Fully support explicit tagging of .Sh and .Ss.

This fixes the offset of two lines in terminal output
and this improves HTML output by putting the id= attribute
and <a> element into the respective <h1> or <h2> element rather
than writing an additional <mark> element.

To that end, introduce node flags NODE_ID (to make the node a link
target, for example by writing an HTML id= attribute or by calling
tag_put()) and NODE_HREF (to make the node a link source, used only
in HTML output, used only to write an <a class="permalink"> element).

In particular:
* In the validator, generalize the concept of the "next node"
such that it also works before .Sh and .Ss.
* If the first argument of .Tg is empty, don't forget to complain
if there are additional arguments, which will be ignored.
* In the terminal formatter, support writing of explicit tags
for all kinds of nodes, not just for .Tg.
* In deroff(), allow nodes to have an explicit string representation
even when they aren't text nodes.  Use this for explicitly tagged
section headers.  Suprisingly, this is sufficient to make HTML
output work, without explicit code changes in the HTML formatter.
* In syntax tree output, display NODE_ID and NODE_HREF.

5 files changed, 64 insertions, 20 deletions

diff --git a/usr.bin/mandoc/mdoc_term.c b/usr.bin/mandoc/mdoc_term.c
index dab449d9da4..54252473702 100644
--- a/usr.bin/mandoc/mdoc_term.c
+++ b/usr.bin/mandoc/mdoc_term.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: mdoc_term.c,v 1.276 2020/02/27 01:25:57 schwarze Exp $ */
+/*	$OpenBSD: mdoc_term.c,v 1.277 2020/02/27 21:38:27 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010, 2012-2020 Ingo Schwarze <schwarze@openbsd.org>
@@ -115,7 +115,6 @@ static	int	  termp_pp_pre(DECL_ARGS);
 static	int	  termp_ss_pre(DECL_ARGS);
 static	int	  termp_sy_pre(DECL_ARGS);
 static	int	  termp_tag_pre(DECL_ARGS);
-static	int	  termp_tg_pre(DECL_ARGS);
 static	int	  termp_under_pre(DECL_ARGS);
 static	int	  termp_vt_pre(DECL_ARGS);
 static	int	  termp_xr_pre(DECL_ARGS);
@@ -242,7 +241,7 @@ static const struct mdoc_term_act mdoc_term_acts[MDOC_MAX - MDOC_Dd] = {
 	{ NULL, termp____post }, /* %Q */
 	{ NULL, termp____post }, /* %U */
 	{ NULL, NULL }, /* Ta */
-	{ termp_tg_pre, NULL }, /* Tg */
+	{ termp_skip_pre, NULL }, /* Tg */
 };
 
 static	int	 fn_prio = TAG_STRONG;
@@ -339,6 +338,10 @@ print_mdoc_node(DECL_ARGS)
 	memset(&npair, 0, sizeof(struct termpair));
 	npair.ppair = pair;
 
+	if (n->flags & NODE_ID)
+		tag_put(n->string == NULL ? n->child->string : n->string,
+		    TAG_MANUAL, p->line);
+
 	/*
 	 * Keeps only work until the end of a line.  If a keep was
 	 * invoked in a prior line, revert it to PREKEEP.
@@ -2064,13 +2067,6 @@ termp_tag_pre(DECL_ARGS)
 }
 
 static int
-termp_tg_pre(DECL_ARGS)
-{
-	tag_put(n->child->string, TAG_MANUAL, p->line);
-	return 0;
-}
-
-static int
 termp_abort_pre(DECL_ARGS)
 {
 	abort();
diff --git a/usr.bin/mandoc/mdoc_validate.c b/usr.bin/mandoc/mdoc_validate.c
index 902294e8d79..033a368cd75 100644
--- a/usr.bin/mandoc/mdoc_validate.c
+++ b/usr.bin/mandoc/mdoc_validate.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: mdoc_validate.c,v 1.293 2020/02/27 01:25:58 schwarze Exp $ */
+/*	$OpenBSD: mdoc_validate.c,v 1.294 2020/02/27 21:38:27 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2012 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2020 Ingo Schwarze <schwarze@openbsd.org>
@@ -1079,21 +1079,32 @@ post_st(POST_ARGS)
 static void
 post_tg(POST_ARGS)
 {
-	struct roff_node	*n, *nch;
+	struct roff_node	*n, *nch, *nn;
 	size_t			len;
 
