Update docos.

CHANGES.md

@@ -1,6 +1,30 @@
 Changes in PDFio
 ================
 
+
+v1.5.3 - 2025-05-03
+-------------------
+
+- Fixed decryption of PDF files "protected" by 40-bit RC4 (Issue #42)
+- Fixed decryption of UTF-16 strings (Issue #42)
+- Fixed decryption of PDF files with large permission values.
+- Fixed support for EncryptMetadata key in the encryption dictionary.
+- Fixed `pdfioObjCopy` and `pdfioPageCopy` to properly identify the source PDF
+  file being used (Issue #125)
+
+
+v1.5.2 - 2025-04-12
+-------------------
+
+- Updated maximum allowed PDF string size to 64k (Issue #117)
+- Updated dictionary reading code to discard duplicate key/value pairs with a
+  warning message (Issue #118)
+- Fixed form detection in `pdfioinfo` example code (Issue #114)
+- Fixed parsing of certain date/time values (Issue #115)
+- Fixed support for empty name values (Issue #116)
+- Fixed range checking in `pdfioImageGetBytesPerLine` (Issue #121)
+
+
 v1.5.1 - 2025-03-28
 -------------------
 

configure

@@ -1,6 +1,6 @@
 #! /bin/sh
 # Guess values for system-dependent variables and create Makefiles.
-# Generated by GNU Autoconf 2.71 for pdfio 1.5.1.
+# Generated by GNU Autoconf 2.71 for pdfio 1.5.3.
 #
 # Report bugs to <https://github.com/michaelrsweet/pdfio/issues>.
 #
@@ -610,8 +610,8 @@ MAKEFLAGS=
 # Identity of this package.
 PACKAGE_NAME='pdfio'
 PACKAGE_TARNAME='pdfio'
-PACKAGE_VERSION='1.5.1'
-PACKAGE_STRING='pdfio 1.5.1'
+PACKAGE_VERSION='1.5.3'
+PACKAGE_STRING='pdfio 1.5.3'
 PACKAGE_BUGREPORT='https://github.com/michaelrsweet/pdfio/issues'
 PACKAGE_URL='https://www.msweet.org/pdfio'
 
@@ -1295,7 +1295,7 @@ if test "$ac_init_help" = "long"; then
   # Omit some internal or obsolete options to make the list less imposing.
   # This message is too long to be a string in the A/UX 3.1 sh.
   cat <<_ACEOF
-\`configure' configures pdfio 1.5.1 to adapt to many kinds of systems.
+\`configure' configures pdfio 1.5.3 to adapt to many kinds of systems.
 
 Usage: $0 [OPTION]... [VAR=VALUE]...
 
@@ -1361,7 +1361,7 @@ fi
 
 if test -n "$ac_init_help"; then
   case $ac_init_help in
-     short | recursive ) echo "Configuration of pdfio 1.5.1:";;
+     short | recursive ) echo "Configuration of pdfio 1.5.3:";;
    esac
   cat <<\_ACEOF
 
@@ -1460,7 +1460,7 @@ fi
 test -n "$ac_init_help" && exit $ac_status
 if $ac_init_version; then
   cat <<\_ACEOF
-pdfio configure 1.5.1
+pdfio configure 1.5.3
 generated by GNU Autoconf 2.71
 
 Copyright (C) 2021 Free Software Foundation, Inc.
@@ -1513,39 +1513,6 @@ fi
 
 } # ac_fn_c_try_compile
 
-# ac_fn_c_check_header_compile LINENO HEADER VAR INCLUDES
-# -------------------------------------------------------
-# Tests whether HEADER exists and can be compiled using the include files in
-# INCLUDES, setting the cache variable VAR accordingly.
-ac_fn_c_check_header_compile ()
-{
-  as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack
-  { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $2" >&5
-printf %s "checking for $2... " >&6; }
-if eval test \${$3+y}
-then :
-  printf %s "(cached) " >&6
-else $as_nop
-  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
-/* end confdefs.h.  */
-$4
-#include <$2>
-_ACEOF
-if ac_fn_c_try_compile "$LINENO"
-then :
-  eval "$3=yes"
-else $as_nop
-  eval "$3=no"
-fi
-rm -f core conftest.err conftest.$ac_objext conftest.beam conftest.$ac_ext
-fi
-eval ac_res=\$$3
-	       { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5
-printf "%s\n" "$ac_res" >&6; }
-  eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno
-
-} # ac_fn_c_check_header_compile
-
 # ac_fn_c_try_link LINENO
 # -----------------------
 # Try to link conftest.$ac_ext, and return whether this succeeded.
@@ -1592,6 +1559,101 @@ fi
   as_fn_set_status $ac_retval
 
 } # ac_fn_c_try_link
+
+# ac_fn_c_check_func LINENO FUNC VAR
+# ----------------------------------
+# Tests whether FUNC exists, setting the cache variable VAR accordingly
+ac_fn_c_check_func ()
+{
+  as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack
+  { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $2" >&5
+printf %s "checking for $2... " >&6; }
+if eval test \${$3+y}
+then :
+  printf %s "(cached) " >&6
+else $as_nop
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+/* Define $2 to an innocuous variant, in case <limits.h> declares $2.
+   For example, HP-UX 11i <limits.h> declares gettimeofday.  */
+#define $2 innocuous_$2
+
+/* System header to define __stub macros and hopefully few prototypes,
+   which can conflict with char $2 (); below.  */
+
+#include <limits.h>
+#undef $2
+
+/* Override any GCC internal prototype to avoid an error.
+   Use char because int might match the return type of a GCC
+   builtin and then its argument prototype would still apply.  */
+#ifdef __cplusplus
+extern "C"
+#endif
+char $2 ();
+/* The GNU C library defines this for functions which it implements
+    to always fail with ENOSYS.  Some functions are actually named
+    something starting with __ and the normal name is an alias.  */
+#if defined __stub_$2 || defined __stub___$2
+choke me
+#endif
+
+int
+main (void)
+{
+return $2 ();
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_c_try_link "$LINENO"
+then :
+  eval "$3=yes"
+else $as_nop
+  eval "$3=no"
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.beam \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+eval ac_res=\$$3
+	       { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5
+printf "%s\n" "$ac_res" >&6; }
+  eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno
+
+} # ac_fn_c_check_func
+
+# ac_fn_c_check_header_compile LINENO HEADER VAR INCLUDES
+# -------------------------------------------------------
+# Tests whether HEADER exists and can be compiled using the include files in
+# INCLUDES, setting the cache variable VAR accordingly.
+ac_fn_c_check_header_compile ()
+{
+  as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack
+  { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $2" >&5
+printf %s "checking for $2... " >&6; }
+if eval test \${$3+y}
+then :
+  printf %s "(cached) " >&6
+else $as_nop
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+$4
+#include <$2>
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"
+then :
+  eval "$3=yes"
+else $as_nop
+  eval "$3=no"
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.beam conftest.$ac_ext
+fi
+eval ac_res=\$$3
+	       { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5
+printf "%s\n" "$ac_res" >&6; }
+  eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno
+
+} # ac_fn_c_check_header_compile
 ac_configure_args_raw=
 for ac_arg
 do
@@ -1616,7 +1678,7 @@ cat >config.log <<_ACEOF
 This file contains any messages produced by compilers while
 running configure, to aid debugging if configure makes a mistake.
 
-It was created by pdfio $as_me 1.5.1, which was
+It was created by pdfio $as_me 1.5.3, which was
 generated by GNU Autoconf 2.71.  Invocation command line was
 
   $ $0$ac_configure_args_raw
@@ -2372,9 +2434,9 @@ ac_compiler_gnu=$ac_cv_c_compiler_gnu
 
 
 
-PDFIO_VERSION="1.5.1"
-PDFIO_VERSION_MAJOR="`echo 1.5.1 | awk -F. '{print $1}'`"
-PDFIO_VERSION_MINOR="`echo 1.5.1 | awk -F. '{printf("%d\n",$2);}'`"
+PDFIO_VERSION="1.5.3"
+PDFIO_VERSION_MAJOR="`echo 1.5.3 | awk -F. '{print $1}'`"
+PDFIO_VERSION_MINOR="`echo 1.5.3 | awk -F. '{printf("%d\n",$2);}'`"
 
 
 
@@ -3877,6 +3939,56 @@ INSTALL="$(pwd)/install-sh"
 printf "%s\n" "using $INSTALL" >&6; }
 
