Add new (2.0) string ufuncs in np.char to UNSUPPORTED_UFUNCS - #15699

Thank you for your contribution to Astropy! 🌌 This checklist is meant to remind the package maintainers who will review this pull request of some common things to look for.

• Do the proposed changes actually accomplish desired goals?
• Do the proposed changes follow the Astropy coding guidelines?
• Are tests added/updated as required? If so, do they follow the Astropy testing guidelines?
• Are docs added/updated as required? If so, do they follow the Astropy documentation guidelines?
• Is rebase and/or squash necessary? If so, please provide the author with appropriate instructions. Also see instructions for rebase and squash.
• Did the CI pass? If no, are the failures related? If you need to run daily and weekly cron jobs as part of the PR, please apply the "Extra CI" label. Codestyle issues can be fixed by the bot.
• Is a change log needed? If yes, did the change log check pass? If no, add the "no-changelog-entry-needed" label. If this is a manual backport, use the "skip-changelog-checks" label unless special changelog handling is necessary.
• Is this a big PR that makes a "What's new?" entry worthwhile and if so, is (1) a "what's new" entry included in this PR and (2) the "whatsnew-needed" label applied?
• Is a milestone set? Milestone must be set but we cannot check for it on Actions; do not let the green checkmark fool you.
• At the time of adding the milestone, if the milestone set requires a backport to release branch(es), apply the appropriate "backport-X.Y.x" label(s) before merge.

Hi @dhomeier, there's already an active PR to resolve this problem (#15691). If you have reasons to reject it can you please let me know there with a review ? Thanks !

@neutrinoceros you're right, this is a separate failure that is not actually covered by CI, see updated comment above; you should be able to reproduce it in your setup by running pip uninstall scipy and then test_quantity_ufuncs.
But since the fix here is rather trivial, feel free to pull it int your PR.

Oh I see, indeed that's a separate problem, so I see no harm in keeping our PRs separate. Thank you !

Indeed I see a difference when running pytest astropy/units/tests/test_quantity_ufuncs.py::TestUfuncHelpers::test_coverage -vvv against main:

E       AssertionError: assert {<ufunc '_rstrip_whitespace'>, <ufunc '_lstrip_chars'>, <ufunc 'startswith'>, <ufunc '_strip_whitespace'>, <ufunc 'isdecimal'>, <ufunc 'isnumeric'>, <ufunc 'endswith'>, <ufunc '_rstrip_chars'>, <ufunc '_lstrip_whitespace'>, <ufunc '_strip_chars'>, <ufunc 'find'>, <ufunc 'str_len'>, <ufunc 'isdigit'>, <ufunc 'count'>, <ufunc 'rfind'>, <ufunc 'isspace'>} == set()
E         Extra items in the left set:
E         <ufunc '_rstrip_whitespace'>
E         <ufunc '_lstrip_chars'>
E         <ufunc 'isnumeric'>
E         <ufunc 'endswith'>
E         <ufunc '_lstrip_whitespace'>
E         <ufunc '_strip_chars'>
E         <ufunc 'isspace'>
E         <ufunc 'startswith'>
E         <ufunc '_strip_whitespace'>
E         <ufunc 'isdecimal'>
E         <ufunc '_rstrip_chars'>
E         <ufunc 'find'>
E         <ufunc 'str_len'>
E         <ufunc 'isdigit'>
E         <ufunc 'count'>
E         <ufunc 'rfind'>
E         Full diff:
E         - set(,
E         - )
E         + {
E         +  <ufunc '_lstrip_chars'>,
E         +  <ufunc '_lstrip_whitespace'>,
E         +  <ufunc '_rstrip_chars'>,
E         +  <ufunc '_rstrip_whitespace'>,
E         +  <ufunc '_strip_chars'>,
E         +  <ufunc '_strip_whitespace'>,
E         +  <ufunc 'count'>,
E         +  <ufunc 'endswith'>,
E         +  <ufunc 'find'>,
E         +  <ufunc 'isdecimal'>,
E         +  <ufunc 'isdigit'>,
E         +  <ufunc 'isnumeric'>,
E         +  <ufunc 'isspace'>,
E         +  <ufunc 'rfind'>,
E         +  <ufunc 'startswith'>,
E         +  <ufunc 'str_len'>,
E         + }

and with this branch

E       AssertionError: assert {<ufunc '_lstrip_chars'>, <ufunc '_rstrip_whitespace'>, <ufunc '_strip_whitespace'>, <ufunc '_strip_chars'>, <ufunc '_lstrip_whitespace'>, <ufunc '_rstrip_chars'>, <ufunc 'str_len'>, <ufunc 'count'>} == set()
E         Extra items in the left set:
E         <ufunc '_lstrip_chars'>
E         <ufunc '_rstrip_whitespace'>
E         <ufunc '_strip_whitespace'>
E         <ufunc '_strip_chars'>
E         <ufunc '_lstrip_whitespace'>
E         <ufunc '_rstrip_chars'>
E         <ufunc 'str_len'>
E         <ufunc 'count'>
E         Full diff:
E         - set(,
E         - )
E         + {
E         +  <ufunc '_lstrip_chars'>,
E         +  <ufunc '_lstrip_whitespace'>,
E         +  <ufunc '_rstrip_chars'>,
E         +  <ufunc '_rstrip_whitespace'>,
E         +  <ufunc '_strip_chars'>,
E         +  <ufunc '_strip_whitespace'>,
E         +  <ufunc 'count'>,
E         +  <ufunc 'str_len'>,
E         + }

but the test fails in all cases ! Do you know what's going on here ? (I rebuilt numpy dev from source right before I wrote this reply)

Yes, just reproduced that after I force-rebuilt numpy from git main to test your branch
(https://pypi.anaconda.org/scientific-python-nightly-wheels/simple gave me something like 2.0.0.dev0+git20231111.9340fca, maybe "nightly" refers to the lunar night...).
So those must also have been added since; count and str_len are also in np.char, but the _strip* ones (numpy/numpy#25152) sit in umath only – wondering if they should be explicitly added to UNSUPPORTED_UFUNCS as they are obviously private.

Numpy "nightly" isn't really published every Earth night.

but the _strip* ones (numpy/numpy#25152) sit in umath only – wondering if they should be explicitly added to UNSUPPORTED_UFUNCS as they are obviously private.

Well, they are the only private ones except one so far:

>>> [u for k, u in np._core.umath.__dict__.items() if isinstance(u, np.ufunc) and k.startswith('_')]
[<ufunc '_ones_like'>, <ufunc '_strip_whitespace'>, <ufunc '_lstrip_whitespace'>, <ufunc '_rstrip_whitespace'>, <ufunc '_strip_chars'>, <ufunc '_lstrip_chars'>, <ufunc '_rstrip_chars'>]

and _ones_like has a helper function.

Thanks for the reviews! Thought about waiting over the weekend for the _strip functions to get in, but since they are not tested anyway it does not matter.

@dhomeier , please remember milestone and backport label next time. Thanks!

@meeseeksdev backport to v6.0.x

Ow, specifically did not set one initially as I did not recall the backport rules for numpy-2 fixes, but then forgot to go through the checklist, sorry.