Add org-src-is-fontify-buffer-p (patch attached)

Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1.1: Type: text/plain, Size: 952 bytes --]

Greetings, org-mode maintainers,

I've become an occasional contributor over on the Emacs side. I thought I'd
give an org-mode contribution a try.

The attached patch came out of the discussion here
https://www.reddit.com/r/emacs/comments/1i3mk6m/disable_eglot_in_orgmode_source_blocks/

I've added the predicate function org-src-is-fontify-buffer-p which can be
used in a prog-mode hook to avoid resource-intensive features such as eglot
inside a fontification buffer. This short example should make it clear.

(defun my/emacs-lisp-hook ()
  (unless (and (featurep 'org) (org-src-is-fontify-buffer-p))
    (eglot-ensure)))

I made a similar patch, submitted via a github pull request, to the
markdown-mode maintainers for the same functionality in their fontification
buffers.

The patch was tested against today's main. I hope you find this useful. Let
me know your feedback.

Thanks for org-mode,

-Stephane Marks

P.S. I am on record with the FSF.

[-- Attachment #1.2: Type: text/html, Size: 2447 bytes --]

[-- Attachment #2: 0001-Add-predicate-function-org-src-is-fontify-buffer-p.patch --]
[-- Type: application/octet-stream, Size: 2717 bytes --]

From 745f9b6720d099c714ce92f8f82f43f66711df4a Mon Sep 17 00:00:00 2001
From: shipmints <shipmints@gmail.com>
Date: Fri, 17 Jan 2025 19:32:17 -0500
Subject: [PATCH] Add predicate function org-src-is-fontify-buffer-p

This can be used in a prog-mode hook to avoid resource-intensive
features such as eglot inside a fontification buffer. A short example
can be found in the function's docstring and in the corresponding NEWS
entry.
---
 etc/ORG-NEWS    |  6 ++++++
 lisp/org-src.el | 20 ++++++++++++++++++++
 2 files changed, 26 insertions(+)

diff --git a/etc/ORG-NEWS b/etc/ORG-NEWS
index adc48f304..0911c3bc9 100644
--- a/etc/ORG-NEWS
+++ b/etc/ORG-NEWS
@@ -200,6 +200,12 @@ take the date as an argument, and generate a list of pairs for
 ~org-datetree-find-create-hierarchy~.  This allows for creating new
 types of datetrees (e.g. for lunar calendars, academic calendars,
 retail 4-4-5 calendars, etc).
+*** Predicate test for fontify buffer
+
+The predicate function ~org-src-is-fontify-buffer-p~ can be used in a
+`prog-mode' hook to avoid resource-intensive features such as `eglot'
+inside a fontification buffer. A short example can be found in the
+function's docstring.
 
 ** New and changed options
 
diff --git a/lisp/org-src.el b/lisp/org-src.el
index 74343bde1..6459fb935 100644
--- a/lisp/org-src.el
+++ b/lisp/org-src.el
@@ -660,6 +660,25 @@ Leave point in edit buffer."
 \f
 ;;; Fontification of source blocks
 
+(defvar-local org-src--is-fontify-buffer nil)
+(put 'org-src--is-fontify-buffer 'permanent-local t) ; needs to survive major-mode housecleaning
+
+;;;###autoload
+(defun org-src-is-fontify-buffer-p (&optional buffer)
+  "Return t if the current buffer is a source block fontify BUFFER.
+
+If BUFFER is nil, the current buffer is used.
+
+This is useful in a `prog-mode' hook to avoid resource-intensive
+features such as `eglot' inside a fontification buffer.
+
+Example:
+  (unless (and (featurep \\='org)
+               (org-src-is-fontify-buffer-p))
+    (eglot-ensure))"
+  (buffer-local-value 'org-src--is-fontify-buffer
+                      (or buffer (current-buffer))))
+
 (defvar org-src-fontify-natively) ; Defined in org.el
 (defun org-src-font-lock-fontify-block (lang start end)
   "Fontify code block between START and END using LANG's syntax.
@@ -678,6 +697,7 @@ as `org-src-fontify-natively' is non-nil."
 	      (erase-buffer)
 	      ;; Add string and a final space to ensure property change.
 	      (insert string " "))
+            (setq org-src--is-fontify-buffer t) ; let the mode know this is a fontify buffer
 	    (unless (eq major-mode lang-mode) (funcall lang-mode))
             (setq native-tab-width tab-width)
             (font-lock-ensure)
-- 
2.47.1

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Ship Mints <shipmints@gmail.com> writes:

> I've added the predicate function org-src-is-fontify-buffer-p which can be
> used in a prog-mode hook to avoid resource-intensive features such as eglot
> inside a fontification buffer. This short example should make it clear.
>
> (defun my/emacs-lisp-hook ()
>   (unless (and (featurep 'org) (org-src-is-fontify-buffer-p))
>     (eglot-ensure)))

Thanks!

