All these return values from utf8_* are confusing, use an enum.

input-keys.c

@@ -169,7 +169,7 @@ input_key(struct window_pane *wp, key_code key, struct mouse_event *m)
 		return;
 	}
 	if (key != KEYC_NONE && justkey > 0x7f && justkey < KEYC_BASE) {
-		if (utf8_split(justkey, &ud) != 0)
+		if (utf8_split(justkey, &ud) != UTF8_DONE)
 			return;
 		if (key & KEYC_ESCAPE)
 			bufferevent_write(wp->event, "\033", 1);

input.c

@@ -1923,7 +1923,7 @@ input_utf8_open(struct input_ctx *ictx)
 {
 	struct utf8_data	*ud = &ictx->utf8data;
 
-	if (!utf8_open(ud, ictx->ch))
+	if (utf8_open(ud, ictx->ch) != UTF8_MORE)
 		log_fatalx("UTF-8 open invalid %#hhx", ictx->ch);
 
 	log_debug("%s %hhu", __func__, ud->size);
@@ -1937,7 +1937,7 @@ input_utf8_add(struct input_ctx *ictx)
 {
 	struct utf8_data	*ud = &ictx->utf8data;
 
-	if (utf8_append(ud, ictx->ch) != 1)
+	if (utf8_append(ud, ictx->ch) != UTF8_MORE)
 		log_fatalx("UTF-8 add invalid %#hhx", ictx->ch);
 
 	log_debug("%s", __func__);
@@ -1951,7 +1951,7 @@ input_utf8_close(struct input_ctx *ictx)
 {
 	struct utf8_data	*ud = &ictx->utf8data;
 
-	if (utf8_append(ud, ictx->ch) != 0)
+	if (utf8_append(ud, ictx->ch) != UTF8_DONE)
 		log_fatalx("UTF-8 close invalid %#hhx", ictx->ch);
 
 	log_debug("%s %hhu '%*s' (width %hhu)", __func__, ud->size,

key-string.c

@@ -144,10 +144,11 @@ key_string_lookup_string(const char *string)
 	static const char	*other = "!#()+,-.0123456789:;<=>?'\r\t";
 	key_code		 key;
 	u_short			 u;
-	int			 size, more;
+	int			 size;
 	key_code		 modifiers;
 	struct utf8_data	 ud;
 	u_int			 i;
+	enum utf8_state		 more;
 
 	/* Is this a hexadecimal value? */
 	if (string[0] == '0' && string[1] == 'x') {
@@ -173,13 +174,12 @@ key_string_lookup_string(const char *string)
 			return (KEYC_NONE);
 	} else {
 		/* Try as a UTF-8 key. */
-		if (utf8_open(&ud, (u_char)*string)) {
+		if ((more = utf8_open(&ud, (u_char)*string)) == UTF8_MORE) {
 			if (strlen(string) != ud.size)
 				return (KEYC_NONE);
-			more = 1;
 			for (i = 1; i < ud.size; i++)
 				more = utf8_append(&ud, (u_char)string[i]);
-			if (more != 0)
+			if (more != UTF8_DONE)
 				return (KEYC_NONE);
 			key = utf8_combine(&ud);
 			return (key | modifiers);
@@ -256,7 +256,7 @@ key_string_lookup_key(key_code key)
 
 	/* Is this a UTF-8 key? */
 	if (key > 127 && key < KEYC_BASE) {
-		if (utf8_split(key, &ud) == 0) {
+		if (utf8_split(key, &ud) == UTF8_DONE) {
 			memcpy(out, ud.data, ud.size);
 			out[ud.size] = '\0';
 			return (out);

screen-write.c

@@ -115,7 +115,7 @@ screen_write_strlen(const char *fmt, ...)
 	struct utf8_data	ud;
 	u_char 	      	       *ptr;
 	size_t			left, size = 0;
-	int			more;
+	enum utf8_state		more;
 
 	va_start(ap, fmt);
 	xvasprintf(&msg, fmt, ap);
@@ -123,17 +123,17 @@ screen_write_strlen(const char *fmt, ...)
 
 	ptr = msg;
 	while (*ptr != '\0') {
-		if (*ptr > 0x7f && utf8_open(&ud, *ptr)) {
+		if (*ptr > 0x7f && utf8_open(&ud, *ptr) == UTF8_MORE) {
 			ptr++;
 
 			left = strlen(ptr);
 			if (left < (size_t)ud.size - 1)
 				break;
-			while ((more = utf8_append(&ud, *ptr)) == 1)
+			while ((more = utf8_append(&ud, *ptr)) == UTF8_MORE)
 				ptr++;
 			ptr++;
 
