Treat filenames as multibyte character strings (according to the current

LC_CTYPE setting) when determining which characters are printable.
This is an often-requested feature.

Use wcwidth() to determine the number of column positions a character
takes up, although there are still a few places left where we assume
1 byte = 1 column position, e.g. line-wrapping when handling the -m option.

The error handling here is somewhat more complicated than usual: we do
our best to show what we can of a filename in the presence of conversion
errors, instead of simply aborting.

4 files changed, 132 insertions, 65 deletions

diff --git a/bin/ls/extern.h b/bin/ls/extern.h
index 9482dad..52b6a85 100644
--- a/bin/ls/extern.h
+++ b/bin/ls/extern.h
@@ -45,6 +45,7 @@ int	 printname(const char *);
 void	 printscol(const DISPLAY *);
 void	 printstream(const DISPLAY *);
 void	 usage(void);
+int	 prn_normal(const char *);
 size_t	 len_octal(const char *, int);
 int	 prn_octal(const char *);
 int	 prn_printable(const char *);
diff --git a/bin/ls/ls.1 b/bin/ls/ls.1
index eb84484..8554f54 100644
--- a/bin/ls/ls.1
+++ b/bin/ls/ls.1
@@ -31,7 +31,7 @@
 .\"     @(#)ls.1	8.7 (Berkeley) 7/29/94
 .\" $FreeBSD$
 .\"
-.Dd March 21, 2004
+.Dd May 2, 2004
 .Dt LS 1
 .Os
 .Sh NAME
@@ -673,7 +673,3 @@ command appeared in
 .Sh BUGS
 To maintain backward compatibility, the relationships between the many
 options are quite complex.
-.Pp
-The
-.Nm
-utility does not recognize multibyte characters in filenames.
diff --git a/bin/ls/print.c b/bin/ls/print.c
index 1d6bec9..eb0258e 100644
--- a/bin/ls/print.c
+++ b/bin/ls/print.c
@@ -147,7 +147,7 @@ printname(const char *name)
 	else if (f_nonprint)
 		return prn_printable(name);
 	else
-		return printf("%s", name);
+		return prn_normal(name);
 }
 
 void
@@ -239,6 +239,7 @@ printstream(const DISPLAY *dp)
 	for (p = dp->list, chcnt = 0; p; p = p->fts_link) {
 		if (p->fts_number == NO_PRINT)
 			continue;
+		/* XXX strlen does not take octal escapes into account. */
 		if (strlen(p->fts_name) + chcnt +
 		    (p->fts_link ? 2 : 0) >= (unsigned)termwidth) {
 			putchar('\n');
diff --git a/bin/ls/util.c b/bin/ls/util.c
index ea305c1..a7b8208 100644
--- a/bin/ls/util.c
+++ b/bin/ls/util.c
@@ -44,25 +44,81 @@ __FBSDID("$FreeBSD$");
 #include <ctype.h>
 #include <err.h>
 #include <fts.h>
+#include <limits.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <wchar.h>
+#include <wctype.h>
 
 #include "ls.h"
 #include "extern.h"
 
 int
+prn_normal(const char *s)
+{
+	mbstate_t mbs;
+	wchar_t wc;
+	int i, n;
+	size_t clen;
+
+	memset(&mbs, 0, sizeof(mbs));
+	n = 0;
+	while ((clen = mbrtowc(&wc, s, MB_LEN_MAX, &mbs)) != 0) {
+		if (clen == (size_t)-2) {
+			n += printf("%s", s);
+			break;
+		}
+		if (clen == (size_t)-1) {
+			memset(&mbs, 0, sizeof(mbs));
+			putchar((unsigned char)*s);
+			s++;
+			n++;
+			continue;
+		}
+		for (i = 0; i < (int)clen; i++)
+			putchar((unsigned char)s[i]);
+		s += clen;
+		n += wcwidth(wc);
+	}
+	return (n);
+}
+
+int
 prn_printable(const char *s)
 {
-	char c;
-	int n;
+	mbstate_t mbs;
+	wchar_t wc;
+	int i, n;
+	size_t clen;
 