 
+
+ac_fn_c_check_func "$LINENO" "timegm" "ac_cv_func_timegm"
+if test "x$ac_cv_func_timegm" = xyes
+then :
+
+
+printf "%s\n" "#define HAVE_TIMEGM 1" >>confdefs.h
+
+    CPPFLAGS="-DHAVE_TIMEGM=1 $CPPFLAGS"
+
+fi
+
+
+{ printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for tm_gmtoff member in tm structure" >&5
+printf %s "checking for tm_gmtoff member in tm structure... " >&6; }
+cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+
+    #include <time.h>
+int
+main (void)
+{
+
+        struct tm t;
+	int o = t.tm_gmtoff;
+
+  ;
+  return 0;
+}
+
+_ACEOF
+if ac_fn_c_try_compile "$LINENO"
+then :
+
+    { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: yes" >&5
+printf "%s\n" "yes" >&6; }
+
+printf "%s\n" "#define HAVE_TM_GMTOFF 1" >>confdefs.h
+
+    CPPFLAGS="-DHAVE_TM_GMTOFF=1 $CPPFLAGS"
+
+else $as_nop
+
+    { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: no" >&5
+printf "%s\n" "no" >&6; }
+
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.beam conftest.$ac_ext
+
+
 if test -n "$ac_tool_prefix"; then
   # Extract the first word of "${ac_tool_prefix}pkg-config", so it can be a program name with args.
 set dummy ${ac_tool_prefix}pkg-config; ac_word=$2
@@ -3998,7 +4110,6 @@ PKGCONFIG_REQUIRES="zlib"
 
 { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for zlib via pkg-config" >&5
 printf %s "checking for zlib via pkg-config... " >&6; }
-
 ac_header= ac_cache=
 for ac_item in $ac_header_c_list
 do
@@ -4988,7 +5099,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
 # report actual input values of CONFIG_FILES etc. instead of their
 # values after options handling.
 ac_log="
-This file was extended by pdfio $as_me 1.5.1, which was
+This file was extended by pdfio $as_me 1.5.3, which was
 generated by GNU Autoconf 2.71.  Invocation command line was
 
   CONFIG_FILES    = $CONFIG_FILES
@@ -5044,7 +5155,7 @@ ac_cs_config_escaped=`printf "%s\n" "$ac_cs_config" | sed "s/^ //; s/'/'\\\\\\\\
 cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
 ac_cs_config='$ac_cs_config_escaped'
 ac_cs_version="\\
-pdfio config.status 1.5.1
+pdfio config.status 1.5.3
 configured by $0, generated by GNU Autoconf 2.71,
   with options \\"\$ac_cs_config\\"
 

configure.ac

@@ -21,7 +21,7 @@ AC_PREREQ([2.70])
 
 
 dnl Package name and version...
-AC_INIT([pdfio], [1.5.1], [https://github.com/michaelrsweet/pdfio/issues], [pdfio], [https://www.msweet.org/pdfio])
+AC_INIT([pdfio], [1.5.3], [https://github.com/michaelrsweet/pdfio/issues], [pdfio], [https://www.msweet.org/pdfio])
 
 PDFIO_VERSION="AC_PACKAGE_VERSION"
 PDFIO_VERSION_MAJOR="`echo AC_PACKAGE_VERSION | awk -F. '{print $1}'`"
@@ -88,6 +88,27 @@ AC_SUBST([INSTALL])
 AC_MSG_RESULT([using $INSTALL])
 
 
+dnl Check for date/time functionality...
+AC_CHECK_FUNC([timegm], [
+    AC_DEFINE([HAVE_TIMEGM], [1], [Do we have the timegm function?])
+    CPPFLAGS="-DHAVE_TIMEGM=1 $CPPFLAGS"
+])
+
+AC_MSG_CHECKING([for tm_gmtoff member in tm structure])
+AC_COMPILE_IFELSE([
+    AC_LANG_PROGRAM([[#include <time.h>]], [[
+        struct tm t;
+	int o = t.tm_gmtoff;
+    ]])
+], [
+    AC_MSG_RESULT([yes])
+    AC_DEFINE([HAVE_TM_GMTOFF], [1], [Have tm_gmtoff member in struct tm?])
+    CPPFLAGS="-DHAVE_TM_GMTOFF=1 $CPPFLAGS"
+], [
+    AC_MSG_RESULT([no])
+])
+
+
 dnl Check for pkg-config, which is used for some other tests later on...
 AC_PATH_TOOL([PKGCONFIG], [pkg-config])
 

doc/pdfio.3

@@ -1,4 +1,4 @@
-.TH pdfio 3 "pdf read/write library" "2025-03-06" "pdf read/write library"
+.TH pdfio 3 "pdf read/write library" "2025-05-03" "pdf read/write library"
 .SH NAME
 pdfio \- pdf read/write library
 .SH Introduction
@@ -325,7 +325,7 @@ where the five arguments to the function are the filename ("myinputfile.pdf"), a
     }
 .fi
 .PP
-The error callback is called for both errors and warnings and accepts the pdfio_file_t pointer, a message string, and the callback pointer value, for example:
+The error callback is called for both errors and warnings and accepts the pdfio_file_t pointer, a message string, and the callback pointer value. It returns true to continue processing the file or false to stop, for example:
 .nf
 
     bool
@@ -335,12 +335,15 @@ The error callback is called for both errors and warnings and accepts the pdfio_
     
       fprintf(stderr, "%s: %s\\n", pdfioFileGetName(pdf), message);
     
-      // Return false to treat warnings as errors
-      return (false);
+      // Return true for warning messages (continue) and false for errors (stop)
+      return (!strncmp(message, "WARNING:", 8));
     }
 .fi
 .PP
 The default error callback (NULL) does the equivalent of the above.
+.PP
+Note: Many errors are unrecoverable, so PDFio ignores the return value from the error callback and always stops processing the PDF file. Warning messages start with the prefix "WARNING:" while errors have no prefix.
+
 .PP
 Each PDF file contains one or more pages. The pdfioFileGetNumPages function returns the number of pages in the file while the pdfioFileGetPage function gets the specified page in the PDF file:
 .nf
@@ -1047,11 +1050,26 @@ The pdfioinfo.c example program opens a PDF file and prints the title, author, c
     {
       const char    *filename;              // PDF filename
       pdfio_file_t  *pdf;                   // PDF file
-      const char    *author;                // Author name
-      time_t        creation_date;          // Creation date
-      struct tm     *creation_tm;           // Creation date/time information
-      char          creation_text[256];     // Creation date/time as a string
-      const char    *title;                 // Title
+      pdfio_dict_t  *catalog;               // Catalog dictionary
+      const char    *author,                // Author name
+                    *creator,               // Creator name
+                    *producer,              // Producer name
+                    *title;                 // Title
+      time_t        creation_date,          // Creation date
+                    modification_date;      // Modification date
+      struct tm     *creation_tm,           // Creation date/time information
+                    *modification_tm;       // Modification date/time information
+      char          creation_text[256],     // Creation date/time as a string
+                    modification_text[256], // Modification date/time human fmt string
+                    range_text[255];        // Page range text
+      size_t        num_pages;              // PDF number of pages
+      bool          has_acroform;           // Does the file have an AcroForm?
+      pdfio_obj_t   *page;                  // Object
+      pdfio_dict_t  *page_dict;             // Object dictionary
+      size_t        cur,                    // Current page index
+                    prev;                   // Previous page index
+      pdfio_rect_t  cur_box,                // Current MediaBox
+                    prev_box;               // Previous MediaBox
     
     
       // Get the filename from the command\-line...
@@ -1064,14 +1082,20 @@ The pdfioinfo.c example program opens a PDF file and prints the title, author, c
       filename = argv[1];
     
       // Open the PDF file with the default callbacks...
-      pdf = pdfioFileOpen(filename, /*password_cb*/NULL, /*password_cbdata*/NULL,
-                          /*error_cb*/NULL, /*error_cbdata*/NULL);
+      pdf = pdfioFileOpen(filename, /*password_cb*/NULL,
+                          /*password_cbdata*/NULL, /*error_cb*/NULL,
+                          /*error_cbdata*/NULL);
       if (pdf == NULL)
         return (1);
     
-      // Get the title and author...
-      author = pdfioFileGetAuthor(pdf);
-      title  = pdfioFileGetTitle(pdf);
+      // Get the title, author, etc...
+      catalog      = pdfioFileGetCatalog(pdf);
+      author       = pdfioFileGetAuthor(pdf);
+      creator      = pdfioFileGetCreator(pdf);
+      has_acroform = pdfioDictGetType(catalog, "AcroForm") != PDFIO_VALTYPE_NONE;
+      num_pages    = pdfioFileGetNumPages(pdf);
+      producer     = pdfioFileGetProducer(pdf);
+      title        = pdfioFileGetTitle(pdf);
     
       // Get the creation date and convert to a string...
       if ((creation_date = pdfioFileGetCreationDate(pdf)) > 0)
@@ -1084,12 +1108,76 @@ The pdfioinfo.c example program opens a PDF file and prints the title, author, c
         snprintf(creation_text, sizeof(creation_text), "\-\- not set \-\-");
       }
     
