Merge branch 'obsd-master'

13 files changed, 187 insertions, 224 deletions

diff --git a/arguments.c b/arguments.c
index 962f3c6a..9466f8f4 100644
--- a/arguments.c
+++ b/arguments.c
@@ -406,11 +406,11 @@ args_string_percentage(const char *value, long long minval, long long maxval,
 		}
 		ll = (curval * ll) / 100;
 		if (ll < minval) {
-			*cause = xstrdup("too large");
+			*cause = xstrdup("too small");
 			return (0);
 		}
 		if (ll > maxval) {
-			*cause = xstrdup("too small");
+			*cause = xstrdup("too large");
 			return (0);
 		}
 	} else {
diff --git a/cmd-parse.y b/cmd-parse.y
index 891f2289..9f36af7e 100644
--- a/cmd-parse.y
+++ b/cmd-parse.y
@@ -26,6 +26,7 @@
 #include <stdlib.h>
 #include <string.h>
 #include <unistd.h>
+#include <wchar.h>
 
 #include "tmux.h"
 
@@ -1251,10 +1252,9 @@ error:
 static int
 yylex_token_escape(char **buf, size_t *len)
 {
-	int			 ch, type, o2, o3;
-	u_int			 size, i, tmp;
-	char			 s[9];
-	struct utf8_data	 ud;
+	int	 ch, type, o2, o3, mlen;
+	u_int	 size, i, tmp;
+	char	 s[9], m[MB_LEN_MAX];
 
 	ch = yylex_getc();
 
@@ -1339,11 +1339,12 @@ unicode:
 		yyerror("invalid \\%c argument", type);
 		return (0);
 	}
-	if (utf8_split(tmp, &ud) != UTF8_DONE) {
+	mlen = wctomb(m, tmp);
+	if (mlen <= 0 || mlen > (int)sizeof m) {
 		yyerror("invalid \\%c argument", type);
 		return (0);
 	}
-	yylex_append(buf, len, ud.data, ud.size);
+	yylex_append(buf, len, m, mlen);
 	return (1);
 }
 
diff --git a/cmd-resize-pane.c b/cmd-resize-pane.c
index 563c95fb..94e060d0 100644
--- a/cmd-resize-pane.c
+++ b/cmd-resize-pane.c
@@ -117,7 +117,7 @@ cmd_resize_pane_exec(struct cmd *self, struct cmdq_item *item)
 	if (args_has(args, 'y')) {
 		y = args_percentage(args, 'y', 0, INT_MAX, w->sy, &cause);
 		if (cause != NULL) {
-			cmdq_error(item, "width %s", cause);
+			cmdq_error(item, "height %s", cause);
 			free(cause);
 			return (CMD_RETURN_ERROR);
 		}
diff --git a/cmd-send-keys.c b/cmd-send-keys.c
index a9ecc807..67af718a 100644
--- a/cmd-send-keys.c
+++ b/cmd-send-keys.c
@@ -91,8 +91,8 @@ cmd_send_keys_inject_string(struct cmdq_item *item, struct cmdq_item *after,
     struct args *args, int i)
 {
 	const char		*s = args->argv[i];
-	struct utf8_data	*ud, *uc;
-	wchar_t			 wc;
+	struct utf8_data	*ud, *loop;
+	utf8_char		 uc;
 	key_code		 key;
 	char			*endptr;
 	long			 n;
@@ -117,10 +117,10 @@ cmd_send_keys_inject_string(struct cmdq_item *item, struct cmdq_item *after,
 	}
 	if (literal) {
 		ud = utf8_fromcstr(s);
-		for (uc = ud; uc->size != 0; uc++) {
-			if (utf8_combine(uc, &wc) != UTF8_DONE)
+		for (loop = ud; loop->size != 0; loop++) {
+			if (utf8_from_data(loop, &uc) != UTF8_DONE)
 				continue;
-			after = cmd_send_keys_inject_key(item, after, wc);
+			after = cmd_send_keys_inject_key(item, after, uc);
 		}
 		free(ud);
 	}
diff --git a/grid.c b/grid.c
index 74b6b7b8..acd1f475 100644
--- a/grid.c
+++ b/grid.c
@@ -76,7 +76,7 @@ grid_need_extended_cell(const struct grid_cell_entry *gce,
 		return (1);
 	if (gc->attr > 0xff)
 		return (1);
-	if (gc->data.size != 1 || gc->data.width != 1)
+	if (gc->data.size > 1 || gc->data.width > 1)
 		return (1);
 	if ((gc->fg & COLOUR_FLAG_RGB) || (gc->bg & COLOUR_FLAG_RGB))
 		return (1);
@@ -114,7 +114,7 @@ grid_extended_cell(struct grid_line *gl, struct grid_cell_entry *gce,
 	gl->flags |= GRID_LINE_EXTENDED;
 
