[ENG-8832] Most recent numbered version of a preprint fails to show for unversioned URL

[opaduchak]

Purpose

Change guid resolve strategy to match expectations of osf users

Changes

• changed on_submit hook to link parent guid to latest version regardles of moderation workflow
• implemented migration job via management command

QA Notes

Please make verification statements inspired by your code and what your code touches.

• Verify
• Verify

What are the areas of risk?

Any concerns/considerations/questions that development raised?

Documentation

Side Effects

Ticket

[cslzchen]

A few questions/thoughts:

• The side effect of this change is the no version URL will lead to page not found if a pending pre-moderation preprints exists

• Does the migration handle the case where a new version is "created but not submitted"? This orphaned preprint has the highest number but should not be considered as the latest.

[cslzchen]

The update looks good but rejected version occurred to me when I looked at the unit test. Probably need to follow up with product on what's the behavior they want.

[cslzchen]

Curious if we should add this to skipped count or let it have its own count?

[opaduchak]

We do not skip this, and keeping separate count is impossible as this count is liked to each preprint.
Here we just iterate through preprint versions and search fro most recent submitted version

[cslzchen]

This is another case that I forgot to ask during my first review. And this is probably a product question too.

• Can you verify that rejected preprints will keep the version number? (Rejection only happens for pre-mod).
• Does product really want the behavior that non-version guid leads to rejected preprints? Or should we change the version to point back to an earlier "latest version" upon rejection?