The syntax of the roff(7) .mc request is quite special

and the roff_onearg() parsing function is too generic,
so provide a dedicated parsing function instead.

This fixes an assertion failure when an \o escape sequence is
passed as the argument; the bug was found by tb@ using afl(1).
It also makes mandoc output more similar to groff in various cases.

9 files changed, 139 insertions, 7 deletions

diff --git a/mandoc.1 b/mandoc.1
index 415769b3..d689fffa 100644
--- a/mandoc.1
+++ b/mandoc.1
@@ -1,4 +1,4 @@
-.\" $Id: mandoc.1,v 1.257 2022/04/24 13:38:46 schwarze Exp $
+.\" $Id: mandoc.1,v 1.258 2022/04/28 16:21:09 schwarze Exp $
 .\"
 .\" Copyright (c) 2012, 2014-2022 Ingo Schwarze <schwarze@openbsd.org>
 .\" Copyright (c) 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -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: April 24 2022 $
+.Dd $Mdocdate: April 28 2022 $
 .Dt MANDOC 1
 .Os
 .Sh NAME
@@ -1754,6 +1754,15 @@ request or a
 layout modifier has an unknown
 .Ar font
 argument.
+.It Sy "ignoring distance argument"
+.Pq roff
+In addition to the margin character, an
+.Ic \&mc
+request has a second argument supposed to represent a distance, but the
+.Nm
+implementation of
+.Ic \&mc
+always ignores the second argument.
 .It Sy "odd number of characters in request"
 .Pq roff
 A
@@ -2124,6 +2133,13 @@ The first argument of a
 request is neither a single ASCII character
 nor a single character escape sequence.
 The request is ignored including all its arguments.
+.It Sy "skipping unusable escape sequence"
+.Pq roff
+The first argument of an
+.Ic mc
+request is neither a single ASCII character
+nor a single character escape sequence.
+All arguments are ignored and printing of a margin character is disabled.
 .It Sy "missing manual name, using \(dq\(dq"
 .Pq mdoc
 The first call to
diff --git a/mandoc.h b/mandoc.h
index 32914deb..1ed0d816 100644
--- a/mandoc.h
+++ b/mandoc.h
@@ -1,4 +1,4 @@
-/* $Id: mandoc.h,v 1.275 2022/04/24 13:38:46 schwarze Exp $ */
+/* $Id: mandoc.h,v 1.276 2022/04/28 16:21:09 schwarze Exp $ */
 /*
  * Copyright (c) 2012-2022 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010, 2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -165,6 +165,7 @@ enum	mandocerr {
 	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_MC_DIST, /* ignoring distance argument: mc ... arg */
 	MANDOCERR_TR_ODD, /* odd number of characters in request: tr char */
 
 	/* related to plain text */
@@ -221,6 +222,7 @@ enum	mandocerr {
 	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_MC_ESC, /* skipping unusable escape sequence: mc arg */
 	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/mandoc_msg.c b/mandoc_msg.c
index 0361fc80..80a49d1d 100644
--- a/mandoc_msg.c
+++ b/mandoc_msg.c
@@ -166,6 +166,7 @@ static	const char *const type_message[MANDOCERR_MAX] = {
 	"invalid Boolean argument",
 	"argument contains two font escapes",
 	"unknown font, skipping request",
+	"ignoring distance argument",
 	"odd number of characters in request",
 
 	/* related to plain text */
@@ -222,6 +223,7 @@ static	const char *const type_message[MANDOCERR_MAX] = {
 	"missing list type, using -item",
 	"argument is not numeric, using 1",
 	"argument is not a character",
+	"skipping unusable escape sequence",
 	"missing manual name, using \"\"",
 	"uname(3) system call failed, using UNKNOWN",
 	"unknown standard specifier",
diff --git a/regress/roff/Makefile b/regress/roff/Makefile
index 9c9580f1..aecb314c 100644
--- a/regress/roff/Makefile
+++ b/regress/roff/Makefile
@@ -1,7 +1,7 @@
-# $OpenBSD: Makefile,v 1.28 2019/01/04 01:06:44 schwarze Exp $
+# $OpenBSD: Makefile,v 1.29 2022/04/28 16:16:46 schwarze Exp $
 
 SUBDIR  = args cond esc scale string
-SUBDIR += br cc ce char de ds ft ig it ll na nr po ps
+SUBDIR += br cc ce char de ds ft ig it ll mc na nr po ps
 SUBDIR += return rm rn shift sp ta ti tr while
 