+      // Get the modification date and convert to a string...
+      if ((modification_date = pdfioFileGetModificationDate(pdf)) > 0)
+      {
+        modification_tm = localtime(&modification_date);
+        strftime(modification_text, sizeof(modification_text), "%c", modification_tm);
+      }
+      else
+      {
+        snprintf(modification_text, sizeof(modification_text), "\-\- not set \-\-");
+      }
+    
       // Print file information to stdout...
       printf("%s:\\n", filename);
-      printf("         Title: %s\\n", title ? title : "\-\- not set \-\-");
-      printf("        Author: %s\\n", author ? author : "\-\- not set \-\-");
-      printf("    Created On: %s\\n", creation_text);
-      printf("  Number Pages: %u\\n", (unsigned)pdfioFileGetNumPages(pdf));
+      printf("           Title: %s\\n", title ? title : "\-\- not set \-\-");
+      printf("          Author: %s\\n", author ? author : "\-\- not set \-\-");
+      printf("         Creator: %s\\n", creator ? creator : "\-\- not set \-\-");
+      printf("        Producer: %s\\n", producer ? producer : "\-\- not set \-\-");
+      printf("      Created On: %s\\n", creation_text);
+      printf("     Modified On: %s\\n", modification_text);
+      printf("         Version: %s\\n", pdfioFileGetVersion(pdf));
+      printf("        AcroForm: %s\\n", has_acroform ? "Yes" : "No");
+      printf(" Number of Pages: %u\\n", (unsigned)num_pages);
+    
+      // Report the MediaBox for all of the pages
+      prev_box.x1 = prev_box.x2 = prev_box.y1 = prev_box.y2 = 0.0;
+    
+      for (cur = 0, prev = 0; cur < num_pages; cur ++)
+      {
+        // Find the MediaBox for this page in the page tree...
+        for (page = pdfioFileGetPage(pdf, cur);
+             page != NULL;
+             page = pdfioDictGetObj(page_dict, "Parent"))
+        {
+          cur_box.x1 = cur_box.x2 = cur_box.y1 = cur_box.y2 = 0.0;
+          page_dict  = pdfioObjGetDict(page);
+    
+          if (pdfioDictGetRect(page_dict, "MediaBox", &cur_box))
+            break;
+        }
+    
+        // If this MediaBox is different from the previous one, show the range of
+        // pages that have that size...
+        if (cur == 0 ||
+            fabs(cur_box.x1 \- prev_box.x1) > 0.01 ||
+            fabs(cur_box.y1 \- prev_box.y1) > 0.01 ||
+            fabs(cur_box.x2 \- prev_box.x2) > 0.01 ||
+            fabs(cur_box.y2 \- prev_box.y2) > 0.01)
+        {
+          if (cur > prev)
+          {
+            snprintf(range_text, sizeof(range_text), "Pages %u\-%u",
+                     (unsigned)(prev + 1), (unsigned)cur);
+            printf("%16s: [%g %g %g %g]\\n", range_text,
+                   prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+          }
+    
+          // Start a new series of pages with the new size...
+          prev     = cur;
+          prev_box = cur_box;
+        }
+      }
+    
+      // Show the last range as needed...
+      if (cur > prev)
+      {
+        snprintf(range_text, sizeof(range_text), "Pages %u\-%u",
+                 (unsigned)(prev + 1), (unsigned)cur);
+        printf("%16s: [%g %g %g %g]\\n", range_text,
+               prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+      }
     
       // Close the PDF file...
       pdfioFileClose(pdf);
@@ -2392,7 +2480,7 @@ PDFIO_ENCRYPTION_RC4_128
 .TP 5
 PDFIO_ENCRYPTION_RC4_40
 .br
-40-bit RC4 encryption (PDF 1.3)
+40-bit RC4 encryption (PDF 1.3, reading only)
 .SS pdfio_filter_e
 Compression/decompression filters for streams
 .TP 5
@@ -2862,6 +2950,8 @@ bool  pdfioArrayRemove (
     size_t n
 );
 .fi
+.PP
+
 .SS pdfioContentClip
 Clip output to the current path.
 .PP
@@ -2998,6 +3088,8 @@ bool  pdfioContentPathEnd (
     pdfio_stream_t *st
 );
 .fi
+.PP
+
 .SS pdfioContentPathLineTo
 Add a straight line to the current path.
 .PP
@@ -3338,6 +3430,8 @@ double  pdfioContentTextMeasure (
 This function measures the given text string "s" and returns its width based
 on "size". The text string must always use the UTF-8 (Unicode) encoding but
 any control characters (such as newlines) are ignored.
+
+
 .SS pdfioContentTextMoveLine
 Move to the next line and offset.
 .PP
@@ -3366,6 +3460,8 @@ bool  pdfioContentTextNewLine (
     pdfio_stream_t *st
 );
 .fi
+.PP
+
 .SS pdfioContentTextNewLineShow
 Move to the next line and show text.
 .PP
@@ -3383,6 +3479,8 @@ This function moves to the next line and then shows some text with optional
 word and character spacing in a PDF content stream. The "unicode" argument
 specifies that the current font maps to full Unicode.  The "s" argument
 specifies a UTF-8 encoded string.
+
+
 .SS pdfioContentTextNewLineShowf
 Show formatted text.
 .PP
@@ -3401,6 +3499,8 @@ This function moves to the next line and shows some formatted text with
 optional word and character spacing in a PDF content stream. The "unicode"
 argument specifies that the current font maps to full Unicode.  The "format"
 argument specifies a UTF-8 encoded \fBprintf\fR-style format string.
+
+
 .SS pdfioContentTextShow
 Show text.
 .PP
@@ -3451,6 +3551,8 @@ bool  pdfioDictClear (
     const char *key
 );
 .fi
+.PP
+
 .SS pdfioDictCopy
 Copy a dictionary to a PDF file.
 .PP
@@ -3523,6 +3625,8 @@ const char * pdfioDictGetKey (
     size_t n
 );
 .fi
+.PP
+
 .SS pdfioDictGetName
 Get a key name value from a dictionary.
 .PP
@@ -3540,6 +3644,8 @@ size_t  pdfioDictGetNumPairs (
     pdfio_dict_t *dict
 );
 .fi
+.PP
+
 .SS pdfioDictGetNumber
 Get a key number value from a dictionary.
 .PP
@@ -3612,6 +3718,8 @@ function "cb":
 
 The iteration continues as long as the callback returns \fBtrue\fR or all keys
 have been iterated.
+
+
 .SS pdfioDictSetArray
 Set a key array in a dictionary.
 .PP
@@ -3767,8 +3875,9 @@ CropBox for pages in the PDF file - if \fBNULL\fR then a default "Universal" siz
 of 8.27x11in (the intersection of US Letter and ISO A4) is used.
 .PP
 The "error_cb" and "error_cbdata" arguments specify an error handler callback
-and its data pointer - if \fBNULL\fR the default error handler is used that
-writes error messages to \fBstderr\fR.
+and its data pointer - if \fBNULL\fR then the default error handler is used that
+writes error messages to \fBstderr\fR.  The error handler callback should return
+\fBtrue\fR to continue writing the PDF file or \fBfalse\fR to stop.
 .SS pdfioFileCreateArrayObj
 Create a new object in a PDF file containing an array.
 .PP
@@ -3926,6 +4035,8 @@ pdfio_obj_t * pdfioFileCreateNameObj (
 .PP
 This function creates a new object with a name value in a PDF file.
 You must call \fIpdfioObjClose\fR to write the object to the file.
+
+
 .SS pdfioFileCreateNumberObj
 Create a new object in a PDF file containing a number.
 .PP
@@ -3938,6 +4049,8 @@ pdfio_obj_t * pdfioFileCreateNumberObj (
 .PP
 This function creates a new object with a number value in a PDF file.
 You must call \fIpdfioObjClose\fR to write the object to the file.
+
+
 .SS pdfioFileCreateObj
 Create a new object in a PDF file.
 .PP
@@ -3986,8 +4099,9 @@ CropBox for pages in the PDF file - if \fBNULL\fR then a default "Universal" siz
 of 8.27x11in (the intersection of US Letter and ISO A4) is used.
 .PP
 The "error_cb" and "error_cbdata" arguments specify an error handler callback
-and its data pointer - if \fBNULL\fR the default error handler is used that
-writes error messages to \fBstderr\fR.
+and its data pointer - if \fBNULL\fR then the default error handler is used that
+writes error messages to \fBstderr\fR.  The error handler callback should return
+\fBtrue\fR to continue writing the PDF file or \fBfalse\fR to stop.
 .PP
 .IP 5
 \fINote\fR: Files created using this API are slightly larger than those
@@ -4016,6 +4130,8 @@ pdfio_obj_t * pdfioFileCreateStringObj (
 .PP
 This function creates a new object with a string value in a PDF file.
 You must call \fIpdfioObjClose\fR to write the object to the file.
+
+
 .SS pdfioFileCreateTemporary
 