+	/* Find the next node. */
 	n = mdoc->last;
+	for (nn = n; nn != NULL; nn = nn->parent) {
+		if (nn->next != NULL) {
+			nn = nn->next;
+			break;
+		}
+	}
+
+	/* Add the default argument, if needed. */
 	nch = n->child;
-	if (nch == NULL && n->next != NULL &&
-	    n->next->child->type == ROFFT_TEXT) {
+	if (nch == NULL && nn != NULL && nn->child->type == ROFFT_TEXT) {
 		mdoc->next = ROFF_NEXT_CHILD;
 		roff_word_alloc(mdoc, n->line, n->pos, n->next->child->string);
 		nch = mdoc->last;
 		nch->flags |= NODE_NOSRC;
 		mdoc->last = n;
 	}
-	if (nch == NULL || *nch->string == '\0') {
+
+	/* Validate the first argument. */
+	if (nch == NULL || *nch->string == '\0')
 		mandoc_msg(MANDOCERR_MACRO_EMPTY, n->line, n->pos, "Tg");
+	if (nch == NULL) {
 		roff_node_delete(mdoc, n);
 		return;
 	}
@@ -1101,14 +1112,42 @@ post_tg(POST_ARGS)
 	if (nch->string[len] != '\0')
 		mandoc_msg(MANDOCERR_TG_SPC, nch->line, nch->pos + len + 1,
 		    "Tg %s", nch->string);
+
+	/* Keep only the first argument. */
 	if (nch->next != NULL) {
 		mandoc_msg(MANDOCERR_ARG_EXCESS, nch->next->line,
 		    nch->next->pos, "Tg ... %s", nch->next->string);
 		while (nch->next != NULL)
 			roff_node_delete(mdoc, nch->next);
 	}
-	if (nch->string[len] != '\0')
+
+	/* Drop the macro if the first argument is invalid. */
+	if (len == 0 || nch->string[len] != '\0') {
 		roff_node_delete(mdoc, n);
+		return;
+	}
+
+	/* By default, write a <mark> element. */
+	n->flags |= NODE_ID;
+	if (nn == NULL)
+		return;
+
+	/* Explicit tagging of specific macros. */
+	switch (nn->tok) {
+	case MDOC_Sh:
+	case MDOC_Ss:
+		if (nn->head->flags & NODE_ID || nn->head->child == NULL)
+			break;
+		n->flags |= NODE_NOPRT;
+		nn->head->flags |= NODE_ID | NODE_HREF;
+		assert(nn->head->string == NULL);
+		nn->head->string = mandoc_strdup(nch->string);
+		break;
+	default:
+		break;
+	}
+	if (n->flags & NODE_NOPRT)
+		n->flags &= ~NODE_ID;
 }
 
 static void
diff --git a/usr.bin/mandoc/roff.c b/usr.bin/mandoc/roff.c
index 9a8e77a9893..31670fc500d 100644
--- a/usr.bin/mandoc/roff.c
+++ b/usr.bin/mandoc/roff.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: roff.c,v 1.242 2020/02/27 01:25:58 schwarze Exp $ */
+/*	$OpenBSD: roff.c,v 1.243 2020/02/27 21:38:27 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2012, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2015, 2017-2020 Ingo Schwarze <schwarze@openbsd.org>
@@ -1171,7 +1171,7 @@ deroff(char **dest, const struct roff_node *n)
 	char	*cp;
 	size_t	 sz;
 
-	if (n->type != ROFFT_TEXT) {
+	if (n->string == NULL) {
 		for (n = n->child; n != NULL; n = n->next)
 			deroff(dest, n);
 		return;
diff --git a/usr.bin/mandoc/roff.h b/usr.bin/mandoc/roff.h
index 24c0e39bc1f..aeb5c68e08a 100644
--- a/usr.bin/mandoc/roff.h
+++ b/usr.bin/mandoc/roff.h
@@ -1,4 +1,4 @@
-/*	$OpenBSD: roff.h,v 1.53 2020/02/27 01:25:58 schwarze Exp $	*/
+/*	$OpenBSD: roff.h,v 1.54 2020/02/27 21:38:27 schwarze Exp $	*/
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2013-2015, 2017-2020 Ingo Schwarze <schwarze@openbsd.org>
@@ -522,6 +522,8 @@ struct	roff_node {
 #define	NODE_NOFILL	 (1 << 8)  /* Fill mode switched off. */
 #define	NODE_NOSRC	 (1 << 9)  /* Generated node, not in input file. */
 #define	NODE_NOPRT	 (1 << 10) /* Shall not print anything. */
+#define	NODE_ID		 (1 << 11) /* Target for deep linking. */
+#define	NODE_HREF	 (1 << 12) /* Link to another place in this page. */
 	int		  prev_font; /* Before entering this node. */
 	int		  aux;     /* Decoded node data, type-dependent. */
 	enum roff_tok	  tok;     /* Request or macro ID. */
diff --git a/usr.bin/mandoc/tree.c b/usr.bin/mandoc/tree.c
index 17ea0f60c91..cc123ccf308 100644
--- a/usr.bin/mandoc/tree.c
+++ b/usr.bin/mandoc/tree.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: tree.c,v 1.52 2020/01/11 16:02:48 schwarze Exp $ */
+/*	$OpenBSD: tree.c,v 1.53 2020/02/27 21:38:27 schwarze Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2011, 2014 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2013-2015, 2017-2020 Ingo Schwarze <schwarze@openbsd.org>
@@ -197,6 +197,13 @@ print_mdoc(const struct roff_node *n, int indent)
 			putchar(')');
 		if (n->flags & NODE_EOS)
 			putchar('.');
+		if (n->flags & NODE_ID) {
+			printf(" ID");
+			if (n->string != NULL)
+				printf("=%s", n->string);
+		}
+		if (n->flags & NODE_HREF)
+			printf(" HREF");
 		if (n->flags & NODE_BROKEN)
 			printf(" BROKEN");
 		if (n->flags & NODE_NOFILL)