 	gee = &gl->extddata[gce->offset];
-	gee->data = utf8_map_big(&gc->data);
+	utf8_from_data(&gc->data, &gee->data);
 	gee->attr = gc->attr;
 	gee->flags = flags;
 	gee->fg = gc->fg;
@@ -496,7 +496,8 @@ grid_get_cell1(struct grid_line *gl, u_int px, struct grid_cell *gc)
 			gc->fg = gee->fg;
 			gc->bg = gee->bg;
 			gc->us = gee->us;
-			utf8_get_big(gee->data, &gc->data);
+			log_debug("!!! %x", gc->flags);
+			utf8_to_data(gee->data, &gc->data);
 		}
 		return;
 	}
@@ -541,6 +542,7 @@ grid_set_cell(struct grid *gd, u_int px, u_int py, const struct grid_cell *gc)
 		gl->cellused = px + 1;
 
 	gce = &gl->celldata[px];
+	if (gc->flags & GRID_FLAG_PADDING) log_debug("!!! padding %d\n", grid_need_extended_cell(gce, gc));
 	if (grid_need_extended_cell(gce, gc))
 		grid_extended_cell(gl, gce, gc);
 	else
@@ -570,7 +572,7 @@ grid_set_cells(struct grid *gd, u_int px, u_int py, const struct grid_cell *gc,
 		gce = &gl->celldata[px + i];
 		if (grid_need_extended_cell(gce, gc)) {
 			gee = grid_extended_cell(gl, gce, gc);
-			gee->data = utf8_set_big(s[i], 1);
+			gee->data = utf8_build_one(s[i], 1);
 		} else
 			grid_store_cell(gce, gc, s[i]);
 	}
diff --git a/input-keys.c b/input-keys.c
index 4b3d2a36..d298395e 100644
--- a/input-keys.c
+++ b/input-keys.c
@@ -468,10 +468,9 @@ input_key(struct screen *s, struct bufferevent *bev, key_code key)
 		return (0);
 	}
 	if (justkey > 0x7f && justkey < KEYC_BASE) {
-		if (utf8_split(justkey, &ud) != UTF8_DONE)
-			return (-1);
 		if (key & KEYC_META)
 			bufferevent_write(bev, "\033", 1);
+		utf8_to_data(justkey, &ud);
 		bufferevent_write(bev, ud.data, ud.size);
 		return (0);
 	}
diff --git a/key-string.c b/key-string.c
index 1efb6e0b..4ee12145 100644
--- a/key-string.c
+++ b/key-string.c
@@ -169,7 +169,7 @@ key_string_lookup_string(const char *string)
 	struct utf8_data	 ud;
 	u_int			 i;
 	enum utf8_state		 more;
-	wchar_t			 wc;
+	utf8_char		 uc;
 
 	/* Is this no key or any key? */
 	if (strcasecmp(string, "None") == 0)
@@ -210,9 +210,9 @@ key_string_lookup_string(const char *string)
 				more = utf8_append(&ud, (u_char)string[i]);
 			if (more != UTF8_DONE)
 				return (KEYC_UNKNOWN);
-			if (utf8_combine(&ud, &wc) != UTF8_DONE)
+			if (utf8_from_data(&ud, &uc) != UTF8_DONE)
 				return (KEYC_UNKNOWN);
-			return (wc|modifiers);
+			return (uc|modifiers);
 		}
 
 		/* Otherwise look the key up in the table. */
@@ -349,12 +349,11 @@ key_string_lookup_key(key_code key, int with_flags)
 
