On 2/23/2023 11:34 AM, Stefan Monnier via Bug reports for GNU Emacs, the 
Swiss army knife of text editors wrote:
> The way I see it, either we still have an inf-loop or your
> (pcomplete-match "\\`--.*=" 0) condition is redundant.
> 
>  From what you say the inf-loop should only manifest if we match neither
> 
>      (pcomplete-match "^--\\([^= \t\n\f]*\\)\\'" 0)
> 
> nor
> 
>      (pcomplete-match "\\`--.*=" 0)
> 
> and indeed:
> 
>      tar --pot\ <TAB>
> 
> seems to hang.

Ah, good point. Instead of '(pcomplete-match "\\`--.*=" 0)', I think we 
want to handle all args starting with "--", but not matching "^--\\([^= 
\t\n\f]*\\)\\'". The goal is to call 'pcomplete-here*' exactly once per 
option.

I did this a slightly different way in the updated patch by moving all 
the "--" cases into a single place (likewise with the "-" options). I 
think this is easier to follow than before, and makes it more obvious 
that we're calling 'pcomplete-here*' the right number of times.