unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
* bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct
@ 2022-02-11  9:45 Ignacio Casso
  2022-02-12  6:55 ` Lars Ingebrigtsen
  0 siblings, 1 reply; 4+ messages in thread
From: Ignacio Casso @ 2022-02-11  9:45 UTC (permalink / raw)
  To: 53938

Hello,

I've noticed a possible bug in Emacs.

If you define a function that uses a cl struct before defining the cl
struct (e.g., because the struct is defined in a package to be autoloaded and the
function in your init file), there is an error when calling the function
later.

Here is a simple elisp file to reproduce it, starting Emacs with "emacs
-Q":

(require 'cl-macs)

(defun change-name (struct new-name)
  (setf (my-struct-name struct) new-name))

(cl-defstruct my-struct
  name)

(setq my-struct (make-my-struct))

(change-name my-struct "My struct")

This produces the following error:

Debugger entered--Lisp error: (void-function \(setf\ my-struct-name\))
  (\(setf\ my-struct-name\) new-name v)
  (let* ((v struct)) (\(setf\ my-struct-name\) new-name v))
  change-name(#s(my-struct :name nil) "asdf")
  eval((change-name my-struct "asdf") nil)
  elisp--eval-last-sexp(nil)
  eval-last-sexp(nil)
  funcall-interactively(eval-last-sexp nil)
  call-interactively(eval-last-sexp nil nil)
  command-execute(eval-last-sexp)

A more realistic example, which produces the same error:

(defun my-package-config ()
  (setf (package-struct-slot package-var) 'new-value))

(eval-after-load 'my-package '(my-package-config))

(cl-defstruct package-struct slot)

(defvar package-var (make-package-struct))

(provide 'my-package)

It does not work either if we use instead a form inside eval-after-load
instead of defining a function, as in this example:

(eval-after-load 'my-package '(setf (package-struct-slot package-var) 'new-value))

However, it does work if we do something like this:

(setq my-package-config '(setf (package-struct-slot package-var) 'new-value))

(eval-after-load 'my-package '(eval my-package-config))


I realize that this might be a design constraint given that this common
lisp feature is "emulated", but maybe it should be better documented or
produce more descriptive errors? It's a really uncommon case though, I
only encountered it trying to configure in a weird way mu4e which uses cl
structs.

Regards,

Ignacio


In GNU Emacs 27.2 (build 1, x86_64-pc-linux-gnu, GTK+ Version 3.24.20)
 of 2022-01-16 built on ignacio-IdeaPad-3-15ADA05
Windowing system distributor 'The X.Org Foundation', version 11.0.12013000
System Description: Ubuntu 20.04.3 LTS





^ permalink raw reply	[flat|nested] 4+ messages in thread

* bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct
  2022-02-11  9:45 bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct Ignacio Casso
@ 2022-02-12  6:55 ` Lars Ingebrigtsen
  2022-02-12 14:45   ` Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors
  0 siblings, 1 reply; 4+ messages in thread
From: Lars Ingebrigtsen @ 2022-02-12  6:55 UTC (permalink / raw)
  To: Ignacio Casso; +Cc: 53938, Stefan Monnier

Ignacio Casso <ignaciocasso@hotmail.com> writes:

> If you define a function that uses a cl struct before defining the cl
> struct (e.g., because the struct is defined in a package to be autoloaded and the
> function in your init file), there is an error when calling the function
> later.
>
> Here is a simple elisp file to reproduce it, starting Emacs with "emacs
> -Q":
>
> (require 'cl-macs)
>
> (defun change-name (struct new-name)
>   (setf (my-struct-name struct) new-name))
>
> (cl-defstruct my-struct
>   name)
>
> (setq my-struct (make-my-struct))
>
> (change-name my-struct "My struct")
>
> This produces the following error:
>
> Debugger entered--Lisp error: (void-function \(setf\ my-struct-name\))
>   (\(setf\ my-struct-name\) new-name v)

If you'd byte-compiled this, then the compiler would say

In end of data:
cl.el: Warning: the function ‘(setf my-struct-name)’ is not known to be
    defined.

So I think that Emacs is probably working as designed here.  But perhaps
Stefan has some comments; added to the CCs.

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





^ permalink raw reply	[flat|nested] 4+ messages in thread

* bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct
  2022-02-12  6:55 ` Lars Ingebrigtsen
@ 2022-02-12 14:45   ` Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors
  2022-02-13  8:22     ` Lars Ingebrigtsen
  0 siblings, 1 reply; 4+ messages in thread
From: Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors @ 2022-02-12 14:45 UTC (permalink / raw)
  To: Lars Ingebrigtsen; +Cc: Ignacio Casso, 53938

Lars Ingebrigtsen [2022-02-12 07:55:46] wrote:
> Ignacio Casso <ignaciocasso@hotmail.com> writes:
>> If you define a function that uses a cl struct before defining the cl
>> struct (e.g., because the struct is defined in a package to be autoloaded and the
>> function in your init file), there is an error when calling the function
>> later.
>>
>> Here is a simple elisp file to reproduce it, starting Emacs with "emacs
>> -Q":
>>
>> (require 'cl-macs)

Please require `cl-lib` rather than `cl-macs` here.

>> (defun change-name (struct new-name)
>>   (setf (my-struct-name struct) new-name))
>>
>> (cl-defstruct my-struct
>>   name)
>>
>> (setq my-struct (make-my-struct))
>>
>> (change-name my-struct "My struct")
>>
>> This produces the following error:
>>
>> Debugger entered--Lisp error: (void-function \(setf\ my-struct-name\))
>>   (\(setf\ my-struct-name\) new-name v)
>
> If you'd byte-compiled this, then the compiler would say
>
> In end of data:
> cl.el: Warning: the function ‘(setf my-struct-name)’ is not known to be
>     defined.
>
> So I think that Emacs is probably working as designed here.  But perhaps
> Stefan has some comments; added to the CCs.

Yes, it's working as designed (because the accessors are supposed to be
inlined).  OTOH, it should work like you want it if you pass
the :noinline option to `cl-defstruct`.


        Stefan






^ permalink raw reply	[flat|nested] 4+ messages in thread

* bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct
  2022-02-12 14:45   ` Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors
@ 2022-02-13  8:22     ` Lars Ingebrigtsen
  0 siblings, 0 replies; 4+ messages in thread
From: Lars Ingebrigtsen @ 2022-02-13  8:22 UTC (permalink / raw)
  To: Stefan Monnier; +Cc: Ignacio Casso, 53938

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

> Yes, it's working as designed (because the accessors are supposed to be
> inlined).  OTOH, it should work like you want it if you pass
> the :noinline option to `cl-defstruct`.

Thanks; then I guess there's nothing to be fixed here, and I'm closing
this bug report.

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





^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2022-02-13  8:22 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-02-11  9:45 bug#53938: 27.2; Defining function that modifies cl-struct before defining cl-struct Ignacio Casso
2022-02-12  6:55 ` Lars Ingebrigtsen
2022-02-12 14:45   ` Stefan Monnier via Bug reports for GNU Emacs, the Swiss army knife of text editors
2022-02-13  8:22     ` Lars Ingebrigtsen

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).