===================================================================
RCS file: /cvs/mandoc/mdoc_validate.c,v
retrieving revision 1.366
retrieving revision 1.379
diff -u -p -r1.366 -r1.379
--- mandoc/mdoc_validate.c	2018/12/30 00:49:55	1.366
+++ mandoc/mdoc_validate.c	2020/02/27 21:43:44	1.379
@@ -1,7 +1,7 @@
-/*	$Id: mdoc_validate.c,v 1.366 2018/12/30 00:49:55 schwarze Exp $ */
+/*	$Id: mdoc_validate.c,v 1.379 2020/02/27 21:43:44 schwarze Exp $ */
 /*
  * Copyright (c) 2008-2012 Kristaps Dzonsons <kristaps@bsd.lv>
- * Copyright (c) 2010-2018 Ingo Schwarze <schwarze@openbsd.org>
+ * Copyright (c) 2010-2020 Ingo Schwarze <schwarze@openbsd.org>
  * Copyright (c) 2010 Joerg Sonnenberger <joerg@netbsd.org>
  *
  * Permission to use, copy, modify, and distribute this software for any
@@ -64,7 +64,7 @@ static	size_t		macro2len(enum roff_tok);
 static	void	 rewrite_macro2len(struct roff_man *, char **);
 static	int	 similar(const char *, const char *);
 
-static	void	 post_abort(POST_ARGS);
+static	void	 post_abort(POST_ARGS) __attribute__((__noreturn__));
 static	void	 post_an(POST_ARGS);
 static	void	 post_an_norm(POST_ARGS);
 static	void	 post_at(POST_ARGS);
@@ -113,6 +113,7 @@ static	void	 post_sm(POST_ARGS);
 static	void	 post_st(POST_ARGS);
 static	void	 post_std(POST_ARGS);
 static	void	 post_sx(POST_ARGS);
+static	void	 post_tg(POST_ARGS);
 static	void	 post_useless(POST_ARGS);
 static	void	 post_xr(POST_ARGS);
 static	void	 post_xx(POST_ARGS);
@@ -238,6 +239,7 @@ static	const v_post mdoc_valids[MDOC_MAX - MDOC_Dd] = 
 	NULL,		/* %Q */
 	NULL,		/* %U */
 	NULL,		/* Ta */
+	post_tg,	/* Tg */
 };
 
 #define	RSORD_MAX 14 /* Number of `Rs' blocks. */
@@ -288,7 +290,7 @@ static	const char * const secnames[SEC__MAX] = {
 
 /* Validate the subtree rooted at mdoc->last. */
 void
-mdoc_node_validate(struct roff_man *mdoc)
+mdoc_validate(struct roff_man *mdoc)
 {
 	struct roff_node *n, *np;
 	const v_post *p;
@@ -319,7 +321,7 @@ mdoc_node_validate(struct roff_man *mdoc)
 
 	mdoc->last = mdoc->last->child;
 	while (mdoc->last != NULL) {
-		mdoc_node_validate(mdoc);
+		mdoc_validate(mdoc);
 		if (mdoc->last == n)
 			mdoc->last = mdoc->last->child;
 		else
@@ -336,9 +338,7 @@ mdoc_node_validate(struct roff_man *mdoc)
 		if (n->sec != SEC_SYNOPSIS ||
 		    (np->tok != MDOC_Cd && np->tok != MDOC_Fd))
 			check_text(mdoc, n->line, n->pos, n->string);
-		if (np->tok != MDOC_Ql && np->tok != MDOC_Dl &&
-		    (np->tok != MDOC_Bd ||
-		     (mdoc->flags & MDOC_LITERAL) == 0) &&
+		if ((n->flags & NODE_NOFILL) == 0 &&
 		    (np->tok != MDOC_It || np->type != ROFFT_HEAD ||
 		     np->parent->parent->norm->Bl.type != LIST_diag))
 			check_text_em(mdoc, n->line, n->pos, n->string);
@@ -411,7 +411,7 @@ check_text(struct roff_man *mdoc, int ln, int pos, cha
 {
 	char		*cp;
 
