hepdata=> select * from relatedtable order by id;
id | table_doi | related_doi
----+--------------------------------+--------------------------------
61 | 10.17182/hepdata.146013.v1/t1 | 10.17182/hepdata.146013.v1/t1
62 | 10.17182/hepdata.146013.v1/t2 | 10.17182/hepdata.146013.v1/t2
63 | 10.17182/hepdata.146013.v1/t3 | 10.17182/hepdata.146013.v1/t3
64 | 10.17182/hepdata.146013.v1/t4 | 10.17182/hepdata.146013.v1/t4
65 | 10.17182/hepdata.146013.v1/t5 | 10.17182/hepdata.146013.v1/t5
where the table_doi and related_doi are identical. Although this situation arises from a misunderstanding by the submitter, the submission code should not create a new RelatedTable object if table_doi = related_doi. Similarly, a new RelatedRecid object should not be created if this_recid = related_recid.
The hepdata-submission and hepdata_lib documentation should also be updated to mention that self-referring related_to_table_dois and related_to_hepdata_records fields should not be used.
By inspecting the database, I found entries like:
where the
table_doi
andrelated_doi
are identical. Although this situation arises from a misunderstanding by the submitter, the submission code should not create a newRelatedTable
object iftable_doi = related_doi
. Similarly, a newRelatedRecid
object should not be created ifthis_recid = related_recid
.The
hepdata-submission
andhepdata_lib
documentation should also be updated to mention that self-referringrelated_to_table_dois
andrelated_to_hepdata_records
fields should not be used.