Re: [Emacs-diffs] master ed29d9f: Add edebug specs for inline.el

Re: [Emacs-diffs] master ed29d9f: Add edebug specs for inline.el

[Stefan Monnier]

>     Add edebug specs for inline.el
>
>     * lisp/emacs-lisp/inline.el (inline-quote)
>     (inline-letevals): Add edebug specs (bug#31051).

I don't understand.  Those macros already have edebug specs (in their
`declare` form).

    (defmacro inline-quote (_exp)
      "Similar to backquote, but quotes code and only accepts , and not ,@."
      (declare (debug t))
      (error "inline-quote can only be used within define-inline"))

    (defmacro inline--leteval (_var-exp &rest _body)
      (declare (indent 1) (debug (sexp &rest body)))
      ;; BEWARE: if we're here it's presumably via macro-expansion of
      ;; inline-letevals, so signal the error in terms of the user's code.
      (error "inline-letevals can only be used within define-inline"))

I can't imagine how adding def-edebug-spec before will make any difference.


        Stefan

Re: [Emacs-diffs] master ed29d9f: Add edebug specs for inline.el

[Lars Ingebrigtsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> I don't understand.  Those macros already have edebug specs (in their
> `declare` form).
>
>     (defmacro inline-quote (_exp)
>       "Similar to backquote, but quotes code and only accepts , and not ,@."
>       (declare (debug t))

Oh, is `debug' for edebug, too?  I assumed it was...  something else...

> I can't imagine how adding def-edebug-spec before will make any difference.

Then me neither, but without the specs the test case fails and with it,
it doesn't fail.  Hm...  Any ideas?  Here's the "emacs -Q" test case:

To reproduce, enter the following code into *scratch*:

;;;  -*- lexical-binding: t -*-

(defvar my-count 0)
(defun my-number ()
  (setq my-count (1+ my-count))
  (message "my-count is %s" my-count)
  my-count)

(define-inline my-divisible-p (n)
  (inline-letevals (n)
    (inline-quote (or (eq 0 (% ,n 2)) (eq 0 (% ,n 3))))))

(defun my-func ()
  (message "the result is %s" (my-divisible-p (my-number))))

(my-func)

Then:

M-x edebug-all-defs RET
M-x eval-buffer RET
g g

Result: Args out of range: [20 48 59 70 71 72], 7


-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: [Emacs-diffs] master ed29d9f: Add edebug specs for inline.el

[Stefan Monnier]

> Oh, is `debug' for edebug,

Yes.

> too?

No: AFAIK it's only used by Edebug so far.

>> I can't imagine how adding def-edebug-spec before will make any difference.
>
> Then me neither, but without the specs the test case fails and with it,
> it doesn't fail.  Hm...  Any ideas?  Here's the "emacs -Q" test case:
>
> To reproduce, enter the following code into *scratch*:

I still get the error with the code in the trunk.


        Stefan

Re: [Emacs-diffs] master ed29d9f: Add edebug specs for inline.el

[Lars Ingebrigtsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> Then me neither, but without the specs the test case fails and with it,
>> it doesn't fail.  Hm...  Any ideas?  Here's the "emacs -Q" test case:
>>
>> To reproduce, enter the following code into *scratch*:
>
> I still get the error with the code in the trunk.

Hm, sounds like ... evaluation order or something is playing a role
here?  I'll revert the patch and reopen the bug.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no