Do you have other examples where the new predicate could be useful apart
from eglot? For eglot specifically, one can do a much simpler test -
check for `buffer-file-name'.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 1541 bytes --]

Eglot was the first request cited. For python programmers that rely on per
project/directory-local or per-buffer virtual environment set up, this can
be heavier than needed just for fontification. I enable eglot manually, but
I do use python virtual environments and would disable initialization in
fontification buffers.

I'd prefer a more generic interface than propagating org internals like the
format of the fortification buffer name. It's what I submitted to the
markdown repo also. The buffer name scheme can change in the future without
impacting this feature.

-Stephane

On Sat, Jan 18, 2025 at 2:27 AM Ihor Radchenko <yantar92@posteo.net> wrote:

> Ship Mints <shipmints@gmail.com> writes:
>
> > I've added the predicate function org-src-is-fontify-buffer-p which can
> be
> > used in a prog-mode hook to avoid resource-intensive features such as
> eglot
> > inside a fontification buffer. This short example should make it clear.
> >
> > (defun my/emacs-lisp-hook ()
> >   (unless (and (featurep 'org) (org-src-is-fontify-buffer-p))
> >     (eglot-ensure)))
>
> Thanks!
>
> Do you have other examples where the new predicate could be useful apart
> from eglot? For eglot specifically, one can do a much simpler test -
> check for `buffer-file-name'.
>
> --
> Ihor Radchenko // yantar92,
> Org mode maintainer,
> Learn more about Org mode at <https://orgmode.org/>.
> Support Org development at <https://liberapay.com/org-mode>,
> or support my work at <https://liberapay.com/yantar92>
>

[-- Attachment #2: Type: text/html, Size: 2540 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Ship Mints <shipmints@gmail.com> writes:

> Eglot was the first request cited. For python programmers that rely on per
> project/directory-local or per-buffer virtual environment set up, this can
> be heavier than needed just for fontification. I enable eglot manually, but
> I do use python virtual environments and would disable initialization in
> fontification buffers.

Ok. Fair enough.

> I'd prefer a more generic interface than propagating org internals like the
> format of the fortification buffer name. It's what I submitted to the
> markdown repo also. The buffer name scheme can change in the future without
> impacting this feature.

From my point of view, the interface you suggest is not generic enough.
It only addresses a single use case - Org fontification.

What about another, more general approach - check if buffer name starts
with space? By convention, buffers with names starting with space are
considered internal.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 1575 bytes --]

Yes, could do, and nothing needed by org-mode or markdown-mode. Could also
test for buffer-file-name nil. Or test both. Let's recommend that simpler
approach and I will rescind my markdown PR in favor of this simpler
suggestion.

On Sat, Jan 18, 2025 at 6:18 AM Ihor Radchenko <yantar92@posteo.net> wrote:

> Ship Mints <shipmints@gmail.com> writes:
>
> > Eglot was the first request cited. For python programmers that rely on
> per
> > project/directory-local or per-buffer virtual environment set up, this
> can
> > be heavier than needed just for fontification. I enable eglot manually,
> but
> > I do use python virtual environments and would disable initialization in
> > fontification buffers.
>
> Ok. Fair enough.
>
> > I'd prefer a more generic interface than propagating org internals like
> the
> > format of the fortification buffer name. It's what I submitted to the
> > markdown repo also. The buffer name scheme can change in the future
> without
> > impacting this feature.
>
> From my point of view, the interface you suggest is not generic enough.
> It only addresses a single use case - Org fontification.
>
> What about another, more general approach - check if buffer name starts
> with space? By convention, buffers with names starting with space are
> considered internal.
>
> --
> Ihor Radchenko // yantar92,
> Org mode maintainer,
> Learn more about Org mode at <https://orgmode.org/>.
> Support Org development at <https://liberapay.com/org-mode>,
> or support my work at <https://liberapay.com/yantar92>
>

[-- Attachment #2: Type: text/html, Size: 2322 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Ship Mints <shipmints@gmail.com> writes:

> Yes, could do, and nothing needed by org-mode or markdown-mode. Could also
> test for buffer-file-name nil. Or test both. Let's recommend that simpler
> approach and I will rescind my markdown PR in favor of this simpler
> suggestion.