 .PP
@@ -4209,8 +4325,18 @@ cancel the open.  If \fBNULL\fR is specified for the callback function and the
 PDF file requires a password, the open will always fail.
 .PP
 The "error_cb" and "error_cbdata" arguments specify an error handler callback
-and its data pointer - if \fBNULL\fR the default error handler is used that
-writes error messages to \fBstderr\fR.
+and its data pointer - if \fBNULL\fR then the default error handler is used that
+writes error messages to \fBstderr\fR.  The error handler callback should return
+\fBtrue\fR to continue reading the PDF file or \fBfalse\fR to stop.
+.PP
+.IP 5
+Note: Error messages starting with "WARNING:" are actually warning
+.IP 5
+messages - the callback should normally return \fBtrue\fR to allow PDFio to
+.IP 5
+try to resolve the issue.  In addition, some errors are unrecoverable and
+.IP 5
+ignore the return value of the error callback.
 .SS pdfioFileSetAuthor
 Set the author for a PDF file.
 .PP
@@ -4389,6 +4515,8 @@ const char * pdfioObjGetName (
     pdfio_obj_t *obj
 );
 .fi
+.PP
+
 .SS pdfioObjGetNumber
 Get the object's number.
 .PP
@@ -4590,6 +4718,11 @@ bool  pdfioStreamPrintf (
     ...
 );
 .fi
+.PP
+This function writes a formatted string to a stream.  In addition to the
+standard \fBprintf\fR format characters, you can use "%H" to format a HTML/XML
+string value, "%N" to format a PDF name value ("/Name"), and "%S" to format
+a PDF string ("(String)") value.
 .SS pdfioStreamPutChar
 Write a single character to a stream.
 .PP

doc/pdfio.html

@@ -1,13 +1,13 @@
 <!DOCTYPE html>
 <html lang="en-US">
 <head>
-<title>PDFio Programming Manual v1.5.0</title>
+<title>PDFio Programming Manual v1.5.3</title>
 <meta http-equiv="Content-Type" content="text/html;charset=utf-8">
 <meta name="generator" content="codedoc v3.8">
 <meta name="author" content="Michael R Sweet">
 <meta name="language" content="en-US">
 <meta name="copyright" content="Copyright © 2021-2025 by Michael R Sweet">
-<meta name="version" content="1.5.0">
+<meta name="version" content="1.5.3">
 <style type="text/css"><!--
 body {
   background: white;
@@ -251,7 +251,7 @@ span.string {
 <body>
 <div class="header">
 <p><img class="title" src="pdfio-512.png"></p>
-<h1 class="title">PDFio Programming Manual v1.5.0</h1>
+<h1 class="title">PDFio Programming Manual v1.5.3</h1>
 <p>Michael R Sweet</p>
 <p>Copyright © 2021-2025 by Michael R Sweet</p>
 </div>
@@ -728,7 +728,7 @@ password_cb(<span class="reserved">void</span> *data, <span class="reserved">con
   <span class="reserved">return</span> (<span class="string">&quot;Password42&quot;</span>);
 }
 </code></pre>
-<p>The error callback is called for both errors and warnings and accepts the <code>pdfio_file_t</code> pointer, a message string, and the callback pointer value, for example:</p>
+<p>The error callback is called for both errors and warnings and accepts the <code>pdfio_file_t</code> pointer, a message string, and the callback pointer value. It returns <code>true</code> to continue processing the file or <code>false</code> to stop, for example:</p>
 <pre><code class="language-c"><span class="reserved">bool</span>
 error_cb(pdfio_file_t *pdf, <span class="reserved">const</span> <span class="reserved">char</span> *message, <span class="reserved">void</span> *data)
 {
@@ -736,11 +736,14 @@ error_cb(pdfio_file_t *pdf, <span class="reserved">const</span> <span class="res
 
   fprintf(stderr, <span class="string">&quot;%s: %s\n&quot;</span>, pdfioFileGetName(pdf), message);
 
-  <span class="comment">// Return false to treat warnings as errors</span>
-  <span class="reserved">return</span> (<span class="reserved">false</span>);
+  <span class="comment">// Return true for warning messages (continue) and false for errors (stop)</span>
+  <span class="reserved">return</span> (!strncmp(message, <span class="string">&quot;WARNING:&quot;</span>, <span class="number">8</span>));
 }
 </code></pre>
 <p>The default error callback (<code>NULL</code>) does the equivalent of the above.</p>
+<blockquote>
+<p>Note: Many errors are unrecoverable, so PDFio ignores the return value from the error callback and always stops processing the PDF file. Warning messages start with the prefix &quot;WARNING:&quot; while errors have no prefix.</p>
+</blockquote>
 <p>Each PDF file contains one or more pages. The <a href="#pdfioFileGetNumPages"><code>pdfioFileGetNumPages</code></a> function returns the number of pages in the file while the <a href="#pdfioFileGetPage"><code>pdfioFileGetPage</code></a> function gets the specified page in the PDF file:</p>
 <pre><code class="language-c">pdfio_file_t *pdf;   <span class="comment">// PDF file</span>
 size_t       i;      <span class="comment">// Looping var</span>
@@ -1165,11 +1168,26 @@ main(<span class="reserved">int</span>  argc,                         <span clas
 {
   <span class="reserved">const</span> <span class="reserved">char</span>    *filename;              <span class="comment">// PDF filename</span>
   pdfio_file_t  *pdf;                   <span class="comment">// PDF file</span>
-  <span class="reserved">const</span> <span class="reserved">char</span>    *author;                <span class="comment">// Author name</span>
-  time_t        creation_date;          <span class="comment">// Creation date</span>
-  <span class="reserved">struct</span> tm     *creation_tm;           <span class="comment">// Creation date/time information</span>
-  <span class="reserved">char</span>          creation_text[<span class="number">256</span>];     <span class="comment">// Creation date/time as a string</span>
-  <span class="reserved">const</span> <span class="reserved">char</span>    *title;                 <span class="comment">// Title</span>
+  pdfio_dict_t  *catalog;               <span class="comment">// Catalog dictionary</span>
+  <span class="reserved">const</span> <span class="reserved">char</span>    *author,                <span class="comment">// Author name</span>
+                *creator,               <span class="comment">// Creator name</span>
+                *producer,              <span class="comment">// Producer name</span>
+                *title;                 <span class="comment">// Title</span>
+  time_t        creation_date,          <span class="comment">// Creation date</span>
+                modification_date;      <span class="comment">// Modification date</span>
+  <span class="reserved">struct</span> tm     *creation_tm,           <span class="comment">// Creation date/time information</span>
+                *modification_tm;       <span class="comment">// Modification date/time information</span>
+  <span class="reserved">char</span>          creation_text[<span class="number">256</span>],     <span class="comment">// Creation date/time as a string</span>
+                modification_text[<span class="number">256</span>], <span class="comment">// Modification date/time human fmt string</span>
+                range_text[<span class="number">255</span>];        <span class="comment">// Page range text</span>
+  size_t        num_pages;              <span class="comment">// PDF number of pages</span>
+  <span class="reserved">bool</span>          has_acroform;           <span class="comment">// Does the file have an AcroForm?</span>
+  pdfio_obj_t   *page;                  <span class="comment">// Object</span>
+  pdfio_dict_t  *page_dict;             <span class="comment">// Object dictionary</span>
+  size_t        cur,                    <span class="comment">// Current page index</span>
+                prev;                   <span class="comment">// Previous page index</span>
+  pdfio_rect_t  cur_box,                <span class="comment">// Current MediaBox</span>
+                prev_box;               <span class="comment">// Previous MediaBox</span>
 
 
   <span class="comment">// Get the filename from the command-line...</span>
@@ -1182,14 +1200,20 @@ main(<span class="reserved">int</span>  argc,                         <span clas
   filename = argv[<span class="number">1</span>];
 
   <span class="comment">// Open the PDF file with the default callbacks...</span>
-  pdf = pdfioFileOpen(filename, <span class="comment">/*password_cb*/</span>NULL, <span class="comment">/*password_cbdata*/</span>NULL,
-                      <span class="comment">/*error_cb*/</span>NULL, <span class="comment">/*error_cbdata*/</span>NULL);
+  pdf = pdfioFileOpen(filename, <span class="comment">/*password_cb*/</span>NULL,
+                      <span class="comment">/*password_cbdata*/</span>NULL, <span class="comment">/*error_cb*/</span>NULL,
+                      <span class="comment">/*error_cbdata*/</span>NULL);
   <span class="reserved">if</span> (pdf == NULL)
     <span class="reserved">return</span> (<span class="number">1</span>);
 
-  <span class="comment">// Get the title and author...</span>
-  author = pdfioFileGetAuthor(pdf);
-  title  = pdfioFileGetTitle(pdf);
+  <span class="comment">// Get the title, author, etc...</span>
+  catalog      = pdfioFileGetCatalog(pdf);
+  author       = pdfioFileGetAuthor(pdf);
+  creator      = pdfioFileGetCreator(pdf);
+  has_acroform = pdfioDictGetType(catalog, <span class="string">&quot;AcroForm&quot;</span>) != PDFIO_VALTYPE_NONE;
+  num_pages    = pdfioFileGetNumPages(pdf);
+  producer     = pdfioFileGetProducer(pdf);
+  title        = pdfioFileGetTitle(pdf);
 
   <span class="comment">// Get the creation date and convert to a string...</span>
   <span class="reserved">if</span> ((creation_date = pdfioFileGetCreationDate(pdf)) &gt; <span class="number">0</span>)
@@ -1202,12 +1226,76 @@ main(<span class="reserved">int</span>  argc,                         <span clas
     snprintf(creation_text, <span class="reserved">sizeof</span>(creation_text), <span class="string">&quot;-- not set --&quot;</span>);
   }
 
