50081: reset global mbstate_t variables when LC_CTYPE changes

3 files changed, 32 insertions, 1 deletions

diff --git a/ChangeLog b/ChangeLog
index da44c2840..80a7f4657 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,8 @@
 2022-04-20  Jun-ichi Takimoto  <takimoto-j@kba.biglobe.ne.jp>
 
+	* 50081: Src/pattern.c, Src/params.c: reset global mbstate_t
+	variables when LC_CTYPE changes
+
 	* 50080: Src/hist.c: add missing STOUC() in casemodify()
 
 2022-04-19  Bart Schaefer  <schaefer@zsh.org>
diff --git a/Src/params.c b/Src/params.c
index 970a207e4..27ea82298 100644
--- a/Src/params.c
+++ b/Src/params.c
@@ -4594,6 +4594,19 @@ static struct localename {
     {NULL, 0}
 };
 
+/* On some systems (at least on NetBSD-9), when LC_CTYPE changes,
+ * global variables (type mbstate_t) used by mbrtowc() etc. need be
+ * reset by clear_mbstate() */
+
+/**/
+static void
+clear_mbstate(void) {
+#ifdef MULTIBYTE_SUPPORT
+    mb_charinit();	/* utils.c */
+    clear_shiftstate();	/* pattern.c */
+#endif
+}
+
 /**/
 static void
 setlang(char *x)
@@ -4616,6 +4629,7 @@ setlang(char *x)
      * that case.
      */
     setlocale(LC_ALL, x ? unmeta(x) : "");
+    clear_mbstate();
     queue_signals();
     for (ln = lc_names; ln->name; ln++)
 	if ((x = getsparam_u(ln->name)) && *x)
@@ -4641,8 +4655,10 @@ lc_allsetfn(Param pm, char *x)
 	    unqueue_signals();
 	}
     }
-    else
+    else {
 	setlocale(LC_ALL, unmeta(x));
+	clear_mbstate();
+    }
 }
 
 /**/
@@ -4679,6 +4695,7 @@ lcsetfn(Param pm, char *x)
 		setlocale(ln->category, unmeta(x));
     }
     unqueue_signals();
+    clear_mbstate();	/* LC_CTYPE may have changed */
 }
 #endif /* USE_LOCALE */
 
@@ -5627,6 +5644,7 @@ endparamscope(void)
 		    setlocale(ln->category, val);
 	    }
 	}
+	clear_mbstate();    /* LC_CTYPE may have changed */
     }
 #endif /* USE_LOCALE */
     unqueue_signals();
diff --git a/Src/pattern.c b/Src/pattern.c
index c0e31b78e..e947d1216 100644
--- a/Src/pattern.c
+++ b/Src/pattern.c
@@ -320,6 +320,14 @@ typedef wint_t patint_t;
  */
 static mbstate_t shiftstate;
 
+/* See clear_mbstate() in params.c for the use of clear_shiftstate() */
+
+/**/
+mod_export void
+clear_shiftstate(void) {
+    memset(&shiftstate, 0, sizeof(shiftstate));
+}
+
 /*
  * Multibyte version: it's (almost) as easy to return the
  * value as not, so do so since we sometimes need it..
@@ -1999,6 +2007,8 @@ charsub(char *x, char *y)
 
 	if (ret == MB_INVALID || ret == MB_INCOMPLETE) {
 	    /* Error.  Treat remainder as single characters */
+	    /* Reset the shift state for next time. */
+	    memset(&shiftstate, 0, sizeof(shiftstate));
 	    return res + (y - x);
 	}