bug#53260: char-syntax differs in interpreter and bytecode [PATCH]

["Mattias Engdegård" <mattiase@acm.org>]

[-- Attachment #1: Type: text/plain, Size: 465 bytes --]

15 jan. 2022 kl. 23.51 skrev Stefan Monnier <monnier@iro.umontreal.ca>:

> Doesn't sound right: char tables are indexed by chars (i.e. Unicode code
> points) not by bytes, so we need to convert the byte into a char
> before indexing.

Sure, I'm happy to do it either way. Chars retrieved from unibyte buffers or strings really should be converted to multibyte before used with char-syntax; unibyte buffers are not very common but strings slightly more so.


[-- Attachment #2: 0001-Fix-Fchar_syntax-for-non-ASCII-in-unibyte-buffers.patch --]
[-- Type: application/octet-stream, Size: 3163 bytes --]

From 2adb5c862232abf126f73d0aa514f5ae8b6babba Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Mattias=20Engdeg=C3=A5rd?= <mattiase@acm.org>
Date: Sun, 16 Jan 2022 11:58:00 +0100
Subject: [PATCH] Fix Fchar_syntax for non-ASCII in unibyte buffers

Fchar_syntax did not convert unibyte characters to multibyte when the
current buffer was unibyte, in contrast to `char-syntax` in
byte-compiled code (bug#53260).

* src/bytecode.c (exec_byte_code): Call out to Fchar_syntax;
the dynamic frequency is too low to justify inlining here, and it
did lead to implementations diverging.
* src/syntax.c (Fchar_syntax): Convert non-ASCII unibyte values to
multibyte.  Remove useless SETUP_BUFFER_SYNTAX_TABLE which has no
effect here.
* test/src/syntax-tests.el (syntax-char-syntax): New test.
---
 src/bytecode.c           |  8 +-------
 src/syntax.c             |  6 +++---
 test/src/syntax-tests.el | 15 +++++++++++++++
 3 files changed, 19 insertions(+), 10 deletions(-)

diff --git a/src/bytecode.c b/src/bytecode.c
index 472992be18..b7e65d05ae 100644
--- a/src/bytecode.c
+++ b/src/bytecode.c
@@ -1167,13 +1167,7 @@ #define DEFINE(name, value) LABEL (name) ,
 	  NEXT;
 
 	CASE (Bchar_syntax):
-	  {
-	    CHECK_CHARACTER (TOP);
-	    int c = XFIXNAT (TOP);
-	    if (NILP (BVAR (current_buffer, enable_multibyte_characters)))
-	      c = make_char_multibyte (c);
-	    XSETFASTINT (TOP, syntax_code_spec[SYNTAX (c)]);
-	  }
+	  TOP = Fchar_syntax (TOP);
 	  NEXT;
 
 	CASE (Bbuffer_substring):
diff --git a/src/syntax.c b/src/syntax.c
index 9df878b8ed..c1e81dfa47 100644
--- a/src/syntax.c
+++ b/src/syntax.c
@@ -1101,10 +1101,10 @@ DEFUN ("char-syntax", Fchar_syntax, Schar_syntax, 1, 1, 0,
 `syntax-after' instead.  */)
   (Lisp_Object character)
 {
-  int char_int;
   CHECK_CHARACTER (character);
-  char_int = XFIXNUM (character);
-  SETUP_BUFFER_SYNTAX_TABLE ();
+  int char_int = XFIXNAT (character);
+  if (NILP (BVAR (current_buffer, enable_multibyte_characters)))
+    char_int = make_char_multibyte (char_int);
   return make_fixnum (syntax_code_spec[SYNTAX (char_int)]);
 }
 
diff --git a/test/src/syntax-tests.el b/test/src/syntax-tests.el
index 3b9f21cde3..501b5e067f 100644
--- a/test/src/syntax-tests.el
+++ b/test/src/syntax-tests.el
@@ -506,4 +506,19 @@ test-from-to-parse-partial-sexp
     (should (parse-partial-sexp 1 1))
     (should-error (parse-partial-sexp 2 1))))
 
+(ert-deftest syntax-char-syntax ()
+  ;; Verify that char-syntax behaves identically in interpreted and
+  ;; byte-compiled code (bug#53260).
+  (let ((cs (byte-compile (lambda (x) (char-syntax x)))))
+    ;; Use a unibyte buffer with a syntax table using symbol syntax
+    ;; for raw byte 128.
+    (with-temp-buffer
+      (set-buffer-multibyte nil)
+      (let ((st (make-syntax-table)))
+        (modify-syntax-entry (unibyte-char-to-multibyte 128) "_" st)
+        (set-syntax-table st)
+        (should (equal (char-syntax 128) ?_))
+        (should (equal (funcall cs 128) ?_))))
+    (list (char-syntax 128) (funcall cs 128))))
+
 ;;; syntax-tests.el ends here
-- 
2.32.0 (Apple Git-132)