Extend libteken to support CJK fullwidth characters.

Introduce a new formatting bit (TF_CJK_RIGHT) that is set when putting a
cell that is the right part of a CJK fullwidth character. This will
allow drivers like vt(9) to support fullwidth characters properly.

emaste@ has a patch to extend vt(9)'s font handling to increase the
number of Unicode -> glyph maps from 2 ({normal,bold)} to 4
({normal,bold} x {left,right}). This will need to use this formatting
bit to determine whether to draw the left or right glyph.

Reviewed by:	emaste

7 files changed, 34 insertions, 20 deletions

diff --git a/sys/dev/syscons/scterm-teken.c b/sys/dev/syscons/scterm-teken.c
index e2e91189ee24..43f16faff982 100644
--- a/sys/dev/syscons/scterm-teken.c
+++ b/sys/dev/syscons/scterm-teken.c
@@ -553,7 +553,14 @@ scteken_putchar(void *arg, const teken_pos_t *tp, teken_char_t c,
 	vm_offset_t p;
 	int cursor, attr;
 
+	/*
+	 * No support for printing right hand sides for CJK fullwidth
+	 * characters. Simply print a space and assume that the left
+	 * hand side describes the entire character.
+	 */
 	attr = scteken_attr(a) << 8;
+	if (a->ta_format & TF_CJK_RIGHT)
+		c = ' ';
 #ifdef TEKEN_UTF8
 	scteken_get_cp437(&c, &attr);
 #endif /* TEKEN_UTF8 */
diff --git a/sys/dev/vt/vt_font.c b/sys/dev/vt/vt_font.c
index eebe74066e69..50bcfe4e4994 100644
--- a/sys/dev/vt/vt_font.c
+++ b/sys/dev/vt/vt_font.c
@@ -87,7 +87,15 @@ vtfont_lookup(const struct vt_font *vf, term_char_t c)
 	uint16_t dst;
 	size_t stride;
 
+	/*
+	 * No support for printing right hand sides for CJK fullwidth
+	 * characters. Simply print a space and assume that the left
+	 * hand side describes the entire character.
+	 */
 	src = TCHAR_CHARACTER(c);
+	if (TCHAR_FORMAT(c) & TF_CJK_RIGHT)
+		src = ' ';
+
 	if (TCHAR_FORMAT(c) & TF_BOLD) {
 		dst = vtfont_bisearch(vf->vf_bold, vf->vf_bold_length, src);
 		if (dst != 0)
diff --git a/sys/kern/subr_terminal.c b/sys/kern/subr_terminal.c
index 91634f4483a2..54636e408c29 100644
--- a/sys/kern/subr_terminal.c
+++ b/sys/kern/subr_terminal.c
@@ -128,7 +128,7 @@ static const teken_attr_t default_message = {
 };
 
 #define	TCHAR_CREATE(c, a)	((c) | \
-	(a)->ta_format << 22 | \
+	(a)->ta_format << 21 | \
 	teken_256to8((a)->ta_fgcolor) << 26 | \
 	teken_256to8((a)->ta_bgcolor) << 29)
 
diff --git a/sys/sys/terminal.h b/sys/sys/terminal.h
index 146b9acdd084..e76190b500a5 100644
--- a/sys/sys/terminal.h
+++ b/sys/sys/terminal.h
@@ -62,15 +62,14 @@ struct tty;
  *
  *  Bits  Meaning
  *  0-20: Character value
- *    21: Unused
- * 22-25: Bold, underline, blink, reverse
+ * 21-25: Bold, underline, blink, reverse, right part of CJK fullwidth character
  * 26-28: Foreground color
  * 29-31: Background color
  */
 
