bug#23130: 25.1.50; `C-h f' etc.: escaping in symbol names

[Drew Adams <drew.adams@oracle.com>]

> >> > The name of the symbol is "foo.", not "foo\.".
> >>
> >> Both refer to the same symbol, as does "\foo.".
> >
> > Not as `symbol-name' values, they don't.
> 
> It's showing a symbol, not the result of calling symbol-name on it.

Yes, we know.

> > The point is to show the symbol _name_, and to do so in a
> > way that is clear to anyone.
> 
> The point is to show a symbol, unambigously.

The point of this bug report is that the symbol's name should be
shown clearly.  The symbol is already printed using prin1 in the
signature: (foo\.)

Show it as Lisp when showing Lisp.  Show it as a symbol name, as
is the Emacs convention, in help text.  And in that case, show it
enclosed in `...'.

> >> > Or better yet, to remove all ambiguity, even for characters
> >> > such as SPC in the function name, enclose the name in `...':
> >>
> >> The name as printed is already unambigous.
> >
> > "bar\ is a Lisp function." is not only unclear; it is incorrect.
> 
> Of course it's incorrect, but describe-function will never generate
> that.

Why don't you read the bug report and try the recipe for yourself,
starting with `emacs -Q'?  That's exactly what `describe-funcction'
"generates":

(defun bar\  () (message "BARRRR"))

`C-h f RET' with point on "bar\ ":

  bar\ is a Lisp function.

  (bar\ )

  Not documented.