Re: [PATCH 2/3] lisp/progmodes/etags.el don't (forward-char) as it's overriden next line

[Konstantin Kharlamov]

В Вс, мар 17, 2019 at 12:12 ДП (AM), Konstantin Kharlamov 
<address@hidden> написал:
> В Сб, мар 16, 2019 at 7:26 ПП (PM), Eli Zaretskii 
> <address@hidden> написал:
> > >  Date: Sat, 16 Mar 2019 18:42:44 +0300
> > >  From: Konstantin Kharlamov <address@hidden>
> > >  Cc: address@hidden
> > >
> > >  >>  -    (and (eq selective-display t)
> > >  >>  -     (looking-at "\^m")
> > >  >>  -     (forward-char 1))
> > >  >>       (beginning-of-line)))
> > >  >
> > >  > Did you actually try this change with selective-display in 
> > > effect?
> > >  > Because I'm not sure it is correct in that case.
> > >  >
> > >  > Thanks.
> > >
> > >  Hmm… I will try a bit later, but the code seems really
> > >  straightforward:
> > >         1. if current character is ^M, then step forward
> > >         2. step to the beginnig of the line
> > >
> > >  The 2 undoes 1.
> >
> > ^M has special meaning in selective-display mode.
>
> Thanks, so, I just tried playing with the mode, and I haven't even 
> found ^M anywhere. I wonder if the ^M being added by the mode is in 
> the past, that just no longer happening.
>
> Specifically, I did:
>
> 1. wrote in scratch buffer
> test
> test
>  test
> 2. evaluated (set-selective-display 1) which transformed the look of 
> the text to
> test...
> 3.α Tried searching for ^M by pressing C-s C-q C-m. It's failing to 
> find anything.
> 3.β Put the caret before the triple dot, as in "test|...", and 
> evaluated (char-after (point)) and (char-after (+ 1 (point))). The 
> first returns "C-j" character, i.e. a newline; the second returns 
> "space character". They're both the exact characters that actually 
> are in text, as if I didn't enable selective-display.
>
> So, even disregarding my patches (because I haven't managed to test 
> this possible corner-case), the need for ^M workarounds is just no 
> longer needed.

That said, would be nice if somebody could confirm my findings, because 
I can not find whether my current Emacs built from vanilla branch or 
the harf-buzz one; and I'm a bit worrying that this could've been 
change specific to harf-buzz branch.