Well. Org or markdown should name the buffers according to the
convention :) Org does it though. So, we should be good.

Closing the patch request.
Canceled.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 1035 bytes --]

For the record, this is what I've recommended and what I will use:

  (defun my/XXX-mode-hook ()
    (when (or buffer-file-name
              (not (string-prefix-p " " (buffer-name))))
      ;; do something expensive
      ))

On Sat, Jan 18, 2025 at 6:36 AM Ihor Radchenko <yantar92@posteo.net> wrote:

> Ship Mints <shipmints@gmail.com> writes:
>
> > Yes, could do, and nothing needed by org-mode or markdown-mode. Could
> also
> > test for buffer-file-name nil. Or test both. Let's recommend that simpler
> > approach and I will rescind my markdown PR in favor of this simpler
> > suggestion.
>
> Well. Org or markdown should name the buffers according to the
> convention :) Org does it though. So, we should be good.
>
> Closing the patch request.
> Canceled.
>
> --
> Ihor Radchenko // yantar92,
> Org mode maintainer,
> Learn more about Org mode at <https://orgmode.org/>.
> Support Org development at <https://liberapay.com/org-mode>,
> or support my work at <https://liberapay.com/yantar92>
>

[-- Attachment #2: Type: text/html, Size: 1984 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 1341 bytes --]

This could be added to the org-mode documentation as a hint for those who
use native fontification. Should I send a patch for that or trivial enough
you'd do it, if you agree?

On Sat, Jan 18, 2025 at 6:54 AM Ship Mints <shipmints@gmail.com> wrote:

> For the record, this is what I've recommended and what I will use:
>
>   (defun my/XXX-mode-hook ()
>     (when (or buffer-file-name
>               (not (string-prefix-p " " (buffer-name))))
>       ;; do something expensive
>       ))
>
> On Sat, Jan 18, 2025 at 6:36 AM Ihor Radchenko <yantar92@posteo.net>
> wrote:
>
>> Ship Mints <shipmints@gmail.com> writes:
>>
>> > Yes, could do, and nothing needed by org-mode or markdown-mode. Could
>> also
>> > test for buffer-file-name nil. Or test both. Let's recommend that
>> simpler
>> > approach and I will rescind my markdown PR in favor of this simpler
>> > suggestion.
>>
>> Well. Org or markdown should name the buffers according to the
>> convention :) Org does it though. So, we should be good.
>>
>> Closing the patch request.
>> Canceled.
>>
>> --
>> Ihor Radchenko // yantar92,
>> Org mode maintainer,
>> Learn more about Org mode at <https://orgmode.org/>.
>> Support Org development at <https://liberapay.com/org-mode>,
>> or support my work at <https://liberapay.com/yantar92>
>>
>

[-- Attachment #2: Type: text/html, Size: 2598 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Jens Schmidt]

On 2025-01-18  12:38, Ihor Radchenko wrote:
> Ship Mints <shipmints@gmail.com> writes:
> 
>> Yes, could do, and nothing needed by org-mode or markdown-mode. Could also
>> test for buffer-file-name nil. Or test both. Let's recommend that simpler
>> approach and I will rescind my markdown PR in favor of this simpler
>> suggestion.
> 
> Well. Org or markdown should name the buffers according to the
> convention :) Org does it though. So, we should be good.
> 
> Closing the patch request.
> Canceled.

Um, sorry for butting in, but I also have an interest of recognizing
Org's special source environments, see here:

  https://list.orgmode.org/9eaf7099554d488d921e64c4b2852a0d@vodafonemail.de/

That thread led to nowhere back then, but I think it still makes
sense to have the various Org source environments detectable in 
a defined and explicit way, and Stephane's request reminded me
about that.

While one could use buffer names also for detecting the source edit
buffers ("*Org Src collateral.org[ shell ]*"), it feels somehow
brittle, since names can change.  So why not extend on Stephane's
request and provide some variable that informs about the various
types of Org source environments?

Like a variable named `org-src-environment-type' bound dynamically
to possible values, e.g. `edit' or `fontify'?

Thanks!  (In particular also for taking the responsibility of Org
maintenance.)

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 1805 bytes --]

Indeed if this is a valid use case for buffer identification, the approach
in the patch seems trivial (to me, buffer locals are more appropriate than
dynamic bindings), if overkill for the purpose I proposed.

