Collect OpenAlex forward citations

[georgerichardson]

Description

This adds:

• A pipeline to collect forward citations from the initial (core) works collected
• Information on how to run the pipeline in the pipelines README
• A getter for forward citations data
• A test for one function in that pipeline

closes #37

Instructions for Reviewer

You can test the pipeline by running it without the production flag python dap_aria_mapping/pipeline/data_collection/openalex_forward_citations.py --datastore=s3 run

Tests can be run using pyest dap_aria_mapping/pipeline/data_collection/tests/test_openalex_forward_citations.py

Please pay special attention the flow generally. I'm not sure having to specify a year is the most logical approach to the collection and some of the parameters aren't tested in test mode. Most of the logic is contained in the flow so can't be easily unit tested. I'm also not sure if the way I'm saving and 'getting' the results is the most intuitive.

In particular, I think the way I'm handling imports during steps and the batched step is a bit of a clunky mix of approaches. Keen to generally get tips on best practice for that.

Checklist:

• [x] I have refactored my code out from notebooks/
• [x] I have checked the code runs
• [x] I have tested the code
• [x] I have run pre-commit and addressed any issues not automatically fixed
• [ ] ~I have merged any new changes from dev~
• [x] I have documented the code
  • [x] Major functions have docstrings
  • [x] Appropriate information has been added to READMEs
• [x] I have explained this PR above
• [x] I have requested a code review

[georgerichardson]

@Jack-Vines not sure quite what I did there but seemed to push a file that was partially updated. Have updated and fixed now. Hoepfully this should work fine.

The original version of this flow took a min and max year, but I decided to simplify it and the output file names by only allowing one year or all. In practice, I'm not sure all years should ever be run in one go as it'd hit the OpenAlex limit. Running the same script repeatedly is also tedious though. Ideas for a future enhancement of a script or scheduling system welcome!