Rudimentary implementation of the roff(7) .char (output glyph

definition) request, used for example by groff_hdtbl(7).

This simplistic implementation may interact incorrectly
with the .tr (input character translation) request.
But come on, you are not only using .char *and* .tr, but you do so
with respect to the same character in the same manual page?

4 files changed, 94 insertions, 6 deletions

diff --git a/usr.bin/mandoc/mandoc.1 b/usr.bin/mandoc/mandoc.1
index cc55b3ceac0..aa19c49d634 100644
--- a/usr.bin/mandoc/mandoc.1
+++ b/usr.bin/mandoc/mandoc.1
@@ -1,4 +1,4 @@
-.\"	$OpenBSD: mandoc.1,v 1.151 2018/08/23 14:16:12 schwarze Exp $
+.\"	$OpenBSD: mandoc.1,v 1.152 2018/08/25 16:43:52 schwarze Exp $
 .\"
 .\" Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
 .\" Copyright (c) 2012, 2014-2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -15,7 +15,7 @@
 .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 .\"
-.Dd $Mdocdate: August 23 2018 $
+.Dd $Mdocdate: August 25 2018 $
 .Dt MANDOC 1
 .Os
 .Sh NAME
@@ -1606,6 +1606,12 @@ or
 .Cm off .
 The invalid argument is moved out of the macro, which leaves the macro
 empty, causing it to toggle the spacing mode.
+.It Sy "argument contains two font escapes"
+.Pq roff
+The second argument of a
+.Ic char
+request contains more than one font escape sequence.
+A wrong font may remain active after using the character.
 .It Sy "unknown font, skipping request"
 .Pq man , tbl
 A
@@ -1959,6 +1965,13 @@ macro fails to specify the list type.
 The argument of a
 .Ic \&ce
 request is not a number.
+.It Sy "argument is not a character"
+.Pq roff
+The first argument of a
+.Ic char
+request is neither a single ASCII character
+nor a single character escape sequence.
+The request is ignored including all its arguments.
 .It Sy "missing manual name, using \(dq\(dq"
 .Pq mdoc
 The first call to
diff --git a/usr.bin/mandoc/mandoc.h b/usr.bin/mandoc/mandoc.h
index 6b39fc056ee..4a46bd89d13 100644
--- a/usr.bin/mandoc/mandoc.h
+++ b/usr.bin/mandoc/mandoc.h
@@ -1,4 +1,4 @@
-/*	$OpenBSD: mandoc.h,v 1.194 2018/08/24 22:56:37 schwarze Exp $ */
+/*	$OpenBSD: mandoc.h,v 1.195 2018/08/25 16:43:52 schwarze Exp $ */
 /*
  * Copyright (c) 2010, 2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -158,6 +158,7 @@ enum	mandocerr {
 	MANDOCERR_LB_BAD, /* unknown library name: Lb ... */
 	MANDOCERR_RS_BAD, /* invalid content in Rs block: macro */
 	MANDOCERR_SM_BAD, /* invalid Boolean argument: macro arg */
+	MANDOCERR_CHAR_FONT, /* argument contains two font escapes */
 	MANDOCERR_FT_BAD, /* unknown font, skipping request: ft font */
 	MANDOCERR_TR_ODD, /* odd number of characters in request: tr char */
 
@@ -212,6 +213,7 @@ enum	mandocerr {
 	MANDOCERR_BD_NOARG, /* skipping display without arguments: Bd */
 	MANDOCERR_BL_NOTYPE, /* missing list type, using -item: Bl */
 	MANDOCERR_CE_NONUM, /* argument is not numeric, using 1: ce ... */
+	MANDOCERR_CHAR_ARG, /* argument is not a character: char ... */
 	MANDOCERR_NM_NONAME, /* missing manual name, using "": Nm */
 	MANDOCERR_OS_UNAME, /* uname(3) system call failed, using UNKNOWN */
 	MANDOCERR_ST_BAD, /* unknown standard specifier: St standard */
diff --git a/usr.bin/mandoc/read.c b/usr.bin/mandoc/read.c
index 2ab9dae312e..e6c81ed3516 100644
--- a/usr.bin/mandoc/read.c
+++ b/usr.bin/mandoc/read.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: read.c,v 1.171 2018/08/24 22:56:37 schwarze Exp $ */
+/*	$OpenBSD: read.c,v 1.172 2018/08/25 16:43:52 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -196,6 +196,7 @@ static	const char * const	mandocerrs[MANDOCERR_MAX] = {
 	"unknown library name",
 	"invalid content in Rs block",
 	"invalid Boolean argument",
+	"argument contains two font escapes",
 	"unknown font, skipping request",
 	"odd number of characters in request",
 
