bug#28254: 26.0.50; SRFI-2 and-let*

[Mark Oteiza]

On 31/08/17 at 10:55pm, npostavs@users.sourceforge.net wrote:
> Mark Oteiza <mvoteiza@udel.edu> writes:
>
> > its behavior overlaps with when-let* but each has things the other
> > lacks.
>
> Can the differences be minimized?  I think having a bunch of tiny
> differences is not good for reader comprehension or code reuse.

Understood.  I was leaning toward putting and-let* in its own file
because of the differences instead of having them all in the same
place.  I'd rather have no and-let* than a gimped and-let* (I shouldn't
have made the alias, in retrospect).

> How much would be lost if we implemented and-let* like this?
>
> (defmacro and-let* (varlist &rest body)
>  `(when-let* ,varlist
>     ,@(or body '(t))))

The (EXPR) part of it, e.g.

 (and-let* ((x ...)
            (y ...)
            ((x > y)))
   ...)

where the body gets executed if x > y.

I didn't try writing it in the style of if-let*--perhaps if-let* could
be extended and all three macros would learn (EXPR).  if-let* already
has the single tuple quirk which differs from most (if not all?) lets in
Elisp.