-			if (more == 0)
+			if (more == UTF8_DONE)
 				size += ud.width;
 		} else {
 			if (*ptr > 0x1f && *ptr < 0x7f)
@@ -178,23 +178,23 @@ screen_write_vnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
 	struct utf8_data	ud;
 	u_char 		       *ptr;
 	size_t		 	left, size = 0;
-	int			more;
+	enum utf8_state		more;
 
 	xvasprintf(&msg, fmt, ap);
 
 	ptr = msg;
 	while (*ptr != '\0') {
-		if (*ptr > 0x7f && utf8_open(&ud, *ptr)) {
+		if (*ptr > 0x7f && utf8_open(&ud, *ptr) == UTF8_MORE) {
 			ptr++;
 
 			left = strlen(ptr);
 			if (left < (size_t)ud.size - 1)
 				break;
-			while ((more = utf8_append(&ud, *ptr)) == 1)
+			while ((more = utf8_append(&ud, *ptr)) == UTF8_MORE)
 				ptr++;
 			ptr++;
 
-			if (more == 0) {
+			if (more == UTF8_DONE) {
 				if (maxlen > 0 &&
 				    size + ud.width > (size_t) maxlen) {
 					while (size < (size_t) maxlen) {
@@ -236,7 +236,7 @@ screen_write_cnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
 	char			*msg;
 	u_char 			*ptr, *last;
 	size_t			 left, size = 0;
-	int			 more;
+	enum utf8_state		 more;
 
 	va_start(ap, fmt);
 	xvasprintf(&msg, fmt, ap);
@@ -260,17 +260,17 @@ screen_write_cnputs(struct screen_write_ctx *ctx, ssize_t maxlen,
 			continue;
 		}
 
-		if (*ptr > 0x7f && utf8_open(&ud, *ptr)) {
+		if (*ptr > 0x7f && utf8_open(&ud, *ptr) == UTF8_MORE) {
 			ptr++;
 
 			left = strlen(ptr);
 			if (left < (size_t)ud.size - 1)
 				break;
-			while ((more = utf8_append(&ud, *ptr)) == 1)
+			while ((more = utf8_append(&ud, *ptr)) == UTF8_MORE)
 				ptr++;
 			ptr++;
 
-			if (more == 0) {
+			if (more == UTF8_DONE) {
 				if (maxlen > 0 &&
 				    size + ud.width > (size_t) maxlen) {
 					while (size < (size_t) maxlen) {

tmux.h

@@ -626,6 +626,11 @@ struct utf8_data {
 
 	u_char	width;	/* 0xff if invalid */
 } __packed;
+enum utf8_state {
+	UTF8_MORE,
+	UTF8_DONE,
+	UTF8_ERROR
+};
 
 /* Grid attributes. */
 #define GRID_ATTR_BRIGHT 0x1
@@ -2191,10 +2196,10 @@ void		 session_renumber_windows(struct session *);
 u_int		 utf8_width(u_int);
 void		 utf8_set(struct utf8_data *, u_char);
 void		 utf8_copy(struct utf8_data *, const struct utf8_data *);
-int		 utf8_open(struct utf8_data *, u_char);
+enum utf8_state	 utf8_open(struct utf8_data *, u_char);
-int		 utf8_append(struct utf8_data *, u_char);
+enum utf8_state	 utf8_append(struct utf8_data *, u_char);
 u_int		 utf8_combine(const struct utf8_data *);
-int		 utf8_split(u_int, struct utf8_data *);
+enum utf8_state	 utf8_split(u_int, struct utf8_data *);
 u_int		 utf8_split2(u_int, u_char *);
 int		 utf8_strvis(char *, const char *, size_t, int);
 char		*utf8_sanitize(const char *);

tty-keys.c

@@ -472,9 +472,10 @@ tty_keys_next(struct tty *tty)
 	const char		*buf;
 	size_t			 len, size;
 	cc_t			 bspace;
-	int			 delay, expired = 0, more;
+	int			 delay, expired = 0;
 	key_code		 key;
 	struct utf8_data	 ud;
+	enum utf8_state		 more;
 	u_int			 i;
 
