Hot fix for polish

[GwennyGit]

Describe the bug

• I: BioCyc URIs/IRIs are missing the 'META:' part which makes them invalid.
• II: If in a model the BioCyc identifiers contain no indication for the BioCyc sub-database like 'META:' for the MetaCyc database this information is also missing from the polished URI/IRI and , thus, causes invalid URIs/IRIs.
• III: Some reported invalid curies in the invalid_curies.tsv make no sense as they have None as database prefix and None as identifier.

Expected behavior

• I & II: The function to generate the URIs/IRIs from bioregistry does not add the 'META:' part. Thus, the implementation in polish needs to handle missing sub-database prefixes for BioCyc URIs/IRIs.
• III: If a database prefix and the corresponding identifier are None this invalid curie should not be reported in the invalid_curies.tsv.

  ---

• [x] Solve issue I
• [x] Solve issue II
• [x] Solve issue III

[GwennyGit]

I changed the implementation of how BioCyc identifiers are treated. The newer version makes more sense with the new setup. However, I also found that the 'bug' was not caused by the 'polish' module itself but rather by the fact that in the AGORA 2 models, it was tested on, the BioCyc identifiers contain no indication for the BioCyc sub-database like 'META:' for the MetaCyc database. I think this issue can be coped with in Polish and, thus, before releasing the newer version I will try to cope with missing sub-database prefixes for BioCyc identifiers.

[GwennyGit]

I found another bug within polish concerning the resulting invalid_curies.tsv file. For a lot of entities None values for the database prefix and the corresponding identifier are printed. Before pushing the changes to main and generating a new release I will investigate and fix this bug.