unofficial mirror of bug-gnu-emacs@gnu.org 
 help / color / mirror / code / Atom feed
From: Theodor Thornhill via "Bug reports for GNU Emacs, the Swiss army knife of text editors" <bug-gnu-emacs@gnu.org>
To: Yuan Fu <casouri@gmail.com>
Cc: 60128@debbugs.gnu.org, eliz@gnu.org,
	Stefan Monnier <monnier@iro.umontreal.ca>
Subject: bug#60128: 30.0.50; [PATCH]: Add treesit-transpose-sexps
Date: Mon, 26 Dec 2022 23:37:40 +0100	[thread overview]
Message-ID: <87wn6dg5aj.fsf@thornhill.no> (raw)
In-Reply-To: <0D9F1198-B5FB-451D-B0E8-DB22910DAB58@gmail.com>

[-- Attachment #1: Type: text/plain, Size: 3083 bytes --]

Yuan Fu <casouri@gmail.com> writes:

>> On Dec 26, 2022, at 12:53 PM, Theodor Thornhill <theo@thornhill.no> wrote:
>> 
>> Theodor Thornhill <theo@thornhill.no> writes:
>> 
>>> Theodor Thornhill <theo@thornhill.no> writes:
>>> 
>>>> Hi there!
>>>> 
>>>> Attached is a patch that enables transpose-sexps for tree-sitter enabled
>>>> modes.
>>>> 
>>>> This function will swap the node _before_ node-at-point with the node
>>>> _after_, so it would do something like:
>>>> 
>>>>       foo a|nd bar => bar and foo|
>>>> 
>>>> or
>>>>       foo(a + 4,| y + c * b, b, d); => foo(y + c * b, a + 4|, b, d);
>>>> 
>>>> It will _not_ try to swap things that are not siblings.  I think that
>>>> makes sense in the case of non-lisp languages, since _most_ places you
>>>> can transpose-sexps you will end up with broken code.
>>>> 
>>> 
>>> from 'transpose-subr-1':
>>> 
>>>  (if (> (cdr pos1) (car pos2)) (error "Don't have two things to
>>>  transpose"))
>>> 
>>> I added this hack into the function in the patch, but I think that
>>> triggering an error is too much.
>>> 
>>>    ;; Hack to trigger the error message in `transpose-subr-1' when we
>>>    ;; don't have siblings to swap.
>>>    (list (cons 0 1) (cons 0 1))))
>>> 
>>> I guess I could just follow suit in my function and do like in the
>>> following patch:
>>> 
>>> Theo
>> 
>> 
>> Considering there is both a bug-report _and_ a discussion around this I
>> guess the best idea is to add the patch to this bug report, and continue
>> discussing this in the report rather than emacs-devel?
>> 
>> What do you think about this patch?
>> 
>> (copied from emacs-devel):
>> It feels a little iffy how to handle the separate return values, but it
>> works.  I'd be super happy for some feedback on how to best solve that,
>> though :)
>
> By separate return values, do you mean the function returns a cons of
> cons? It seems fine to me. Though I think the docstring could be more
> specific. Like saying return a cons (REGION . REGION), where REGION is
> (BEG . END), where BEG and END...
>

I updated the patch adressing your comment.  The current "protocol" used
on the master branch implies that the regions are calculated by actually
moving inside the buffer and running (point).  Because we don't need
that with the ast in hand I'm trying to support a new protocol, where we
just supply the data we need.  It's a little difficult to "surgically"
modify the behavior, so I'm very open to suggestions, but I believe it
should work properly as the patch stands.

>> 
>> Also, I made the treesit-transpose-sexps a little better imo, in that we
>> only find named nodes to swap, but use every available node for the
>> entry. We rarely, if ever want to swap the unnamed nodes.
>> 
>> Eli, does this require a NEWS entry or more documentation?
>
> IMHO a backend/helper function shouldn’t signal a user-error, it’s
> better to return nil and let the command to signal errors.

Yeah, I agree.  Adjusted the patch.  Let me know what you think!

Theo


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: 0001-Add-treesit-transpose-sexps-bug-60128.patch --]
[-- Type: text/x-diff, Size: 8824 bytes --]

From 74a7980758b929c991e317ea281b7cd4a097fdff Mon Sep 17 00:00:00 2001
From: Theodor Thornhill <theo@thornhill.no>
Date: Sun, 25 Dec 2022 20:11:59 +0100
Subject: [PATCH] Add treesit-transpose-sexps (bug#60128)

We don't really need to rely on forward-sexp to define what to
transpose.  In tree-sitter we can consider siblings as "balanced
expressions", and swap them without doing any movement to calculate
where the siblings in question are.

