From: Mattias Engdegård <mattiase@acm.org>
Date: Fri, 3 Jun 2022 08:12:24 +0000 (+0200)
Subject: Let ?\LF signal an error (bug#55738)
X-Git-Tag: emacs-29.0.90~1910^2~265
X-Git-Url: http://git.eshelyaron.com/gitweb/?a=commitdiff_plain;h=28622d4dd0347227a28b7b25c674437239a00a06;p=emacs.git

Let ?\LF signal an error (bug#55738)

As suggested by Stefan Monnier.

* src/lread.c (read_escape):
Signal an error for ?\LF since it cannot reasonably be intended.
* test/src/lread-tests.el (lread-escaped-lf): Update test.
* etc/NEWS: Announce.
---

diff --git a/etc/NEWS b/etc/NEWS
index f9409bb24f8..54bc6d80e1a 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -1922,6 +1922,9 @@ It was previously only run by 'clone-indirect-buffer' and
 called by both of these, the hook is now run by all 3 of these
 functions.
 
+---
+** '?\' at the end of a line now signals an error.
+Previously it produced a nonsense value, -1, that was never intended.
 
 * Lisp Changes in Emacs 29.1
 
diff --git a/src/lread.c b/src/lread.c
index 4b7d38a8e6c..1d20470a8bf 100644
--- a/src/lread.c
+++ b/src/lread.c
@@ -2664,6 +2664,10 @@ read_escape (Lisp_Object readcharfun)
     case 'v':
       return '\v';
 
+    case '\n':
+      /* ?\LF is an error; it's probably a user mistake.  */
+      error ("Invalid escape character syntax");
+
     case 'M':
       c = READCHAR;
       if (c != '-')
diff --git a/test/src/lread-tests.el b/test/src/lread-tests.el
index 99eec9d5487..f190f14781e 100644
--- a/test/src/lread-tests.el
+++ b/test/src/lread-tests.el
@@ -318,8 +318,8 @@ literals (Bug#20852)."
                  '(## . 2))))
 
 (ert-deftest lread-escaped-lf ()
-  ;; ?\LF should produce LF (only inside string literals do we ignore \LF).
-  (should (equal (read-from-string "?\\\n") '(?\n . 3)))
+  ;; ?\LF should signal an error; \LF is ignored inside string literals.
+  (should-error (read-from-string "?\\\n x"))
   (should (equal (read-from-string "\"a\\\nb\"") '("ab" . 6))))
 
 ;;; lread-tests.el ends here