Clarify -man -T[x]html handling of `br' within `B'.

Consolidated node unlinking in -man.
Conclude nested next-line scope issues noted by Ingo Schwarze.

5 files changed, 95 insertions, 57 deletions

diff --git a/libman.h b/libman.h
index ad33e44f..28dc5ffa 100644
--- a/libman.h
+++ b/libman.h
@@ -1,4 +1,4 @@
-/*	$Id: libman.h,v 1.24 2010/03/22 05:59:32 kristaps Exp $ */
+/*	$Id: libman.h,v 1.25 2010/03/22 14:03:03 kristaps Exp $ */
 /*
  * Copyright (c) 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -45,6 +45,7 @@ enum	merr {
 	WMSEC,
 	WDATE,
 	WLNSCOPE,
+	WLNSCOPE2,
 	WTSPACE,
 	WTQUOTE,
 	WNODATA,
@@ -94,6 +95,7 @@ int		  man_body_alloc(struct man *, int, int, int);
 int		  man_elem_alloc(struct man *, int, int, int);
 void		  man_node_free(struct man_node *);
 void		  man_node_freelist(struct man_node *);
+void		  man_node_unlink(struct man *, struct man_node *);
 void		  man_hash_init(void);
 int		  man_hash_find(const char *);
 int		  man_macroend(struct man *);
diff --git a/man.c b/man.c
index d823a5d1..2aee2cb5 100644
--- a/man.c
+++ b/man.c
@@ -1,4 +1,4 @@
-/*	$Id: man.c,v 1.50 2010/03/22 05:59:32 kristaps Exp $ */
+/*	$Id: man.c,v 1.51 2010/03/22 14:03:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -35,6 +35,7 @@ const	char *const __man_merrnames[WERRMAX] = {
 	"invalid manual section", /* WMSEC */
 	"invalid date format", /* WDATE */
 	"scope of prior line violated", /* WLNSCOPE */
+	"over-zealous prior line scope violation", /* WLNSCOPE2 */
 	"trailing whitespace", /* WTSPACE */
 	"unterminated quoted parameter", /* WTQUOTE */
 	"document has no body", /* WNODATA */
@@ -535,41 +536,37 @@ man_pmacro(struct man *m, int ln, char *buf)
 			goto err;
 
 	/* 
-	 * Remove prior ELINE macro, as a macro is clobbering it by
-	 * being invoked without prior text.  Note that NSCOPED macros
-	 * do not close out ELINE macros, as they print no text.
+	 * Remove prior ELINE macro, as it's being clobbering by a new
+	 * macro.  Note that NSCOPED macros do not close out ELINE
+	 * macros---they don't print text---so we let those slip by.
 	 */
 
-	if (m->flags & MAN_ELINE && 
-			! (MAN_NSCOPED & man_macros[c].flags)) {
+	if ( ! (MAN_NSCOPED & man_macros[c].flags) &&
+			m->flags & MAN_ELINE) {
+		assert(MAN_TEXT != m->last->type);
+
+		/*
+		 * This occurs in the following construction:
+		 *   .B
+		 *   .br
+		 *   .B
+		 *   .br
+		 *   I hate man macros.
+		 * Flat-out disallow this madness.
+		 */
+		if (MAN_NSCOPED & man_macros[m->last->tok].flags)
+			return(man_perr(m, ln, ppos, WLNSCOPE));
+
 		n = m->last;
+
+		assert(n);
 		assert(NULL == n->child);
 		assert(0 == n->nchild);
+
 		if ( ! man_nwarn(m, n, WLNSCOPE))
 			return(0);
 
-		/* FIXME: when called as in:
-		 *
-		 * .B
-		 * .br
-		 * .B
-		 * .br
-		 * hello
-		 */
-
-		if (n->prev) {
-			assert(n != n->parent->child);
-			assert(n == n->prev->next);
-			n->prev->next = NULL;
-			m->last = n->prev;
-			m->next = MAN_NEXT_SIBLING;
-		} else {
-			assert(n == n->parent->child);
-			n->parent->child = NULL;
-			m->last = n->parent;
-			m->next = MAN_NEXT_CHILD;
-		}
-
+		man_node_unlink(m, n);
 		man_node_free(n);
 		m->flags &= ~MAN_ELINE;
 	}
@@ -671,3 +668,28 @@ man_err(struct man *m, int line, int pos, int iserr, enum merr type)
 
 	return(man_vwarn(m, line, pos, p));
 }
+
+
+void
+man_node_unlink(struct man *m, struct man_node *n)
+{
+
+	if (n->prev) {
+		n->prev->next = n->next;
+		if (m->last == n) {
+			assert(NULL == n->next);
+			m->last = n->prev;
+			m->next = MAN_NEXT_SIBLING;
+		}
+	} else {
+		n->parent->child = n->next;
+		if (m->last == n) {
+			assert(NULL == n->next);
+			m->last = n->parent;
+			m->next = MAN_NEXT_CHILD;
+		}
+	}
+
+	if (n->next)
+		n->next->prev = n->prev;
+}
diff --git a/man_action.c b/man_action.c
index 2b8eba9f..d481f103 100644
--- a/man_action.c
+++ b/man_action.c
@@ -1,4 +1,4 @@
-/*	$Id: man_action.c,v 1.25 2010/01/01 17:14:27 kristaps Exp $ */
+/*	$Id: man_action.c,v 1.26 2010/03/22 14:03:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -178,24 +178,8 @@ post_TH(struct man *m)
 	if (n && (n = n->next))
 		m->meta.vol = mandoc_strdup(n->string);
 