On Sat, Jan 18, 2025 at 7:16 AM Jens Schmidt <jschmidt4gnu@vodafonemail.de>
wrote:

> On 2025-01-18  12:38, Ihor Radchenko wrote:
> > Ship Mints <shipmints@gmail.com> writes:
> >
> >> Yes, could do, and nothing needed by org-mode or markdown-mode. Could
> also
> >> test for buffer-file-name nil. Or test both. Let's recommend that
> simpler
> >> approach and I will rescind my markdown PR in favor of this simpler
> >> suggestion.
> >
> > Well. Org or markdown should name the buffers according to the
> > convention :) Org does it though. So, we should be good.
> >
> > Closing the patch request.
> > Canceled.
>
> Um, sorry for butting in, but I also have an interest of recognizing
> Org's special source environments, see here:
>
>
> https://list.orgmode.org/9eaf7099554d488d921e64c4b2852a0d@vodafonemail.de/
>
> That thread led to nowhere back then, but I think it still makes
> sense to have the various Org source environments detectable in
> a defined and explicit way, and Stephane's request reminded me
> about that.
>
> While one could use buffer names also for detecting the source edit
> buffers ("*Org Src collateral.org[ shell ]*"), it feels somehow
> brittle, since names can change.  So why not extend on Stephane's
> request and provide some variable that informs about the various
> types of Org source environments?
>
> Like a variable named `org-src-environment-type' bound dynamically
> to possible values, e.g. `edit' or `fontify'?
>
> Thanks!  (In particular also for taking the responsibility of Org
> maintenance.)
>
>

[-- Attachment #2: Type: text/html, Size: 2588 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Ship Mints <shipmints@gmail.com> writes:

> This could be added to the org-mode documentation as a hint for those who
> use native fontification. Should I send a patch for that or trivial enough
> you'd do it, if you agree?

I am not sure if this problem is so common that we need to put it into
the manual.

I think creating a WORG page that collects performance hacks and
suggestions could be a good alternative.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Jens Schmidt <jschmidt4gnu@vodafonemail.de> writes:

> Um, sorry for butting in, but I also have an interest of recognizing
> Org's special source environments, see here:
>
>   https://list.orgmode.org/9eaf7099554d488d921e64c4b2852a0d@vodafonemail.de/
>
> That thread led to nowhere back then, but I think it still makes
> sense to have the various Org source environments detectable in 
> a defined and explicit way, and Stephane's request reminded me
> about that.

AFAIR, your request was not about detecting "src edit" buffers, but about
detecting when a command is called in those buffers in very specific
scenario - by `org-babel-do-in-edit-buffer', while keeping the normal
behavior when editing those buffers interactively.

> While one could use buffer names also for detecting the source edit
> buffers ("*Org Src collateral.org[ shell ]*"), it feels somehow
> brittle, since names can change.  So why not extend on Stephane's
> request and provide some variable that informs about the various
> types of Org source environments?

How would that help your original request?

> Like a variable named `org-src-environment-type' bound dynamically
> to possible values, e.g. `edit' or `fontify'?

I am reluctant. This will force us to make sure that we arrange this
variable appropriately in every single temporary buffer Org creates for
various purposes. In other words, it will increase maintenance burden.

Unless strongly justified, I am not going to do it.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ship Mints]

[-- Attachment #1: Type: text/plain, Size: 871 bytes --]

Makes sense. I'm not set up to contribute to WORG, perhaps you can do it,
if you don't mind?

On Sat, Jan 18, 2025 at 7:39 AM Ihor Radchenko <yantar92@posteo.net> wrote:

> Ship Mints <shipmints@gmail.com> writes:
>
> > This could be added to the org-mode documentation as a hint for those who
> > use native fontification. Should I send a patch for that or trivial
> enough
> > you'd do it, if you agree?
>
> I am not sure if this problem is so common that we need to put it into
> the manual.
>
> I think creating a WORG page that collects performance hacks and
> suggestions could be a good alternative.
>
> --
> Ihor Radchenko // yantar92,
> Org mode maintainer,
> Learn more about Org mode at <https://orgmode.org/>.
> Support Org development at <https://liberapay.com/org-mode>,
> or support my work at <https://liberapay.com/yantar92>
>

[-- Attachment #2: Type: text/html, Size: 1592 bytes --]

Re: Add org-src-is-fontify-buffer-p (patch attached)

[Ihor Radchenko]

Ship Mints <shipmints@gmail.com> writes:

> Makes sense. I'm not set up to contribute to WORG, perhaps you can do it,
> if you don't mind?

I can add it to my todo list. But that means not soon.

-- 
Ihor Radchenko // yantar92,
Org mode maintainer,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>