Give -Thtml and -Txhtml the gift of recognising escapes when calculating

widths (e.g., `Bl -tag -width "\s[blahblah]bar").  This has long since
been done for -Tascii but escaped noticed with -T[x]html.

3 files changed, 45 insertions, 8 deletions

diff --git a/html.c b/html.c
index b1ba0dd2..8fb81c51 100644
--- a/html.c
+++ b/html.c
@@ -1,4 +1,4 @@
-/*	$Id: html.c,v 1.137 2011/04/30 22:24:31 kristaps Exp $ */
+/*	$Id: html.c,v 1.138 2011/05/14 16:28:23 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2011 Ingo Schwarze <schwarze@openbsd.org>
@@ -31,6 +31,7 @@
 #include <unistd.h>
 
 #include "mandoc.h"
+#include "libmandoc.h"
 #include "out.h"
 #include "html.h"
 #include "main.h"
@@ -294,6 +295,41 @@ print_metaf(struct html *h, enum mandoc_esc deco)
 			print_otag(h, TAG_I, 0, NULL);
 }
 
+int
+html_strlen(const char *cp)
+{
+	int		 ssz, sz;
+	const char	*seq, *p;
+
+	/*
+	 * Account for escaped sequences within string length
+	 * calculations.  This follows the logic in term_strlen() as we
+	 * must calculate the width of produced strings.
+	 * Assume that characters are always width of "1".  This is
+	 * hacky, but it gets the job done for approximation of widths.
+	 */
+
+	sz = 0;
+	while (NULL != (p = strchr(cp, '\\'))) {
+		sz += (int)(p - cp);
+		++cp;
+		switch (mandoc_escape(&cp, &seq, &ssz)) {
+		case (ESCAPE_ERROR):
+			return(sz);
+		case (ESCAPE_PREDEF):
+			sz++;
+			break;
+		case (ESCAPE_SPECIAL):
+			sz++;
+			break;
+		default:
+			break;
+		}
+	}
+
+	assert(sz >= 0);
+	return(sz + strlen(cp));
+}
 
 static int
 print_encode(struct html *h, const char *p, int norecurse)
diff --git a/html.h b/html.h
index 89dd7c63..35fb6c67 100644
--- a/html.h
+++ b/html.h
@@ -1,4 +1,4 @@
-/*	$Id: html.h,v 1.41 2011/04/29 22:18:12 kristaps Exp $ */
+/*	$Id: html.h,v 1.42 2011/05/14 16:28:23 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -155,6 +155,7 @@ void		  bufncat(struct html *, const char *, size_t);
 void		  bufinit(struct html *);
 
 void		  html_idcat(char *, const char *, int);
+int		  html_strlen(const char *);
 
 __END_DECLS
 
diff --git a/mdoc_html.c b/mdoc_html.c
index 652e8536..73cf1182 100644
--- a/mdoc_html.c
+++ b/mdoc_html.c
@@ -1,4 +1,4 @@
-/*	$Id: mdoc_html.c,v 1.165 2011/04/23 09:10:50 kristaps Exp $ */
+/*	$Id: mdoc_html.c,v 1.166 2011/05/14 16:28:23 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -288,7 +288,7 @@ a2width(const char *p, struct roffsu *su)
 
 	if ( ! a2roffsu(p, su, SCALE_MAX)) {
 		su->unit = SCALE_BU;
-		su->scale = (int)strlen(p);
+		su->scale = html_strlen(p);
 	}
 }
 
@@ -355,7 +355,7 @@ a2offs(const char *p, struct roffsu *su)
 		SCALE_HS_INIT(su, INDENT * 2);
 	else if ( ! a2roffsu(p, su, SCALE_MAX)) {
 		su->unit = SCALE_BU;
-		su->scale = (int)strlen(p);
+		su->scale = html_strlen(p);
 	}
 }
 
@@ -703,7 +703,7 @@ mdoc_nm_pre(MDOC_ARGS)
 {
 	struct htmlpair	 tag;
 	struct roffsu	 su;
-	size_t		 len;
+	int		 len;
 
 	switch (n->type) {
 	case (MDOC_ELEM):
@@ -731,10 +731,10 @@ mdoc_nm_pre(MDOC_ARGS)
 
 	for (len = 0, n = n->child; n; n = n->next)
 		if (MDOC_TEXT == n->type)
-			len += strlen(n->string);
+			len += html_strlen(n->string);
 
 	if (0 == len && m->name)
-		len = strlen(m->name);
+		len = html_strlen(m->name);
 
 	SCALE_HS_INIT(&su, (double)len);
 	bufcat_su(h, "width", &su);