* lisp/simple.el (transpose-sexps-function): New defvar-local.
(transpose-sexps): Use the new defvar-local if available.
(transpose-subr): Check whether the mover function returns a cons of
conses, then run transpose-subr-1 on the position-pairs.
* lisp/treesit.el (treesit-transpose-sexps): New function.
---
 lisp/simple.el  | 106 +++++++++++++++++++++++++++++-------------------
 lisp/treesit.el |  28 ++++++++++++-
 2 files changed, 91 insertions(+), 43 deletions(-)

diff --git a/lisp/simple.el b/lisp/simple.el
index 4551b749d5..99dbfaea9f 100644
--- a/lisp/simple.el
+++ b/lisp/simple.el
@@ -8438,6 +8438,21 @@ transpose-words
   (interactive "*p")
   (transpose-subr 'forward-word arg))
 
+(defvar-local transpose-sexps-function nil
+  "If non-nil, `transpose-sexps' delegates to this function.
+
+This function takes one argument ARG, a number as provided
+through running `transpose-sexps'.  It supports the following two
+return values:
+
+1. A cons (REGION . REGION), where REGION is (BEG . END) and BEG
+and END are buffer positions.
+
+2. A cons (BEG . END), where BEG and END are buffer positions.
+
+If the second return value is chosen for this function, it is
+expected to behave similarly to `forward-sexp' and friends.")
+
 (defun transpose-sexps (arg &optional interactive)
   "Like \\[transpose-chars] (`transpose-chars'), but applies to sexps.
 Unlike `transpose-words', point must be between the two sexps and not
@@ -8454,36 +8469,37 @@ transpose-sexps
           (transpose-sexps arg nil)
         (scan-error (user-error "Not between two complete sexps")))
     (transpose-subr
-     (lambda (arg)
-       ;; Here we should try to simulate the behavior of
-       ;; (cons (progn (forward-sexp x) (point))
-       ;;       (progn (forward-sexp (- x)) (point)))
-       ;; Except that we don't want to rely on the second forward-sexp
-       ;; putting us back to where we want to be, since forward-sexp-function
-       ;; might do funny things like infix-precedence.
-       (if (if (> arg 0)
-	       (looking-at "\\sw\\|\\s_")
-	     (and (not (bobp))
-		  (save-excursion
-                    (forward-char -1)
-                    (looking-at "\\sw\\|\\s_"))))
-	   ;; Jumping over a symbol.  We might be inside it, mind you.
-	   (progn (funcall (if (> arg 0)
-			       'skip-syntax-backward 'skip-syntax-forward)
-			   "w_")
-		  (cons (save-excursion (forward-sexp arg) (point)) (point)))
-         ;; Otherwise, we're between sexps.  Take a step back before jumping
-         ;; to make sure we'll obey the same precedence no matter which
-         ;; direction we're going.
-         (funcall (if (> arg 0) 'skip-syntax-backward 'skip-syntax-forward)
-                  " .")
-         (cons (save-excursion (forward-sexp arg) (point))
-	       (progn (while (or (forward-comment (if (> arg 0) 1 -1))
-			         (not (zerop (funcall (if (> arg 0)
-							  'skip-syntax-forward
-						        'skip-syntax-backward)
-						      ".")))))
-		      (point)))))
+     (if transpose-sexps-function transpose-sexps-function
+       (lambda (arg)
+         ;; Here we should try to simulate the behavior of
+         ;; (cons (progn (forward-sexp x) (point))
+         ;;       (progn (forward-sexp (- x)) (point)))
+         ;; Except that we don't want to rely on the second forward-sexp
+         ;; putting us back to where we want to be, since forward-sexp-function
+         ;; might do funny things like infix-precedence.
+         (if (if (> arg 0)
+	         (looking-at "\\sw\\|\\s_")
+	       (and (not (bobp))
+	            (save-excursion
+                      (forward-char -1)
+                      (looking-at "\\sw\\|\\s_"))))
+             ;; Jumping over a symbol.  We might be inside it, mind you.
+	     (progn (funcall (if (> arg 0)
+			         #'skip-syntax-backward #'skip-syntax-forward)
+			     "w_")
+	            (cons (save-excursion (forward-sexp arg) (point)) (point)))
+           ;; Otherwise, we're between sexps.  Take a step back before jumping
+           ;; to make sure we'll obey the same precedence no matter which
+           ;; direction we're going.
+           (funcall (if (> arg 0) #'skip-syntax-backward #'skip-syntax-forward)
+                    " .")
+           (cons (save-excursion (forward-sexp arg) (point))
+	         (progn (while (or (forward-comment (if (> arg 0) 1 -1))
+			           (not (zerop (funcall (if (> arg 0)
+						            #'skip-syntax-forward
+						          #'skip-syntax-backward)
+						        ".")))))
+		        (point))))))
      arg 'special)))
 
