Do not write duplicate id= attributes, they violate HTML syntax.

Append suffixes for disambiguation.  Issue first reported by Jakub
Klinkovsky <j dot l dot k at gmx dot com> (Arch Linux).

4 files changed, 56 insertions, 29 deletions

diff --git a/usr.bin/mandoc/html.c b/usr.bin/mandoc/html.c
index 211eca10bbf..8afcc21b68c 100644
--- a/usr.bin/mandoc/html.c
+++ b/usr.bin/mandoc/html.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: html.c,v 1.98 2018/05/21 01:10:06 schwarze Exp $ */
+/*	$OpenBSD: html.c,v 1.99 2018/05/25 20:23:39 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2011-2015, 2017, 2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -20,6 +20,7 @@
 #include <assert.h>
 #include <ctype.h>
 #include <stdarg.h>
+#include <stddef.h>
 #include <stdio.h>
 #include <stdint.h>
 #include <stdlib.h>
@@ -27,6 +28,7 @@
 #include <unistd.h>
 
 #include "mandoc_aux.h"
+#include "mandoc_ohash.h"
 #include "mandoc.h"
 #include "roff.h"
 #include "out.h"
@@ -115,6 +117,9 @@ static	const char	*const roffscales[SCALE_MAX] = {
 	"ex", /* SCALE_FS */
 };
 
+/* Avoid duplicate HTML id= attributes. */
+static	struct ohash	 id_unique;
+
 static	void	 a2width(const char *, struct roffsu *);
 static	void	 print_byte(struct html *, char);
 static	void	 print_endword(struct html *);
@@ -142,6 +147,8 @@ html_alloc(const struct manoutput *outopts)
 	if (outopts->fragment)
 		h->oflags |= HTML_FRAGMENT;
 
+	mandoc_ohash_init(&id_unique, 4, 0);
+
 	return h;
 }
 
@@ -150,15 +157,22 @@ html_free(void *p)
 {
 	struct tag	*tag;
 	struct html	*h;
+	char		*cp;
+	unsigned int	 slot;
 
 	h = (struct html *)p;
-
 	while ((tag = h->tag) != NULL) {
 		h->tag = tag->next;
 		free(tag);
 	}
-
 	free(h);
+
+	cp = ohash_first(&id_unique, &slot);
+	while (cp != NULL) {
+		free(cp);
+		cp = ohash_next(&id_unique, &slot);
+	}
+	ohash_delete(&id_unique);
 }
 
 void
@@ -255,10 +269,12 @@ print_metaf(struct html *h, enum mandoc_esc deco)
 }
 
 char *
-html_make_id(const struct roff_node *n)
+html_make_id(const struct roff_node *n, int unique)
 {
 	const struct roff_node	*nch;
-	char			*buf, *cp;
+	char			*buf, *bufs, *cp;
+	unsigned int		 slot;
+	int			 suffix;
 
 	for (nch = n->child; nch != NULL; nch = nch->next)
 		if (nch->type != ROFFT_TEXT)
@@ -275,6 +291,30 @@ html_make_id(const struct roff_node *n)
 		if (*cp == ' ')
 			*cp = '_';
 
+	if (unique == 0)
+		return buf;
+
+	/* Avoid duplicate HTML id= attributes. */
+
+	bufs = NULL;
+	suffix = 1;
+	slot = ohash_qlookup(&id_unique, buf);
+	cp = ohash_find(&id_unique, slot);
+	if (cp != NULL) {
+		while (cp != NULL) {
+			free(bufs);
+			if (++suffix > 127) {
+				free(buf);
+				return NULL;
+			}
+			mandoc_asprintf(&bufs, "%s_%d", buf, suffix);
+			slot = ohash_qlookup(&id_unique, bufs);
+			cp = ohash_find(&id_unique, slot);
+		}
+		free(buf);
+		buf = bufs;
+	}
+	ohash_insert(&id_unique, slot, buf);
 	return buf;
 }
 
diff --git a/usr.bin/mandoc/html.h b/usr.bin/mandoc/html.h
index d44224fe64c..7768fb99f43 100644
--- a/usr.bin/mandoc/html.h
+++ b/usr.bin/mandoc/html.h
@@ -1,4 +1,4 @@
-/*	$OpenBSD: html.h,v 1.51 2018/05/09 00:45:33 schwarze Exp $ */
+/*	$OpenBSD: html.h,v 1.52 2018/05/25 20:23:39 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2017, 2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -133,5 +133,5 @@ void		  print_eqn(struct html *, const struct eqn_box *);
 void		  print_paragraph(struct html *);
 void		  print_endline(struct html *);
 
