further work on unicodization

Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>

7 files changed, 146 insertions, 41 deletions

diff --git a/TODO_unicode b/TODO_unicode
new file mode 100644
index 0000000..c29fd93
--- /dev/null
+++ b/TODO_unicode
@@ -0,0 +1,45 @@
+Already fixed applets:
+cal
+lsmod
+df
+dumpleases
+
+Applets which may need unicode handling (more extensive than sanitizing
+of filenames in error messages):
+
+ls - uses unicode_strlen, not scrlen
+expand, unexpand - uses unicode_strlen, not scrlen
+ash, hush through lineedit - uses unicode_strlen, not scrlen
+top - need to sanitize process args
+ps - need to sanitize process args
+less
+more
+vi
+ed
+cut
+awk
+sed
+tr
+grep egrep fgrep
+fold
+sort
+head, tail
+catv - "display nonprinting chars" - what this could mean for unicode?
+wc
+chat
+dumpkmap
+last - just line up columns
+man
+microcom
+strings
+watch
+
+Unsure, may need fixing:
+
+hostname - do we really want to protect against bad chars in it?
+patch
+addgroup, adduser, delgroup, deluser
+telnet
+telnetd
+od
+printf
diff --git a/coreutils/cal.c b/coreutils/cal.c
index 5ecb913..207fa96 100644
--- a/coreutils/cal.c
+++ b/coreutils/cal.c
@@ -135,7 +135,7 @@ int cal_main(int argc UNUSED_PARAM, char **argv)
 			if (julian)
 				*hp++ = ' ';
 			{
-				char *two_wchars = unicode_cut_nchars(2, buf);
+				char *two_wchars = unicode_conv_to_printable_fixedwidth(NULL, buf, 2);
 				strcpy(hp, two_wchars);
 				free(two_wchars);
 			}
diff --git a/coreutils/df.c b/coreutils/df.c
index ae68f08..4b23faa 100644
--- a/coreutils/df.c
+++ b/coreutils/df.c
@@ -114,9 +114,6 @@ int df_main(int argc UNUSED_PARAM, char **argv)
 	while (1) {
 		const char *device;
 		const char *mount_point;
-#if ENABLE_FEATURE_ASSUME_UNICODE
-		size_t dev_len;
-#endif
 
 		if (mount_table) {
 			mount_entry = getmntent(mount_table);
@@ -178,11 +175,15 @@ int df_main(int argc UNUSED_PARAM, char **argv)
 #endif
 
 #if ENABLE_FEATURE_ASSUME_UNICODE
-			dev_len = unicode_strlen(device);
-			if (dev_len > 20) {
-				printf("%s\n%20s", device, "");
-			} else {
-				printf("%s%*s", device, 20 - (int)dev_len, "");
+			{
+				uni_stat_t uni_stat;
+				char *uni_dev = unicode_conv_to_printable(&uni_stat, device);
+				if (uni_stat.unicode_width > 20) {
+					printf("%s\n%20s", uni_dev, "");
+				} else {
+					printf("%s%*s", uni_dev, 20 - (int)uni_stat.unicode_width, "");
+				}
+				free(uni_dev);
 			}
 #else
 			if (printf("\n%-20s" + 1, device) > 20)
diff --git a/include/unicode.h b/include/unicode.h
index f1a252c..f32e565 100644
--- a/include/unicode.h
+++ b/include/unicode.h
@@ -14,15 +14,25 @@ enum {
 #if !ENABLE_FEATURE_ASSUME_UNICODE
 
 # define unicode_strlen(string) strlen(string)
-# define unicode_scrlen(string) TODO
 # define unicode_status UNICODE_OFF
 # define init_unicode() ((void)0)
 
 #else
 
 size_t FAST_FUNC unicode_strlen(const char *string);
-char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src);
-unsigned FAST_FUNC unicode_padding_to_width(unsigned width, const char *src);
+enum {
+	UNI_FLAG_PAD = (1 << 0),
+};
+typedef struct uni_stat_t {
+	unsigned byte_count;
+	unsigned unicode_count;
+	unsigned unicode_width;
+} uni_stat_t;
+//UNUSED: unsigned FAST_FUNC unicode_padding_to_width(unsigned width, const char *src);
+//UNUSED: char* FAST_FUNC unicode_conv_to_printable2(uni_stat_t *stats, const char *src, unsigned width, int flags);
+char* FAST_FUNC unicode_conv_to_printable(uni_stat_t *stats, const char *src);
+char* FAST_FUNC unicode_conv_to_printable_maxwidth(uni_stat_t *stats, const char *src, unsigned maxwidth);
+char* FAST_FUNC unicode_conv_to_printable_fixedwidth(uni_stat_t *stats, const char *src, unsigned width);
 