 	/* Is this a UTF-8 key? */
 	if (key > 127 && key < KEYC_BASE) {
-		if (utf8_split(key, &ud) == UTF8_DONE) {
-			off = strlen(out);
-			memcpy(out + off, ud.data, ud.size);
-			out[off + ud.size] = '\0';
-			goto out;
-		}
+		utf8_to_data(key, &ud);
+		off = strlen(out);
+		memcpy(out + off, ud.data, ud.size);
+		out[off + ud.size] = '\0';
+		goto out;
 	}
 
 	/* Invalid keys are errors. */
diff --git a/screen-write.c b/screen-write.c
index 062a2929..68c3aa76 100644
--- a/screen-write.c
+++ b/screen-write.c
@@ -39,7 +39,7 @@ static const struct grid_cell *screen_write_combine(struct screen_write_ctx *,
 		    const struct utf8_data *, u_int *);
 
 static const struct grid_cell screen_write_pad_cell = {
-	{ { 0 }, 0, 0, 0 }, 0, GRID_FLAG_PADDING, 0, 8, 8
+	{ { 0 }, 0, 0, 0 }, 0, GRID_FLAG_PADDING, 8, 8, 0
 };
 
 struct screen_write_collect_item {
diff --git a/status.c b/status.c
index 93ac70df..ad4d6cda 100644
--- a/status.c
+++ b/status.c
@@ -1275,8 +1275,7 @@ process_key:
 append_key:
 	if (key <= 0x1f || key >= KEYC_BASE)
 		return (0);
-	if (utf8_split(key, &tmp) != UTF8_DONE)
-		return (0);
+	utf8_to_data(key, &tmp);
 
 	c->prompt_buffer = xreallocarray(c->prompt_buffer, size + 2,
 	    sizeof *c->prompt_buffer);
diff --git a/tmux.h b/tmux.h
index 87d9dd3c..9cea480a 100644
--- a/tmux.h
+++ b/tmux.h
@@ -27,7 +27,6 @@
 #include <stdarg.h>
 #include <stdio.h>
 #include <termios.h>
-#include <wchar.h>
 
 #ifdef HAVE_UTEMPTER
 #include <utempter.h>
@@ -598,10 +597,13 @@ struct msg_write_close {
 #define ALL_MOUSE_MODES (MODE_MOUSE_STANDARD|MODE_MOUSE_BUTTON|MODE_MOUSE_ALL)
 #define MOTION_MOUSE_MODES (MODE_MOUSE_BUTTON|MODE_MOUSE_ALL)
 
+/* A single UTF-8 character. */
+typedef u_int utf8_char;
+
 /*
- * A single UTF-8 character. UTF8_SIZE must be big enough to hold combining
+ * An expanded UTF-8 character. UTF8_SIZE must be big enough to hold combining
  * characters as well. It can't be more than 32 bytes without changes to how
- * big characters are stored.
+ * characters are stored.
  */
 #define UTF8_SIZE 21
 struct utf8_data {
@@ -675,7 +677,7 @@ struct grid_cell {
 
 /* Grid extended cell entry. */
 struct grid_extd_entry {
-	u_int			data;
+	utf8_char		data;
 	u_short			attr;
 	u_char			flags;
 	int			fg;
@@ -2891,15 +2893,13 @@ u_int		 session_group_attached_count(struct session_group *);
 void		 session_renumber_windows(struct session *);
 
