[pyos][design] Not one, not two, but four different methods to get and parse crossref metadata

[sneakers-the-rat]

Part of: pyOpenSci/software-submission#267 (comment)

• _verify_doi_and_get_metadata -

  OneCite/onecite/pipeline.py

  Line 365 in 12b1dea

  def _verify_doi_and_get_metadata(self, doi: str) -> Optional[Dict]:

• _resolve_doi_via_crossref_title -

  OneCite/onecite/pipeline.py

  Line 1655 in 12b1dea

  def _resolve_doi_via_crossref_title(self, candidate_title: str, original_query: str) -> Optional[Dict]:

• _search_crossref -

  OneCite/onecite/pipeline.py

  Line 1705 in 12b1dea

  def _search_crossref(self, query: str, limit: int = 15) -> List[Dict]:

• _get_crossref_metadata -

  OneCite/onecite/pipeline.py

  Line 2578 in 12b1dea

  def _get_crossref_metadata(self, doi: str) -> Optional[Dict]:

Each of which is slightly different. There should be one and only one method of getting, and especially parsing crossref metadata. The methods appear in both the identifier and the enrichment module, so lucky us, we get to hit crossref typically three times for every item - once in the "crossref search strategy 1", another for "crossref search strategy 2," and again during enrichment.