Accomodate for groff's crappy behaviour wherein an unrecognised

single-character escape (and ONLY this type of escape) will map back
into itself:

       "If a backslash is followed by a character that does not
	constitute a defined escape sequence the backslash is silently
        ignored and the  character maps to itself."

(From groff.7.)

Found by Jason McIntyre.

5 files changed, 27 insertions, 15 deletions

diff --git a/html.c b/html.c
index 5e19b09c..070562fb 100644
--- a/html.c
+++ b/html.c
@@ -1,4 +1,4 @@
-/*	$Id: html.c,v 1.107 2010/07/16 22:33:30 kristaps Exp $ */
+/*	$Id: html.c,v 1.108 2010/07/21 20:35:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -88,7 +88,8 @@ static	const char	*const htmlattrs[ATTR_MAX] = {
 	"summary",
 };
 
-static	void		  print_spec(struct html *, const char *, size_t);
+static	void		  print_spec(struct html *, enum roffdeco,
+				const char *, size_t);
 static	void		  print_res(struct html *, const char *, size_t);
 static	void		  print_ctag(struct html *, enum htmltag);
 static	void		  print_doctype(struct html *);
@@ -215,7 +216,7 @@ print_gen_head(struct html *h)
 
 
 static void
-print_spec(struct html *h, const char *p, size_t len)
+print_spec(struct html *h, enum roffdeco d, const char *p, size_t len)
 {
 	int		 cp;
 	const char	*rhs;
@@ -224,6 +225,9 @@ print_spec(struct html *h, const char *p, size_t len)
 	if ((cp = chars_spec2cp(h->symtab, p, len)) > 0) {
 		printf("&#%d;", cp);
 		return;
+	} else if (-1 == cp && DECO_SSPECIAL == d) {
+		fwrite(p, 1, len, stdout);
+		return;
 	} else if (-1 == cp)
 		return;
 
@@ -342,8 +346,10 @@ print_encode(struct html *h, const char *p, int norecurse)
 		case (DECO_RESERVED):
 			print_res(h, seq, sz);
 			break;
+		case (DECO_SSPECIAL):
+			/* FALLTHROUGH */
 		case (DECO_SPECIAL):
-			print_spec(h, seq, sz);
+			print_spec(h, deco, seq, sz);
 			break;
 		case (DECO_PREVIOUS):
 			/* FALLTHROUGH */
diff --git a/mandoc.c b/mandoc.c
index c001b637..69972c04 100644
--- a/mandoc.c
+++ b/mandoc.c
@@ -1,4 +1,4 @@
-/*	$Id: mandoc.c,v 1.24 2010/07/18 22:55:06 kristaps Exp $ */
+/*	$Id: mandoc.c,v 1.25 2010/07/21 20:35:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -314,7 +314,7 @@ mandoc_eos(const char *p, size_t sz, int enclosed)
 	 */
 
 	found = 0;
-	for (q = p + sz - 1; q >= p; q--) {
+	for (q = p + (int)sz - 1; q >= p; q--) {
 		switch (*q) {
 		case ('\"'):
 			/* FALLTHROUGH */
diff --git a/out.c b/out.c
index 1077aaed..62fda5cf 100644
--- a/out.c
+++ b/out.c
@@ -1,4 +1,4 @@
-/*	$Id: out.c,v 1.20 2010/07/19 07:53:40 kristaps Exp $ */
+/*	$Id: out.c,v 1.21 2010/07/21 20:35:03 kristaps Exp $ */
 /*
  * Copyright (c) 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -278,7 +278,7 @@ a2roffdeco(enum roffdeco *d, const char **word, size_t *sz)
 		*d = DECO_NOSPACE;
 		return(i);
 	default:
-		*d = DECO_SPECIAL;
+		*d = DECO_SSPECIAL;
 		i--;
 		lim = 1;
 		break;
diff --git a/out.h b/out.h
index fd1e5bd8..13665e46 100644
--- a/out.h
+++ b/out.h
@@ -1,4 +1,4 @@
-/*	$Id: out.h,v 1.13 2010/07/18 22:55:06 kristaps Exp $ */
+/*	$Id: out.h,v 1.14 2010/07/21 20:35:03 kristaps Exp $ */
 /*
  * Copyright (c) 2009 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -37,8 +37,9 @@ enum	roffscale {
 
 enum	roffdeco {
 	DECO_NONE,
-	DECO_SPECIAL,
-	DECO_RESERVED,
+	DECO_SPECIAL, /* special character */
+	DECO_SSPECIAL, /* single-char special */
+	DECO_RESERVED, /* reserved word */
 	DECO_BOLD,
 	DECO_ITALIC,
 	DECO_ROMAN,
diff --git a/term.c b/term.c
index de8c3380..dd7aac50 100644
--- a/term.c
+++ b/term.c
@@ -1,4 +1,4 @@
-/*	$Id: term.c,v 1.162 2010/07/17 12:01:43 kristaps Exp $ */
+/*	$Id: term.c,v 1.163 2010/07/21 20:35:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010 Ingo Schwarze <schwarze@openbsd.org>
@@ -34,7 +34,8 @@
 #include "term.h"
 #include "main.h"
 
-static	void		  spec(struct termp *, const char *, size_t);
+static	void		  spec(struct termp *, enum roffdeco,
+				const char *, size_t);
 static	void		  res(struct termp *, const char *, size_t);
 static	void		  buffera(struct termp *, const char *, size_t);
 static	void		  bufferc(struct termp *, char);
@@ -360,7 +361,7 @@ term_vspace(struct termp *p)
 
 
 static void
-spec(struct termp *p, const char *word, size_t len)
+spec(struct termp *p, enum roffdeco d, const char *word, size_t len)
 {
 	const char	*rhs;
 	size_t		 sz;
@@ -368,6 +369,8 @@ spec(struct termp *p, const char *word, size_t len)
 	rhs = chars_spec2str(p->symtab, word, len, &sz);
 	if (rhs) 
 		encode(p, rhs, sz);
+	else if (DECO_SSPECIAL == d)
+		encode(p, word, len);
 }
 
 
@@ -519,7 +522,9 @@ term_word(struct termp *p, const char *word)
 			res(p, seq, ssz);
 			break;
 		case (DECO_SPECIAL):
-			spec(p, seq, ssz);
+			/* FALLTHROUGH */
+		case (DECO_SSPECIAL):
+			spec(p, deco, seq, ssz);
 			break;
 		case (DECO_BOLD):
 			term_fontrepl(p, TERMFONT_BOLD);