From e997256c0986b92663c1c5f6e5d5d8a4e8a6eadf Mon Sep 17 00:00:00 2001
From: ben <ben@cda61777-01e9-0310-a592-d414129be87e>
Date: Sun, 2 Dec 2007 18:38:31 +0000
Subject: [PATCH] Cope with TrueType fonts with slightly broken cmaps, just
 ignoring code points that can't be resolved (apart from warning about it).

git-svn-id: svn://svn.tartarus.org/sgt/halibut@7800 cda61777-01e9-0310-a592-d414129be87e
---
 error.c         | 10 +++++++++-
 halibut.h       |  1 +
 in_sfnt.c       | 20 ++++++++++++++++----
 inputs/test.but |  1 +
 4 files changed, 27 insertions(+), 5 deletions(-)

diff --git a/error.c b/error.c
index b6c2007..2a2c6cd 100644
--- a/error.c
+++ b/error.c
@@ -18,7 +18,7 @@ static void do_error(int code, va_list ap) {
     char c;
     int i, j;
     char *sp, *sp2;
-    wchar_t *wsp, *wsp2;
+    wchar_t *wsp, *wsp2, wc;
     filepos fpos, fpos2, *fposp;
     int flags = 0;
 
@@ -374,6 +374,14 @@ static void do_error(int code, va_list ap) {
 	sprintf(error, "font has an invalid header");
 	flags = FILEPOS;
 	break;	
+      case err_sfntbadglyph:
+	fpos = *va_arg(ap, filepos *);
+	wc = va_arg(ap, wchar_t);
+	sprintf(error,
+		"warning: character U+%04X references an non-existent glyph",
+		wc);
+	flags = FILEPOS;
+	break;
       case err_whatever:
 	sp = va_arg(ap, char *);
         vsprintf(error, sp, ap);
diff --git a/halibut.h b/halibut.h
index 183a0cb..4f433e1 100644
--- a/halibut.h
+++ b/halibut.h
@@ -260,6 +260,7 @@ enum {
     err_sfntnounicmap,		       /* sfnt has no UCS-2 cmap */
     err_sfnttablevers,		       /* sfnt table version unknown */
     err_sfntbadhdr,		       /* sfnt has bad header */
+    err_sfntbadglyph,		       /* sfnt cmap references bad glyph */
     err_whatever                       /* random error of another type */
 };
 
diff --git a/in_sfnt.c b/in_sfnt.c
index 73479b1..fc37dc0 100644
--- a/in_sfnt.c
+++ b/in_sfnt.c
@@ -885,19 +885,31 @@ void sfnt_getmap(font_info *fi) {
 			for (k = startCode[j]; k <= endCode[j]; k++) {
 			    idx = (k + idDelta[j]) & 0xffff;
 			    if (idx != 0) {
-				if (idx > sf->nglyphs) goto bad;
+				if (idx > sf->nglyphs)  {
+				    error(err_sfntbadglyph, &sf->pos,
+					  (wchar_t)k);
+				    continue;
+				}
 				fi->bmp[k] = sfnt_indextoglyph(sf, idx);
 			    }
 			}
 		    } else {
 			unsigned startidx = idRangeOffset[j]/2 - segcount + j;
 			for (k = startCode[j]; k <= endCode[j]; k++) {
-			    if (startidx + k - startCode[j] >= nglyphindex)
-				goto bad;
+			    if (startidx + k - startCode[j] >=
+				nglyphindex) {
+				error(err_sfntbadglyph, &sf->pos,
+				      (wchar_t)k);
+				continue;
+			    }
 			    idx = glyphIndexArray[startidx + k - startCode[j]];
 			    if (idx != 0) {
 				idx = (idx + idDelta[j]) & 0xffff;
-				if (idx > sf->nglyphs) goto bad;
+				if (idx > sf->nglyphs) {
+				    error(err_sfntbadglyph, &sf->pos,
+					  (wchar_t)k);
+				    continue;
+				}
 				fi->bmp[k] = sfnt_indextoglyph(sf, idx);
 			    }
 			}
diff --git a/inputs/test.but b/inputs/test.but
index 69c1c82..95b8346 100644
--- a/inputs/test.but
+++ b/inputs/test.but
@@ -13,6 +13,7 @@ backend has to treat it specially in order to pass it to pdfmark.
 \cfg{info-dir-entry}{Florbles}{test.but}{Subsection of Halibut test
 document}{sub-sub}
 \cfg{info-section-underline}{2}{~}
+\cfg{info-charset}{utf-8}
 \cfg{man-headnumbers}{true}
 
 \cfg{contents}{Contents(edited)}
-- 
2.11.0