 (defun transpose-lines (arg)
@@ -8509,19 +8525,23 @@ transpose-lines
 ;; FIXME document SPECIAL.
 (defun transpose-subr (mover arg &optional special)
   "Subroutine to do the work of transposing objects.
-Works for lines, sentences, paragraphs, etc.  MOVER is a function that
-moves forward by units of the given object (e.g. `forward-sentence',
-`forward-paragraph').  If ARG is zero, exchanges the current object
-with the one containing mark.  If ARG is an integer, moves the
-current object past ARG following (if ARG is positive) or
-preceding (if ARG is negative) objects, leaving point after the
-current object."
-  (let ((aux (if special mover
-	       (lambda (x)
-		 (cons (progn (funcall mover x) (point))
-		       (progn (funcall mover (- x)) (point))))))
-	pos1 pos2)
+Works for lines, sentences, paragraphs, etc.  MOVER is either a
+function that moves forward by units of the given
+object (e.g. `forward-sentence', `forward-paragraph'), or a
+function that calculates a cons of two position-pairs.  If ARG is
+zero, exchanges the current object with the one containing mark.
+If ARG is an integer, moves the current object past ARG
+following (if ARG is positive) or preceding (if ARG is negative)
+objects, leaving point after the current object."
+  (let* ((aux (if special mover
+	        (lambda (x)
+		  (cons (progn (funcall mover x) (point))
+		        (progn (funcall mover (- x)) (point))))))
+	 (pos1 (save-excursion (funcall aux arg)))
+         pos2)
     (cond
+     ((and (consp (car pos1)) (consp (cdr pos1)))
+      (transpose-subr-1 (car pos1) (cdr pos1)))
      ((= arg 0)
       (save-excursion
 	(setq pos1 (funcall aux 1))
@@ -8542,6 +8562,8 @@ transpose-subr
       (goto-char (+ (car pos2) (- (cdr pos1) (car pos1))))))))
 
 (defun transpose-subr-1 (pos1 pos2)
+  (unless (and pos1 pos2)
+    (error "Don't have two things to transpose"))
   (when (> (car pos1) (cdr pos1)) (setq pos1 (cons (cdr pos1) (car pos1))))
   (when (> (car pos2) (cdr pos2)) (setq pos2 (cons (cdr pos2) (car pos2))))
   (when (> (car pos1) (car pos2))
diff --git a/lisp/treesit.el b/lisp/treesit.el
index cefbed1a16..2bd7f71f2f 100644
--- a/lisp/treesit.el
+++ b/lisp/treesit.el
@@ -1582,6 +1582,31 @@ treesit-search-forward-goto
       (goto-char current-pos)))
     node))
 
+(defun treesit-transpose-sexps (&optional arg)
+  "Tree-sitter `transpose-sexps' function.
+Arg is the same as in `transpose-sexps'.
+
+Locate the node closest to POINT, and transpose that node with
+its sibling node ARG nodes away.
+
+Return a pair of positions as described by
+`transpose-sexps-function' for use in `transpose-subr' and
+friends."
+  (let* ((parent (treesit-node-parent (treesit-node-at (point))))
+         (child (treesit-node-child parent 0 t)))
+    (named-let loop ((prev child)
+                     (next (treesit-node-child
+                            parent (+ arg (treesit-node-index child t))
+                            t)))
+      (when (and prev next)
+        (if (< (point) (treesit-node-end next))
+            (cons (cons (treesit-node-start prev)
+                        (treesit-node-end prev))
+                  (cons (treesit-node-start next)
+                        (treesit-node-end next)))
+          (loop (treesit-node-next-sibling prev t)
+                (treesit-node-next-sibling next t)))))))
+
 ;;; Navigation, defun, things
 ;;
 ;; Emacs lets you define "things" by a regexp that matches the type of
@@ -2111,7 +2136,8 @@ treesit-major-mode-setup
   ;; Defun name.
   (when treesit-defun-name-function
     (setq-local add-log-current-defun-function
-                #'treesit-add-log-current-defun)))
+                #'treesit-add-log-current-defun))
+  (setq-local transpose-sexps-function #'treesit-transpose-sexps))
 
 ;;; Debugging
 
-- 
2.34.1


  reply	other threads:[~2022-12-26 22:37 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-12-16 20:04 bug#60128: 30.0.50; [PATCH]: Add treesit-transpose-sexps Theodor Thornhill via Bug reports for GNU Emacs, the Swiss army knife of text editors
2022-12-17 12:52 ` Theodor Thornhill via Bug reports for GNU Emacs, the Swiss army knife of text editors
2022-12-26 20:53   ` Theodor Thornhill via Bug reports for GNU Emacs, the Swiss army knife of text editors
2022-12-26 21:26     ` Yuan Fu
2022-12-26 22:37       ` Theodor Thornhill via Bug reports for GNU Emacs, the Swiss army knife of text editors [this message]
2022-12-27 11:56     ` Eli Zaretskii
2023-01-07 23:18 ` Yuan Fu
2023-01-08 11:56   ` Eli Zaretskii
2023-01-08 12:13     ` Theodor Thornhill via Bug reports for GNU Emacs, the Swiss army knife of text editors
2023-09-05 15:56       ` Stefan Kangas

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87wn6dg5aj.fsf@thornhill.no \
    --to=bug-gnu-emacs@gnu.org \
    --cc=60128@debbugs.gnu.org \
    --cc=casouri@gmail.com \
    --cc=eliz@gnu.org \
    --cc=monnier@iro.umontreal.ca \
    --cc=theo@thornhill.no \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).