ENH: Add strip/lstrip/rstrip ufuncs for unicode and bytes - #25152

Is there a way to do the equivalent of this with a promoter?

We would need to either dig deeper into the resolve_dtypes_raw (it's plausible) and/or figure out a way to make arguments truly optional.
Unfortunately, I am not sure I can condone adding such a hacky old-style resolver. The only reason I didn't start deleting them, is that I am not sure if we have consensus on going ahead with gh-25009 before releasing 2.0 (it might become necessary, although not the part with actually removing things).

I am not sure if we have consensus on going ahead

Can we discuss this at a community meeting?

Do you have any updates on this from the community meeting? Are there any action points for me related to this?

No, I brought it up briefly, but didn't have a clear plan of action. I would even prefer two ufuncs for now (as is done for np.clip, which is different but in the same boat to some degree) over adding this hack, TBH. (I guess strip_whitespace.)

Of course it would be cooler to implement a generic solution, but I can't help with a clear plan of action for that atm, unfortunately.

Maybe we should make a, hopefully brief, meeting with @ngoldbaum to see if we can think of a better approach.

Sorry I didn't respond to the ping last week. Happy to have a chat, although I'm off Thursday and Friday for thanksgiving.

Let's schedule something for early next week?

@seberg @ngoldbaum This is ready to go on my side.

To follow up on the earlier discussion: Sebastian, Lysandros, and I had a video call where we agreed this was the best approach. I'm planning to merge this once the tests finish. Thanks @lysnikolaou!