 /* utf8.c */
-u_int		 utf8_set_big(char, u_int);
-u_int		 utf8_map_big(const struct utf8_data *);
-void		 utf8_get_big(u_int, struct utf8_data *);
+utf8_char	 utf8_build_one(char, u_int);
+enum utf8_state	 utf8_from_data(const struct utf8_data *, utf8_char *);
+void		 utf8_to_data(utf8_char, struct utf8_data *);
 void		 utf8_set(struct utf8_data *, u_char);
 void		 utf8_copy(struct utf8_data *, const struct utf8_data *);
 enum utf8_state	 utf8_open(struct utf8_data *, u_char);
 enum utf8_state	 utf8_append(struct utf8_data *, u_char);
-enum utf8_state	 utf8_combine(const struct utf8_data *, wchar_t *);
-enum utf8_state	 utf8_split(wchar_t, struct utf8_data *);
 int		 utf8_isvalid(const char *);
 int		 utf8_strvis(char *, const char *, size_t, int);
 int		 utf8_stravis(char **, const char *, int);
diff --git a/tty-keys.c b/tty-keys.c
index bbd4bb5f..2e375807 100644
--- a/tty-keys.c
+++ b/tty-keys.c
@@ -578,8 +578,8 @@ tty_keys_next1(struct tty *tty, const char *buf, size_t len, key_code *key,
 	struct tty_key		*tk, *tk1;
 	struct utf8_data	 ud;
 	enum utf8_state		 more;
+	utf8_char		 uc;
 	u_int			 i;
-	wchar_t			 wc;
 
 	log_debug("%s: next key is %zu (%.*s) (expired=%d)", c->name, len,
 	    (int)len, buf, expired);
@@ -611,12 +611,12 @@ tty_keys_next1(struct tty *tty, const char *buf, size_t len, key_code *key,
 		if (more != UTF8_DONE)
 			return (-1);
 
-		if (utf8_combine(&ud, &wc) != UTF8_DONE)
+		if (utf8_from_data(&ud, &uc) != UTF8_DONE)
 			return (-1);
-		*key = wc;
+		*key = uc;
 
 		log_debug("%s: UTF-8 key %.*s %#llx", c->name, (int)ud.size,
-		    buf, *key);
+		    ud.data, *key);
 		return (0);
 	}
 
diff --git a/utf8.c b/utf8.c
index efe8679d..353c90ba 100644
--- a/utf8.c
+++ b/utf8.c
@@ -26,153 +26,157 @@
 
 #include "tmux.h"
 
-static int	utf8_width(wchar_t);
-
-struct utf8_big_item {
-	u_int			index;
-	RB_ENTRY(utf8_big_item)	entry;
+struct utf8_item {
+	u_int			offset;
+	RB_ENTRY(utf8_item)	entry;
 
 	char			data[UTF8_SIZE];
 	u_char			size;
 };
-RB_HEAD(utf8_big_tree, utf8_big_item);
+RB_HEAD(utf8_tree, utf8_item);
 
 static int
-utf8_big_cmp(struct utf8_big_item *bi1, struct utf8_big_item *bi2)
+utf8_cmp(struct utf8_item *ui1, struct utf8_item *ui2)
 {
-	if (bi1->size < bi2->size)
+	if (ui1->size < ui2->size)
 		return (-1);
-	if (bi1->size > bi2->size)
+	if (ui1->size > ui2->size)
 		return (1);
-	return (memcmp(bi1->data, bi2->data, bi1->size));
+	return (memcmp(ui1->data, ui2->data, ui1->size));
 }
-RB_GENERATE_STATIC(utf8_big_tree, utf8_big_item, entry, utf8_big_cmp);
-static struct utf8_big_tree utf8_big_tree = RB_INITIALIZER(utf8_big_tree);
+RB_GENERATE_STATIC(utf8_tree, utf8_item, entry, utf8_cmp);
+static struct utf8_tree utf8_tree = RB_INITIALIZER(utf8_tree);
 
-static struct utf8_big_item *utf8_big_list;
-static u_int utf8_big_list_size;
-static u_int utf8_big_list_used;
+static struct utf8_item *utf8_list;
+static u_int		 utf8_list_size;
+static u_int		 utf8_list_used;
 
-union utf8_big_map {
-	u_int	value;
+union utf8_map {
+	utf8_char	uc;
 	struct {
 		u_char	flags;
-#define UTF8_BIG_SIZE 0x1f
-#define UTF8_BIG_WIDTH2 0x20
+#define UTF8_FLAG_SIZE 0x1f
+#define UTF8_FLAG_WIDTH2 0x20
 
 		u_char	data[3];
 	};
 } __packed;
 
