bug#20871: 25.0.50; fill-single-char-nobreak-p does not recognize a single-letter word when it is preceded by an open paren

[Eli Zaretskii]

> From: Michał Nazarewicz <mina86@mina86.com>
> Date: Fri, 16 Aug 2019 16:30:55 +0100
> Cc: Mattias Engdegård <mattiase@acm.org>, mbork@mbork.pl, 
>       20871@debbugs.gnu.org
> 
> On Fri, 16 Aug 2019 at 13:53, Eli Zaretskii <eliz@gnu.org> wrote:
> > Michal says that the original fill-single-char-nobreak-p function
> > was supposed to handle the Polish case as well.  If that's true,
> > perhaps we should deprecate fill-polish-nobreak-p.
> 
> I’m in favour of deprecating one of the functions, yes.

OK, but see below.

> > However, I do wonder how can a general function handle these cases,
> > since in general there's no problem in breaking a line after
> > single-letter words.
> 
> We could try and look at `tildify-mode’ for inspiration and especially
> ‘tildify-space-pattern’¹ which is:
> 
>     "[,:;(][ \t]*[a]\\|\\<[AIKOSUVWZikosuvwz]"

My problem is conceptual rather than practical.  Since in , e.g.,
English it is okay to break a line after single-letter words, whereas
in Polish it is not, I wonder how can we have a single function
satisfy both requirements.  We would need some option, and then we
would need to decide what is the trigger for changing the value of
that option -- it could be the user, or the language environment, or
maybe something else.

tildify.el explicitly says that its defaults are for a specific
language, so I don't think it solves the problem that bothers me, as
described above.  This is why I originally suggested a separate
function -- having that is equivalent to having an option which
determines a behavior that depends on the language.

I'm also okay with extending tildify.el to support more than just
Czech rules, but that's a separate issue.