?