 	/* Get key buffer. */
@@ -539,17 +540,16 @@ first_key:
 	}
 
 	/* Is this valid UTF-8? */
-	if (utf8_open(&ud, (u_char)*buf)) {
+	if ((more = utf8_open(&ud, (u_char)*buf) == UTF8_MORE)) {
 		size = ud.size;
 		if (len < size) {
 			if (expired)
 				goto discard_key;
 			goto partial_key;
 		}
-		more = 1;
 		for (i = 1; i < size; i++)
 			more = utf8_append(&ud, (u_char)buf[i]);
-		if (more != 0)
+		if (more != UTF8_DONE)
 			goto discard_key;
 		key = utf8_combine(&ud);
 		log_debug("UTF-8 key %.*s %#llx", (int)size, buf, key);
@@ -656,7 +656,7 @@ tty_keys_mouse(struct tty *tty, const char *buf, size_t len, size_t *size)
 	struct utf8_data	 ud;
 	u_int			 i, value, x, y, b, sgr_b;
 	u_char			 sgr_type, c;
-	int			 more;
+	enum utf8_state		 more;
 
 	/*
 	 * Standard mouse sequences are \033[M followed by three characters
@@ -697,14 +697,14 @@ tty_keys_mouse(struct tty *tty, const char *buf, size_t len, size_t *size)
 				return (1);
 
 			if (tty->mode & MODE_MOUSE_UTF8) {
-				if (utf8_open(&ud, buf[*size])) {
+				if (utf8_open(&ud, buf[*size]) == UTF8_MORE) {
 					if (ud.size != 2)
 						return (-1);
 					(*size)++;
 					if (len <= *size)
 						return (1);
 					more = utf8_append(&ud, buf[*size]);
-					if (more != 0)
+					if (more != UTF8_DONE)
 						return (-1);
 					value = utf8_combine(&ud);
 				} else

utf8.c

@@ -381,10 +381,8 @@ utf8_copy(struct utf8_data *to, const struct utf8_data *from)
  * 11000010-11011111 C2-DF start of 2-byte sequence
  * 11100000-11101111 E0-EF start of 3-byte sequence
  * 11110000-11110100 F0-F4 start of 4-byte sequence
- *
- * Returns 1 if more UTF-8 to come, 0 if not UTF-8.
  */
-int
+enum utf8_state
 utf8_open(struct utf8_data *ud, u_char ch)
 {
 	memset(ud, 0, sizeof *ud);
@@ -395,18 +393,13 @@ utf8_open(struct utf8_data *ud, u_char ch)
 	else if (ch >= 0xf0 && ch <= 0xf4)
 		ud->size = 4;
 	else
-		return (0);
+		return (UTF8_ERROR);
 	utf8_append(ud, ch);
-	return (1);
+	return (UTF8_MORE);
 }
 
-/*
+/* Append character to UTF-8, closing if finished. */
- * Append character to UTF-8, closing if finished.
+enum utf8_state
- *
- * Returns 1 if more UTF-8 data to come, 0 if finished and valid, -1 if
- * finished and invalid.
- */
-int
 utf8_append(struct utf8_data *ud, u_char ch)
 {
 	if (ud->have >= ud->size)
@@ -419,12 +412,12 @@ utf8_append(struct utf8_data *ud, u_char ch)
 
 	ud->data[ud->have++] = ch;
 	if (ud->have != ud->size)
-		return (1);
+		return (UTF8_MORE);
 
 	if (ud->width == 0xff)
-		return (-1);
+		return (UTF8_ERROR);
 	ud->width = utf8_width(utf8_combine(ud));
-	return (0);
+	return (UTF8_DONE);
 }
 
 /* Build UTF-8 width tree. */
@@ -501,7 +494,7 @@ utf8_combine(const struct utf8_data *ud)
 }
 
 /* Split 32-bit Unicode into UTF-8. */
-int
+enum utf8_state
 utf8_split(u_int uc, struct utf8_data *ud)
 {
 	if (uc < 0x7f) {
@@ -523,9 +516,9 @@ utf8_split(u_int uc, struct utf8_data *ud)
 		ud->data[2] = 0x80 | ((uc >> 6) & 0x3f);
 		ud->data[3] = 0x80 | (uc & 0x3f);
 	} else
-		return (-1);
+		return (UTF8_ERROR);
 	ud->width = utf8_width(uc);
-	return (0);
+	return (UTF8_DONE);
 }
 