-static const union utf8_big_map utf8_big_space1 = {
+static const union utf8_map utf8_space1 = {
 	.flags = 1,
 	.data = " "
 };
-static const union utf8_big_map utf8_big_space2 = {
-	.flags = UTF8_BIG_WIDTH2|2,
+static const union utf8_map utf8_space2 = {
+	.flags = UTF8_FLAG_WIDTH2|2,
 	.data = "  "
 };
 
-/* Get a big item by index. */
-static struct utf8_big_item *
-utf8_get_big_item(const char *data, size_t size)
+/* Get a UTF-8 item by offset. */
+static struct utf8_item *
+utf8_get_item(const char *data, size_t size)
 {
-	struct utf8_big_item bi;
+	struct utf8_item	ui;
+
+	memcpy(ui.data, data, size);
+	ui.size = size;
 
-	memcpy(bi.data, data, size);
-	bi.size = size;
+	return (RB_FIND(utf8_tree, &utf8_tree, &ui));
+}
 
-	return (RB_FIND(utf8_big_tree, &utf8_big_tree, &bi));
+/* Expand UTF-8 list. */
+static int
+utf8_expand_list(void)
+{
+	if (utf8_list_size == 0xffffff)
+		return (-1);
+	if (utf8_list_size == 0)
+		utf8_list_size = 256;
+	else if (utf8_list_size > 0x7fffff)
+		utf8_list_size = 0xffffff;
+	else
+		utf8_list_size *= 2;
+	utf8_list = xreallocarray(utf8_list, utf8_list_size, sizeof *utf8_list);
+	return (0);
 }
 
-/* Add a big item. */
+/* Add a UTF-8 item. */
 static int
-utf8_put_big_item(const char *data, size_t size, u_int *index)
+utf8_put_item(const char *data, size_t size, u_int *offset)
 {
-	struct utf8_big_item	*bi;
+	struct utf8_item	*ui;
 
-	bi = utf8_get_big_item(data, size);
-	if (bi != NULL) {
-		*index = bi->index;
+	ui = utf8_get_item(data, size);
+	if (ui != NULL) {
+		*offset = ui->offset;
 		log_debug("%s: have %.*s at %u", __func__, (int)size, data,
-		    *index);
+		    *offset);
 		return (0);
 	}
 
-	if (utf8_big_list_used == utf8_big_list_size) {
-		if (utf8_big_list_size == 0xffffff)
-			return (-1);
-		if (utf8_big_list_size == 0)
-			utf8_big_list_size = 256;
-		else if (utf8_big_list_size > 0x7fffff)
-			utf8_big_list_size = 0xffffff;
-		else
-			utf8_big_list_size *= 2;
-		utf8_big_list = xreallocarray(utf8_big_list, utf8_big_list_size,
-		    sizeof *utf8_big_list);
-	}
-	*index = utf8_big_list_used++;
+	if (utf8_list_used == utf8_list_size && utf8_expand_list() != 0)
+		return (-1);
+	*offset = utf8_list_used++;
 
-	bi = &utf8_big_list[*index];
-	bi->index = *index;
-	memcpy(bi->data, data, size);
-	bi->size = size;
-	RB_INSERT(utf8_big_tree, &utf8_big_tree, bi);
+	ui = &utf8_list[*offset];
+	ui->offset = *offset;
+	memcpy(ui->data, data, size);
+	ui->size = size;
+	RB_INSERT(utf8_tree, &utf8_tree, ui);
 
-	log_debug("%s: added %.*s at %u", __func__, (int)size, data, *index);
+	log_debug("%s: added %.*s at %u", __func__, (int)size, data, *offset);
 	return (0);
 }
 
-/* Get UTF-8 as index into buffer. */
-u_int
-utf8_map_big(const struct utf8_data *ud)
+/* Get UTF-8 character from data. */
+enum utf8_state
+utf8_from_data(const struct utf8_data *ud, utf8_char *uc)
 {
-	union utf8_big_map	 m = { .value = 0 };
-	u_int			 o;
-	const char		*data = ud->data;
-	size_t			 size = ud->size;
+	union utf8_map	 m = { .uc = 0 };
+	u_int		 offset;
 
 	if (ud->width != 1 && ud->width != 2)
-		return (utf8_big_space1.value);
+		return (utf8_space1.uc);
 
