bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

[Andrea Corallo via Bug reports for GNU Emacs, the Swiss army knife of text editors]

[-- Attachment #1: Type: text/plain, Size: 363 bytes --]

Hi,

debugging a native-comp regression on the branch I'm working on I came
to this.

In `python-indent-region' we are comparing raw syntax descriptors using
`eq' but I think we should use `equal' (as we do underneath in
`python-fill-string').

I'm quite convinced the fix is correct but; should this go also in 27
where the code is the same?

Regards

  Andrea


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Compare-raw-syntax-descriptors-with-equal-in-python-.patch --]
[-- Type: text/x-diff, Size: 1061 bytes --]

From 9163724fcd8eea5c9e2908ecb38d90406669a8d3 Mon Sep 17 00:00:00 2001
From: Andrea Corallo <akrl@sdf.org>
Date: Sat, 19 Dec 2020 23:51:36 +0100
Subject: [PATCH] * Compare raw syntax descriptors with equal in
 `python-indent-region'.

	* lisp/progmodes/python.el (python-indent-region): Compare raw
	syntax descriptors with equal.
---
 lisp/progmodes/python.el | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lisp/progmodes/python.el b/lisp/progmodes/python.el
index d75944a702..dc2efdfcf7 100644
--- a/lisp/progmodes/python.el
+++ b/lisp/progmodes/python.el
@@ -1259,7 +1259,7 @@ Called from a program, START and END specify the region to indent."
                    ;; Don't mess with strings, unless it's the
                    ;; enclosing set of quotes or a docstring.
                    (or (not (python-syntax-context 'string))
-                       (eq
+                       (equal
                         (syntax-after
                          (+ (1- (point))
                             (current-indentation)
-- 
2.20.1

bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

[Stefan Kangas]

Andrea Corallo <akrl@sdf.org> writes:

> In `python-indent-region' we are comparing raw syntax descriptors using
> `eq' but I think we should use `equal' (as we do underneath in
> `python-fill-string').
>
> I'm quite convinced the fix is correct but; should this go also in 27
> where the code is the same?

That was in December, and there were no followups.

We are one week away from Emacs 27.2, so I don't think we want this on
emacs-27 right now.  If you think the fix is correct, I don't see why we
couldn't push it there after Emacs 27.2 is released.

Should we perhaps start by getting the fix pushed to master?

>>From 9163724fcd8eea5c9e2908ecb38d90406669a8d3 Mon Sep 17 00:00:00 2001
> From: Andrea Corallo <akrl@sdf.org>
> Date: Sat, 19 Dec 2020 23:51:36 +0100
> Subject: [PATCH] * Compare raw syntax descriptors with equal in
>  `python-indent-region'.
>
> 	* lisp/progmodes/python.el (python-indent-region): Compare raw
> 	syntax descriptors with equal.
> ---
>  lisp/progmodes/python.el | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/lisp/progmodes/python.el b/lisp/progmodes/python.el
> index d75944a702..dc2efdfcf7 100644
> --- a/lisp/progmodes/python.el
> +++ b/lisp/progmodes/python.el
> @@ -1259,7 +1259,7 @@ Called from a program, START and END specify the region to indent."
>                     ;; Don't mess with strings, unless it's the
>                     ;; enclosing set of quotes or a docstring.
>                     (or (not (python-syntax-context 'string))
> -                       (eq
> +                       (equal
>                          (syntax-after
>                           (+ (1- (point))
>                              (current-indentation)

bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

[Andrea Corallo via Bug reports for GNU Emacs, the Swiss army knife of text editors]

Stefan Kangas <stefan@marxist.se> writes:

> Andrea Corallo <akrl@sdf.org> writes:
>
>> In `python-indent-region' we are comparing raw syntax descriptors using
>> `eq' but I think we should use `equal' (as we do underneath in
>> `python-fill-string').
>>
>> I'm quite convinced the fix is correct but; should this go also in 27
>> where the code is the same?
>
> That was in December, and there were no followups.
>
> We are one week away from Emacs 27.2, so I don't think we want this on
> emacs-27 right now.  If you think the fix is correct, I don't see why we
> couldn't push it there after Emacs 27.2 is released.
>
> Should we perhaps start by getting the fix pushed to master?

Uhmm, I thought the patch was correct but re-testing it this is now
breaking the `python-indent-region-5' test.

Maybe there's a reason why we should use eq there or maybe this is
unveiling a real bug?  Not sure, I guess some expert of that code should
probably have a look.

  Andrea

bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

[Lars Ingebrigtsen]

Andrea Corallo <akrl@sdf.org> writes:

> Uhmm, I thought the patch was correct but re-testing it this is now
> breaking the `python-indent-region-5' test.
>
> Maybe there's a reason why we should use eq there or maybe this is
> unveiling a real bug?  Not sure, I guess some expert of that code should
> probably have a look.

The code in question is:

                   ;; Don't mess with strings, unless it's the
                   ;; enclosing set of quotes or a docstring.
                   (or (not (python-syntax-context 'string))
                       (eq
                        (syntax-after
                         (+ (1- (point))
                            (current-indentation)
                            (python-syntax-count-quotes (char-after) (point))))
                        (string-to-syntax "|"))
                       (python-info-docstring-p))

That `eq' form will never return anything other than nil, so removing it
gives us identical results.  So I'm not quite sure what it's trying to
do -- it's probably meant to be `equal', but that will (as you've found
out) lead to mis-indenting some code.

So I've added Fabián to the CCs; perhaps he has some comments.

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

bug#45328: 28.0.50; [PATCH] Compare raw syntax descriptors with equal in `python-indent-region'

[Lars Ingebrigtsen]

Andrea Corallo <akrl@sdf.org> writes:

> Uhmm, I thought the patch was correct but re-testing it this is now
> breaking the `python-indent-region-5' test.
>
> Maybe there's a reason why we should use eq there or maybe this is
> unveiling a real bug?  Not sure, I guess some expert of that code should
> probably have a look.

I tried applying Andrea's patch to Emacs 29, and it didn't lead to any
test failures that I can see (and the patch looks "obviously correct"),
so I've now pushed it to Emacs 29.

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