===================================================================
RCS file: /cvs/mandoc/mandocdb.c,v
retrieving revision 1.273
retrieving revision 1.275
diff -u -p -r1.273 -r1.275
--- mandoc/mandocdb.c	2024/05/14 18:52:31	1.273
+++ mandoc/mandocdb.c	2025/06/05 12:33:41	1.275
@@ -1,6 +1,6 @@
-/* $Id: mandocdb.c,v 1.273 2024/05/14 18:52:31 schwarze Exp $ */
+/* $Id: mandocdb.c,v 1.275 2025/06/05 12:33:41 schwarze Exp $ */
 /*
- * Copyright (c) 2011-2021 Ingo Schwarze <schwarze@openbsd.org>
+ * Copyright (c) 2011-2021, 2024, 2025 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2011, 2012 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2016 Ed Maste <emaste@freebsd.org>
  *
@@ -151,6 +151,8 @@ static	int	 parse_mdoc_Fn(struct mpage *, const struct
 			const struct roff_node *);
 static	int	 parse_mdoc_Fo(struct mpage *, const struct roff_meta *,
 			const struct roff_node *);
+static	int	 parse_mdoc_Lb(struct mpage *, const struct roff_meta *,
+			const struct roff_node *);
 static	int	 parse_mdoc_Nd(struct mpage *, const struct roff_meta *,
 			const struct roff_node *);
 static	int	 parse_mdoc_Nm(struct mpage *, const struct roff_meta *,
@@ -173,7 +175,7 @@ static	void	 say(const char *, const char *, ...)
 			__attribute__((__format__ (__printf__, 2, 3)));
 static	int	 set_basedir(const char *, int);
 static	int	 treescan(void);
-static	size_t	 utf8(unsigned int, char [7]);
+static	size_t	 utf8(unsigned int, char[5]);
 
 static	int		 nodb; /* no database changes */
 static	int		 mparse_options; /* abort the parse early */
@@ -299,7 +301,7 @@ static	const struct mdoc_handler mdoc_handlers[MDOC_MA
 	{ NULL, 0, 0 },  /* Hf */
 	{ NULL, 0, 0 },  /* Fr */
 	{ NULL, 0, 0 },  /* Ud */
-	{ NULL, TYPE_Lb, NODE_NOSRC },  /* Lb */
+	{ parse_mdoc_Lb, 0, 0 },  /* Lb */
 	{ NULL, 0, 0 },  /* Lp */
 	{ NULL, TYPE_Lk, 0 },  /* Lk */
 	{ NULL, TYPE_Mt, NODE_NOSRC },  /* Mt */
@@ -1754,6 +1756,25 @@ parse_mdoc_Fo(struct mpage *mpage, const struct roff_m
 }
 
 static int
+parse_mdoc_Lb(struct mpage *mpage, const struct roff_meta *meta,
+	const struct roff_node *n)
+{
+	char *cp;
+
+	for (n = n->child; n != NULL; n = n->next) {
+		if (n->flags & NODE_NOSRC)
+			continue;
+		cp = n->string;
+		if (n->sec == SEC_SYNOPSIS)
+			mandoc_asprintf(&cp, "lib%s", cp);
+		putkey(mpage, cp, TYPE_Lb);
+		if (n->sec == SEC_SYNOPSIS)
+			free(cp);
+	}
+	return 0;
+}
+
+static int
 parse_mdoc_Va(struct mpage *mpage, const struct roff_meta *meta,
 	const struct roff_node *n)
 {
@@ -1907,49 +1928,35 @@ putkeys(const struct mpage *mpage, char *cp, size_t sz
  * Take a Unicode codepoint and produce its UTF-8 encoding.
  * This isn't the best way to do this, but it works.
  * The magic numbers are from the UTF-8 packaging.
- * They're not as scary as they seem: read the UTF-8 spec for details.
+ * Read the UTF-8 spec or the utf8(7) manual page for details.
  */
 static size_t
-utf8(unsigned int cp, char out[7])
+utf8(unsigned int cp, char out[5])
 {
 	size_t		 rc;
 
-	rc = 0;
-	if (cp <= 0x0000007F) {
+	if (cp <= 0x7f) {
 		rc = 1;
 		out[0] = (char)cp;
-	} else if (cp <= 0x000007FF) {
+	} else if (cp <= 0x7ff) {
 		rc = 2;
 		out[0] = (cp >> 6  & 31) | 192;
 		out[1] = (cp       & 63) | 128;
-	} else if (cp <= 0x0000FFFF) {
+	} else if (cp >= 0xd800 && cp <= 0xdfff) {
+		rc = 0; /* reject UTF-16 surrogate */
+	} else if (cp <= 0xffff) {
 		rc = 3;
 		out[0] = (cp >> 12 & 15) | 224;
 		out[1] = (cp >> 6  & 63) | 128;
 		out[2] = (cp       & 63) | 128;
-	} else if (cp <= 0x001FFFFF) {
+	} else if (cp <= 0x10ffff) {
 		rc = 4;
 		out[0] = (cp >> 18 &  7) | 240;
 		out[1] = (cp >> 12 & 63) | 128;
 		out[2] = (cp >> 6  & 63) | 128;
 		out[3] = (cp       & 63) | 128;
-	} else if (cp <= 0x03FFFFFF) {
-		rc = 5;
-		out[0] = (cp >> 24 &  3) | 248;
-		out[1] = (cp >> 18 & 63) | 128;
-		out[2] = (cp >> 12 & 63) | 128;
-		out[3] = (cp >> 6  & 63) | 128;
-		out[4] = (cp       & 63) | 128;
-	} else if (cp <= 0x7FFFFFFF) {
-		rc = 6;
-		out[0] = (cp >> 30 &  1) | 252;
-		out[1] = (cp >> 24 & 63) | 128;
-		out[2] = (cp >> 18 & 63) | 128;
-		out[3] = (cp >> 12 & 63) | 128;
-		out[4] = (cp >> 6  & 63) | 128;
-		out[5] = (cp       & 63) | 128;
 	} else
-		return 0;
+		rc = 0;
 
 	out[rc] = '\0';
 	return rc;