 .include "../Makefile.sub"
diff --git a/regress/roff/mc/Makefile b/regress/roff/mc/Makefile
new file mode 100644
index 00000000..445af7d3
--- /dev/null
+++ b/regress/roff/mc/Makefile
@@ -0,0 +1,6 @@
+# $OpenBSD: Makefile,v 1.1 2022/04/28 16:16:46 schwarze Exp $
+
+REGRESS_TARGETS	 = args
+LINT_TARGETS	 = args
+
+.include <bsd.regress.mk>
diff --git a/regress/roff/mc/args.in b/regress/roff/mc/args.in
new file mode 100644
index 00000000..4e9b245f
--- /dev/null
+++ b/regress/roff/mc/args.in
@@ -0,0 +1,29 @@
+.\" $OpenBSD: args.in,v 1.1 2022/04/28 16:16:46 schwarze Exp $
+.TH MC-ARGS 1 "April 28, 2022"
+.SH NAME
+mc-args \- arguments to the .mc request
+.SH DESCRIPTION
+.ll 50n
+.nf
+initial text
+.mc |suffix
+ASCII character in the margin
+.mc \CXX
+invalid escape sequence
+.mc \!
+unsupported escape sequence
+.mc \a
+ignored escape sequence
+.mc \(ba
+special character in the margin
+.mc \[integral]
+special character represented as string
+.mc \fR
+font escape in the margin
+.mc \N'124'
+numbered character in the margin
+.mc \[u007C]suffix
+Unicode character in the margin
+.mc \o'o/'
+overstriking in the margin
+.ll
diff --git a/regress/roff/mc/args.out_ascii b/regress/roff/mc/args.out_ascii
new file mode 100644
index 00000000..d1a56be6
--- /dev/null
+++ b/regress/roff/mc/args.out_ascii
@@ -0,0 +1,19 @@
+MC-ARGS(1)                  General Commands Manual                 MC-ARGS(1)
+
+NNAAMMEE
+       mc-args - arguments to the .mc request
+
+DDEESSCCRRIIPPTTIIOONN
+       initial text
+       ASCII character in the margin               |
+       invalid escape sequence
+       unsupported escape sequence
+       ignored escape sequence
+       special character in the margin             |
+       special character represented as string     <integral>
+       font escape in the margin
+       numbered character in the margin            |
+       Unicode character in the margin             |
+       overstriking in the margin
+
+OpenBSD                         April 28, 2022                      MC-ARGS(1)
diff --git a/regress/roff/mc/args.out_lint b/regress/roff/mc/args.out_lint
new file mode 100644
index 00000000..56ed4e28
--- /dev/null
+++ b/regress/roff/mc/args.out_lint
@@ -0,0 +1,9 @@
+mandoc: args.in:9:6: WARNING: ignoring distance argument: mc ... suffix
+mandoc: args.in:11:5: WARNING: invalid escape sequence: \C
+mandoc: args.in:11:5: ERROR: skipping unusable escape sequence: mc \C
+mandoc: args.in:13:5: UNSUPP: unsupported escape sequence: \!
+mandoc: args.in:13:5: ERROR: skipping unusable escape sequence: mc \!
+mandoc: args.in:15:5: ERROR: skipping unusable escape sequence: mc \a
+mandoc: args.in:21:5: ERROR: skipping unusable escape sequence: mc \fR
+mandoc: args.in:25:13: WARNING: ignoring distance argument: mc ... suffix
+mandoc: args.in:27:5: ERROR: skipping unusable escape sequence: mc \o'o/'
diff --git a/roff.c b/roff.c
index 328855c0..7bed9429 100644
--- a/roff.c
+++ b/roff.c
@@ -1,4 +1,4 @@
-/* $Id: roff.c,v 1.383 2022/04/24 17:40:22 schwarze Exp $ */
+/* $Id: roff.c,v 1.384 2022/04/28 16:21:10 schwarze Exp $ */
 /*
  * Copyright (c) 2010-2015, 2017-2022 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -227,6 +227,7 @@ static	int		 roff_line_ignore(ROFF_ARGS);
 static	void		 roff_man_alloc1(struct roff_man *);
 static	void		 roff_man_free1(struct roff_man *);
 static	int		 roff_manyarg(ROFF_ARGS);
+static	int		 roff_mc(ROFF_ARGS);
 static	int		 roff_noarg(ROFF_ARGS);
 static	int		 roff_nop(ROFF_ARGS);
 static	int		 roff_nr(ROFF_ARGS);
@@ -379,7 +380,7 @@ static	struct roffmac	 roffs[TOKEN_NONE] = {
 	{ roff_noarg, NULL, NULL, 0 },  /* fi */
 	{ roff_onearg, NULL, NULL, 0 },  /* ft */
 	{ roff_onearg, NULL, NULL, 0 },  /* ll */
-	{ roff_onearg, NULL, NULL, 0 },  /* mc */
+	{ roff_mc, NULL, NULL, 0 },  /* mc */
 	{ roff_noarg, NULL, NULL, 0 },  /* nf */
 	{ roff_onearg, NULL, NULL, 0 },  /* po */
 	{ roff_onearg, NULL, NULL, 0 },  /* rj */
@@ -3732,6 +3733,54 @@ roff_eo(ROFF_ARGS)
 }
 
 static int
+roff_mc(ROFF_ARGS)
+{
+	struct roff_node	*n;
+	char			*cp;
+
+	/* Parse the first argument. */
+
+	cp = buf->buf + pos;
+	if (*cp != '\0')
+		cp++;
+	if (buf->buf[pos] == '\\') {
+		switch (mandoc_escape((const char **)&cp, NULL, NULL)) {
+		case ESCAPE_SPECIAL:
+		case ESCAPE_UNICODE:
+		case ESCAPE_NUMBERED:
+			break;
+		default:
+			*cp = '\0';
+			mandoc_msg(MANDOCERR_MC_ESC, ln, pos,
+			    "mc %s", buf->buf + pos);
+			buf->buf[pos] = '\0';
+			break;
+		}
+	}
+
+	/* Ignore additional arguments. */
+
+	while (*cp == ' ')
+		*cp++ = '\0';
+	if (*cp != '\0') {
+		mandoc_msg(MANDOCERR_MC_DIST, ln, (int)(cp - buf->buf),
+		    "mc ... %s", cp);
+		*cp = '\0';
+	}
+
+	/* Create the .mc node. */
+
+	roff_elem_alloc(r->man, ln, ppos, tok);
+	n = r->man->last;
+	if (buf->buf[pos] != '\0')
+		roff_word_alloc(r->man, ln, pos, buf->buf + pos);
+	n->flags |= NODE_LINE | NODE_VALID | NODE_ENDED;
+	r->man->last = n;
+	r->man->next = ROFF_NEXT_SIBLING;
+	return ROFF_IGN;
+}
+
+static int
 roff_nop(ROFF_ARGS)
 {
 	while (buf->buf[pos] == ' ')