From 3e52d29d3db263b73e601c30a30d8e687795cce9 Mon Sep 17 00:00:00 2001
From: Ingo Schwarze <schwarze@cvs.openbsd.org>
Date: Mon, 20 Apr 2020 12:59:26 +0000
Subject: In fragment identifiers, use ~%d for ordinal suffixes, and reserve
 the character '~' for that purpose.

Bug found by validator.w3.org in openssl(1), which contains both a
tag "tls1_2" and a second instance of a tag "tls1", which also resulted
in "tls1_2", causing a clash.  Now, the second instance of "tls1" is
rendered as "tls1~2" instead, employing the newly reserved '~'.
---
 regress/usr.bin/mandoc/man/IP/literal.out_html | 6 +++---
 regress/usr.bin/mandoc/man/TP/literal.out_html | 2 +-
 regress/usr.bin/mandoc/man/TP/vert.out_html    | 2 +-
 regress/usr.bin/mandoc/mdoc/Bd/nested.out_html | 4 ++--
 regress/usr.bin/mandoc/mdoc/Sh/tag.out_html    | 2 +-
 usr.bin/mandoc/html.c                          | 7 ++++---
 6 files changed, 12 insertions(+), 11 deletions(-)

diff --git a/regress/usr.bin/mandoc/man/IP/literal.out_html b/regress/usr.bin/mandoc/man/IP/literal.out_html
index 3b9cc429a3c..aa0536e795c 100644
--- a/regress/usr.bin/mandoc/man/IP/literal.out_html
+++ b/regress/usr.bin/mandoc/man/IP/literal.out_html
@@ -8,7 +8,7 @@ literal
 text
 </pre>
 <dl class="Bl-tag">
-  <dt id="tag_2"><a class="permalink" href="#tag_2">tag</a></dt>
+  <dt id="tag~2"><a class="permalink" href="#tag~2">tag</a></dt>
   <dd>
     <pre>
 indented
@@ -32,7 +32,7 @@ literal
 text
 </pre>
 <dl class="Bl-tag">
-  <dt id="tag_3"><a class="permalink" href="#tag_3">tag</a></dt>
+  <dt id="tag~3"><a class="permalink" href="#tag~3">tag</a></dt>
   <dd>
     <pre>
 indented
@@ -48,7 +48,7 @@ text
   out of indented paragraph</a></h2>
 <p class="Pp">regular text</p>
 <dl class="Bl-tag">
-  <dt id="tag_4"><a class="permalink" href="#tag_4">tag</a></dt>
+  <dt id="tag~4"><a class="permalink" href="#tag~4">tag</a></dt>
   <dd>indented regular text
     <pre>
 indented
diff --git a/regress/usr.bin/mandoc/man/TP/literal.out_html b/regress/usr.bin/mandoc/man/TP/literal.out_html
index 8b818ed8c68..914c7b5c6ef 100644
--- a/regress/usr.bin/mandoc/man/TP/literal.out_html
+++ b/regress/usr.bin/mandoc/man/TP/literal.out_html
@@ -8,7 +8,7 @@ literal
 text
 </pre>
 <dl class="Bl-tag">
-  <dt id="tag_2"><a class="permalink" href="#tag_2">tag</a></dt>
+  <dt id="tag~2"><a class="permalink" href="#tag~2">tag</a></dt>
   <dd>
     <pre>
 indented
diff --git a/regress/usr.bin/mandoc/man/TP/vert.out_html b/regress/usr.bin/mandoc/man/TP/vert.out_html
index 1124ed33dd7..4499432b1a5 100644
--- a/regress/usr.bin/mandoc/man/TP/vert.out_html
+++ b/regress/usr.bin/mandoc/man/TP/vert.out_html
@@ -4,6 +4,6 @@
 <dl class="Bl-tag">
   <dt id="tag"><a class="permalink" href="#tag">tag</a></dt>
   <dd>text</dd>
-  <dt id="tag_2"><a class="permalink" href="#tag_2">tag</a></dt>
+  <dt id="tag~2"><a class="permalink" href="#tag~2">tag</a></dt>
   <dd>text</dd>
 </dl>
diff --git a/regress/usr.bin/mandoc/mdoc/Bd/nested.out_html b/regress/usr.bin/mandoc/mdoc/Bd/nested.out_html
index bc71c2700f2..9476189599a 100644
--- a/regress/usr.bin/mandoc/mdoc/Bd/nested.out_html
+++ b/regress/usr.bin/mandoc/mdoc/Bd/nested.out_html
@@ -8,10 +8,10 @@
 outer text</div>
 regular text
 <div class="Bd Pp
-  Bd-indent" id="outer_2"><a class="permalink" href="#outer_2">outer</a> text
+  Bd-indent" id="outer~2"><a class="permalink" href="#outer~2">outer</a> text
   (4n)
 <div class="Bd Pp
-  Bd-indent" id="inner_2"><a class="permalink" href="#inner_2">inner</a> text
+  Bd-indent" id="inner~2"><a class="permalink" href="#inner~2">inner</a> text
   (2n)</div>
 outer text
 <dl class="Bl-tag">
diff --git a/regress/usr.bin/mandoc/mdoc/Sh/tag.out_html b/regress/usr.bin/mandoc/mdoc/Sh/tag.out_html
index d99c6d4da69..0dd39398470 100644
--- a/regress/usr.bin/mandoc/mdoc/Sh/tag.out_html
+++ b/regress/usr.bin/mandoc/mdoc/Sh/tag.out_html
@@ -2,7 +2,7 @@
 </section>
 </section>
 <section class="Sh">
-<h1 class="Sh" id="DESCRIPTION_2"><a class="permalink" href="#DESCRIPTION_2">DESCRIPTION</a></h1>
+<h1 class="Sh" id="DESCRIPTION~2"><a class="permalink" href="#DESCRIPTION~2">DESCRIPTION</a></h1>
 <p class="Pp">Text in duplicate description section.</p>
 </section>
 <section class="Sh">
diff --git a/usr.bin/mandoc/html.c b/usr.bin/mandoc/html.c
index b95c609ae79..8892009d9f6 100644
--- a/usr.bin/mandoc/html.c
+++ b/usr.bin/mandoc/html.c
@@ -1,4 +1,4 @@
-/* $OpenBSD: html.c,v 1.140 2020/04/19 15:15:54 schwarze Exp $ */
+/* $OpenBSD: html.c,v 1.141 2020/04/20 12:59:24 schwarze Exp $ */
 /*
  * Copyright (c) 2011-2015, 2017-2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2008-2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
@@ -381,11 +381,12 @@ html_make_id(const struct roff_node *n, int unique)
 	 * permitted in URL-fragment strings according to the
 	 * explicit list at:
 	 * https://url.spec.whatwg.org/#url-fragment-string
+	 * In addition, reserve '~' for ordinal suffixes.
 	 */
 
 	for (cp = buf; *cp != '\0'; cp++)
 		if (isalnum((unsigned char)*cp) == 0 &&
-		    strchr("!$&'()*+,-./:;=?@_~", *cp) == NULL)
+		    strchr("!$&'()*+,-./:;=?@_", *cp) == NULL)
 			*cp = '_';
 
 	if (unique == 0)
@@ -405,7 +406,7 @@ html_make_id(const struct roff_node *n, int unique)
 
 	if (entry->ord > 1) {
 		cp = buf;
-		mandoc_asprintf(&buf, "%s_%d", cp, entry->ord);
+		mandoc_asprintf(&buf, "%s~%d", cp, entry->ord);
 		free(cp);
 	}
 	return buf;
-- 
cgit v1.2.3