From 1547b04e66e0f01ed6795f95c9111bdae3a078f2 Mon Sep 17 00:00:00 2001
From: tg <tg@mirbsd.org>
Date: Sun, 5 Nov 2006 12:11:14 +0000
Subject: [PATCH] add new "set -o utf8-hack", currently no effect set
 automatically on startup if we have locale functions (on MirOS)

---
 Makefile |  4 ++--
 check.t  |  4 ++--
 main.c   | 23 ++++++++++++++++++++++-
 misc.c   |  3 ++-
 sh.h     |  5 +++--
 5 files changed, 31 insertions(+), 8 deletions(-)

diff --git a/Makefile b/Makefile
index bcf6dd0..3170e45 100644
--- a/Makefile
+++ b/Makefile
@@ -1,9 +1,9 @@
-# $MirOS: src/bin/mksh/Makefile,v 1.12 2006/08/26 20:30:27 tg Exp $
+# $MirOS: src/bin/mksh/Makefile,v 1.13 2006/11/05 12:11:13 tg Exp $
 
 PROG=		mksh
 SRCS=		alloc.c edit.c eval.c exec.c expr.c funcs.c histrap.c \
 		jobs.c lex.c main.c misc.c shf.c syn.c tree.c var.c
-CPPFLAGS+=	-DHAVE_ARC4RANDOM -DHAVE_ARC4RANDOM_PUSH
+CPPFLAGS+=	-DHAVE_ARC4RANDOM -DHAVE_ARC4RANDOM_PUSH -DHAVE_LANGSTUFF
 CDIAGFLAGS+=	-Wno-cast-qual
 
 LINKS+=		${BINDIR}/${PROG} ${BINDIR}/sh
diff --git a/check.t b/check.t
index 54906e1..4dbdee2 100644
--- a/check.t
+++ b/check.t
@@ -1,4 +1,4 @@
-# $MirOS: src/bin/mksh/check.t,v 1.67 2006/09/30 02:13:20 tg Exp $
+# $MirOS: src/bin/mksh/check.t,v 1.68 2006/11/05 12:11:13 tg Exp $
 # $OpenBSD: bksl-nl.t,v 1.2 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: history.t,v 1.5 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: read.t,v 1.3 2003/03/10 03:48:16 david Exp $
@@ -3872,5 +3872,5 @@ category: pdksh
 stdin:
 	echo $KSH_VERSION
 expected-stdout:
-	@(#)MIRBSD KSH R28 2006/09/30
+	@(#)MIRBSD KSH R28 2006/11/05
 ---
diff --git a/main.c b/main.c
index 94c81e9..4da821c 100644
--- a/main.c
+++ b/main.c
@@ -6,7 +6,12 @@
 #define	EXTERN
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/main.c,v 1.49 2006/09/30 02:13:21 tg Exp $");
+#if (HAVE_LANGSTUFF - 0)
+#include <langinfo.h>
+#include <locale.h>
+#endif
+
+__RCSID("$MirOS: src/bin/mksh/main.c,v 1.50 2006/11/05 12:11:14 tg Exp $");
 
 extern char **environ;
 
@@ -60,6 +65,9 @@ main(int argc, char *argv[])
 	pid_t ppid;
 	struct tbl *vp;
 	struct stat s_stdin;
+#if (HAVE_LANGSTUFF - 0)
+	const char *cc;
+#endif
 
 	/* make sure argv[] is sane */
 	if (!*argv) {
@@ -134,6 +142,19 @@ main(int argc, char *argv[])
 	setstr(vp, def_path, KSH_RETURN_ERROR);
 
 
+#if (HAVE_LANGSTUFF - 0)
+	/* Check if we're in an UTF-8 locale */
+	cc = setlocale(LC_CTYPE, "");
+#ifdef CODESET
+	if (strcasecmp(cc, "UTF-8") && strcasecmp(cc, "utf8") &&
+	    strcasecmp(cc, "CESU-8") && strcasecmp(cc, "cesu8"))
+		cc = nl_langinfo(CODESET);
+#endif
+	if (!strcasecmp(cc, "UTF-8") || !strcasecmp(cc, "utf8") ||
+	    !strcasecmp(cc, "CESU-8") || !strcasecmp(cc, "cesu8"))
+		Flag(FUTFHACK) = 1;
+#endif
+
 	/* Turn on nohup by default for now - will change to off
 	 * by default once people are aware of its existence
 	 * (at&t ksh does not have a nohup option - it always sends
diff --git a/misc.c b/misc.c
index 8250e18..a94b4a9 100644
--- a/misc.c
+++ b/misc.c
@@ -3,7 +3,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.16 2006/08/24 20:32:53 tg Exp $\t"
+__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.17 2006/11/05 12:11:14 tg Exp $\t"
 	MKSH_SH_H_ID);
 
 short chtypes[UCHAR_MAX+1];	/* type bits for unsigned char */
@@ -142,6 +142,7 @@ const struct option options[] = {
 	{ "sh",		  0,		OF_ANY }, /* non-standard */
 	{ "stdin",	's',	    OF_CMDLINE }, /* pseudo non-standard */
 	{ "trackall",	'h',		OF_ANY },
+	{ "utf8-hack",	  0,		OF_ANY }, /* non-standard */
 	{ "verbose",	'v',		OF_ANY },
 	{ "vi",		  0,		OF_ANY },
 	{ "viraw",	  0,		OF_ANY }, /* no effect */
diff --git a/sh.h b/sh.h
index ac5a014..07575b9 100644
--- a/sh.h
+++ b/sh.h
@@ -8,8 +8,8 @@
 /*	$OpenBSD: c_test.h,v 1.4 2004/12/20 11:34:26 otto Exp $	*/
 /*	$OpenBSD: tty.h,v 1.5 2004/12/20 11:34:26 otto Exp $	*/
 
-#define MKSH_SH_H_ID "$MirOS: src/bin/mksh/sh.h,v 1.47 2006/10/03 19:46:14 tg Exp $"
-#define MKSH_VERSION "R28 2006/09/30"
+#define MKSH_SH_H_ID "$MirOS: src/bin/mksh/sh.h,v 1.48 2006/11/05 12:11:14 tg Exp $"
+#define MKSH_VERSION "R28 2006/11/05"
 
 #include <sys/cdefs.h>
 #if !defined(__RCSID) || !defined(__SCCSID)
@@ -283,6 +283,7 @@ enum sh_flag {
 	FSH,		/* -o sh (dummy, for pdksh compatibility) */
 	FSTDIN,		/* -s: (invocation) parse stdin */
 	FTRACKALL,	/* -h: create tracked aliases for all commands */
+	FUTFHACK,	/* utf-8 hack for command line editing */
 	FVERBOSE,	/* -v: echo input */
 	FVI,		/* vi command editing */
 	FVIRAW,		/* always read in raw mode (ignored) */