Regexps and hard newlines

Regexps and hard newlines

[Richard M. Stallman]

What do people think of the idea of making . match soft newlines?
That would fix problems where various features get confused
by longlines.el.

Re: Regexps and hard newlines

[Kevin Rodgers]

Richard M. Stallman wrote:
> What do people think of the idea of making . match soft newlines?
> That would fix problems where various features get confused
> by longlines.el.

Brilliant!

-- 
Kevin Rodgers

Re: Regexps and hard newlines

[David Kastrup]

"Richard M. Stallman" <rms@gnu.org> writes:

> What do people think of the idea of making . match soft newlines?
> That would fix problems where various features get confused
> by longlines.el.

I think that a soft newline, most particularly so in the context of
longlines.el, should pretty much be a space with a display property of
newline on it, so search and replace would not even see it as anything
but space.

If that's not doable, it should at the very least be in all the same
categories as a normal space when doing regexp (and possibly other)
matches.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Regexps and hard newlines

[Richard M. Stallman]

    I think that a soft newline, most particularly so in the context of
    longlines.el, should pretty much be a space with a display property of
    newline on it, so search and replace would not even see it as anything
    but space.

That could be too drastic--I don't think that commands that
normally operate on lines should ignore soft newlines.
C-a and C-e now have special commands to implement them,
but there is also C-k, C-x C-t, and probably many more.

The change I suggested, affecting only `,' in regexps, might also be
too drastic.  That's why I am asking to see if people can think of
problems it could cause.