-	if (MDOC_LITERAL & mdoc->flags)
+	if (mdoc->last->flags & NODE_NOFILL)
 		return;
 
 	for (cp = p; NULL != (p = strchr(p, '\t')); p++)
@@ -1092,6 +1092,80 @@ post_st(POST_ARGS)
 }
 
 static void
+post_tg(POST_ARGS)
+{
+	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 && 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;
+	}
+
+	/* 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;
+	}
+	len = strcspn(nch->string, " \t");
+	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);
+	}
+
+	/* 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
 post_obsolete(POST_ARGS)
 {
 	struct roff_node *n;
@@ -1188,11 +1262,17 @@ post_fname(POST_ARGS)
 	size_t			 pos;
 
 	n = mdoc->last->child;
-	pos = strcspn(n->string, "()");
-	cp = n->string + pos;
-	if ( ! (cp[0] == '\0' || (cp[0] == '(' && cp[1] == '*')))
-		mandoc_msg(MANDOCERR_FN_PAREN, n->line, n->pos + pos,
-		    "%s", n->string);
+	cp = n->string;
+	if (*cp == '(') {
+		if (cp[strlen(cp + 1)] == ')')
+			return;
+		pos = 0;
+	} else {
+		pos = strcspn(cp, "()");
+		if (cp[pos] == '\0')
+			return;
+	}
+	mandoc_msg(MANDOCERR_FN_PAREN, n->line, n->pos + pos, "%s", cp);
 }
 
 static void
@@ -1583,7 +1663,7 @@ post_it(POST_ARGS)
 			mandoc_msg(MANDOCERR_BL_COL, nit->line, nit->pos,
 			    "%d columns, %d cells", cols, i);
 		else if (nit->head->next->child != NULL &&
-		    nit->head->next->child->line > nit->line)
+		    nit->head->next->child->flags & NODE_LINE)
 			mandoc_msg(MANDOCERR_IT_NOARG,
 			    nit->line, nit->pos, "Bl -column It");
 		break;
@@ -1719,8 +1799,7 @@ post_bl_head(POST_ARGS)
 static void
 post_bl(POST_ARGS)
 {
-	struct roff_node	*nparent, *nprev; /* of the Bl block */
-	struct roff_node	*nblock, *nbody;  /* of the Bl */
+	struct roff_node	*nbody;           /* of the Bl */
 	struct roff_node	*nchild, *nnext;  /* of the Bl body */
 	const char		*prev_Er;
 	int			 order;
@@ -1741,88 +1820,73 @@ post_bl(POST_ARGS)
 	if (nbody->end != ENDBODY_NOT)
 		return;
 