 # if ENABLE_LOCALE_SUPPORT
 
diff --git a/libbb/unicode.c b/libbb/unicode.c
index 878af84..4e7e3a9 100644
--- a/libbb/unicode.c
+++ b/libbb/unicode.c
@@ -246,29 +246,45 @@ size_t FAST_FUNC unicode_strlen(const char *string)
 	return width;
 }
 
-char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src)
+static char* FAST_FUNC unicode_conv_to_printable2(uni_stat_t *stats, const char *src, unsigned width, int flags)
 {
 	char *dst;
 	unsigned dst_len;
+	unsigned uni_count;
+	unsigned uni_width;
 
 	if (unicode_status != UNICODE_ON) {
-		char *d = dst = xmalloc(width + 1);
-		while ((int)--width >= 0) {
-			unsigned char c = *src;
-			if (c == '\0') {
-				do
-					*d++ = ' ';
-				while ((int)--width >= 0);
-				break;
+		char *d;
+		if (flags & UNI_FLAG_PAD) {
+			d = dst = xmalloc(width + 1);
+			while ((int)--width >= 0) {
+				unsigned char c = *src;
+				if (c == '\0') {
+					do
+						*d++ = ' ';
+					while ((int)--width >= 0);
+					break;
+				}
+				*d++ = (c >= ' ' && c < 0x7f) ? c : '?';
+				src++;
+			}
+			*d = '\0';
+		} else {
+			d = dst = xstrndup(src, width);
+			while (*d) {
+				unsigned char c = *d;
+				if (c < ' ' || c >= 0x7f)
+					*d = '?';
+				d++;
 			}
-			*d++ = (c >= ' ' && c < 0x7f) ? c : '?';
-			src++;
 		}
-		*d = '\0';
+		if (stats)
+			stats->byte_count = stats->unicode_count = (d - dst);
 		return dst;
 	}
 
 	dst = NULL;
+	uni_count = uni_width = 0;
 	dst_len = 0;
 	while (1) {
 		int w;
@@ -301,7 +317,7 @@ char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src)
 			/* src = NULL: invalid sequence is seen,
 			 * else: wc is set, src is advanced to next mb char
 			 */
-			if (src1) {/* no error */
+			if (src1) { /* no error */
 				if (wc == 0) /* end-of-string */
 					break;
 				src = src1;
@@ -315,8 +331,8 @@ char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src)
 			goto subst;
 		w = wcwidth(wc);
 		if ((ENABLE_UNICODE_COMBINING_WCHARS && w < 0) /* non-printable wchar */
-		 || (!ENABLE_UNICODE_COMBINING_WCHARS && wc <= 0)
-		 || (!ENABLE_UNICODE_WIDE_WCHARS && wc > 1)
+		 || (!ENABLE_UNICODE_COMBINING_WCHARS && w <= 0)
+		 || (!ENABLE_UNICODE_WIDE_WCHARS && w > 1)
 		) {
  subst:
 			wc = CONFIG_SUBST_WCHAR;
@@ -331,6 +347,8 @@ char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src)
 			break;
 		}
 
+		uni_count++;
+		uni_width += w;
 		dst = xrealloc(dst, dst_len + MB_CUR_MAX);
 #if ENABLE_LOCALE_SUPPORT
 		{
@@ -343,15 +361,37 @@ char* FAST_FUNC unicode_cut_nchars(unsigned width, const char *src)
 	}
 
 	/* Pad to remaining width */
-	dst = xrealloc(dst, dst_len + width + 1);
-	while ((int)--width >= 0) {
-		dst[dst_len++] = ' ';
+	if (flags & UNI_FLAG_PAD) {
+		dst = xrealloc(dst, dst_len + width + 1);
+		uni_count += width;
+		uni_width += width;
+		while ((int)--width >= 0) {
+			dst[dst_len++] = ' ';
+		}
 	}
 	dst[dst_len] = '\0';
+	if (stats) {
+		stats->byte_count = dst_len;
+		stats->unicode_count = uni_count;
+		stats->unicode_width = uni_width;
+	}
 
 	return dst;
 }
+char* FAST_FUNC unicode_conv_to_printable(uni_stat_t *stats, const char *src)
+{
+	return unicode_conv_to_printable2(stats, src, INT_MAX, 0);
+}
+char* FAST_FUNC unicode_conv_to_printable_maxwidth(uni_stat_t *stats, const char *src, unsigned maxwidth)
+{
+	return unicode_conv_to_printable2(stats, src, maxwidth, 0);
+}
+char* FAST_FUNC unicode_conv_to_printable_fixedwidth(uni_stat_t *stats, const char *src, unsigned width)
+{
+	return unicode_conv_to_printable2(stats, src, width, UNI_FLAG_PAD);
+}
 