+  <span class="comment">// Get the modification date and convert to a string...</span>
+  <span class="reserved">if</span> ((modification_date = pdfioFileGetModificationDate(pdf)) &gt; <span class="number">0</span>)
+  {
+    modification_tm = localtime(&amp;modification_date);
+    strftime(modification_text, <span class="reserved">sizeof</span>(modification_text), <span class="string">&quot;%c&quot;</span>, modification_tm);
+  }
+  <span class="reserved">else</span>
+  {
+    snprintf(modification_text, <span class="reserved">sizeof</span>(modification_text), <span class="string">&quot;-- not set --&quot;</span>);
+  }
+
   <span class="comment">// Print file information to stdout...</span>
   printf(<span class="string">&quot;%s:\n&quot;</span>, filename);
-  printf(<span class="string">&quot;         Title: %s\n&quot;</span>, title ? title : <span class="string">&quot;-- not set --&quot;</span>);
-  printf(<span class="string">&quot;        Author: %s\n&quot;</span>, author ? author : <span class="string">&quot;-- not set --&quot;</span>);
-  printf(<span class="string">&quot;    Created On: %s\n&quot;</span>, creation_text);
-  printf(<span class="string">&quot;  Number Pages: %u\n&quot;</span>, (<span class="reserved">unsigned</span>)pdfioFileGetNumPages(pdf));
+  printf(<span class="string">&quot;           Title: %s\n&quot;</span>, title ? title : <span class="string">&quot;-- not set --&quot;</span>);
+  printf(<span class="string">&quot;          Author: %s\n&quot;</span>, author ? author : <span class="string">&quot;-- not set --&quot;</span>);
+  printf(<span class="string">&quot;         Creator: %s\n&quot;</span>, creator ? creator : <span class="string">&quot;-- not set --&quot;</span>);
+  printf(<span class="string">&quot;        Producer: %s\n&quot;</span>, producer ? producer : <span class="string">&quot;-- not set --&quot;</span>);
+  printf(<span class="string">&quot;      Created On: %s\n&quot;</span>, creation_text);
+  printf(<span class="string">&quot;     Modified On: %s\n&quot;</span>, modification_text);
+  printf(<span class="string">&quot;         Version: %s\n&quot;</span>, pdfioFileGetVersion(pdf));
+  printf(<span class="string">&quot;        AcroForm: %s\n&quot;</span>, has_acroform ? <span class="string">&quot;Yes&quot;</span> : <span class="string">&quot;No&quot;</span>);
+  printf(<span class="string">&quot; Number of Pages: %u\n&quot;</span>, (<span class="reserved">unsigned</span>)num_pages);
+
+  <span class="comment">// Report the MediaBox for all of the pages</span>
+  prev_box.x1 = prev_box.x2 = prev_box.y1 = prev_box.y2 = <span class="number">0.0</span>;
+
+  <span class="reserved">for</span> (cur = <span class="number">0</span>, prev = <span class="number">0</span>; cur &lt; num_pages; cur ++)
+  {
+    <span class="comment">// Find the MediaBox for this page in the page tree...</span>
+    <span class="reserved">for</span> (page = pdfioFileGetPage(pdf, cur);
+         page != NULL;
+         page = pdfioDictGetObj(page_dict, <span class="string">&quot;Parent&quot;</span>))
+    {
+      cur_box.x1 = cur_box.x2 = cur_box.y1 = cur_box.y2 = <span class="number">0.0</span>;
+      page_dict  = pdfioObjGetDict(page);
+
+      <span class="reserved">if</span> (pdfioDictGetRect(page_dict, <span class="string">&quot;MediaBox&quot;</span>, &amp;cur_box))
+        <span class="reserved">break</span>;
+    }
+
+    <span class="comment">// If this MediaBox is different from the previous one, show the range of</span>
+    <span class="comment">// pages that have that size...</span>
+    <span class="reserved">if</span> (cur == <span class="number">0</span> ||
+        fabs(cur_box.x1 - prev_box.x1) &gt; <span class="number">0.01</span> ||
+        fabs(cur_box.y1 - prev_box.y1) &gt; <span class="number">0.01</span> ||
+        fabs(cur_box.x2 - prev_box.x2) &gt; <span class="number">0.01</span> ||
+        fabs(cur_box.y2 - prev_box.y2) &gt; <span class="number">0.01</span>)
+    {
+      <span class="reserved">if</span> (cur &gt; prev)
+      {
+        snprintf(range_text, <span class="reserved">sizeof</span>(range_text), <span class="string">&quot;Pages %u-%u&quot;</span>,
+                 (<span class="reserved">unsigned</span>)(prev + <span class="number">1</span>), (<span class="reserved">unsigned</span>)cur);
+        printf(<span class="string">&quot;%16s: [%g %g %g %g]\n&quot;</span>, range_text,
+               prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+      }
+
+      <span class="comment">// Start a new series of pages with the new size...</span>
+      prev     = cur;
+      prev_box = cur_box;
+    }
+  }
+
+  <span class="comment">// Show the last range as needed...</span>
+  <span class="reserved">if</span> (cur &gt; prev)
+  {
+    snprintf(range_text, <span class="reserved">sizeof</span>(range_text), <span class="string">&quot;Pages %u-%u&quot;</span>,
+             (<span class="reserved">unsigned</span>)(prev + <span class="number">1</span>), (<span class="reserved">unsigned</span>)cur);
+    printf(<span class="string">&quot;%16s: [%g %g %g %g]\n&quot;</span>, range_text,
+           prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+  }
 
