Empty list of OAI-PMH sets to harvest due to exclusion for some centre registry endpoints

[twagoo]

After implementation of #26, endpoints that have only set(s) defined that are excluded can end up having no records harvested at all if there are also records to be harvested without a specified set.

A concrete example is the HZSK endpoint. It does not specify sets but the centre registry defines a set of the 'WebLicht' type. The harvester applies the following filtering:

[WebLicht]
⬇
[]

The result of this can be seen in the endpoint specific log:

2020-07-16T18:04:16,689 INFO [HZSK Repository] Worker - Processing provider[HZSK Repository (only set(s): ) @ http://corpora.uni-hamburg.de:8080/oai/provider] using scenario[ListRecords], incremental[false], timeout[60] and retry[count=2,delays=[10000]]

only set(s): implies that nothing is harvested. Solution: if only an empty list of sets remains, set harvesting should not take place.

[menzowindhouwer]

Does your commit already fix the problem? Otherwise I can take it along to the work I'm currently doing, i.e. bringing all (HuC/CLARIN) branches together in a new 1.2 release ...

[twagoo]

@menzowindhouwer

Does your commit already fix the problem?

I haven't checked it yet. Will need to test, ideally write some unit tests. Maybe I'll manage today or tomorrow, but if you see a chance to look at this in the meantime - feel free :) Either way I'll keep you posted!

[menzowindhouwer]

I'm working on the tests right now, so I'll add one. See https://github.com/menzowindhouwer/oai-harvest-manager/tree/develop