+#ifdef UNUSED
 unsigned FAST_FUNC unicode_padding_to_width(unsigned width, const char *src)
 {
 	if (unicode_status != UNICODE_ON) {
@@ -382,3 +422,4 @@ unsigned FAST_FUNC unicode_padding_to_width(unsigned width, const char *src)
 			return 0;
 	}
 }
+#endif
diff --git a/modutils/lsmod.c b/modutils/lsmod.c
index cc6b616..50621c2 100644
--- a/modutils/lsmod.c
+++ b/modutils/lsmod.c
@@ -46,9 +46,6 @@ int lsmod_main(int argc UNUSED_PARAM, char **argv UNUSED_PARAM)
 #if ENABLE_FEATURE_LSMOD_PRETTY_2_6_OUTPUT
 	char *token[4];
 	parser_t *parser = config_open("/proc/modules");
-# if ENABLE_FEATURE_ASSUME_UNICODE
-	size_t name_len;
-# endif
 	init_unicode();
 
 	printf("%-24sSize  Used by", "Module");
@@ -64,9 +61,13 @@ int lsmod_main(int argc UNUSED_PARAM, char **argv UNUSED_PARAM)
 			} else
 				token[3] = (char *) "";
 # if ENABLE_FEATURE_ASSUME_UNICODE
-			name_len = unicode_strlen(token[0]);
-			name_len = (name_len > 19) ? 0 : 19 - name_len;
-			printf("%s%*s %8s %2s %s\n", token[0], name_len, "", token[1], token[2], token[3]);
+			{
+				uni_stat_t uni_stat;
+				char *uni_name = unicode_conv_to_printable(&uni_stat, token[0]);
+				unsigned pad_len = (uni_stat.unicode_width > 19) ? 0 : 19 - uni_stat.unicode_width;
+				printf("%s%*s %8s %2s %s\n", uni_name, pad_len, "", token[1], token[2], token[3]);
+				free(uni_name);
+			}
 # else
 			printf("%-19s %8s %2s %s\n", token[0], token[1], token[2], token[3]);
 # endif
@@ -78,9 +79,13 @@ int lsmod_main(int argc UNUSED_PARAM, char **argv UNUSED_PARAM)
 			// so trimming the trailing char is just what we need!
 			token[3][strlen(token[3])-1] = '\0';
 # if ENABLE_FEATURE_ASSUME_UNICODE
-			name_len = unicode_strlen(token[0]);
-			name_len = (name_len > 19) ? 0 : 19 - name_len;
-			printf("%s%*s %8s %2s %s\n", token[0], name_len, "", token[1], token[2], token[3]);
+			{
+				uni_stat_t uni_stat;
+				char *uni_name = unicode_conv_to_printable(&uni_stat, token[0]);
+				unsigned pad_len = (uni_stat.unicode_width > 19) ? 0 : 19 - uni_stat.unicode_width;
+				printf("%s%*s %8s %2s %s\n", uni_name, pad_len, "", token[1], token[2], token[3]);
+				free(uni_name);
+			}
 # else
 			printf("%-19s %8s %2s %s\n", token[0], token[1], token[2], token[3]);
 # endif
diff --git a/networking/udhcp/dumpleases.c b/networking/udhcp/dumpleases.c
index d8f5da7..eab9713 100644
--- a/networking/udhcp/dumpleases.c
+++ b/networking/udhcp/dumpleases.c
@@ -71,8 +71,11 @@ int dumpleases_main(int argc UNUSED_PARAM, char **argv)
 		/* actually, 15+1 and 19+1, +1 is a space between columns */
 		/* lease.hostname is char[20] and is always NUL terminated */
 #if ENABLE_FEATURE_ASSUME_UNICODE
-		printf(" %-16s%s%*s", inet_ntoa(addr), lease.hostname,
-			20 - (int)unicode_strlen(lease.hostname), "");
+		{
+			char *uni_name = unicode_conv_to_printable_fixedwidth(NULL, lease.hostname, 20);
+			printf(" %-16s%s", inet_ntoa(addr), uni_name);
+			free(uni_name);
+		}
 #else
 		printf(" %-16s%-20s", inet_ntoa(addr), lease.hostname);
 #endif