bug#35564: [PATCH v5] Tweak dired warning about "wildcard" characters

[Kévin Le Gouguec]

Eli Zaretskii <eliz@gnu.org> writes:

> I'll let other participants of this long discussion to chime in, but
> in general I'd like to postpone this till after the emacs-27 branch is
> cut (hopefully, very soon), as this constitutes a significant behavior
> change, AFAIU.

I won't insist too much for this to land on Emacs 27, since a) the "bug"
it fixes is fairly minor, b) I know everyone's plate is quite full, and
c) I mostly use the master branch anyway, so it's not like I'll have to
wait to benefit.


For the record though, I'll point out a few reasons why I think it
should be "safe" to include this in the upcoming release:

- the changes are fairly limited in scope: they only affect
  dired-do-shell-command;

- in the simplest case, the UI change is minor: it turns this message:

  > Confirm--do you mean to use ‘?’ as a wildcard? (y or n)

  into this one:

  > 1 occurrence of ‘?’ will not be substituted.  Proceed? (y, n, ?) ?

  (or, with read-multiple-choice:)

  > 1 occurrence of ‘?’ will not be substituted.  Proceed? (_y_es, _n_o, toggle 
_d_etails, _?_): 

- the "riskiest" refactoring changes have been handled by Noam[1] and
  are partially covered by unit tests.


Most of the lengthy discussion was about finding the right balance
between message correctness and verbosity; hopefully the eventual
behavior change is not that significant.


> Thanks.

Thank you for your time.


[1] 
https://debbugs.gnu.org/cgi/bugreport.cgi?att=2;filename=0002-Dedup-dired-aux-isolated-char-searching-Bug-35564.patch;bug=35564;msg=202