 typedef uint32_t term_char_t;
 #define	TCHAR_CHARACTER(c)	((c) & 0x1fffff)
-#define	TCHAR_FORMAT(c)		(((c) >> 22) & 0xf)
+#define	TCHAR_FORMAT(c)		(((c) >> 21) & 0x1f)
 #define	TCHAR_FGCOLOR(c)	(((c) >> 26) & 0x7)
 #define	TCHAR_BGCOLOR(c)	((c) >> 29)
 
diff --git a/sys/teken/demo/teken_demo.c b/sys/teken/demo/teken_demo.c
index 16bb4b7b198a..08323dc1a400 100644
--- a/sys/teken/demo/teken_demo.c
+++ b/sys/teken/demo/teken_demo.c
@@ -86,9 +86,10 @@ printchar(const teken_pos_t *p)
 	assert(p->tp_row < NROWS);
 	assert(p->tp_col < NCOLS);
 
-	getyx(stdscr, y, x);
-
 	px = &buffer[p->tp_col][p->tp_row];
+	/* No need to print right hand side of CJK character manually. */
+	if (px->a.ta_format & TF_CJK_RIGHT)
+		return;
 
 	/* Convert Unicode to UTF-8. */
 	if (px->c < 0x80) {
@@ -118,8 +119,8 @@ printchar(const teken_pos_t *p)
 
 	bkgdset(attr | COLOR_PAIR(teken_256to8(px->a.ta_fgcolor) +
 	      8 * teken_256to8(px->a.ta_bgcolor)));
+	getyx(stdscr, y, x);
 	mvaddstr(p->tp_row, p->tp_col, str);
-
 	move(y, x);
 }
 
diff --git a/sys/teken/teken.h b/sys/teken/teken.h
index a9bca98e9669..27d1ce404740 100644
--- a/sys/teken/teken.h
+++ b/sys/teken/teken.h
@@ -41,10 +41,11 @@
 typedef uint32_t teken_char_t;
 typedef unsigned short teken_unit_t;
 typedef unsigned char teken_format_t;
-#define	TF_BOLD		0x01
-#define	TF_UNDERLINE	0x02
-#define	TF_BLINK	0x04
-#define	TF_REVERSE	0x08
+#define	TF_BOLD		0x01	/* Bold character. */
+#define	TF_UNDERLINE	0x02	/* Underline character. */
+#define	TF_BLINK	0x04	/* Blinking character. */
+#define	TF_REVERSE	0x08	/* Reverse rendered character. */
+#define	TF_CJK_RIGHT	0x10	/* Right-hand side of CJK character. */
 typedef unsigned char teken_color_t;
 #define	TC_BLACK	0
 #define	TC_RED		1
diff --git a/sys/teken/teken_subr.h b/sys/teken/teken_subr.h
index f25ab6aa34d0..be3ce8abfe83 100644
--- a/sys/teken/teken_subr.h
+++ b/sys/teken/teken_subr.h
@@ -791,21 +791,19 @@ teken_subr_do_putchar(teken_t *t, const teken_pos_t *tp, teken_char_t c,
 		teken_funcs_copy(t, &ctr, &ctp);
 	}
 
+	teken_funcs_putchar(t, tp, c, &t->t_curattr);
+
 	if (width == 2 && tp->tp_col + 1 < t->t_winsize.tp_col) {
 		teken_pos_t tp2;
+		teken_attr_t attr;
 
-		/*
-		 * Store a space behind double width characters before
-		 * actually printing them. This prevents artifacts when
-		 * the consumer doesn't render it using double width
-		 * glyphs.
-		 */
+		/* Print second half of CJK fullwidth character. */
 		tp2.tp_row = tp->tp_row;
 		tp2.tp_col = tp->tp_col + 1;
-		teken_funcs_putchar(t, &tp2, BLANK, &t->t_curattr);
+		attr = t->t_curattr;
+		attr.ta_format |= TF_CJK_RIGHT;
+		teken_funcs_putchar(t, &tp2, c, &attr);
 	}
-
-	teken_funcs_putchar(t, tp, c, &t->t_curattr);
 }
 
 static void