Re-train paper ranking model with data from curated_papers.tsv

[nagutm]

This pull request implements functionality in the paper ranking model to include data from the curated_papers.tsv file when re-training the model and excluding any curated publications from re-appearing in the predictions.tsv file.

[codecov]

Codecov Report

Attention: Patch coverage is 84.61538% with 4 lines in your changes missing coverage. Please review.

Project coverage is 46.62%. Comparing base (8950e70) to head (e94d7e9).
Report is 171 commits behind head on main.

Files with missing lines	Patch %	Lines
src/bioregistry/analysis/paper_ranking.py	84.61%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1248      +/-   ##
==========================================
+ Coverage   42.51%   46.62%   +4.10%     
==========================================
  Files         117      118       +1     
  Lines        8327     8286      -41     
  Branches     1963     1362     -601     
==========================================
+ Hits         3540     3863     +323     
+ Misses       4582     4238     -344     
+ Partials      205      185      -20     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[cthoyt]

there's an (more?)itertools function for chunking that will make this code more idiomatic

[nagutm]

The latest commit introduces a unit test for the paper_ranking model to serve as a smoke test. The test ensures that the pipeline runs successfully without errors and generates the expected output files when provided with valid inputs.

It executes the paper_ranking CLI pipeline via the CliRunner to ensure that

• pipeline runs successfully with exit code 0.
• Output files evaluation.tsv, importances.tsv, and predictions.tsv are generated in the expected output directory.