* bug#23947: Regression in EIEIO
@ 2016-07-11 17:27 Stefan Monnier
2016-07-11 17:42 ` Clément Pit--Claudel
0 siblings, 1 reply; 5+ messages in thread
From: Stefan Monnier @ 2016-07-11 17:27 UTC (permalink / raw)
To: 23947
Package: Emacs
Version: 25.0.95
I finally tracked down why GNU ELPA's "rudel" package wasn't working
with Emacs-25, and it turns out to be a combination of two bugs
in the new EIEIO code.
Rudel seems to be the heaviest user of EIEIO so far (e.g. the only user
of the :c3 method invocation order, since the eieio-core code had
a plain bug which made it signal an error whenever that was used).
The patch below lets Rudel setup a shared buffer, so at least the core
functionality now works. To my eyes, this is a very safe patch which
just fixes an incompatibility and an error I introduced in the new EIEIO
code when moving to cl-generic, so I think it should go to emacs-25.
Any objection?
Stefan
2016-07-11 Stefan Monnier <monnier@iro.umontreal.ca>
* lisp/emacs-lisp/eieio-compat.el (eieio--defmethod): Better obey
EIEIO's `no-applicable-method' calling convention.
* lisp/emacs-lisp/eieio-core.el (eieio--class-precedence-c3): `class'
is not a symbol but a class object.
diff --git a/lisp/emacs-lisp/eieio-compat.el b/lisp/emacs-lisp/eieio-compat.el
index 6d4798b..6aba8a3 100644
--- a/lisp/emacs-lisp/eieio-compat.el
+++ b/lisp/emacs-lisp/eieio-compat.el
@@ -188,7 +188,8 @@ eieio--defmethod
(`no-applicable-method
(setq method 'cl-no-applicable-method)
(setq specializers `(generic ,@specializers))
- (lambda (generic arg &rest args) (apply code arg generic args)))
+ (lambda (generic arg &rest args)
+ (apply code arg (cl--generic-name generic) (cons arg args))))
(_ code))))
(cl-generic-define-method
method (unless (memq kind '(nil :primary)) (list kind))
diff --git a/lisp/emacs-lisp/eieio-core.el b/lisp/emacs-lisp/eieio-core.el
index 631e4a4..223c2a6 100644
--- a/lisp/emacs-lisp/eieio-core.el
+++ b/lisp/emacs-lisp/eieio-core.el
@@ -976,7 +976,7 @@ eieio--class/struct-parents
(defun eieio--class-precedence-c3 (class)
"Return all parents of CLASS in c3 order."
- (let ((parents (eieio--class-parents (cl--find-class class))))
+ (let ((parents (eieio--class-parents class)))
(eieio--c3-merge-lists
(list class)
(append
@@ -1101,7 +1101,7 @@ eieio--generic-subclass-specializers
(list eieio--generic-subclass-generalizer))
\f
-;;;### (autoloads nil "eieio-compat" "eieio-compat.el" "6aca3c1b5f751a01331761da45fc4f5c")
+;;;### (autoloads nil "eieio-compat" "eieio-compat.el" "dba4205b1a0d7133f1311d975b4d0ebe")
;;; Generated autoloads from eieio-compat.el
(autoload 'eieio--defalias "eieio-compat" "\
^ permalink raw reply related [flat|nested] 5+ messages in thread
* bug#23947: Regression in EIEIO
2016-07-11 17:27 bug#23947: Regression in EIEIO Stefan Monnier
@ 2016-07-11 17:42 ` Clément Pit--Claudel
2016-07-12 13:06 ` Stefan Monnier
0 siblings, 1 reply; 5+ messages in thread
From: Clément Pit--Claudel @ 2016-07-11 17:42 UTC (permalink / raw)
To: 23947
[-- Attachment #1.1: Type: text/plain, Size: 222 bytes --]
On 2016-07-11 19:27, Stefan Monnier wrote:
> + (apply code arg (cl--generic-name generic) (cons arg args))))
Hmm, why does `code' receive `arg' twice? (once as the first arg and once as the third one)
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply [flat|nested] 5+ messages in thread
* bug#23947: Regression in EIEIO
2016-07-11 17:42 ` Clément Pit--Claudel
@ 2016-07-12 13:06 ` Stefan Monnier
2016-07-12 21:13 ` Clément Pit--Claudel
2016-07-14 19:09 ` Stefan Monnier
0 siblings, 2 replies; 5+ messages in thread
From: Stefan Monnier @ 2016-07-12 13:06 UTC (permalink / raw)
To: Clément Pit--Claudel; +Cc: 23947
>> + (apply code arg (cl--generic-name generic) (cons arg args))))
> Hmm, why does `code' receive `arg' twice? (once as the first arg and
> once as the third one)
Because that's how EIEIO's `no-applicable-method' was defined.
More specifically, it takes (obj method &rest args) where `obj' is the
first arg that was passed to the method and `args' is the complete list
of arguments that were passed to the method.
Stefan
^ permalink raw reply [flat|nested] 5+ messages in thread
* bug#23947: Regression in EIEIO
2016-07-12 13:06 ` Stefan Monnier
@ 2016-07-12 21:13 ` Clément Pit--Claudel
2016-07-14 19:09 ` Stefan Monnier
1 sibling, 0 replies; 5+ messages in thread
From: Clément Pit--Claudel @ 2016-07-12 21:13 UTC (permalink / raw)
To: Stefan Monnier; +Cc: 23947
[-- Attachment #1.1: Type: text/plain, Size: 523 bytes --]
On 07/12/2016 03:06 PM, Stefan Monnier wrote:
>>> + (apply code arg (cl--generic-name generic) (cons arg args))))
>
>> Hmm, why does `code' receive `arg' twice? (once as the first arg and
>> once as the third one)
>
> Because that's how EIEIO's `no-applicable-method' was defined.
> More specifically, it takes (obj method &rest args) where `obj' is the
> first arg that was passed to the method and `args' is the complete list
> of arguments that were passed to the method.
Thanks Stefan :)
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply [flat|nested] 5+ messages in thread
* bug#23947: Regression in EIEIO
2016-07-12 13:06 ` Stefan Monnier
2016-07-12 21:13 ` Clément Pit--Claudel
@ 2016-07-14 19:09 ` Stefan Monnier
1 sibling, 0 replies; 5+ messages in thread
From: Stefan Monnier @ 2016-07-14 19:09 UTC (permalink / raw)
To: 23947-done
Version: 25.1
Installed, thanks,
Stefan
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2016-07-14 19:09 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-07-11 17:27 bug#23947: Regression in EIEIO Stefan Monnier
2016-07-11 17:42 ` Clément Pit--Claudel
2016-07-12 13:06 ` Stefan Monnier
2016-07-12 21:13 ` Clément Pit--Claudel
2016-07-14 19:09 ` Stefan Monnier
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).