   <span class="comment">// Close the PDF file...</span>
   pdfioFileClose(pdf);
@@ -2661,7 +2749,7 @@ size_t pdfioArrayGetSize(<a href="#pdfio_array_t">pdfio_array_t</a> *a);</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Value type</p>
-<h3 class="function"><a id="pdfioArrayRemove">pdfioArrayRemove</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioArrayRemove">pdfioArrayRemove</a></h3>
 <p class="description">Remove an array entry.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioArrayRemove(<a href="#pdfio_array_t">pdfio_array_t</a> *a, size_t n);</p>
@@ -2866,7 +2954,7 @@ using the <a href="#pdfioPageDictAddImage"><code>pdfioPageDictAddImage</code></a
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description"><code>true</code> on success, <code>false</code> on failure</p>
-<h3 class="function"><a id="pdfioContentPathEnd">pdfioContentPathEnd</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.1&#160;</span><a id="pdfioContentPathEnd">pdfioContentPathEnd</a></h3>
 <p class="description">Clear the current path.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioContentPathEnd(<a href="#pdfio_stream_t">pdfio_stream_t</a> *st);</p>
@@ -3356,7 +3444,7 @@ are 0, a solid line is drawn.</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description"><code>true</code> on success, <code>false</code> on failure</p>
-<h3 class="function"><a id="pdfioContentTextMeasure">pdfioContentTextMeasure</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioContentTextMeasure">pdfioContentTextMeasure</a></h3>
 <p class="description">Measure a text string and return its width.</p>
 <p class="code">
 <span class="reserved">double</span> pdfioContentTextMeasure(<a href="#pdfio_obj_t">pdfio_obj_t</a> *font, <span class="reserved">const</span> <span class="reserved">char</span> *s, <span class="reserved">double</span> size);</p>
@@ -3374,7 +3462,9 @@ are 0, a solid line is drawn.</p>
 <h4 class="discussion">Discussion</h4>
 <p class="discussion">This function measures the given text string &quot;s&quot; and returns its width based
 on &quot;size&quot;. The text string must always use the UTF-8 (Unicode) encoding but
-any control characters (such as newlines) are ignored.</p>
+any control characters (such as newlines) are ignored.
+
+</p>
 <h3 class="function"><a id="pdfioContentTextMoveLine">pdfioContentTextMoveLine</a></h3>
 <p class="description">Move to the next line and offset.</p>
 <p class="code">
@@ -3405,7 +3495,7 @@ any control characters (such as newlines) are ignored.</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description"><code>true</code> on success, <code>false</code> on failure</p>
-<h3 class="function"><a id="pdfioContentTextNewLine">pdfioContentTextNewLine</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioContentTextNewLine">pdfioContentTextNewLine</a></h3>
 <p class="description">Move to the next line.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioContentTextNewLine(<a href="#pdfio_stream_t">pdfio_stream_t</a> *st);</p>
@@ -3416,7 +3506,7 @@ any control characters (such as newlines) are ignored.</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description"><code>true</code> on success, <code>false</code> on failure</p>
-<h3 class="function"><a id="pdfioContentTextNewLineShow">pdfioContentTextNewLineShow</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioContentTextNewLineShow">pdfioContentTextNewLineShow</a></h3>
 <p class="description">Move to the next line and show text.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioContentTextNewLineShow(<a href="#pdfio_stream_t">pdfio_stream_t</a> *st, <span class="reserved">double</span> ws, <span class="reserved">double</span> cs, <span class="reserved">bool</span> unicode, <span class="reserved">const</span> <span class="reserved">char</span> *s);</p>
@@ -3439,8 +3529,10 @@ any control characters (such as newlines) are ignored.</p>
 <p class="discussion">This function moves to the next line and then shows some text with optional
 word and character spacing in a PDF content stream. The &quot;unicode&quot; argument
 specifies that the current font maps to full Unicode.  The &quot;s&quot; argument
-specifies a UTF-8 encoded string.</p>
-<h3 class="function"><a id="pdfioContentTextNewLineShowf">pdfioContentTextNewLineShowf</a></h3>
+specifies a UTF-8 encoded string.
+
+</p>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioContentTextNewLineShowf">pdfioContentTextNewLineShowf</a></h3>
 <p class="description">Show formatted text.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioContentTextNewLineShowf(<a href="#pdfio_stream_t">pdfio_stream_t</a> *st, <span class="reserved">double</span> ws, <span class="reserved">double</span> cs, <span class="reserved">bool</span> unicode, <span class="reserved">const</span> <span class="reserved">char</span> *format, ...);</p>
@@ -3465,7 +3557,9 @@ specifies a UTF-8 encoded string.</p>
 <p class="discussion">This function moves to the next line and shows some formatted text with
 optional word and character spacing in a PDF content stream. The &quot;unicode&quot;
 argument specifies that the current font maps to full Unicode.  The &quot;format&quot;
-argument specifies a UTF-8 encoded <code>printf</code>-style format string.</p>
+argument specifies a UTF-8 encoded <code>printf</code>-style format string.
+
+</p>
 <h3 class="function"><a id="pdfioContentTextShow">pdfioContentTextShow</a></h3>
 <p class="description">Show text.</p>
 <p class="code">
@@ -3528,7 +3622,7 @@ argument specifies an array of UTF-8 encoded strings.</p>
 <p class="discussion">This function shows some formatted text in a PDF content stream. The
 &quot;unicode&quot; argument specifies that the current font maps to full Unicode.
 The &quot;format&quot; argument specifies a UTF-8 encoded <code>printf</code>-style format string.</p>
-<h3 class="function"><a id="pdfioDictClear">pdfioDictClear</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioDictClear">pdfioDictClear</a></h3>
 <p class="description">Remove a key/value pair from a dictionary.</p>
 <p class="code">
 <span class="reserved">bool</span> pdfioDictClear(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict, <span class="reserved">const</span> <span class="reserved">char</span> *key);</p>
@@ -3632,7 +3726,7 @@ time_t pdfioDictGetDate(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict, <span cl
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Value</p>
-<h3 class="function"><a id="pdfioDictGetKey">pdfioDictGetKey</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioDictGetKey">pdfioDictGetKey</a></h3>
 <p class="description">Get the key for the specified pair.</p>
 <p class="code">
 <span class="reserved">const</span> <span class="reserved">char</span> *pdfioDictGetKey(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict, size_t n);</p>
@@ -3658,7 +3752,7 @@ time_t pdfioDictGetDate(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict, <span cl
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Value</p>
-<h3 class="function"><a id="pdfioDictGetNumPairs">pdfioDictGetNumPairs</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioDictGetNumPairs">pdfioDictGetNumPairs</a></h3>
 <p class="description">Get the number of key/value pairs in a dictionary.</p>
 <p class="code">
 size_t pdfioDictGetNumPairs(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict);</p>
@@ -3736,7 +3830,7 @@ size_t pdfioDictGetNumPairs(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict);</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Value type</p>
-<h3 class="function"><a id="pdfioDictIterateKeys">pdfioDictIterateKeys</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.1&#160;</span><a id="pdfioDictIterateKeys">pdfioDictIterateKeys</a></h3>
 <p class="description">Iterate the keys in a dictionary.</p>
 <p class="code">
 <span class="reserved">void</span> pdfioDictIterateKeys(<a href="#pdfio_dict_t">pdfio_dict_t</a> *dict, <a href="#pdfio_dict_cb_t">pdfio_dict_cb_t</a> cb, <span class="reserved">void</span> *cb_data);</p>
@@ -3763,7 +3857,9 @@ my_dict_cb(pdfio_dict_t *dict, const char *key, void *cb_data)
 </pre>
 
 The iteration continues as long as the callback returns <code>true</code> or all keys
-have been iterated.</p>
+have been iterated.
+
+</p>
 <h3 class="function"><a id="pdfioDictSetArray">pdfioDictSetArray</a></h3>
 <p class="description">Set a key array in a dictionary.</p>
 <p class="code">
@@ -3991,8 +4087,9 @@ CropBox for pages in the PDF file - if <code>NULL</code> then a default &quot;Un
 of 8.27x11in (the intersection of US Letter and ISO A4) is used.<br>
