GNU bug report logs - #61640
30.0.50; Help Elisp xref recognize defclass parent classes

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 61640 in the body.
You can then email your comments to 61640 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: bug-gnu-emacs <at> gnu.org
Subject: 30.0.50; Help Elisp xref recognize defclass parent classes
Date: Sun, 19 Feb 2023 14:08:42 -0800

[Message part 1 (text/plain, inline)]

In a form such as this:

(defclass ebdb-field-id (ebdb-field-labeled ebdb-field-obfuscated ebdb-field-user)
  ((label-list :initform 'ebdb-id-label-list)
   ...
   ))

All the symbols in

(ebdb-field-labeled ebdb-field-obfuscated ebdb-field-user)

should be recognized by `elisp--xref-infer-namespace' as 'function.
Right now, this is treated like a function call (ie like '(function
variable variable)).

I can't promise that the attached patch is exactly the right thing, but
it ought to be pretty close.

Thanks,
Eric

[elispxrefdefclass.diff (text/x-patch, attachment)]

Message #8 received at 61640 <at> debbugs.gnu.org (full text, mbox):

From: Dmitry Gutov <dgutov <at> yandex.ru>
To: Eric Abrahamsen <eric <at> ericabrahamsen.net>, 61640 <at> debbugs.gnu.org
Subject: Re: bug#61640: 30.0.50; Help Elisp xref recognize defclass parent
 classes
Date: Mon, 20 Feb 2023 00:42:25 +0200

Hi!

On 20/02/2023 00:08, Eric Abrahamsen wrote:
> In a form such as this:
> 
> (defclass ebdb-field-id (ebdb-field-labeled ebdb-field-obfuscated ebdb-field-user)
>    ((label-list :initform 'ebdb-id-label-list)
>     ...
>     ))
> 
> All the symbols in
> 
> (ebdb-field-labeled ebdb-field-obfuscated ebdb-field-user)
> 
> should be recognized by `elisp--xref-infer-namespace' as 'function.
> Right now, this is treated like a function call (ie like '(function
> variable variable)).
> 
> I can't promise that the attached patch is exactly the right thing, but
> it ought to be pretty close.

Thanks for the patch. Could you try including a corresponding test?

There are examples in test/lisp/progmodes/elisp-mode-tests.el, see the 
ones using the helper xref-elisp-deftest, starting with line 395.

Message #11 received at 61640 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: Dmitry Gutov <dgutov <at> yandex.ru>
Cc: 61640 <at> debbugs.gnu.org
Subject: Re: bug#61640: 30.0.50; Help Elisp xref recognize defclass parent
 classes
Date: Sun, 19 Feb 2023 15:45:37 -0800

Dmitry Gutov <dgutov <at> yandex.ru> writes:

> Hi!
>
> On 20/02/2023 00:08, Eric Abrahamsen wrote:
>> In a form such as this:
>> (defclass ebdb-field-id (ebdb-field-labeled ebdb-field-obfuscated
>> ebdb-field-user)
>>    ((label-list :initform 'ebdb-id-label-list)
>>     ...
>>     ))
>> All the symbols in
>> (ebdb-field-labeled ebdb-field-obfuscated ebdb-field-user)
>> should be recognized by `elisp--xref-infer-namespace' as 'function.
>> Right now, this is treated like a function call (ie like '(function
>> variable variable)).
>> I can't promise that the attached patch is exactly the right thing,
>> but
>> it ought to be pretty close.
>
> Thanks for the patch. Could you try including a corresponding test?
>
> There are examples in test/lisp/progmodes/elisp-mode-tests.el, see the
> ones using the helper xref-elisp-deftest, starting with line 395.

There's a lot going on in there!

The tests using xref-elisp-deftest look like they're mostly aimed at
correctly finding definitions, which isn't the problem here: the
defclass is always found correctly *if* xref infers the namespace as
'function.

Shouldn't this instead be an addition to `elisp-mode-infer-namespace',
like with p8 below?