-	if (size > UTF8_BIG_SIZE)
+	if (ud->size > UTF8_FLAG_SIZE)
 		goto fail;
-	if (size == 1)
-		return (utf8_set_big(data[0], 1));
+	if (ud->size == 1)
+		return (utf8_build_one(ud->data[0], 1));
 
-	m.flags = size;
+	m.flags = ud->size;
 	if (ud->width == 2)
-		m.flags |= UTF8_BIG_WIDTH2;
+		m.flags |= UTF8_FLAG_WIDTH2;
 
-	if (size <= 3) {
-		memcpy(&m.data, data, size);
-		return (m.value);
+	if (ud->size <= 3)
+		memcpy(m.data, ud->data, ud->size);
+	else {
+		if (utf8_put_item(ud->data, ud->size, &offset) != 0)
+			goto fail;
+		m.data[0] = (offset & 0xff);
+		m.data[1] = (offset >> 8) & 0xff;
+		m.data[2] = (offset >> 16);
 	}
-
-	if (utf8_put_big_item(data, size, &o) != 0)
-		goto fail;
-	m.data[0] = (o & 0xff);
-	m.data[1] = (o >> 8) & 0xff;
-	m.data[2] = (o >> 16);
-	return (m.value);
+	*uc = m.uc;
+	return (UTF8_DONE);
 
 fail:
 	if (ud->width == 1)
-		return (utf8_big_space1.value);
-	return (utf8_big_space2.value);
+		*uc = utf8_space1.uc;
+	else
+		*uc = utf8_space2.uc;
+	return (UTF8_ERROR);
 }
 
-/* Get UTF-8 from index into buffer. */
+/* Get UTF-8 data from character. */
 void
-utf8_get_big(u_int v, struct utf8_data *ud)
+utf8_to_data(utf8_char uc, struct utf8_data *ud)
 {
-	union utf8_big_map	 m = { .value = v };
-	struct utf8_big_item	*bi;
-	u_int			 o;
+	union utf8_map		 m = { .uc = uc };
+	struct utf8_item	*ui;
+	u_int			 offset;
 
 	memset(ud, 0, sizeof *ud);
-	ud->size = ud->have = (m.flags & UTF8_BIG_SIZE);
-	if (m.flags & UTF8_BIG_WIDTH2)
+	ud->size = ud->have = (m.flags & UTF8_FLAG_SIZE);
+	if (m.flags & UTF8_FLAG_WIDTH2)
 		ud->width = 2;
 	else
 		ud->width = 1;
@@ -182,24 +186,24 @@ utf8_get_big(u_int v, struct utf8_data *ud)
 		return;
 	}
 
-	o = ((u_int)m.data[2] << 16)|((u_int)m.data[1] << 8)|m.data[0];
-	if (o >= utf8_big_list_used)
+	offset = ((u_int)m.data[2] << 16)|((u_int)m.data[1] << 8)|m.data[0];
+	if (offset >= utf8_list_used)
 		memset(ud->data, ' ', ud->size);
 	else {
-		bi = &utf8_big_list[o];
-		memcpy(ud->data, bi->data, ud->size);
+		ui = &utf8_list[offset];
+		memcpy(ud->data, ui->data, ud->size);
 	}
 }
 
-/* Get big value for UTF-8 single character. */
+/* Get UTF-8 character from a single ASCII character. */
 u_int
-utf8_set_big(char c, u_int width)
+utf8_build_one(char c, u_int width)
 {
-	union utf8_big_map	m = { .flags = 1, .data[0] = c };
+	union utf8_map	m = { .flags = 1, .data[0] = c };
 
 	if (width == 2)
-		m.flags |= UTF8_BIG_WIDTH2;
-	return (m.value);
+		m.flags |= UTF8_FLAG_WIDTH2;
+	return (m.uc);
 }
 
 /* Set a single character. */
@@ -224,6 +228,30 @@ utf8_copy(struct utf8_data *to, const struct utf8_data *from)
 		to->data[i] = '\0';
 }
 