 <br>
 The &quot;error_cb&quot; and &quot;error_cbdata&quot; arguments specify an error handler callback
-and its data pointer - if <code>NULL</code> the default error handler is used that
-writes error messages to <code>stderr</code>.</p>
+and its data pointer - if <code>NULL</code> then the default error handler is used that
+writes error messages to <code>stderr</code>.  The error handler callback should return
+<code>true</code> to continue writing the PDF file or <code>false</code> to stop.</p>
 <h3 class="function"><a id="pdfioFileCreateArrayObj">pdfioFileCreateArrayObj</a></h3>
 <p class="description">Create a new object in a PDF file containing an array.</p>
 <p class="code">
@@ -4175,7 +4272,7 @@ image on the page.<br>
 Note: Currently PNG support is limited to grayscale, RGB, or indexed files
 without interlacing or alpha.  Transparency (masking) based on color/index
 is supported.</blockquote>
-<h3 class="function"><a id="pdfioFileCreateNameObj">pdfioFileCreateNameObj</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioFileCreateNameObj">pdfioFileCreateNameObj</a></h3>
 <p class="description">Create a new object in a PDF file containing a name.</p>
 <p class="code">
 <a href="#pdfio_obj_t">pdfio_obj_t</a> *pdfioFileCreateNameObj(<a href="#pdfio_file_t">pdfio_file_t</a> *pdf, <span class="reserved">const</span> <span class="reserved">char</span> *name);</p>
@@ -4190,8 +4287,10 @@ is supported.</blockquote>
 <p class="description">New object</p>
 <h4 class="discussion">Discussion</h4>
 <p class="discussion">This function creates a new object with a name value in a PDF file.
-You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.</p>
-<h3 class="function"><a id="pdfioFileCreateNumberObj">pdfioFileCreateNumberObj</a></h3>
+You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.
+
+</p>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioFileCreateNumberObj">pdfioFileCreateNumberObj</a></h3>
 <p class="description">Create a new object in a PDF file containing a number.</p>
 <p class="code">
 <a href="#pdfio_obj_t">pdfio_obj_t</a> *pdfioFileCreateNumberObj(<a href="#pdfio_file_t">pdfio_file_t</a> *pdf, <span class="reserved">double</span> number);</p>
@@ -4206,7 +4305,9 @@ You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write t
 <p class="description">New object</p>
 <h4 class="discussion">Discussion</h4>
 <p class="discussion">This function creates a new object with a number value in a PDF file.
-You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.</p>
+You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.
+
+</p>
 <h3 class="function"><a id="pdfioFileCreateObj">pdfioFileCreateObj</a></h3>
 <p class="description">Create a new object in a PDF file.</p>
 <p class="code">
@@ -4267,8 +4368,9 @@ CropBox for pages in the PDF file - if <code>NULL</code> then a default &quot;Un
 of 8.27x11in (the intersection of US Letter and ISO A4) is used.<br>
 <br>
 The &quot;error_cb&quot; and &quot;error_cbdata&quot; arguments specify an error handler callback
-and its data pointer - if <code>NULL</code> the default error handler is used that
-writes error messages to <code>stderr</code>.<br>
+and its data pointer - if <code>NULL</code> then the default error handler is used that
+writes error messages to <code>stderr</code>.  The error handler callback should return
+<code>true</code> to continue writing the PDF file or <code>false</code> to stop.<br>
 <br>
 </p><blockquote>
 <em>Note</em>: Files created using this API are slightly larger than those
@@ -4287,7 +4389,7 @@ stored as indirect object references.</blockquote>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Contents stream</p>
-<h3 class="function"><a id="pdfioFileCreateStringObj">pdfioFileCreateStringObj</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.2&#160;</span><a id="pdfioFileCreateStringObj">pdfioFileCreateStringObj</a></h3>
 <p class="description">Create a new object in a PDF file containing a string.</p>
 <p class="code">
 <a href="#pdfio_obj_t">pdfio_obj_t</a> *pdfioFileCreateStringObj(<a href="#pdfio_file_t">pdfio_file_t</a> *pdf, <span class="reserved">const</span> <span class="reserved">char</span> *string);</p>
@@ -4302,7 +4404,9 @@ stored as indirect object references.</blockquote>
 <p class="description">New object</p>
 <h4 class="discussion">Discussion</h4>
 <p class="discussion">This function creates a new object with a string value in a PDF file.
-You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.</p>
+You must call <a href="#pdfioObjClose"><code>pdfioObjClose</code></a> to write the object to the file.
+
+</p>
 <h3 class="function"><a id="pdfioFileCreateTemporary">pdfioFileCreateTemporary</a></h3>
 <p class="description"></p>
 <p class="code">
@@ -4585,8 +4689,15 @@ cancel the open.  If <code>NULL</code> is specified for the callback function an
 PDF file requires a password, the open will always fail.<br>
 <br>
 The &quot;error_cb&quot; and &quot;error_cbdata&quot; arguments specify an error handler callback
-and its data pointer - if <code>NULL</code> the default error handler is used that
-writes error messages to <code>stderr</code>.</p>
+and its data pointer - if <code>NULL</code> then the default error handler is used that
+writes error messages to <code>stderr</code>.  The error handler callback should return
+<code>true</code> to continue reading the PDF file or <code>false</code> to stop.<br>
+<br>
+</p><blockquote>
+Note: Error messages starting with &quot;WARNING:&quot; are actually warning
+messages - the callback should normally return <code>true</code> to allow PDFio to
+try to resolve the issue.  In addition, some errors are unrecoverable and
+ignore the return value of the error callback.</blockquote>
 <h3 class="function"><a id="pdfioFileSetAuthor">pdfioFileSetAuthor</a></h3>
 <p class="description">Set the author for a PDF file.</p>
 <p class="code">
@@ -4702,7 +4813,7 @@ size_t pdfioImageGetBytesPerLine(<a href="#pdfio_obj_t">pdfio_obj_t</a> *obj);</
 <td class="description">Image object</td></tr>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
-<p class="description">Number of bytes per line</p>
+<p class="description">Number of bytes per line or <code>0</code> on error</p>
 <h3 class="function"><a id="pdfioImageGetHeight">pdfioImageGetHeight</a></h3>
 <p class="description">Get the height of an image object.</p>
 <p class="code">
@@ -4807,7 +4918,7 @@ size_t pdfioObjGetLength(<a href="#pdfio_obj_t">pdfio_obj_t</a> *obj);</p>
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description">Length in bytes or <code>0</code> for none</p>
-<h3 class="function"><a id="pdfioObjGetName">pdfioObjGetName</a></h3>
+<h3 class="function"><span class="info">&#160;PDFio v1.4&#160;</span><a id="pdfioObjGetName">pdfioObjGetName</a></h3>
 <p class="description">Get the name value associated with an object.</p>
 <p class="code">
 <span class="reserved">const</span> <span class="reserved">char</span> *pdfioObjGetName(<a href="#pdfio_obj_t">pdfio_obj_t</a> *obj);</p>
@@ -5081,6 +5192,11 @@ ssize_t pdfioStreamPeek(<a href="#pdfio_stream_t">pdfio_stream_t</a> *st, <span
 </tbody></table>
 <h4 class="returnvalue">Return Value</h4>
 <p class="description"><code>true</code> on success, <code>false</code> on failure</p>
+<h4 class="discussion">Discussion</h4>
+<p class="discussion">This function writes a formatted string to a stream.  In addition to the
+standard <code>printf</code> format characters, you can use &quot;%H&quot; to format a HTML/XML
+string value, &quot;%N&quot; to format a PDF name value (&quot;/Name&quot;), and &quot;%S&quot; to format
+a PDF string (&quot;(String)&quot;) value.</p>
 <h3 class="function"><a id="pdfioStreamPutChar">pdfioStreamPutChar</a></h3>
 <p class="description">Write a single character to a stream.</p>
 <p class="code">
@@ -5314,7 +5430,7 @@ typedef enum <a href="#pdfio_valtype_e">pdfio_valtype_e</a> pdfio_valtype_t;
 <tr><th>PDFIO_ENCRYPTION_AES_128 </th><td class="description">128-bit AES encryption (PDF 1.6)</td></tr>
 <tr><th>PDFIO_ENCRYPTION_NONE </th><td class="description">No encryption</td></tr>
 <tr><th>PDFIO_ENCRYPTION_RC4_128 </th><td class="description">128-bit RC4 encryption (PDF 1.4)</td></tr>
-<tr><th>PDFIO_ENCRYPTION_RC4_40 </th><td class="description">40-bit RC4 encryption (PDF 1.3)</td></tr>
+<tr><th>PDFIO_ENCRYPTION_RC4_40 </th><td class="description">40-bit RC4 encryption (PDF 1.3, reading only)</td></tr>
 </tbody></table>
 <h3 class="enumeration"><a id="pdfio_filter_e">pdfio_filter_e</a></h3>
 <p class="description">Compression/decompression filters for streams</p>

