bug#6227: Color isearch regexp submatches differently

[Eli Zaretskii]

> From: Lars Ingebrigtsen <larsi@gnus.org>
> Cc: juri@jurta.org,  lennart.borgman@gmail.com,  6227@debbugs.gnu.org
> Date: Sun, 20 Sep 2020 11:41:10 +0200
> 
> Eli Zaretskii <eliz@gnu.org> writes:
> 
> >> Anybody else got any comments on this?
> >
> > Comment #1: Do we really want to turn this feature on by default?
> 
> I don't see why not.  It's just a more detailed visualisation.

But with slightly different colors.  How do we know long-time users of
Isearch will understand what they mean, unless they deliberately turn
on this option?

> > Comment #2: This seems to effectively disable the feature on displays
> > that have fewer than 88 colors.  Is that intentional?  If so, why
> > doesn't the documentation say so?
> 
> I'm guessing it's just cargo-culting off of the isearch face:
> 
> (defface isearch
>   '((((class color) (min-colors 88) (background light))

I don't understand: the 'isearch' face has definitions for all kinds
of displays, even for monochrome ones:

  (defface isearch
    '((((class color) (min-colors 88) (background light))
       ;; The background must not be too dark, for that means
       ;; the character is hard to see when the cursor is there.
       (:background "magenta3" :foreground "lightskyblue1"))
      (((class color) (min-colors 88) (background dark))
       (:background "palevioletred2" :foreground "brown4"))
      (((class color) (min-colors 16))
       (:background "magenta4" :foreground "cyan1"))
      (((class color) (min-colors 8))
       (:background "magenta4" :foreground "cyan1"))
      (t (:inverse-video t)))