@@ -250,6 +251,7 @@ static	const char * const	mandocerrs[MANDOCERR_MAX] = {
 	"skipping display without arguments",
 	"missing list type, using -item",
 	"argument is not numeric, using 1",
+	"argument is not a character",
 	"missing manual name, using \"\"",
 	"uname(3) system call failed, using UNKNOWN",
 	"unknown standard specifier",
diff --git a/usr.bin/mandoc/roff.c b/usr.bin/mandoc/roff.c
index 57ff4a137db..d6fc50bb27f 100644
--- a/usr.bin/mandoc/roff.c
+++ b/usr.bin/mandoc/roff.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: roff.c,v 1.212 2018/08/24 22:56:37 schwarze Exp $ */
+/*	$OpenBSD: roff.c,v 1.213 2018/08/25 16:43:52 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2015, 2017, 2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -174,6 +174,7 @@ static	int		 roff_br(ROFF_ARGS);
 static	int		 roff_cblock(ROFF_ARGS);
 static	int		 roff_cc(ROFF_ARGS);
 static	int		 roff_ccond(struct roff *, int, int);
+static	int		 roff_char(ROFF_ARGS);
 static	int		 roff_cond(ROFF_ARGS);
 static	int		 roff_cond_text(ROFF_ARGS);
 static	int		 roff_cond_sub(ROFF_ARGS);
@@ -395,7 +396,7 @@ static	struct roffmac	 roffs[TOKEN_NONE] = {
 	{ roff_insec, NULL, NULL, 0 },  /* cf */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* cflags */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* ch */
-	{ roff_unsupp, NULL, NULL, 0 },  /* char */
+	{ roff_char, NULL, NULL, 0 },  /* char */
 	{ roff_unsupp, NULL, NULL, 0 },  /* chop */
 	{ roff_line_ignore, NULL, NULL, 0 },  /* class */
 	{ roff_insec, NULL, NULL, 0 },  /* close */
@@ -3320,6 +3321,76 @@ roff_cc(ROFF_ARGS)
 }
 
 static int
+roff_char(ROFF_ARGS)
+{
+	const char	*p, *kp, *vp;
+	size_t		 ksz, vsz;
+	int		 font;
+
+	/* Parse the character to be replaced. */
+
+	kp = buf->buf + pos;
+	p = kp + 1;
+	if (*kp == '\0' || (*kp == '\\' &&
+	     mandoc_escape(&p, NULL, NULL) != ESCAPE_SPECIAL) ||
+	    (*p != ' ' && *p != '\0')) {
+		mandoc_vmsg(MANDOCERR_CHAR_ARG, r->parse,
+		    ln, pos, "char %s", kp);
+		return ROFF_IGN;
+	}
+	ksz = p - kp;
+	while (*p == ' ')
+		p++;
+
+	/*
+	 * If the replacement string contains a font escape sequence,
+	 * we have to restore the font at the end.
+	 */
+
+	vp = p;
+	vsz = strlen(p);
+	font = 0;
+	while (*p != '\0') {
+		if (*p++ != '\\')
+			continue;
+		switch (mandoc_escape(&p, NULL, NULL)) {
+		case ESCAPE_FONT:
+		case ESCAPE_FONTROMAN:
+		case ESCAPE_FONTITALIC:
+		case ESCAPE_FONTBOLD:
+		case ESCAPE_FONTBI:
+		case ESCAPE_FONTPREV:
+			font++;
+			break;
+		default:
+			break;
+		}
+	}
+	if (font > 1)
+		mandoc_msg(MANDOCERR_CHAR_FONT, r->parse,
+		    ln, vp - buf->buf, vp);
+
+	/*
+	 * Approximate the effect of .char using the .tr tables.
+	 * XXX In groff, .char and .tr interact differently.
+	 */
+
+	if (ksz == 1) {
+		if (r->xtab == NULL)
+			r->xtab = mandoc_calloc(128, sizeof(*r->xtab));
+		assert((unsigned int)*kp < 128);
+		free(r->xtab[(int)*kp].p);
+		r->xtab[(int)*kp].sz = mandoc_asprintf(&r->xtab[(int)*kp].p,
+		    "%s%s", vp, font ? "\fP" : "");
+	} else {
+		roff_setstrn(&r->xmbtab, kp, ksz, vp, vsz, 0);
+		if (font)
+			roff_setstrn(&r->xmbtab, kp, ksz, "\\fP", 3, 1);
+	}
+	return ROFF_IGN;
+}
+
+static int
 roff_ec(ROFF_ARGS)
 {
 	const char	*p;