-	/* 
-	 * The end document shouldn't have the prologue macros as part
-	 * of the syntax tree (they encompass only meta-data).  
-	 */
-
-	if (m->last->parent->child == m->last) {
-		m->last->parent->child = NULL;
-		n = m->last;
-		m->last = m->last->parent;
-		m->next = MAN_NEXT_CHILD;
-	} else {
-		assert(m->last->prev);
-		m->last->prev->next = NULL;
-		n = m->last;
-		m->last = m->last->prev;
-		m->next = MAN_NEXT_SIBLING;
-	}
-
+	n = m->last;
+	man_node_unlink(m, n);
 	man_node_freelist(n);
 	return(1);
 }
diff --git a/man_html.c b/man_html.c
index b5e69343..495a5fca 100644
--- a/man_html.c
+++ b/man_html.c
@@ -1,4 +1,4 @@
-/*	$Id: man_html.c,v 1.27 2010/03/22 05:59:32 kristaps Exp $ */
+/*	$Id: man_html.c,v 1.28 2010/03/22 14:03:03 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009 Kristaps Dzonsons <kristaps@kth.se>
  *
@@ -181,6 +181,12 @@ print_man_node(MAN_ARGS)
 
 	bufinit(h);
 
+	/*
+	 * FIXME: embedded elements within next-line scopes (e.g., `br'
+	 * within an empty `B') will cause formatting to be forgotten
+	 * due to scope closing out.
+	 */
+
 	switch (n->type) {
 	case (MAN_ROOT):
 		child = man_root_pre(m, n, h);
@@ -567,6 +573,8 @@ man_IP_pre(MAN_ARGS)
 	SCALE_HS_INIT(&su, INDENT);
 	width = 0;
 
+	/* Width is the last token. */
+
 	if (MAN_IP == n->tok && NULL != nn)
 		if (NULL != (nn = nn->next)) {
 			for ( ; nn->next; nn = nn->next)
@@ -574,11 +582,14 @@ man_IP_pre(MAN_ARGS)
 			width = a2width(nn, &su);
 		}
 
+	/* Width is the first token. */
+
 	if (MAN_TP == n->tok && NULL != nn) {
+		/* Skip past non-text children. */
 		while (nn && MAN_TEXT != nn->type)
 			nn = nn->next;
-		/* FIXME: sync with pre_TP(), man_term.c */
-		width = a2width(nn, &su);
+		if (nn)
+			width = a2width(nn, &su);
 	}
 
 	if (MAN_BLOCK == n->type) {
@@ -604,12 +615,19 @@ man_IP_pre(MAN_ARGS)
 	PAIR_STYLE_INIT(&tag, h);
 	print_otag(h, TAG_DIV, 1, &tag);
 
-	/* With a length string, manually omit the last child. */
+	/*
+	 * Without a length string, we can print all of our children.
+	 */
 
 	if ( ! width)
 		return(1);
 
-	/* FIXME: sync with pre_TP(), man_term.c */
+	/*
+	 * When a length has been specified, we need to carefully print
+	 * our child context:  IP gets all children printed but the last
+	 * (the width), while TP gets all children printed but the first
+	 * (the width).
+	 */
 
 	if (MAN_IP == n->tok)
 		for (nn = n->child; nn->next; nn = nn->next)
diff --git a/mandoc.1 b/mandoc.1
index 64152e8b..5ab24126 100644
--- a/mandoc.1
+++ b/mandoc.1
@@ -1,4 +1,4 @@
-.\"	$Id: mandoc.1,v 1.50 2010/01/29 14:39:38 kristaps Exp $
+.\"	$Id: mandoc.1,v 1.51 2010/03/22 14:03:03 kristaps Exp $
 .\"
 .\" Copyright (c) 2009 Kristaps Dzonsons <kristaps@kth.se>
 .\"
@@ -14,7 +14,7 @@
 .\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
 .\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
 .\"
-.Dd $Mdocdate: January 29 2010 $
+.Dd $Mdocdate: March 22 2010 $
 .Dt MANDOC 1
 .Os
 .
@@ -502,8 +502,8 @@ and
 .Fl T Ns Ar xhtml
 CSS2 styling used for
 .Fl m Ns Ar doc
-input lists does not render properly in brain-dead browsers, such as
-Internet Explorer 6 and earlier.
+input lists does not render properly in older browsers, such as Internet
+Explorer 6 and earlier.
 .Pp
 In
 .Fl T Ns Ar html
@@ -525,3 +525,15 @@ font size escape documented in
 .Xr mdoc 7
 and
 .Xr man 7 .
+.Pp
+Nesting elements within next-line element scopes of
+.Fl m Ar Ns an ,
+such as
+.Sq br
+within an empty
+.Sq B ,
+will confuse
+.Fl T Ns Ar html
+and
+.Fl T Ns Ar xhtml
+and cause it to forget the formatting.