bug#35286: 26.2; indent-sexp broken

[Noam Postavsky]

Leo Liu <sdl.web@gmail.com> writes:

> Treating things between parentheses as sexp [...] is not lisp-specific

This part makes sense to me.

> using indent-sexp to indent it is not lisp-specific

This is what surprises me.  How does
lisp-indent-calc-next/calculate-lisp-indent give correct results for
non-lisp?  I think using prog-indent-sexp for non-lisp would make more
sense.

Anyway, the following patch seems to work with your examples.

>From 0739d40a3c1aa5690442f84cc7bf5fa093f5b06e Mon Sep 17 00:00:00 2001
From: Noam Postavsky <npostavs@gmail.com>
Date: Mon, 15 Apr 2019 18:49:57 -0400
Subject: [PATCH] Be more careful about indent-sexp going over eol (Bug#35286)

* lisp/emacs-lisp/lisp-mode.el (indent-sexp): Only go over multiple
sexps if the end of line is within a sexp.
---
 lisp/emacs-lisp/lisp-mode.el            | 22 ++++++++++++++--------
 test/lisp/emacs-lisp/lisp-mode-tests.el | 12 ++++++++++++
 2 files changed, 26 insertions(+), 8 deletions(-)

diff --git a/lisp/emacs-lisp/lisp-mode.el b/lisp/emacs-lisp/lisp-mode.el
index 57f57175c5..74bf0c87c5 100644
--- a/lisp/emacs-lisp/lisp-mode.el
+++ b/lisp/emacs-lisp/lisp-mode.el
@@ -1205,19 +1205,25 @@ (defun indent-sexp (&optional endpos)
                     ;; Get error now if we don't have a complete sexp
                     ;; after point.
                     (save-excursion
+                      (forward-sexp 1)
                       (let ((eol (line-end-position)))
-                        (forward-sexp 1)
                         ;; We actually look for a sexp which ends
                         ;; after the current line so that we properly
                         ;; indent things like #s(...).  This might not
                         ;; be needed if Bug#15998 is fixed.
-                        (condition-case ()
-                            (while (and (< (point) eol) (not (eobp)))
-                              (forward-sexp 1))
-                          ;; But don't signal an error for incomplete
-                          ;; sexps following the first complete sexp
-                          ;; after point.
-                          (scan-error nil)))
+                        (when (and (< (point) eol)
+                                   ;; Check if eol is within a sexp.
+                                   (> (nth 0 (save-excursion
+                                               (parse-partial-sexp
+                                                (point) eol)))
+                                      0))
+                          (condition-case ()
+                              (while (< (point) eol)
+                                (forward-sexp 1))
+                            ;; But don't signal an error for incomplete
+                            ;; sexps following the first complete sexp
+                            ;; after point.
+                            (scan-error nil))))
                       (point)))))
     (save-excursion
       (while (let ((indent (lisp-indent-calc-next parse-state))
diff --git a/test/lisp/emacs-lisp/lisp-mode-tests.el 
b/test/lisp/emacs-lisp/lisp-mode-tests.el
index a6370742ab..3782bad315 100644
--- a/test/lisp/emacs-lisp/lisp-mode-tests.el
+++ b/test/lisp/emacs-lisp/lisp-mode-tests.el
@@ -136,6 +136,18 @@ (ert-deftest indent-sexp-cant-go ()
     (indent-sexp)
     (should (equal (buffer-string) "(())"))))
 
+(ert-deftest indent-sexp-stop-before-eol-comment ()
+  "`indent-sexp' shouldn't look for more sexps after an eol comment."
+  ;; See https://debbugs.gnu.org/cgi/bugreport.cgi?bug=35286.
+  (with-temp-buffer
+    (emacs-lisp-mode)
+    (let ((str "() ;;\n  x"))
+      (insert str)
+      (goto-char (point-min))
+      (indent-sexp)
+      ;; The "x" is in the next sexp, so it shouldn't get indented.
+      (should (equal (buffer-string) str)))))
+
 (ert-deftest lisp-indent-region ()
   "Test basics of `lisp-indent-region'."
   (with-temp-buffer
-- 
2.11.0