From: Kristaps Dzonsons <kristaps@bsd.lv>
Date: Tue, 17 May 2011 11:55:08 +0000 (+0000)
Subject: Flip on printing `?' at Unicode codepoints in -Tascii, -Tpdf, and -Tps.
X-Git-Tag: VERSION_1_11_3~34
X-Git-Url: https://git.cameronkatri.com/mandoc.git/commitdiff_plain/d72b93d8133ec9ce52e458f06774fdf6db5b77cc?ds=inline

Flip on printing `?' at Unicode codepoints in -Tascii, -Tpdf, and -Tps.
The reasoning behind printing SOMETHING at a Unicode codepoint is
because the input is not "wrong" (we suppress printing of "wrong"
things).  It's just that ASCII can't handle it.
---

diff --git a/term.c b/term.c
index 0ea561cb..bfedefb0 100644
--- a/term.c
+++ b/term.c
@@ -1,4 +1,4 @@
-/*	$Id: term.c,v 1.191 2011/05/15 22:29:50 kristaps Exp $ */
+/*	$Id: term.c,v 1.192 2011/05/17 11:55:08 kristaps Exp $ */
 /*
  * Copyright (c) 2008, 2009, 2010 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010, 2011 Ingo Schwarze <schwarze@openbsd.org>
@@ -451,6 +451,9 @@ term_word(struct termp *p, const char *word)
 			break;
 
 		switch (esc) {
+		case (ESCAPE_UNICODE):
+			encode(p, "?", 1);
+			break;
 		case (ESCAPE_NUMBERED):
 			if ('\0' != (c = mchars_num2char(seq, sz)))
 				encode(p, &c, 1);
@@ -584,6 +587,7 @@ term_strlen(const struct termp *p, const char *cp)
 		for (i = 0; i < rsz; i++)
 			sz += (*p->width)(p, *cp++);
 
+		c = 0;
 		switch (*cp) {
 		case ('\\'):
 			cp++;
@@ -591,8 +595,12 @@ term_strlen(const struct termp *p, const char *cp)
 			switch (mandoc_escape(&cp, &seq, &ssz)) {
 			case (ESCAPE_ERROR):
 				return(sz);
+			case (ESCAPE_UNICODE):
+				c = '?';
+				/* FALLTHROUGH */
 			case (ESCAPE_NUMBERED):
-				c = mchars_num2char(seq, ssz);
+				if ('\0' != c)
+					c = mchars_num2char(seq, ssz);
 				if ('\0' != c)
 					sz += (*p->width)(p, c);
 				break;