Closed tacaswell closed 4 months ago
I do not understand why the 3.8 test works on main (as pip install sphinx<=7.2
fails on py38 for me locally....)
Yes, I was seeing a handful of errors from sphinx (I did not dig too deep) but given the amount of other smoke from sphinx 7.3 (and that tests passed on main with 7.2.6) I took a guess and pinned back.
@jarrodmillman this one is ready to go but I'm not sure what's going on with the CI status. I tried to merge but the only thing it allows is "enabling" merging once these jobs finish. I'm happy to go in and fixup the CI to disallow this but I don't to meddle if it might be violating some policy. Personally I'd be in favor of removing the "Required" tag on these jobs, especially since the dependencies can get quite complicated for numpydoc's support windows.
oh, I see what happened...the version pinning changed the job name which changed to miss the rules...
@jarrodmillman Would it be possible to get a release with this in it? I'm happy to do the leg work if needed.
Looks like the release action has a && github.actor == 'jarrodmillman'
in the conditional so @jarrodmillman I'll let you take care of it 🙂
Please test https://pypi.org/project/numpydoc/1.8.0rc1/. I ran into an issue #577 when I tried to release 1.8.0rc0. Has anyone run into that before?
The value returned by the coroutine is carried on the
StopIteration
Exception that is raised when exhausted.This is a follow up to https://github.com/numpy/numpydoc/pull/145 which allows documenting. This came up in the discussions then (https://github.com/numpy/numpydoc/issues/47#issuecomment-344104267), but I missed that it was forbidden!