bug#60464: 29.0.60; Regression - pcomplete-arg fails with argument 'last

[Gregory Heytings]

> > Stefan and Eli, does that patch look right to you?
>
> I don't consider myself an expert on pcomplete, sorry.  I will defer to 
> Jim and Stefan.  (Although I must say that I'm not sure I see a bug 
> here.)
>
> However, I do have two questions:
>
> . is this suggested for master or for the emacs-29 branch?
> . if the latter, then which past version of Emacs handled this correctly?

It's a sequel of dafa6d6bad, so it should probably go to emacs-29.  I 
can't really answer your second question with 100% accuracy, but I believe 
Emacs 28 behaved correctly in the cases discussed in this bug and the 
earlier ones (bug#59956 and bug#60021).