bug#25482: Fwd: Re: bug#25482: 26.0.50; Allow setting `query-replace-from-to-separator` to nil

[Thierry Volpiatto <thierry.volpiatto@gmail.com>]

Thierry Volpiatto <thierry.volpiatto@gmail.com> writes:

> Eli Zaretskii <eliz@gnu.org> writes:
>
>> Sorry, I don't think I follow: you should be able to set this variable
>> to " -> ", a string, to get the old behavior and avoid non-ASCII
>> strings in your minibuffer-history.
>
> No, you don't have the same behavior, you endup with a non ascii string
> which goes to minibuffer-history, so the behavior is not as emacs-24.5
> if you set this variable to " -> ".
>
>> So why did you need the value of nil?
>
> Because when it is nil the string is not added to history, just like in
> 24.5.
>
>>> +    (when (stringp query-replace-from-to-separator)
>>> +      (setq query-replace-from-to-separator
>>> +            (propertize (if (char-displayable-p
>>> +                             (string-to-char query-replace-from-to-separator))
>>
>> This doesn't look right: string-to-char will return the blank
>> character, which is the first character of " → ", and I don't think
>> you meant that.
>
> Yes, you are right, thanks, however see how the actual code is wrong,
> what if someone set the separator to another unicode string.
>
>> In any case, this looks more complex than it has to be, because you
>> for some reason want to have the nil value.  Please elaborate on the
>> need for that.
>
> To retrieve the same behavior as in emacs-24.5, see above.


-- 
Thierry