Uniqueness of OAI base URL in OHS

We are using OHS version 2.3.1 and we are aggregating content from a number of repositories, including consortia who share an OAI base URL but use individual sets for their collections.

We want to create an archive for each institution, so that they can be represented and filtered separately in our search interface for the user. However, we cannot use the GUI to create an entry for each institution, as the field ‘OAI Base URL’ in the archive has to be unique. We can’t enter the set suffix as a qualifier, as this is not validated by OHS.

Can we create each archive in MySQL instead? How can we recreate the input method of the GUI - which database tables do we need to fill? If we create multiple database entries that way, each having the same OAI base URL, will this ‘break’ the GUI if we try to use it subsequently?

We received the following advice from PKP upon asking this question to support, but they also suggested we should post here, as I don’t think the below fully answers my question above.

“The relevant constraint is going to be whether or not records appear in multiple sets. If each record only appears in a single set, then it should be possible to relax the uniqueness constraint on OAI URLs. However, if records appear in several sets, then there will be collisions and OHS will not behave well.”

Can anyone here advise?

Many thanks.

Hi @ucl,

It’s a requirement of the OAI-PMH specification that OAI base URLs be unique, unfortunately – I don’t think OHS will behave well even if you enter the data manually in MySQL. I would recommend following the spec and making sure that the base URLs are unique.

Alec Smecher
Public Knowledge Project Team