Re: Lisp primitives and their calling of the change hooks

[Alan Mackenzie]

Hello, Clément.

On Wed, Jan 10, 2018 at 17:06:57 -0500, Clément Pit-Claudel wrote:
> On 2018-01-10 14:48, Alan Mackenzie wrote:
> > At the moment, non-balancing b/a-c-f are used only in primitives which
> > can't change the length of the buffer - things like
> > translate-region-internal and upcase-region.

> Sorry if I missed previous parts of the discussion that make the
> following irrelevant: in a buffer containing "Straße" a call to
> (buffer-size) returns 6, and a call to (buffer-size) after C-x h M-x
> upcase-region returns 7 (because the text becomes STRASSE).  Does this
> contradict your assertion above? Does it matter?

Yes, it does contradict my assertion, and it doesn't matter much, except
to me, because it makes my arguments weaker. ;-(  But I ought to have
thought of that myself, seeing how often and how far I walk along the
things in daily life.

> (Additionally, and likely unrelatedly, the region boundaries after
> running upcase-region on that example seem wrong — they don't include
> the 'E').

That sounds like a bug indeed.  But it's too late now for me to look at
it at the moment.

> Clément.

-- 
Alan Mackenzie (Nuremberg, Germany).