Documented that `\s' and `\f' don't work in HTML mode (and why).

Added support for recognising the many forms of `\s' (doesn't yet render).

3 files changed, 103 insertions, 17 deletions

diff --git a/html.c b/html.c
index 094cc86e..d64385a5 100644
--- a/html.c
+++ b/html.c
@@ -270,21 +270,19 @@ print_escape(struct html *h, const char **p)
 			return;
 		}
 
-		switch (*wp) {
-		case ('B'):
-			/* TODO */
-			break;
-		case ('I'):
-			/* TODO */
-			break;
-		case ('P'):
-			/* FALLTHROUGH */
-		case ('R'):
-			/* TODO */
-			break;
-		default:
-			break;
-		}
+		/* 
+		 * These aren't supported, as they're symmetry-breaking
+		 * constructs that don't play well with hierarchical
+		 * mark-up.  Consider:
+		 *
+		 * \fBHello.
+		 * .PP
+		 * World.
+		 *
+		 * The style started before "Hello" wouldn't be able to
+		 * propogate into the next `PP' because we'd exit the
+		 * current paragraph's scope.
+		 */
 
 		*p = wp;
 		return;
diff --git a/mandoc.1 b/mandoc.1
index 0aebd7eb..1e6dd4b3 100644
--- a/mandoc.1
+++ b/mandoc.1
@@ -419,3 +419,14 @@ the maximum size of an element attribute is determined by
 which is usually 1024 bytes.  Be aware of this when setting long link
 formats with
 .Fl O Ns Ar man=fmt .
+.Pp
+The
+.Fl T Ns Ar html
+utility doesn't support the
+.Sq \ef
+and
+.Sq \es
+text decorations documented in
+.Xr mdoc 7
+and
+.Xr man 7 .
diff --git a/mandoc.c b/mandoc.c
index 1099ca4d..58c2d273 100644
--- a/mandoc.c
+++ b/mandoc.c
@@ -35,7 +35,9 @@ static int	 a2time(time_t *, const char *, const char *);
 int
 mandoc_special(const char *p)
 {
-	int		 c;
+	int		 terminator;	/* Terminator for \s. */
+	int		 lim;		/* Limit for N in \s. */
+	int		 c, i;
 	
 	if ('\\' != *p++)
 		return(0);
@@ -74,9 +76,84 @@ mandoc_special(const char *p)
 	case ('e'):
 		return(2);
 	case ('f'):
-		if (0 == *++p || ! isgraph((u_char)*p))
+		if ('\0' == *++p || ! isgraph((u_char)*p))
 			return(0);
 		return(3);
+	case ('s'):
+		if ('\0' == *++p)
+			return(2);
+
+		c = 2;
+		terminator = 0;
+		lim = 1;
+
+		if (*p == '\'') {
+			lim = 0;
+			terminator = 1;
+			++p;
+			++c;
+		} else if (*p == '[') {
+			lim = 0;
+			terminator = 2;
+			++p;
+			++c;
+		} else if (*p == '(') {
+			lim = 2;
+			terminator = 3;
+			++p;
+			++c;
+		}
+
+		if (*p == '+' || *p == '-') {
+			++p;
+			++c;
+		}
+
+		if (*p == '\'') {
+			if (terminator)
+				return(0);
+			lim = 0;
+			terminator = 1;
+			++p;
+			++c;
+		} else if (*p == '[') {
+			if (terminator)
+				return(0);
+			lim = 0;
+			terminator = 2;
+			++p;
+			++c;
+		} else if (*p == '(') {
+			if (terminator)
+				return(0);
+			lim = 2;
+			terminator = 3;
+			++p;
+			++c;
+		}
+
+		/* TODO: needs to handle floating point. */
+
+		if ( ! isdigit((u_char)*p))
+			return(0);
+
+		for (i = 0; isdigit((u_char)*p); i++) {
+			if (lim && i >= lim)
+				break;
+			++p;
+			++c;
+		}
+
+		if (terminator && terminator < 3) {
+			if (1 == terminator && *p != '\'')
+				return(0);
+			if (2 == terminator && *p != ']')
+				return(0);
+			++p;
+			++c;
+		}
+
+		return(c);
 	case ('*'):
 		if (0 == *++p || ! isgraph((u_char)*p))
 			return(0);