 /* Split a two-byte UTF-8 character. */
@@ -551,26 +544,24 @@ utf8_strvis(char *dst, const char *src, size_t len, int flag)
 {
 	struct utf8_data	 ud;
 	const char		*start, *end;
-	int			 more;
+	enum utf8_state		 more;
 	size_t			 i;
 
 	start = dst;
 	end = src + len;
 
 	while (src < end) {
-		if (utf8_open(&ud, *src)) {
+		if ((more = utf8_open(&ud, *src)) == UTF8_MORE) {
-			more = 1;
+			while (++src < end && more == UTF8_MORE)
-			while (++src < end && more == 1)
 				more = utf8_append(&ud, *src);
-			if (more == 0) {
+			if (more == UTF8_DONE) {
 				/* UTF-8 character finished. */
 				for (i = 0; i < ud.size; i++)
 					*dst++ = ud.data[i];
 				continue;
-			} else if (ud.have > 0) {
-				/* Not a complete, valid UTF-8 character. */
-				src -= ud.have;
 			}
+			/* Not a complete, valid UTF-8 character. */
+			src -= ud.have;
 		}
 		if (src < end - 1)
 			dst = vis(dst, src[0], flag, src[1]);
@@ -593,7 +584,7 @@ utf8_sanitize(const char *src)
 {
 	char			*dst;
 	size_t			 n;
-	int			 more;
+	enum utf8_state		 more;
 	struct utf8_data	 ud;
 	u_int			 i;
 
@@ -602,11 +593,10 @@ utf8_sanitize(const char *src)
 	n = 0;
 	while (*src != '\0') {
 		dst = xreallocarray(dst, n + 1, sizeof *dst);
-		if (utf8_open(&ud, *src)) {
+		if ((more = utf8_open(&ud, *src)) == UTF8_MORE) {
-			more = 1;
+			while (*++src != '\0' && more == UTF8_MORE)
-			while (*++src != '\0' && more == 1)
 				more = utf8_append(&ud, *src);
-			if (more != 1) {
+			if (more == UTF8_DONE) {
 				dst = xreallocarray(dst, n + ud.width,
 				    sizeof *dst);
 				for (i = 0; i < ud.width; i++)
@@ -617,6 +607,8 @@ utf8_sanitize(const char *src)
 		}
 		if (*src > 0x1f && *src < 0x7f)
 			dst[n++] = *src;
+		else
+			dst[n++] = '_';
 		src++;
 	}
 
@@ -634,27 +626,24 @@ utf8_fromcstr(const char *src)
 {
 	struct utf8_data	*dst;
 	size_t			 n;
-	int			 more;
+	enum utf8_state		 more;
 
 	dst = NULL;
 
 	n = 0;
 	while (*src != '\0') {
 		dst = xreallocarray(dst, n + 1, sizeof *dst);
-		if (utf8_open(&dst[n], *src)) {
+		if ((more = utf8_open(&dst[n], *src)) == UTF8_MORE) {
-			more = 1;
+			while (*++src != '\0' && more == UTF8_MORE)
-			while (*++src != '\0' && more == 1)
 				more = utf8_append(&dst[n], *src);
-			if (more != 1) {
+			if (more == UTF8_DONE) {
 				n++;
 				continue;
 			}
 			src -= dst[n].have;
 		}
-		if (*src > 0x1f && *src < 0x7f) {
+		utf8_set(&dst[n], *src);
-			utf8_set(&dst[n], *src);
+		n++;
-			n++;
-		}
 		src++;
 	}
 
@@ -690,21 +679,20 @@ utf8_cstrwidth(const char *s)
 {
 	struct utf8_data	tmp;
 	u_int			width;
-	int			more;
+	enum utf8_state		more;
 
 	width = 0;
 	while (*s != '\0') {
-		if (utf8_open(&tmp, *s)) {
+		if ((more = utf8_open(&tmp, *s)) == UTF8_MORE) {
-			more = 1;
+			while (*++s != '\0' && more == UTF8_MORE)
-			while (*++s != '\0' && more == 1)
 				more = utf8_append(&tmp, *s);
-			if (more != 1) {
+			if (more == UTF8_DONE) {
 				width += tmp.width;
 				continue;
 			}
 			s -= tmp.have;
 		}
-		if (*s > 0x1f && *s < 0x7f)
+		if (*s > 0x1f && *s != 0x7f)
 			width++;
 		s++;
 	}