From mboxrd@z Thu Jan 1 00:00:00 1970 Path: news.gmane.org!.POSTED!not-for-mail From: Philipp Stephani
> From: Philipp Stephani <p.stephani2@gmail.com>
> Date: Sat, 13 May 2017 16:44:08 +0200
> Cc: Philipp Stephani <phst@google.com>
>
> Instead of a lambda, create a new type containing all data required to=
> call the function, and support it in the evaluator.=C2=A0 Because this= type
> now also needs to store the function documentation, it is too big for<= br> > Lisp_Misc; use a pseudovector instead.=C2=A0 That also has the nice be= nefit
> that we don't have to add special support to the garbage collector= .
Thanks.
> +#ifdef HAVE_MODULES
> +=C2=A0 else if (MODULE_FUNCTIONP (fun))
> +=C2=A0 =C2=A0 doc =3D XMODULE_FUNCTION (fun)->documentation;
> +#endif
I wonder whether it would be cleaner to have MODULE_FUNCTIONP defined
even when HAVE_MODULES is not: it looks like it could save us quite a
few #ifdef's.
> -DEFUN ("internal--module-call", Finternal_module_call, Sint= ernal_module_call, 1, MANY, 0,
> -=C2=A0 =C2=A0 =C2=A0 =C2=A0doc: /* Internal function to call a module= function.
> -ENVOBJ is a save pointer to a module_fun_env structure.
> -ARGLIST is a list of arguments passed to SUBRPTR.
> -usage: (module-call ENVOBJ &rest ARGLIST)=C2=A0 =C2=A0*/)
> -=C2=A0 (ptrdiff_t nargs, Lisp_Object *arglist)
> +Lisp_Object
> +funcall_module (const struct Lisp_Module_Function *const envptr,
> +=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 ptrdiff_t nar= gs, Lisp_Object *arglist)
>=C2=A0 {
I don't think we can simply remove a function that was already present<= br> in one or more Emacs releases.=C2=A0 We need to provide a backward
compatibility layer, and definitely also mention this in NEWS.This is purely internal, and we can change implemen= tation details at will. I've added a NEWS entry, though.=C2=A0