-	nchild = nbody->child;
-	if (nchild == NULL) {
-		mandoc_msg(MANDOCERR_BLK_EMPTY,
-		    nbody->line, nbody->pos, "Bl");
-		return;
+	/*
+	 * Up to the first item, move nodes before the list,
+	 * but leave transparent nodes where they are
+	 * if they precede an item.
+	 * The next non-transparent node is kept in nchild.
+	 * It only needs to be updated after a non-transparent
+	 * node was moved out, and at the very beginning
+	 * when no node at all was moved yet.
+	 */
+
+	nchild = mdoc->last;
+	for (;;) {
+		if (nchild == mdoc->last)
+			nchild = roff_node_child(nbody);
+		if (nchild == NULL) {
+			mdoc->last = nbody;
+			mandoc_msg(MANDOCERR_BLK_EMPTY,
+			    nbody->line, nbody->pos, "Bl");
+			return;
+		}
+		if (nchild->tok == MDOC_It) {
+			mdoc->last = nbody;
+			break;
+		}
+		mandoc_msg(MANDOCERR_BL_MOVE, nbody->child->line,
+		    nbody->child->pos, "%s", roff_name[nbody->child->tok]);
+		if (nbody->parent->prev == NULL) {
+			mdoc->last = nbody->parent->parent;
+			mdoc->next = ROFF_NEXT_CHILD;
+		} else {
+			mdoc->last = nbody->parent->prev;
+			mdoc->next = ROFF_NEXT_SIBLING;
+		}
+		roff_node_relink(mdoc, nbody->child);
 	}
+
+	/*
+	 * We have reached the first item,
+	 * so moving nodes out is no longer possible.
+	 * But in .Bl -column, the first rows may be implicit,
+	 * that is, they may not start with .It macros.
+	 * Such rows may be followed by nodes generated on the
+	 * roff level, for example .TS.
+	 * Wrap such roff nodes into an implicit row.
+	 */
+
 	while (nchild != NULL) {
-		nnext = nchild->next;
-		if (nchild->tok == MDOC_It ||
-		    (nchild->tok == MDOC_Sm &&
-		     nnext != NULL && nnext->tok == MDOC_It)) {
-			nchild = nnext;
+		if (nchild->tok == MDOC_It) {
+			nchild = roff_node_next(nchild);
 			continue;
 		}
-
-		/*
-		 * In .Bl -column, the first rows may be implicit,
-		 * that is, they may not start with .It macros.
-		 * Such rows may be followed by nodes generated on the
-		 * roff level, for example .TS, which cannot be moved
-		 * out of the list.  In that case, wrap such roff nodes
-		 * into an implicit row.
-		 */
-
-		if (nchild->prev != NULL) {
-			mdoc->last = nchild;
+		nnext = nchild->next;
+		mdoc->last = nchild->prev;
+		mdoc->next = ROFF_NEXT_SIBLING;
+		roff_block_alloc(mdoc, nchild->line, nchild->pos, MDOC_It);
+		roff_head_alloc(mdoc, nchild->line, nchild->pos, MDOC_It);
+		mdoc->next = ROFF_NEXT_SIBLING;
+		roff_body_alloc(mdoc, nchild->line, nchild->pos, MDOC_It);
+		while (nchild->tok != MDOC_It) {
+			roff_node_relink(mdoc, nchild);
+			if (nnext == NULL)
+				break;
+			nchild = nnext;
+			nnext = nchild->next;
 			mdoc->next = ROFF_NEXT_SIBLING;
-			roff_block_alloc(mdoc, nchild->line,
-			    nchild->pos, MDOC_It);
-			roff_head_alloc(mdoc, nchild->line,
-			    nchild->pos, MDOC_It);
-			mdoc->next = ROFF_NEXT_SIBLING;
-			roff_body_alloc(mdoc, nchild->line,
-			    nchild->pos, MDOC_It);
-			while (nchild->tok != MDOC_It) {
-				roff_node_relink(mdoc, nchild);
-				if ((nchild = nnext) == NULL)
-					break;
-				nnext = nchild->next;
-				mdoc->next = ROFF_NEXT_SIBLING;
-			}
-			mdoc->last = nbody;
-			continue;
 		}
-
-		mandoc_msg(MANDOCERR_BL_MOVE, nchild->line, nchild->pos,
-		    "%s", roff_name[nchild->tok]);
-
-		/*
-		 * Move the node out of the Bl block.
-		 * First, collect all required node pointers.
-		 */
-
-		nblock  = nbody->parent;
-		nprev   = nblock->prev;
-		nparent = nblock->parent;
-
-		/*
-		 * Unlink this child.
-		 */
-
-		nbody->child = nnext;
-		if (nnext == NULL)
-			nbody->last  = NULL;
-		else
-			nnext->prev = NULL;
-
-		/*
-		 * Relink this child.
-		 */
-
-		nchild->parent = nparent;
-		nchild->prev   = nprev;
-		nchild->next   = nblock;
-
-		nblock->prev = nchild;
-		if (nprev == NULL)
-			nparent->child = nchild;
-		else
-			nprev->next = nchild;
-
-		nchild = nnext;
+		mdoc->last = nbody;
 	}
 
 	if (mdoc->meta.os_e != MANDOC_OS_NETBSD)
@@ -1900,35 +1964,12 @@ post_sm(POST_ARGS)
 static void
 post_root(POST_ARGS)
 {
-	const char *openbsd_arch[] = {
-		"alpha", "amd64", "arm64", "armv7", "hppa", "i386",
-		"landisk", "loongson", "luna88k", "macppc", "mips64",
-		"octeon", "sgi", "socppc", "sparc64", NULL
-	};
-	const char *netbsd_arch[] = {
-		"acorn26", "acorn32", "algor", "alpha", "amiga",
-		"arc", "atari",
-		"bebox", "cats", "cesfic", "cobalt", "dreamcast",
-		"emips", "evbarm", "evbmips", "evbppc", "evbsh3", "evbsh5",
-		"hp300", "hpcarm", "hpcmips", "hpcsh", "hppa",
-		"i386", "ibmnws", "luna68k",
-		"mac68k", "macppc", "mipsco", "mmeye", "mvme68k", "mvmeppc",
-		"netwinder", "news68k", "newsmips", "next68k",
-		"pc532", "playstation2", "pmax", "pmppc", "prep",
-		"sandpoint", "sbmips", "sgimips", "shark",
-		"sparc", "sparc64", "sun2", "sun3",
-		"vax", "walnut", "x68k", "x86", "x86_64", "xen", NULL
-        };
-	const char **arches[] = { NULL, netbsd_arch, openbsd_arch };
-
 	struct roff_node *n;
-	const char **arch;
 
