bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

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

[-- Attachment #1.1: Type: text/plain, Size: 386 bytes --]

Please find attached a patch which fixes a minor issue with narrowing to
fields during comint input fontification. The issue didn't cause any
visible bugs that I know of, I only noticed it during experimentation
with tree-sitter integration.

Not that my copyright paperwork renewal is currently in progress, but I
hope that the number of lines in this patch fall below the threshold.


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1.2: 0001-comint-fontify-input-Fix-field-boundary-issue.patch --]
[-- Type: text/x-patch, Size: 1406 bytes --]

From 5e2effcaf8d4b1f1cc835fae910e8d85abb7be3c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Miha=20Rihtar=C5=A1i=C4=8D?= <miha@kamnitnik.top>
Date: Sat, 26 Nov 2022 23:01:24 +0100
Subject: [PATCH] comint-fontify-input: Fix field boundary issue

* lisp/comint.el (comint--intersect-regions): Don't call field-end if
we are on a field boundary already.
---
 lisp/comint.el | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/lisp/comint.el b/lisp/comint.el
index ddb3464891..f47e6089f2 100644
--- a/lisp/comint.el
+++ b/lisp/comint.el
@@ -4122,9 +4122,15 @@ comint--intersect-regions
         (save-restriction
           (let ((beg2 beg1)
                 (end2 end1))
-            (when (= beg2 beg)
+            (when (and (= beg2 beg)
+                       (> beg2 (point-min))
+                       (eq is-output
+                           (eq (get-text-property (1- beg2) 'field) 'output)))
               (setq beg2 (field-beginning beg2)))
-            (when (= end2 end)
+            (when (and (= end2 end)
+                       (< end2 (point-max))
+                       (eq is-output
+                           (eq (get-text-property (1+ end2) 'field) 'output)))
               (setq end2 (field-end end2)))
             ;; Narrow to the whole field surrounding the region
             (narrow-to-region beg2 end2))
-- 
2.38.1


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 861 bytes --]

bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

[Eli Zaretskii]

> Date: Sun, 27 Nov 2022 10:23:57 +0100
> From: miha--- via "Bug reports for GNU Emacs,
>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
> 
> Please find attached a patch which fixes a minor issue with narrowing to
> fields during comint input fontification. The issue didn't cause any
> visible bugs that I know of, I only noticed it during experimentation
> with tree-sitter integration.

Thanks.  Do you have a recipe for demonstrating the problem which you are
trying to fix?

> Not that my copyright paperwork renewal is currently in progress, but I
> hope that the number of lines in this patch fall below the threshold.

Since when your previous assignment became outdated?  I need to know that to
make sure the cumulative amount of changes we already received from you
doesn't exceed the limit.

bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

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

[-- Attachment #1: Type: text/plain, Size: 2554 bytes --]

Eli Zaretskii <eliz@gnu.org> writes:

>> Date: Sun, 27 Nov 2022 10:23:57 +0100
>> From: miha--- via "Bug reports for GNU Emacs,
>>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
>> 
>> Please find attached a patch which fixes a minor issue with narrowing to
>> fields during comint input fontification. The issue didn't cause any
>> visible bugs that I know of, I only noticed it during experimentation
>> with tree-sitter integration.
>
> Thanks.  Do you have a recipe for demonstrating the problem which you are
> trying to fix?

Instrument comint input fontification with this diff:

diff --git a/lisp/comint.el b/lisp/comint.el
index f47e6089f2..c83da95ad2 100644
--- a/lisp/comint.el
+++ b/lisp/comint.el
@@ -4083,7 +4083,10 @@ comint--fontify-input-fontify-region
            nil (lambda (beg end)
                  (unless (get-text-property
                           beg 'comint--fontify-input-inhibit-fontification)
-                   (font-lock-fontify-region beg end verbose)))
+                   (font-lock-fontify-region beg end verbose)
+                   ;; Output contents of the visible portion of the
+                   ;; buffer to *Messages*.
+                   (message "%S" (buffer-substring-no-properties (point-min) (point-max)))))
            beg end)))
     (`((jit-lock-bounds ,beg1 . ,_) . (jit-lock-bounds ,_ . ,end1))
      (setq beg (min beg beg1))

Now open up an M-x shell and type "date RET" into the buffer to make
your shell output the current date and an new prompt. Press "C-p C-p
C-e" to move point to "date" and change it into "datee".