-	for (n = 0; (c = *s) != '\0'; ++s, ++n)
-		if (isprint((unsigned char)c))
-			putchar(c);
-		else
+	memset(&mbs, 0, sizeof(mbs));
+	n = 0;
+	while ((clen = mbrtowc(&wc, s, MB_LEN_MAX, &mbs)) != 0) {
+		if (clen == (size_t)-1) {
+			putchar('?');
+			s++;
+			n++;
+			memset(&mbs, 0, sizeof(mbs));
+			continue;
+		}
+		if (clen == (size_t)-2) {
 			putchar('?');
-	return n;
+			n++;
+			break;
+		}
+		if (!iswprint(wc)) {
+			putchar('?');
+			s += clen;
+			n++;
+			continue;
+		}
+		for (i = 0; i < (int)clen; i++)
+			putchar((unsigned char)s[i]);
+		s += clen;
+		n += wcwidth(wc);
+	}
+	return (n);
 }
 
 /*
@@ -81,70 +137,83 @@ prn_printable(const char *s)
 size_t
 len_octal(const char *s, int len)
 {
-	size_t r = 0;
+	mbstate_t mbs;
+	wchar_t wc;
+	size_t clen, r;
 
-	while (len--)
-		if (isprint((unsigned const char)*s++)) r++; else r += 4;
-	return r;
+	memset(&mbs, 0, sizeof(mbs));
+	r = 0;
+	while (len != 0 && (clen = mbrtowc(&wc, s, len, &mbs)) != 0) {
+		if (clen == (size_t)-1) {
+			r += 4;
+			s++;
+			len--;
+			memset(&mbs, 0, sizeof(mbs));
+			continue;
+		}
+		if (clen == (size_t)-2) {
+			r += 4 * len;
+			break;
+		}
+		if (iswprint(wc))
+			r++;
+		else
+			r += 4 * clen;
+		s += clen;
+	}
+	return (r);
 }
 
 int
 prn_octal(const char *s)
 {
-        unsigned char ch;
-	int len = 0;
+	static const char esc[] = "\\\\\"\"\aa\bb\ff\nn\rr\tt\vv";
+	const char *p;
+	mbstate_t mbs;
+	wchar_t wc;
+	size_t clen;
+	unsigned char ch;
+	int goodchar, i, len, prtlen;
 	
-        while ((ch = (unsigned char)*s++)) {
-	        if (isprint(ch) && (ch != '\"') && (ch != '\\'))
-		        putchar(ch), len++;
-	        else if (f_octal_escape) {
-	                putchar('\\');
-		        switch (ch) {
-			case '\\':
-			        putchar('\\');
-				break;
-			case '\"':
-			        putchar('"');
-				break;
-			case '\a':
-			        putchar('a');
-				break;
-			case '\b':
-			        putchar('b');
-				break;
-			case '\f':
-			        putchar('f');
-				break;
-			case '\n':
-			        putchar('n');
-				break;
-			case '\r':
-			        putchar('r');
-				break;
-			case '\t':
-			        putchar('t');
-				break;
-			case '\v':
-			        putchar('v');
-				break;
- 		        default:
+	memset(&mbs, 0, sizeof(mbs));
+	len = 0;
+	while ((clen = mbrtowc(&wc, s, MB_LEN_MAX, &mbs)) != 0) {
+		goodchar = clen != (size_t)-1 && clen != (size_t)-2;
+		if (goodchar && iswprint(wc) && wc != L'\"' && wc != L'\\') {
+			for (i = 0; i < (int)clen; i++)
+				putchar((unsigned char)s[i]);
+			len += wcwidth(wc);
+		} else if (goodchar && f_octal_escape && wc >= 0 &&
+		    wc <= (wchar_t)UCHAR_MAX &&
+		    (p = strchr(esc, (char)wc)) != NULL) {
+			putchar('\\');
+			putchar(p[1]);
+			len += 2;
+		} else {
+			if (goodchar)
+				prtlen = clen;
+			else if (clen == (size_t)-1)
+				prtlen = 1;
+			else
+				prtlen = strlen(s);
+			for (i = 0; i < prtlen; i++) {
+				ch = (unsigned char)s[i];
+				putchar('\\');
 		                putchar('0' + (ch >> 6));
 		                putchar('0' + ((ch >> 3) & 7));
 		                putchar('0' + (ch & 7));
-		                len += 2;
-			        break;
-		        }
-		        len += 2;
-	        }
-		else {
-			putchar('\\');
-	                putchar('0' + (ch >> 6));
-	                putchar('0' + ((ch >> 3) & 7));
-	                putchar('0' + (ch & 7));
-	                len += 4;
+				len += 4;
+			}
 		}
+		if (clen == (size_t)-2)
+			break;
+		if (clen == (size_t)-1) {
+			memset(&mbs, 0, sizeof(mbs));
+			s++;
+		} else
+			s += clen;
 	}
-	return len;
+	return (len);
 }
 
 void