Python 3.13 migrator fails with KeyError: 'patch' in (recently added) ops code

[jakirkham]

Am seeing the following error with the polars migration in this bot job:

Traceback (most recent call last):
  File "/home/runner/work/cf-scripts/cf-scripts/cf-scripts/conda_forge_tick/auto_tick.py", line 721, in _run_migrator_on_feedstock_branch
    migrator_uid, pr_json = run_with_tmpdir(
                            ^^^^^^^^^^^^^^^^
  File "/home/runner/work/cf-scripts/cf-scripts/cf-scripts/conda_forge_tick/auto_tick.py", line 445, in run_with_tmpdir
    return run(
           ^^^^
  File "/home/runner/work/cf-scripts/cf-scripts/cf-scripts/conda_forge_tick/auto_tick.py", line 547, in run
    rerender_info = _run_rerender(
                    ^^^^^^^^^^^^^^
  File "/home/runner/work/cf-scripts/cf-scripts/cf-scripts/conda_forge_tick/auto_tick.py", line 252, in _run_rerender
    rerender_msg = rerender_feedstock(str(context.local_clone_dir), timeout=900)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/runner/work/cf-scripts/cf-scripts/cf-scripts/conda_forge_tick/rerender_feedstock.py", line 24, in rerender_feedstock
    return _rerender(feedstock_dir, timeout=timeout, use_container=use_container)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/runner/micromamba/envs/cf-scripts/lib/python3.11/site-packages/conda_forge_feedstock_ops/rerender.py", line 48, in rerender
    return rerender_containerized(
           ^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/runner/micromamba/envs/cf-scripts/lib/python3.11/site-packages/conda_forge_feedstock_ops/rerender.py", line 116, in rerender_containerized
    if data["commit_message"] is not None and data["patch"] is not None:
                                              ~~~~^^^^^^^^^
KeyError: 'patch'

However the error doesn't seem to indicate an issue with polars

Instead it leads back to the code in this repo. It looks like the line with the error was added last week with PR: #22

[jakirkham]

cc @beckermr

[jakirkham]

Another failure is molmod in this job. Perhaps that is an easier case to debug

Also nnetsauce in this job

[beckermr]

This has already been fixed. We're waiting for the bot to retry failures. It does this basically last.

[jakirkham]

Where was this fixed?

[beckermr]

See the discussion here: regro/cf-graph-countyfair#35

I rebuilt the container by hand to fix the issue and the sync is now working.

[beckermr]

This is fixed. I moved the lock file back to cf-scripts.

[jakirkham]

Thanks Matt! 🙏