-char		 *html_make_id(const struct roff_node *);
+char		 *html_make_id(const struct roff_node *, int);
 int		  html_strlen(const char *);
diff --git a/usr.bin/mandoc/man_html.c b/usr.bin/mandoc/man_html.c
index c7f7ea123d6..49d89429551 100644
--- a/usr.bin/mandoc/man_html.c
+++ b/usr.bin/mandoc/man_html.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: man_html.c,v 1.102 2018/05/08 21:42:11 schwarze Exp $ */
+/*	$OpenBSD: man_html.c,v 1.103 2018/05/25 20:23:39 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2013,2014,2015,2017,2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -426,11 +426,10 @@ man_SH_pre(MAN_ARGS)
 	char	*id;
 
 	if (n->type == ROFFT_HEAD) {
-		id = html_make_id(n);
+		id = html_make_id(n, 1);
 		print_otag(h, TAG_H1, "cTi", "Sh", id);
 		if (id != NULL)
 			print_otag(h, TAG_A, "chR", "permalink", id);
-		free(id);
 	}
 	return 1;
 }
@@ -496,11 +495,10 @@ man_SS_pre(MAN_ARGS)
 	char	*id;
 
 	if (n->type == ROFFT_HEAD) {
-		id = html_make_id(n);
+		id = html_make_id(n, 1);
 		print_otag(h, TAG_H2, "cTi", "Ss", id);
 		if (id != NULL)
 			print_otag(h, TAG_A, "chR", "permalink", id);
-		free(id);
 	}
 	return 1;
 }
diff --git a/usr.bin/mandoc/mdoc_html.c b/usr.bin/mandoc/mdoc_html.c
index 072dcb0df64..d1e43c0ffa4 100644
--- a/usr.bin/mandoc/mdoc_html.c
+++ b/usr.bin/mandoc/mdoc_html.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: mdoc_html.c,v 1.178 2018/05/21 01:10:06 schwarze Exp $ */
+/*	$OpenBSD: mdoc_html.c,v 1.179 2018/05/25 20:23:39 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2014,2015,2016,2017,2018 Ingo Schwarze <schwarze@openbsd.org>
@@ -500,7 +500,7 @@ cond_id(const struct roff_node *n)
 	     (n->parent->tok == MDOC_Xo &&
 	      n->parent->parent->prev == NULL &&
 	      n->parent->parent->parent->tok == MDOC_It)))
-		return html_make_id(n);
+		return html_make_id(n, 1);
 	return NULL;
 }
 
@@ -511,11 +511,10 @@ mdoc_sh_pre(MDOC_ARGS)
 
 	switch (n->type) {
 	case ROFFT_HEAD:
-		id = html_make_id(n);
+		id = html_make_id(n, 1);
 		print_otag(h, TAG_H1, "cTi", "Sh", id);
 		if (id != NULL)
 			print_otag(h, TAG_A, "chR", "permalink", id);
-		free(id);
 		break;
 	case ROFFT_BODY:
 		if (n->sec == SEC_AUTHORS)
@@ -535,11 +534,10 @@ mdoc_ss_pre(MDOC_ARGS)
 	if (n->type != ROFFT_HEAD)
 		return 1;
 
-	id = html_make_id(n);
+	id = html_make_id(n, 1);
 	print_otag(h, TAG_H2, "cTi", "Ss", id);
 	if (id != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
-	free(id);
 	return 1;
 }
 
@@ -551,7 +549,6 @@ mdoc_fl_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Fl", id);
-	free(id);
 
 	print_text(h, "\\-");
 	if (!(n->child == NULL &&
@@ -571,7 +568,6 @@ mdoc_cm_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Cm", id);
-	free(id);
 	return 1;
 }
 
@@ -880,7 +876,7 @@ mdoc_sx_pre(MDOC_ARGS)
 {
 	char	*id;
 
-	id = html_make_id(n);
+	id = html_make_id(n, 0);
 	print_otag(h, TAG_A, "cThR", "Sx", id);
 	free(id);
 	return 1;
@@ -1028,7 +1024,6 @@ mdoc_dv_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Dv", id);
-	free(id);
 	return 1;
 }
 
@@ -1040,7 +1035,6 @@ mdoc_ev_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Ev", id);
-	free(id);
 	return 1;
 }
 
@@ -1053,12 +1047,11 @@ mdoc_er_pre(MDOC_ARGS)
 	    (n->parent->tok == MDOC_It ||
 	     (n->parent->tok == MDOC_Bq &&
 	      n->parent->parent->parent->tok == MDOC_It)) ?
-	    html_make_id(n) : NULL;
+	    html_make_id(n, 1) : NULL;
 
 	if (id != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Er", id);
-	free(id);
 	return 1;
 }
 
@@ -1409,7 +1402,6 @@ mdoc_ic_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "cTi", "Ic", id);
-	free(id);
 	return 1;
 }
 
@@ -1462,7 +1454,6 @@ mdoc_ms_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_SPAN, "cTi", "Ms", id);
-	free(id);
 	return 1;
 }
 
@@ -1503,7 +1494,6 @@ mdoc_no_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_SPAN, "ci", "No", id);
-	free(id);
 	return 1;
 }
 
@@ -1515,7 +1505,6 @@ mdoc_li_pre(MDOC_ARGS)
 	if ((id = cond_id(n)) != NULL)
 		print_otag(h, TAG_A, "chR", "permalink", id);
 	print_otag(h, TAG_CODE, "ci", "Li", id);
-	free(id);
 	return 1;
 }