Assign replications to multiple references

[LukasWallrich]

How do we deal with published reports from individual studies - e.g. this ML5 case:
https://journals.sagepub.com/doi/full/10.1177/2515245920917334

They should be linked to both so that we can show all originals for ML5 - but also link the specific replication publication to the original.

This also relates to preprints, which we should be able to link to their targets without duplicates in database.

We might need a "also published as" column for that - but need to remember that this is not symmetric. To be discussed and fixed soon, as it has implications across all interfaces (including pubpeer comments)

[LukasWallrich]

preprint - published (notify about published, but also link from preprint to original)
published - retracted published version (notify about published, but also link from preprint to original)
for many study papers, prioritise the most specific published version, other in url_r

[LukasWallrich]

alt_id_o and alt_id_r which can be dois, other stable identifiers, urls

[LukasWallrich]

Only 10.1017/xps.2022.35 is the canonical link - replace the SCORE report and put that plus the score meta-paper into alt_id (https://osf.io/ybzev)

Then add the SCORE meta-paper to all SCORE replications as alt_id