Dublincore parser to replace arxiv parser

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Comparison is base (1ebe706) 89.80% compared to head (fc88789) 89.78%.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #84 +/- ## ========================================== - Coverage 89.80% 89.78% -0.02% ========================================== Files 25 25 Lines 2619 2615 -4 ========================================== - Hits 2352 2348 -4 Misses 267 267 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[seasidesparrow]

Minor change to dubcore: the PoS records have a <dc:publisher> tag, in this case "Sissa Medialab". I'd like to capture that because there's space for it in base.py (in output["publication"]["publisher"])

I think if you add a few lines to look for dc publisher and then populate self.base_metadata["publisher"], that should be sufficient.