  (elisp-mode-test--with-buffer
      (concat "(list {p1}alpha {p2}beta)\n"
              "(progn {p3}gamma {p4}delta)\n"
              "(lambda ({p5}epsilon {p6}zeta) {p7}eta)\n"
              "(defclass child-class (parent-1 {p8}parent-2)\n")
    (should (equal (elisp--xref-infer-namespace p1) 'variable))
    (should (equal (elisp--xref-infer-namespace p2) 'variable))
    (should (equal (elisp--xref-infer-namespace p3) 'variable))
    (should (equal (elisp--xref-infer-namespace p4) 'variable))
    (should (equal (elisp--xref-infer-namespace p5) 'variable))
    (should (equal (elisp--xref-infer-namespace p6) 'variable))
    (should (equal (elisp--xref-infer-namespace p7) 'variable))
    (should (equal (elisp--xref-infer-namespace p8) 'function)))

Obviously just guessing here...

Message #16 received at 61640-done <at> debbugs.gnu.org (full text, mbox):

From: Dmitry Gutov <dgutov <at> yandex.ru>
To: Eric Abrahamsen <eric <at> ericabrahamsen.net>
Cc: 61640-done <at> debbugs.gnu.org
Subject: Re: bug#61640: 30.0.50; Help Elisp xref recognize defclass parent
 classes
Date: Mon, 20 Feb 2023 03:05:31 +0200

Version: 30.1

On 20/02/2023 01:45, Eric Abrahamsen wrote:
> Shouldn't this instead be an addition to `elisp-mode-infer-namespace',
> like with p8 below?
> 
>    (elisp-mode-test--with-buffer
>        (concat "(list {p1}alpha {p2}beta)\n"
>                "(progn {p3}gamma {p4}delta)\n"
>                "(lambda ({p5}epsilon {p6}zeta) {p7}eta)\n"
>                "(defclass child-class (parent-1 {p8}parent-2)\n")
>      (should (equal (elisp--xref-infer-namespace p1) 'variable))
>      (should (equal (elisp--xref-infer-namespace p2) 'variable))
>      (should (equal (elisp--xref-infer-namespace p3) 'variable))
>      (should (equal (elisp--xref-infer-namespace p4) 'variable))
>      (should (equal (elisp--xref-infer-namespace p5) 'variable))
>      (should (equal (elisp--xref-infer-namespace p6) 'variable))
>      (should (equal (elisp--xref-infer-namespace p7) 'variable))
>      (should (equal (elisp--xref-infer-namespace p8) 'function)))
> 
> Obviously just guessing here...

It seems you're right, thanks.

I've put the test case separately and pushed the combined change to 
master, commit cac13e36054.

That fixes the bug in my testing, but please do tell if something else 
comes up.

Message #19 received at 61640 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: 61640 <at> debbugs.gnu.org
Cc: dgutov <at> yandex.ru
Subject: Re: bug#61640: 30.0.50; Help Elisp xref recognize defclass parent
 classes
Date: Sun, 19 Feb 2023 17:43:48 -0800

Dmitry Gutov <dgutov <at> yandex.ru> writes:

> Version: 30.1
>
> On 20/02/2023 01:45, Eric Abrahamsen wrote:
>> Shouldn't this instead be an addition to `elisp-mode-infer-namespace',
>> like with p8 below?
>>    (elisp-mode-test--with-buffer
>>        (concat "(list {p1}alpha {p2}beta)\n"
>>                "(progn {p3}gamma {p4}delta)\n"
>>                "(lambda ({p5}epsilon {p6}zeta) {p7}eta)\n"
>>                "(defclass child-class (parent-1 {p8}parent-2)\n")
>>      (should (equal (elisp--xref-infer-namespace p1) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p2) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p3) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p4) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p5) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p6) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p7) 'variable))
>>      (should (equal (elisp--xref-infer-namespace p8) 'function)))
>> Obviously just guessing here...
>
> It seems you're right, thanks.
>
> I've put the test case separately and pushed the combined change to
> master, commit cac13e36054.
>
> That fixes the bug in my testing, but please do tell if something else
> comes up.

Thanks!

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.