Fixed problem with UTF-8 BOM

.gitignore

@@ -1,3 +1,4 @@
 build
 epub2txt
 *.o
+*.epub

Makefile

@@ -1,4 +1,4 @@
-VERSION := 2.03
+VERSION := 2.04
 CC      := gcc
 CFLAGS  := -Wall -fPIC -fPIE 
 LDFLAGS := -pie -s

README.md

@@ -1,6 +1,6 @@
 # epub2txt -- Extract text from EPUB documents  
 
-Version 2.03, January 2022 
+Version 2.04, April 2022 
 
 ## What is this? 
 
@@ -232,6 +232,7 @@ even approximately, in ASCII.
 
 Date | Change
 -----|-------
+2.04, Apr 2022 | Improved handling of UTF-8 BOMs 
 2.03, Jan 2022 | Fixed a buffer overrun bug 
 2.02, May 2020 | Updated XML parser 
 2.01, January 2019 | Various bug fixes

man1/epub2txt.1

@@ -1,9 +1,9 @@
-.\" Copyright (C) 2013-19 Kevin Boone 
+.\" Copyright (C) 2013-22 Kevin Boone 
 .\" Permission is granted to any individual or institution to use, copy, or
 .\" redistribute this software so long as all of the original files are
 .\" included, and that this copyright notice is retained.
 .\"
-.TH epub2txt 1 "May 2020"
+.TH epub2txt 1 "April 2022"
 .SH NAME
 epub2txt \- Extract text from EPUB documents 
 .SH SYNOPSIS

src/main.c

@@ -136,7 +136,7 @@ int main (int argc, char **argv)
   if (show_version)
     {
     printf (APPNAME " version " VERSION "\n");
-    printf ("Copyright (c)2013-2020 Kevin Boone and contributors\n");
+    printf ("Copyright (c)2013-2022 Kevin Boone and contributors\n");
     printf ("Distributed under the terms of the GNU Public Licence, v3.0\n");
     exit (0);
     }

src/string.c

@@ -250,9 +250,21 @@ BOOL string_create_from_utf8_file (const char *filename,
     fstat (f, &sb);
     int64_t size = sb.st_size;
     char *buff = malloc (size + 2);
-    read (f, buff, size);
     self->str = buff; 
-    self->str[size] = 0;
+
+    // Read the first three characters, to check for a UTF-8 byte-order-mark
+    read (f, buff, 3);
+    if (buff[0] == (char)0xEF && buff[1] == (char)0xBB && buff[2] == (char)0xBF)
+      {
+      read (f, buff, size - 3);
+      self->str[size - 3] = 0;
+      }
+    else
+      {
+      read (f, buff + 3, size - 3);
+      self->str[size] = 0;
+      }
+
     *result = self;
     ok = TRUE;
     }

src/wstring.c

@@ -92,7 +92,12 @@ BOOL wstring_create_from_utf8_file (const char *filename,
     close (f);
     buff[n] = 0;
 
-    self->str = wstring_convert_utf8_to_utf32 (buff);
+    // Might need to skip a UTF-8 BOM when reading file
+    if (buff[0] == (char)0xEF && buff[1] == (char)0xBB && buff[2] == (char)0xBF)
+      self->str = wstring_convert_utf8_to_utf32 (buff + 3);
+    else
+      self->str = wstring_convert_utf8_to_utf32 (buff);
+
     free (buff);
 
     *result = self;