Open up *Messages* and find the following strings:

"date"

"date
"

"datee
Sun Nov 27 01:05:27 PM CET 2022
~$ "

These strings represent the context in which shell fontification was
performed on the input. The last one includes "Sun Nov 27 01:05:27 PM
CET 2022\n~$ ", which is a mistake that this patch attempts to fix.
However, in pretty much all cases, this didn't result in any user
visible inaccuracies, since shell fontification of a command rarely
depends on the contents of the lines that follow it.

>> Not that my copyright paperwork renewal is currently in progress, but I
>> hope that the number of lines in this patch fall below the threshold.
>
> Since when your previous assignment became outdated?  I need to know that to
> make sure the cumulative amount of changes we already received from you
> doesn't exceed the limit.

24. of October this year.

Thanks, best regards.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 861 bytes --]

bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

[Eli Zaretskii]

> From: <miha@kamnitnik.top>
> Cc: 59626@debbugs.gnu.org
> Date: Sun, 27 Nov 2022 13:22:50 +0100
> 
> > Thanks.  Do you have a recipe for demonstrating the problem which you are
> > trying to fix?
> 
> Instrument comint input fontification with this diff:
> 
> diff --git a/lisp/comint.el b/lisp/comint.el
> index f47e6089f2..c83da95ad2 100644
> --- a/lisp/comint.el
> +++ b/lisp/comint.el
> @@ -4083,7 +4083,10 @@ comint--fontify-input-fontify-region
>             nil (lambda (beg end)
>                   (unless (get-text-property
>                            beg 'comint--fontify-input-inhibit-fontification)
> -                   (font-lock-fontify-region beg end verbose)))
> +                   (font-lock-fontify-region beg end verbose)
> +                   ;; Output contents of the visible portion of the
> +                   ;; buffer to *Messages*.
> +                   (message "%S" (buffer-substring-no-properties (point-min) (point-max)))))
>             beg end)))
>      (`((jit-lock-bounds ,beg1 . ,_) . (jit-lock-bounds ,_ . ,end1))
>       (setq beg (min beg beg1))
> 
> Now open up an M-x shell and type "date RET" into the buffer to make
> your shell output the current date and an new prompt. Press "C-p C-p
> C-e" to move point to "date" and change it into "datee".
> 
> Open up *Messages* and find the following strings:
> 
> "date"
> 
> "date
> "
> 
> "datee
> Sun Nov 27 01:05:27 PM CET 2022
> ~$ "
> 
> These strings represent the context in which shell fontification was
> performed on the input. The last one includes "Sun Nov 27 01:05:27 PM
> CET 2022\n~$ ", which is a mistake that this patch attempts to fix.
> However, in pretty much all cases, this didn't result in any user
> visible inaccuracies, since shell fontification of a command rarely
> depends on the contents of the lines that follow it.

Thanks, I will study this.

> >> Not that my copyright paperwork renewal is currently in progress, but I
> >> hope that the number of lines in this patch fall below the threshold.
> >
> > Since when your previous assignment became outdated?  I need to know that to
> > make sure the cumulative amount of changes we already received from you
> > doesn't exceed the limit.
> 
> 24. of October this year.

Then it's okay to install this without an assignment (when we decide to
install it).

bug#59626: 29.0.50; [PATCH] comint-fontify-input: Fix field boundary issue

[Eli Zaretskii]

> Date: Sun, 27 Nov 2022 10:23:57 +0100
> From: miha--- via "Bug reports for GNU Emacs,
>  the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
> 
> Please find attached a patch which fixes a minor issue with narrowing to
> fields during comint input fontification. The issue didn't cause any
> visible bugs that I know of, I only noticed it during experimentation
> with tree-sitter integration.
> 
> Not that my copyright paperwork renewal is currently in progress, but I
> hope that the number of lines in this patch fall below the threshold.

Thanks, installed on the master branch, and closing the bug.