From 8c64f063a1603a4396fc3a3a61188d37f00d48ae Mon Sep 17 00:00:00 2001
From: wiz <wiz@NetBSD.org>
Date: Tue, 5 Jun 2007 17:48:19 +0000
Subject: [PATCH] Apply SODA Nuriyuki's patch for multibyte support. Based on
 soda-revised2.patch (in private mail) with a change from soda on
 tech-userlevel.

Ok christos@.
---
 bin/ls/util.c | 87 +++++++++++++++++++++++++++++++++++++++++++++------
 1 file changed, 77 insertions(+), 10 deletions(-)

diff --git a/bin/ls/util.c b/bin/ls/util.c
index dec71d750b2f..9cf586b61f0b 100644
--- a/bin/ls/util.c
+++ b/bin/ls/util.c
@@ -1,4 +1,4 @@
-/*	$NetBSD: util.c,v 1.30 2006/12/14 14:15:26 christos Exp $	*/
+/*	$NetBSD: util.c,v 1.31 2007/06/05 17:48:19 wiz Exp $	*/
 
 /*
  * Copyright (c) 1989, 1993, 1994
@@ -37,14 +37,13 @@
 #if 0
 static char sccsid[] = "@(#)util.c	8.5 (Berkeley) 4/28/95";
 #else
-__RCSID("$NetBSD: util.c,v 1.30 2006/12/14 14:15:26 christos Exp $");
+__RCSID("$NetBSD: util.c,v 1.31 2007/06/05 17:48:19 wiz Exp $");
 #endif
 #endif /* not lint */
 
 #include <sys/types.h>
 #include <sys/stat.h>
 
-#include <ctype.h>
 #include <err.h>
 #include <fts.h>
 #include <limits.h>
@@ -52,6 +51,8 @@ __RCSID("$NetBSD: util.c,v 1.30 2006/12/14 14:15:26 christos Exp $");
 #include <stdlib.h>
 #include <string.h>
 #include <vis.h>
+#include <wchar.h>
+#include <wctype.h>
 
 #include "ls.h"
 #include "extern.h"
@@ -84,17 +85,83 @@ safe_print(const char *src)
 		/* NOTREACHED */
 }
 
+/*
+ * The reasons why we don't use putwchar(wc) here are:
+ * - If wc == L'\0', we need to restore the initial shift state, but
+ *   the C language standard doesn't say that putwchar(L'\0') does.
+ * - It isn't portable to mix a wide-oriented function (i.e. getwchar)
+ *   with byte-oriented functions (printf et al.) in same FILE.
+ */
+static int
+printwc(wchar_t wc, mbstate_t *pst)
+{
+	size_t size;
+	char buf[MB_LEN_MAX];
+
+	size = wcrtomb(buf, wc, pst);
+	if (size == (size_t)-1) /* This shouldn't happen, but for sure */
+		return 0;
+	if (wc == L'\0') {
+		/* The following condition must be always true, but for sure */
+		if (size > 0 && buf[size - 1] == '\0')
+			--size;
+	}
+	if (size > 0)
+		fwrite(buf, 1, size, stdout);
+	return wc == L'\0' ? 0 : wcwidth(wc);
+}
+
 int
 printescaped(const char *src)
 {
-	unsigned char c;
-	int n;
+	int n = 0;
+	mbstate_t src_state, stdout_state;
+	/* The following +1 is to pass '\0' at the end of src to mbrtowc(). */
+	const char *endptr = src + strlen(src) + 1;
 
-	for (n = 0; (c = *src) != '\0'; ++src, ++n)
-		if (isprint(c))
-			(void)putchar(c);
-		else
-			(void)putchar('?');
+	/*
+	 * We have to reset src_state each time in this function, because
+	 * the codeset of src pathname may not match with current locale.
+	 * Note that if we pass NULL instead of src_state to mbrtowc(),
+	 * there is no way to reset the state.
+	 */
+	memset(&src_state, 0, sizeof(src_state));
+	memset(&stdout_state, 0, sizeof(stdout_state));
+	while (src < endptr) {
+		wchar_t wc;
+		size_t rv, span = endptr - src;
+
+#if 0
+/* soda says:
+ * Comment this out, because if there are redundant escape sequences
+ * which exceed 32 bytes, our current implementation doesn't display
+ * the pathname correctly with above.
+ */
+		if (span > MB_CUR_MAX)
+			span = MB_CUR_MAX;
+#endif
+		rv = mbrtowc(&wc, src, span, &src_state);
+		if (rv == 0) { /* assert(wc == L'\0'); */
+			/* The following may output a shift sequence. */
+			n += printwc(wc, &stdout_state);
+			break;
+		}
+		if (rv == (size_t)-1) { /* probably errno == EILSEQ */
+			n += printwc(L'?', &stdout_state);
+			/* try to skip 1byte, because there is no better way */
+			src++;
+			memset(&src_state, 0, sizeof(src_state));
+		} else if (rv == (size_t)-2) {
+			if (span < MB_CUR_MAX) { /* incomplete char */
+				n += printwc(L'?', &stdout_state);
+				break;
+			}
+			src += span; /* a redundant shift sequence? */
+		} else {
+			n += printwc(iswprint(wc) ? wc : L'?', &stdout_state);
+			src += rv;
+		}
+	}
 	return n;
 }