Improve handling of ambiguous overstrike sequences. A sequence of _\b_

can either mean an underlined underscore or a bold underscore. This
ambiguity can be 'resolved' by takeing the state of the surrounding text
into account. If surrounded by bold text, the result should probably be
bold and likewise for underlined. less(1) previously only looked at the
preceding text and ul(1) didn't examine the context at all.

tweaks and ok schwarze
ok tb (on a previous version of the diff)

3 files changed, 43 insertions, 13 deletions

diff --git a/usr.bin/less/less.h b/usr.bin/less/less.h
index a67401e9bf9..a7fe3220534 100644
--- a/usr.bin/less/less.h
+++ b/usr.bin/less/less.h
@@ -149,6 +149,7 @@ struct textlist {
 #define	AT_ANSI		(1 << 4)  /* Content-supplied "ANSI" escape sequence */
 #define	AT_BINARY	(1 << 5)  /* LESS*BINFMT representation */
 #define	AT_HILITE	(1 << 6)  /* Internal highlights (e.g., for search) */
+#define	AT_INDET	(1 << 7)  /* Indeterminate: either bold or underline */
 
 #define	CONTROL(c)	((c)&037)
 
diff --git a/usr.bin/less/line.c b/usr.bin/less/line.c
index fe2fe9ff1f5..ebf4b1b5b97 100644
--- a/usr.bin/less/line.c
+++ b/usr.bin/less/line.c
@@ -32,7 +32,6 @@ off_t highest_hilite;		/* Pos of last hilite in file found so far */
 static int curr;		/* Index into linebuf */
 static int column;	/* Printable length, accounting for backspaces, etc. */
 static int overstrike;		/* Next char should overstrike previous char */
-static int last_overstrike = AT_NORMAL;
 static int is_null_line;	/* There is no current line */
 static int lmargin;		/* Left margin */
 static char pendc;
@@ -127,7 +126,6 @@ prewind(void)
 	column = 0;
 	cshift = 0;
 	overstrike = 0;
-	last_overstrike = AT_NORMAL;
 	mbc_buf_len = 0;
 	is_null_line = 0;
 	pendc = '\0';
@@ -515,10 +513,6 @@ store_char(LWCHAR ch, char a, char *rep, off_t pos)
 	char cs;
 	int matches;
 
-	w = (a & (AT_UNDERLINE|AT_BOLD));	/* Pre-use w.  */
-	if (w != AT_NORMAL)
-		last_overstrike = w;
-
 	if (is_hilited(pos, pos+1, 0, &matches)) {
 		/*
 		 * This character should be highlighted.
@@ -808,10 +802,12 @@ do_append(LWCHAR ch, char *rep, off_t pos)
 			if (ch == '_') {
 				if ((a & (AT_BOLD|AT_UNDERLINE)) != AT_NORMAL)
 					a |= (AT_BOLD|AT_UNDERLINE);
-				else if (last_overstrike != AT_NORMAL)
-					a |= last_overstrike;
-				else
+				else if (curr > 0 && attr[curr - 1] & AT_UNDERLINE)
+					a |= AT_UNDERLINE;
+				else if (curr > 0 && attr[curr - 1] & AT_BOLD)
 					a |= AT_BOLD;
+				else
+					a |= AT_INDET;
 			} else {
 				a |= AT_BOLD;
 			}
@@ -825,10 +821,13 @@ do_append(LWCHAR ch, char *rep, off_t pos)
 		/* Else we replace prev_ch, but we keep its attributes.  */
 	} else if (overstrike < 0) {
 		if (is_composing_char(ch) ||
-		    is_combining_char(get_wchar(linebuf + curr), ch))
+		    is_combining_char(get_wchar(linebuf + curr), ch)) {
 			/* Continuation of the same overstrike.  */
-			a = last_overstrike;
-		else
+			if (curr > 0)
+				a = attr[curr - 1] & (AT_UNDERLINE | AT_BOLD);
+			else
+				a = AT_NORMAL;
+		} else
 			overstrike = 0;
 	}
 
@@ -894,6 +893,8 @@ pflushmbc(void)
 void
 pdone(int endline, int forw)
 {
+	int i;
+
 	(void) pflushmbc();
 
 	if (pendc && (pendc != '\r' || !endline))
@@ -904,6 +905,16 @@ pdone(int endline, int forw)
 		 */
 		(void) do_append(pendc, NULL, pendpos);
 
+	for (i = curr - 1; i >= 0; i--) {
+		if (attr[i] & AT_INDET) {
+			attr[i] &= ~AT_INDET;
+			if (i < curr - 1 && attr[i + 1] & AT_BOLD)
+				attr[i] |= AT_BOLD;
+			else
+				attr[i] |= AT_UNDERLINE;
+		}
+	}
+
 	/*
 	 * Make sure we've shifted the line, if we need to.
 	 */
diff --git a/usr.bin/ul/ul.c b/usr.bin/ul/ul.c
index 847d1013373..d5f894b4b61 100644
--- a/usr.bin/ul/ul.c
+++ b/usr.bin/ul/ul.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: ul.c,v 1.20 2016/01/18 17:34:26 schwarze Exp $	*/
+/*	$OpenBSD: ul.c,v 1.21 2016/03/26 08:59:29 natano Exp $	*/
 /*	$NetBSD: ul.c,v 1.3 1994/12/07 00:28:24 jtc Exp $	*/
 
 /*
@@ -55,6 +55,7 @@
 #define	SUBSC	004	/* Dim | Ul */
 #define	UNDERL	010	/* Ul */
 #define	BOLD	020	/* Bold */
+#define	INDET	040	/* Indeterminate: either Bold or Ul */
 
 int	must_use_uc, must_overstrike;
 char	*CURS_UP, *CURS_RIGHT, *CURS_LEFT,
@@ -266,6 +267,13 @@ mfilter(FILE *f)
 			if (obuf[col].c_char == L'\0') {
 				obuf[col].c_char = L'_';
 				obuf[col].c_width = 1;
+			} else if (obuf[col].c_char == L'_') {
+				if (obuf[col - 1].c_mode & UNDERL)
+					obuf[col].c_mode |= UNDERL | mode;
+				else if (obuf[col - 1].c_mode & BOLD)
+					obuf[col].c_mode |= BOLD | mode;
+				else
+					obuf[col].c_mode |= INDET | mode;
 			} else
 				obuf[col].c_mode |= UNDERL | mode;
 			/* FALLTHROUGH */
@@ -326,6 +334,16 @@ flushln(void)
 	int lastmode, i;
 	int hadmodes = 0;
 
+	for (i = maxcol; i > 0; i--) {
+		if (obuf[i].c_mode & INDET) {
+			obuf[i].c_mode &= ~INDET;
+			if (i < maxcol && obuf[i + 1].c_mode & BOLD)
+				obuf[i].c_mode |= BOLD;
+			else
+				obuf[i].c_mode |= UNDERL;
+		}
+	}
+
 	lastmode = NORMAL;
 	for (i = 1; i < maxcol; i++) {
 		if (obuf[i].c_mode != lastmode) {