+/* Get width of Unicode character. */
+static enum utf8_state
+utf8_width(struct utf8_data *ud, int *width)
+{
+	wchar_t	wc;
+
+	switch (mbtowc(&wc, ud->data, ud->size)) {
+	case -1:
+		log_debug("UTF-8 %.*s, mbtowc() %d", (int)ud->size, ud->data,
+		    errno);
+		mbtowc(NULL, NULL, MB_CUR_MAX);
+		return (UTF8_ERROR);
+	case 0:
+		return (UTF8_ERROR);
+	}
+	*width = wcwidth(wc);
+	if (*width < 0 || *width > 0xff) {
+		log_debug("UTF-8 %.*s, wcwidth() %d", (int)ud->size, ud->data,
+		    *width);
+		return (UTF8_ERROR);
+	}
+	return (UTF8_DONE);
+}
+
 /*
  * Open UTF-8 sequence.
  *
@@ -251,7 +279,6 @@ utf8_open(struct utf8_data *ud, u_char ch)
 enum utf8_state
 utf8_append(struct utf8_data *ud, u_char ch)
 {
-	wchar_t	wc;
 	int	width;
 
 	if (ud->have >= ud->size)
@@ -268,10 +295,7 @@ utf8_append(struct utf8_data *ud, u_char ch)
 
 	if (ud->width == 0xff)
 		return (UTF8_ERROR);
-
-	if (utf8_combine(ud, &wc) != UTF8_DONE)
-		return (UTF8_ERROR);
-	if ((width = utf8_width(wc)) < 0)
+	if (utf8_width(ud, &width) != UTF8_DONE)
 		return (UTF8_ERROR);
 	ud->width = width;
 
@@ -310,50 +334,6 @@ utf8_width(wchar_t wc)
 	return (width);
 }
 
-/* Combine UTF-8 into Unicode. */
-enum utf8_state
-utf8_combine(const struct utf8_data *ud, wchar_t *wc)
-{
-#ifdef HAVE_UTF8PROC
-	switch (utf8proc_mbtowc(wc, ud->data, ud->size)) {
-#else
-	switch (mbtowc(wc, ud->data, ud->size)) {
-#endif
-	case -1:
-		log_debug("UTF-8 %.*s, mbtowc() %d", (int)ud->size, ud->data,
-		    errno);
-		mbtowc(NULL, NULL, MB_CUR_MAX);
-		return (UTF8_ERROR);
-	case 0:
-		return (UTF8_ERROR);
-	default:
-		return (UTF8_DONE);
-	}
-}
-
-/* Split Unicode into UTF-8. */
-enum utf8_state
-utf8_split(wchar_t wc, struct utf8_data *ud)
-{
-	char	s[MB_LEN_MAX];
-	int	slen;
-
-#ifdef HAVE_UTF8PROC
-	slen = utf8proc_wctomb(s, wc);
-#else
-	slen = wctomb(s, wc);
-#endif
-	if (slen <= 0 || slen > (int)sizeof ud->data)
-		return (UTF8_ERROR);
-
-	memcpy(ud->data, s, slen);
-	ud->size = slen;
-
-	ud->width = utf8_width(wc);
-	return (UTF8_DONE);
-}
-
-/*
  * Encode len characters from src into dst, which is guaranteed to have four
  * bytes available for each character from src (for \abc or UTF-8) plus space
  * for \0.
@@ -362,13 +342,10 @@ int
 utf8_strvis(char *dst, const char *src, size_t len, int flag)
 {
 	struct utf8_data	 ud;
-	const char		*start, *end;
+	const char		*start = dst, *end = src + len;
 	enum utf8_state		 more;
 	size_t			 i;
 
-	start = dst;
-	end = src + len;
-
 	while (src < end) {
 		if ((more = utf8_open(&ud, *src)) == UTF8_MORE) {
 			while (++src < end && more == UTF8_MORE)
@@ -394,7 +371,6 @@ utf8_strvis(char *dst, const char *src, size_t len, int flag)
 			dst = vis(dst, src[0], flag, '\0');
 		src++;
 	}
-
 	*dst = '\0';
 	return (dst - start);
 }
@@ -417,9 +393,9 @@ utf8_stravis(char **dst, const char *src, int flag)
 int
 utf8_isvalid(const char *s)
 {
-	struct utf8_data	 ud;
-	const char		*end;
-	enum utf8_state		 more;
+	struct utf8_data ud;
+	const char	*end;
+	enum utf8_state	 more;
 