 	/* Add missing prologue data. */
 
 	if (mdoc->meta.date == NULL)
-		mdoc->meta.date = mdoc->quick ? mandoc_strdup("") :
-		    mandoc_normdate(mdoc, NULL, 0, 0);
+		mdoc->meta.date = mandoc_normdate(NULL, NULL);
 
 	if (mdoc->meta.title == NULL) {
 		mandoc_msg(MANDOCERR_DT_NOTITLE, 0, 0, "EOF");
@@ -1948,22 +1989,18 @@ post_root(POST_ARGS)
 		    "(OpenBSD)" : "(NetBSD)");
 
 	if (mdoc->meta.arch != NULL &&
-	    (arch = arches[mdoc->meta.os_e]) != NULL) {
-		while (*arch != NULL && strcmp(*arch, mdoc->meta.arch))
-			arch++;
-		if (*arch == NULL) {
-			n = mdoc->meta.first->child;
-			while (n->tok != MDOC_Dt ||
-			    n->child == NULL ||
-			    n->child->next == NULL ||
-			    n->child->next->next == NULL)
-				n = n->next;
-			n = n->child->next->next;
-			mandoc_msg(MANDOCERR_ARCH_BAD, n->line, n->pos,
-			    "Dt ... %s %s", mdoc->meta.arch,
-			    mdoc->meta.os_e == MANDOC_OS_OPENBSD ?
-			    "(OpenBSD)" : "(NetBSD)");
-		}
+	    arch_valid(mdoc->meta.arch, mdoc->meta.os_e) == 0) {
+		n = mdoc->meta.first->child;
+		while (n->tok != MDOC_Dt ||
+		    n->child == NULL ||
+		    n->child->next == NULL ||
+		    n->child->next->next == NULL)
+			n = n->next;
+		n = n->child->next->next;
+		mandoc_msg(MANDOCERR_ARCH_BAD, n->line, n->pos,
+		    "Dt ... %s %s", mdoc->meta.arch,
+		    mdoc->meta.os_e == MANDOC_OS_OPENBSD ?
+		    "(OpenBSD)" : "(NetBSD)");
 	}
 
 	/* Check that we begin with a proper `Sh'. */
@@ -2486,20 +2523,20 @@ post_ignpar(POST_ARGS)
 static void
 post_prevpar(POST_ARGS)
 {
-	struct roff_node *n;
+	struct roff_node *n, *np;
 
 	n = mdoc->last;
-	if (NULL == n->prev)
-		return;
 	if (n->type != ROFFT_ELEM && n->type != ROFFT_BLOCK)
 		return;
+	if ((np = roff_node_prev(n)) == NULL)
+		return;
 
 	/*
 	 * Don't allow `Pp' prior to a paragraph-type
 	 * block: `Pp' or non-compact `Bd' or `Bl'.
 	 */
 
-	if (n->prev->tok != MDOC_Pp && n->prev->tok != ROFF_br)
+	if (np->tok != MDOC_Pp && np->tok != ROFF_br)
 		return;
 	if (n->tok == MDOC_Bl && n->norm->Bl.comp)
 		return;
@@ -2508,9 +2545,9 @@ post_prevpar(POST_ARGS)
 	if (n->tok == MDOC_It && n->parent->norm->Bl.comp)
 		return;
 
-	mandoc_msg(MANDOCERR_PAR_SKIP, n->prev->line, n->prev->pos,
-	    "%s before %s", roff_name[n->prev->tok], roff_name[n->tok]);
-	roff_node_delete(mdoc, n->prev);
+	mandoc_msg(MANDOCERR_PAR_SKIP, np->line, np->pos,
+	    "%s before %s", roff_name[np->tok], roff_name[n->tok]);
+	roff_node_delete(mdoc, np);
 }
 
 static void
@@ -2530,7 +2567,6 @@ static void
 post_dd(POST_ARGS)
 {
 	struct roff_node *n;
-	char		 *datestr;
 
 	n = mdoc->last;
 	n->flags |= NODE_NOPRT;
@@ -2547,21 +2583,10 @@ post_dd(POST_ARGS)
 		mandoc_msg(MANDOCERR_PROLOG_ORDER,
 		    n->line, n->pos, "Dd after Os");
 
-	if (n->child == NULL || n->child->string[0] == '\0') {
-		mdoc->meta.date = mdoc->quick ? mandoc_strdup("") :
-		    mandoc_normdate(mdoc, NULL, n->line, n->pos);
-		return;
-	}
-
-	datestr = NULL;
-	deroff(&datestr, n);
-	if (mdoc->quick)
-		mdoc->meta.date = datestr;
-	else {
-		mdoc->meta.date = mandoc_normdate(mdoc,
-		    datestr, n->line, n->pos);
-		free(datestr);
-	}
+	if (mdoc->quick && n != NULL)
+		mdoc->meta.date = mandoc_strdup("");
+	else
+		mdoc->meta.date = mandoc_normdate(n->child, n);
 }
 
 static void