add active PRs to black ignore

[marinegor]

Fixes #

Changes made in this Pull Request:

• add files present in open PRs to the list of extend-exclude for black

PR Checklist

• Tests?
• Docs?
• CHANGELOG updated?
• Issue raised/referenced?

Developers certificate of origin

• I certify that this contribution is covered by the LGPLv2.1+ license as defined in our LICENSE and adheres to the Developer Certificate of Origin.

---

📚 Documentation preview 📚: https://mdanalysis--4815.org.readthedocs.build/en/4815/

[marinegor]

cc @RMeli -- as we discussed, I finally managed to automate creation of the list of files in active PRs. Not sure if it works as necessary (e.g. the comment lines in the list worry me, and I assume that everything after the first comment is not accounted for) -- could you double-check if it works for you? Or perhaps there's a better way to configure these files?

[marinegor]

The code I used to generate the lists:

#!/bin/bash

echo '' > active_files.lst
for num in $(gh pr list -L 200 | awk '{print $1}'); do
  echo Working on $num...
  echo "# ${num} $(gh pr view ${num} --json title --jq .title)" >> active_files.lst
  gh pr diff  --name-only ${num} >> active_files.lst
  echo '# ---------------------' >> active_files.lst
done

cat active_files.lst | grep -E '^(#|package/.*.py)' > active_files_package.lst
cat active_files.lst | grep -E '^(#|testuite/.*.py)' > active_files_testsuite.lst

[RMeli]

Thanks @marinegor, looks useful! We will have to fix CI first, and merge the lib formatting which is already mainly done.

[RMeli]

@marinegor I had a look at your script. I think looking at 200 PRs is a bit pessimistic. I think it might be worth including only PRs that have been updated (JSON filed: updatedAt) in the last few months. Maybe 2 months is enough (or even too much)? If a PR has been stale for a while, there will be likely conflicts anyways and we can help on a case-by-case basis.

[marinegor]

@RMeli 2 months it is, updated

[RMeli]

Thanks @marinegor, can you please share the script to filter by updatedAt? (Sorry, I'm being lazy here...)

[RMeli]

Please also fix the conflicts and exclude the files of the bits that we already formatted (lib/ for instance).

[marinegor]

@RMeli the script:

#!/bin/bash

rm active_files.lst
for num in \
  $(gh pr list -L 200 --json number,updatedAt,createdAt,title \
  | jq '[.[] | select(.updatedAt > ("2024-10-02"))]' \
  | jq '.[] | .number');
do
  echo Working on $num...
  echo "# ${num} $(gh pr view ${num} --json title --jq .title)" >> active_files.lst
  gh pr diff  --name-only ${num} >> active_files.lst
  echo '# ---------------------' >> active_files.lst
done

cat active_files.lst | grep -E '^(#|package/.*.py)' > active_files_package.lst
cat active_files.lst | grep -E '^(#|testsuite/.*.py)' > active_files_testsuite.lst

and then I manually move the lists to pyproject.toml's

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.63%. Comparing base (a27591c) to head (e3a98e4).
Report is 9 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4815      +/-   ##
===========================================
- Coverage    93.68%   93.63%   -0.05%     
===========================================
  Files          177      189      +12     
  Lines        21743    22845    +1102     
  Branches      3055     3064       +9     
===========================================
+ Hits         20370    21391    +1021     
- Misses         927     1002      +75     
- Partials       446      452       +6     

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

[RMeli]

As discussed I closed #4518. The situation looks much better.

PRs that still do a reasonable amount of changes are

• Fix for issue #4655 Removing mutable data structures and function calls as default arguments in the entire codebase #4810 @RMeli
• Implementing gemmi-based mmcif reader (with easy extension to PDB/PDBx and mmJSON) #4712 @marinegor
• Deprecate ts argument to _read_next_timestep #4334 @hmacdope

#4810 should be easy to fix and incorporate quickly. But I think if you can update the exclusion list, we can merge this and start formatting what is left.

[RMeli]

Please do save the active_files_*.lst files, so that we have the list of PRs to keep track of and we can do the formatting as soon as a PR is merged.

[RMeli]

@marinegor any chance this can be updated in the next few days? I plan to work on the reminder of the code base during Christmas break.

[marinegor]

@RMeli sure, just ping me like 48 hours in advance, top.

Also, I could add the script to the maintainer directory of the package, though don't really think it's a good idea -- but ping me if it is.

[RMeli]

I think about now is more and less 48 hours in advance to my holidays 😛

I think the script is very specific for this transition, so it is good to have it in the PR comments, but I don't think there is a real need to have it within the repo.

[marinegor]

@RMeli done!

I meant that perhaps it's good to not do that again if a PR gets updated, but then I just decided to update the script:

#!/bin/bash

rm active_files.lst
for num in \
  $(gh pr list -L 200 --json number,updatedAt,createdAt,title \
  | jq '[.[] | select(.updatedAt > ("2024-10-02"))]' \
  | jq '.[] | .number');
do
  echo Working on $num...
  echo "# ${num} $(gh pr view ${num} --json title --jq .title)" >> active_files.lst
  gh pr diff  --name-only ${num} >> active_files.lst
  echo '# ---------------------' >> active_files.lst
done


cat active_files.lst | sed '/#/d' | sed -e 's/\.py$/\\.py/' | sed -e 's/^\(.*\)\(\n\1\)\+$/\1/' | grep -E '^(#|package/.*.py)' | sed '1!s/^/| /' > active_files_package.lst
cat active_files.lst | sed '/#/d' | sed -e 's/\.py$/\\.py/' | sed -e 's/^\(.*\)\(\n\1\)\+$/\1/' | grep -E '^(#|testsuite/.*.py)' | sed '1!s/^/| /' > active_files_testsuite.lst

now the files that it outputs can be directly pasted after the __pycache__ line in the respective pyproject.toml file.

[RMeli]

Thanks @marinegor, I'm going to merge this. The files are useful so that we can remove the different exclusions when PRs are merged. I'll make a note to remove them once formatting is completed.