 	end = s + strlen(s);
 	while (s < end) {
@@ -445,15 +421,12 @@ utf8_isvalid(const char *s)
 char *
 utf8_sanitize(const char *src)
 {
-	char			*dst;
-	size_t			 n;
-	enum utf8_state		 more;
-	struct utf8_data	 ud;
-	u_int			 i;
-
-	dst = NULL;
+	char		*dst = NULL;
+	size_t		 n = 0;
+	enum utf8_state	 more;
+	struct utf8_data ud;
+	u_int		 i;
 
-	n = 0;
 	while (*src != '\0') {
 		dst = xreallocarray(dst, n + 1, sizeof *dst);
 		if ((more = utf8_open(&ud, *src)) == UTF8_MORE) {
@@ -474,7 +447,6 @@ utf8_sanitize(const char *src)
 			dst[n++] = '_';
 		src++;
 	}
-
 	dst = xreallocarray(dst, n + 1, sizeof *dst);
 	dst[n] = '\0';
 	return (dst);
@@ -496,9 +468,8 @@ u_int
 utf8_strwidth(const struct utf8_data *s, ssize_t n)
 {
 	ssize_t	i;
-	u_int	width;
+	u_int	width = 0;
 
-	width = 0;
 	for (i = 0; s[i].size != 0; i++) {
 		if (n != -1 && n == i)
 			break;
@@ -514,13 +485,10 @@ utf8_strwidth(const struct utf8_data *s, ssize_t n)
 struct utf8_data *
 utf8_fromcstr(const char *src)
 {
-	struct utf8_data	*dst;
-	size_t			 n;
+	struct utf8_data	*dst = NULL;
+	size_t			 n = 0;
 	enum utf8_state		 more;
 
-	dst = NULL;
-
-	n = 0;
 	while (*src != '\0') {
 		dst = xreallocarray(dst, n + 1, sizeof *dst);
 		if ((more = utf8_open(&dst[n], *src)) == UTF8_MORE) {
@@ -536,7 +504,6 @@ utf8_fromcstr(const char *src)
 		n++;
 		src++;
 	}
-
 	dst = xreallocarray(dst, n + 1, sizeof *dst);
 	dst[n].size = 0;
 	return (dst);
@@ -546,18 +513,14 @@ utf8_fromcstr(const char *src)
 char *
 utf8_tocstr(struct utf8_data *src)
 {
-	char	*dst;
-	size_t	 n;
+	char	*dst = NULL;
+	size_t	 n = 0;
 
-	dst = NULL;
-
-	n = 0;
 	for(; src->size != 0; src++) {
 		dst = xreallocarray(dst, n + src->size, 1);
 		memcpy(dst + n, src->data, src->size);
 		n += src->size;
 	}
-
 	dst = xreallocarray(dst, n + 1, 1);
 	dst[n] = '\0';
 	return (dst);
@@ -595,7 +558,7 @@ utf8_padcstr(const char *s, u_int width)
 {
 	size_t	 slen;
 	char	*out;
-	u_int	  n, i;
+	u_int	 n, i;
 
 	n = utf8_cstrwidth(s);
 	if (n >= width)
@@ -616,7 +579,7 @@ utf8_rpadcstr(const char *s, u_int width)
 {
 	size_t	 slen;
 	char	*out;
-	u_int	  n, i;
+	u_int	 n, i;
 
 	n = utf8_cstrwidth(s);
 	if (n >= width)
diff --git a/window-copy.c b/window-copy.c
index b869fc83..cfa91df9 100644
--- a/window-copy.c
+++ b/window-copy.c
@@ -2571,7 +2571,7 @@ window_copy_cellstring(const struct grid_line *gl, u_int px, size_t *size,
 		return (&gce->data.data);
 	}
 
-	utf8_get_big(gl->extddata[gce->offset].data, &ud);
+	utf8_to_data(gl->extddata[gce->offset].data, &ud);
 	*size = ud.size;
 	*allocated = 1;