doc/pdfio.md

@@ -343,8 +343,8 @@ password_cb(void *data, const char *filename)
 ```
 
 The error callback is called for both errors and warnings and accepts the
-`pdfio_file_t` pointer, a message string, and the callback pointer value, for
-example:
+`pdfio_file_t` pointer, a message string, and the callback pointer value. It
+returns `true` to continue processing the file or `false` to stop, for example:
 
 ```c
 bool
@@ -354,13 +354,17 @@ error_cb(pdfio_file_t *pdf, const char *message, void *data)
 
   fprintf(stderr, "%s: %s\n", pdfioFileGetName(pdf), message);
 
-  // Return false to treat warnings as errors
-  return (false);
+  // Return true for warning messages (continue) and false for errors (stop)
+  return (!strncmp(message, "WARNING:", 8));
 }
 ```
 
 The default error callback (`NULL`) does the equivalent of the above.
 
+> Note: Many errors are unrecoverable, so PDFio ignores the return value from
+> the error callback and always stops processing the PDF file.  Warning messages
+> start with the prefix "WARNING:" while errors have no prefix.
+
 Each PDF file contains one or more pages.  The [`pdfioFileGetNumPages`](@@)
 function returns the number of pages in the file while the
 [`pdfioFileGetPage`](@@) function gets the specified page in the PDF file:
@@ -889,11 +893,26 @@ main(int  argc,                         // I - Number of command-line arguments
 {
   const char    *filename;              // PDF filename
   pdfio_file_t  *pdf;                   // PDF file
-  const char    *author;                // Author name
-  time_t        creation_date;          // Creation date
-  struct tm     *creation_tm;           // Creation date/time information
-  char          creation_text[256];     // Creation date/time as a string
-  const char    *title;                 // Title
+  pdfio_dict_t  *catalog;               // Catalog dictionary
+  const char    *author,                // Author name
+                *creator,               // Creator name
+                *producer,              // Producer name
+                *title;                 // Title
+  time_t        creation_date,          // Creation date
+                modification_date;      // Modification date
+  struct tm     *creation_tm,           // Creation date/time information
+                *modification_tm;       // Modification date/time information
+  char          creation_text[256],     // Creation date/time as a string
+                modification_text[256], // Modification date/time human fmt string
+                range_text[255];        // Page range text
+  size_t        num_pages;              // PDF number of pages
+  bool          has_acroform;           // Does the file have an AcroForm?
+  pdfio_obj_t   *page;                  // Object
+  pdfio_dict_t  *page_dict;             // Object dictionary
+  size_t        cur,                    // Current page index
+                prev;                   // Previous page index
+  pdfio_rect_t  cur_box,                // Current MediaBox
+                prev_box;               // Previous MediaBox
 
 
   // Get the filename from the command-line...
@@ -906,14 +925,20 @@ main(int  argc,                         // I - Number of command-line arguments
   filename = argv[1];
 
   // Open the PDF file with the default callbacks...
-  pdf = pdfioFileOpen(filename, /*password_cb*/NULL, /*password_cbdata*/NULL,
-                      /*error_cb*/NULL, /*error_cbdata*/NULL);
+  pdf = pdfioFileOpen(filename, /*password_cb*/NULL,
+                      /*password_cbdata*/NULL, /*error_cb*/NULL,
+                      /*error_cbdata*/NULL);
   if (pdf == NULL)
     return (1);
 
-  // Get the title and author...
-  author = pdfioFileGetAuthor(pdf);
-  title  = pdfioFileGetTitle(pdf);
+  // Get the title, author, etc...
+  catalog      = pdfioFileGetCatalog(pdf);
+  author       = pdfioFileGetAuthor(pdf);
+  creator      = pdfioFileGetCreator(pdf);
+  has_acroform = pdfioDictGetType(catalog, "AcroForm") != PDFIO_VALTYPE_NONE;
+  num_pages    = pdfioFileGetNumPages(pdf);
+  producer     = pdfioFileGetProducer(pdf);
+  title        = pdfioFileGetTitle(pdf);
 
   // Get the creation date and convert to a string...
   if ((creation_date = pdfioFileGetCreationDate(pdf)) > 0)
@@ -926,12 +951,76 @@ main(int  argc,                         // I - Number of command-line arguments
     snprintf(creation_text, sizeof(creation_text), "-- not set --");
   }
 
+  // Get the modification date and convert to a string...
+  if ((modification_date = pdfioFileGetModificationDate(pdf)) > 0)
+  {
+    modification_tm = localtime(&modification_date);
+    strftime(modification_text, sizeof(modification_text), "%c", modification_tm);
+  }
+  else
+  {
+    snprintf(modification_text, sizeof(modification_text), "-- not set --");
+  }
+
   // Print file information to stdout...
   printf("%s:\n", filename);
-  printf("         Title: %s\n", title ? title : "-- not set --");
-  printf("        Author: %s\n", author ? author : "-- not set --");
-  printf("    Created On: %s\n", creation_text);
-  printf("  Number Pages: %u\n", (unsigned)pdfioFileGetNumPages(pdf));
+  printf("           Title: %s\n", title ? title : "-- not set --");
+  printf("          Author: %s\n", author ? author : "-- not set --");
+  printf("         Creator: %s\n", creator ? creator : "-- not set --");
+  printf("        Producer: %s\n", producer ? producer : "-- not set --");
+  printf("      Created On: %s\n", creation_text);
+  printf("     Modified On: %s\n", modification_text);
+  printf("         Version: %s\n", pdfioFileGetVersion(pdf));
+  printf("        AcroForm: %s\n", has_acroform ? "Yes" : "No");
+  printf(" Number of Pages: %u\n", (unsigned)num_pages);
+
+  // Report the MediaBox for all of the pages
+  prev_box.x1 = prev_box.x2 = prev_box.y1 = prev_box.y2 = 0.0;
+
+  for (cur = 0, prev = 0; cur < num_pages; cur ++)
+  {
+    // Find the MediaBox for this page in the page tree...
+    for (page = pdfioFileGetPage(pdf, cur);
+         page != NULL;
+         page = pdfioDictGetObj(page_dict, "Parent"))
+    {
+      cur_box.x1 = cur_box.x2 = cur_box.y1 = cur_box.y2 = 0.0;
+      page_dict  = pdfioObjGetDict(page);
+
+      if (pdfioDictGetRect(page_dict, "MediaBox", &cur_box))
+        break;
+    }
+
+    // If this MediaBox is different from the previous one, show the range of
+    // pages that have that size...
+    if (cur == 0 ||
+        fabs(cur_box.x1 - prev_box.x1) > 0.01 ||
+        fabs(cur_box.y1 - prev_box.y1) > 0.01 ||
+        fabs(cur_box.x2 - prev_box.x2) > 0.01 ||
+        fabs(cur_box.y2 - prev_box.y2) > 0.01)
+    {
+      if (cur > prev)
+      {
+        snprintf(range_text, sizeof(range_text), "Pages %u-%u",
+                 (unsigned)(prev + 1), (unsigned)cur);
+        printf("%16s: [%g %g %g %g]\n", range_text,
+               prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+      }
+
+      // Start a new series of pages with the new size...
+      prev     = cur;
+      prev_box = cur_box;
+    }
+  }
+
+  // Show the last range as needed...
+  if (cur > prev)
+  {
+    snprintf(range_text, sizeof(range_text), "Pages %u-%u",
+             (unsigned)(prev + 1), (unsigned)cur);
+    printf("%16s: [%g %g %g %g]\n", range_text,
+           prev_box.x1, prev_box.y1, prev_box.x2, prev_box.y2);
+  }
 
   // Close the PDF file...
   pdfioFileClose(pdf);

examples/pdfioinfo.c

@@ -68,7 +68,7 @@ main(int  argc,				// I - Number of command-line arguments
   catalog      = pdfioFileGetCatalog(pdf);
   author       = pdfioFileGetAuthor(pdf);
   creator      = pdfioFileGetCreator(pdf);
-  has_acroform = pdfioDictGetObj(catalog, "AcroForm") != NULL ? true : false;
+  has_acroform = pdfioDictGetType(catalog, "AcroForm") != PDFIO_VALTYPE_NONE;
   num_pages    = pdfioFileGetNumPages(pdf);
   producer     = pdfioFileGetProducer(pdf);
   title        = pdfioFileGetTitle(pdf);

examples/pdfiomerge.c

@@ -8,8 +8,8 @@
 //
 // Usage:
 //
-//   ./pdfmerge [-o OUTPUT.pdf] INPUT.pdf [... INPUT.pdf]
-//   ./pdfmerge INPUT.pdf [... INPUT.pdf] >OUTPUT.pdf
+//   ./pdfiomerge [-o OUTPUT.pdf] INPUT.pdf [... INPUT.pdf]
+//   ./pdfiomerge INPUT.pdf [... INPUT.pdf] >OUTPUT.pdf
 //
 
 #include <pdfio.h>
@@ -47,7 +47,7 @@ main(int  argc,				// I - Number of command-line arguments
     }
     else if (!strncmp(argv[i], "--", 2))
     {
-      fprintf(stderr, "pdfmerge: Unknown option '%s'.\n", argv[i]);
+      fprintf(stderr, "pdfiomerge: Unknown option '%s'.\n", argv[i]);
       return (usage(stderr));
     }
     else if (argv[i][0] == '-')
@@ -59,14 +59,14 @@ main(int  argc,				// I - Number of command-line arguments
           case 'o' : // -o OUTPUT.pdf
               if (outpdf)
               {
-                fputs("pdfmerge: Only one output file can be specified.\n", stderr);
+                fputs("pdfiomerge: Only one output file can be specified.\n", stderr);
                 return (usage(stderr));
               }
 
               i ++;
               if (i >= argc)
               {
-                fputs("pdfmerge: Missing output filename after '-o'.\n", stderr);
+                fputs("pdfiomerge: Missing output filename after '-o'.\n", stderr);
                 return (usage(stderr));
               }
 
@@ -75,7 +75,7 @@ main(int  argc,				// I - Number of command-line arguments
               break;
 
           default :
-              fprintf(stderr, "pdfmerge: Unknown option '-%c'.\n", *opt);
+              fprintf(stderr, "pdfiomerge: Unknown option '-%c'.\n", *opt);
               return (usage(stderr));
         }
       }

makesrcdist

@@ -32,11 +32,11 @@ if test $(grep AC_INIT configure.ac | awk '{print $2}') != "[$version],"; then
     status=1
 fi
 
-if test $(head -4 CHANGES.md | tail -1 | awk '{print $1}') != "v$version"; then
+if test $(head -5 CHANGES.md | tail -1 | awk '{print $1}') != "v$version"; then
     echo "Still need to update CHANGES.md version number."
     status=1
 fi
-if test $(head -4 CHANGES.md | tail -1 | awk '{print $3}') = "YYYY-MM-DD"; then
+if test $(head -5 CHANGES.md | tail -1 | awk '{print $3}') = "YYYY-MM-DD"; then
     echo "Still need to update CHANGES.md release date."
     status=1
 fi

pdfio-array.c

@@ -640,6 +640,8 @@ _pdfioArrayRead(pdfio_file_t   *pdf,	// I - PDF file
 //
 // 'pdfioArrayRemove()' - Remove an array entry.
 //
+// @since PDFio v1.4@
+//
 
 bool					// O - `true` on success, `false` otherwise
 pdfioArrayRemove(pdfio_array_t *a,	// I - Array

pdfio-common.c

@@ -47,7 +47,7 @@ _pdfioFileConsume(pdfio_file_t *pdf,	// I - PDF file
 // `false` to halt.
 //
 
-bool					// O - `false` to stop
+bool					// O - `false` to stop, `true` to continue
 _pdfioFileDefaultError(
     pdfio_file_t *pdf,			// I - PDF file
     const char   *message,		// I - Error message
@@ -57,7 +57,7 @@ _pdfioFileDefaultError(
 
   fprintf(stderr, "%s: %s\n", pdf->filename, message);
 
-  return (false);
+  return (!strncmp(message, "WARNING:", 8));
 }
 
 
@@ -134,19 +134,20 @@ _pdfioFileGetChar(pdfio_file_t *pdf)	// I - PDF file
 bool					// O - `true` on success, `false` on error
 _pdfioFileGets(pdfio_file_t *pdf,	// I - PDF file
                char         *buffer,	// I - Line buffer
-	       size_t       bufsize)	// I - Size of line buffer
+	       size_t       bufsize,	// I - Size of line buffer
+	       bool         discard)	// I - OK to discard excess line chars?
 {
   bool	eol = false;			// End of line?
   char	*bufptr = buffer,		// Pointer into buffer
 	*bufend = buffer + bufsize - 1;	// Pointer to end of buffer
 
 
-  PDFIO_DEBUG("_pdfioFileGets(pdf=%p, buffer=%p, bufsize=%lu) bufpos=%ld, buffer=%p, bufptr=%p, bufend=%p, offset=%lu\n", pdf, buffer, (unsigned long)bufsize, (long)pdf->bufpos, pdf->buffer, pdf->bufptr, pdf->bufend, (unsigned long)(pdf->bufpos + (pdf->bufptr - pdf->buffer)));
+  PDFIO_DEBUG("_pdfioFileGets(pdf=%p, buffer=%p, bufsize=%lu, discard=%s) bufpos=%ld, buffer=%p, bufptr=%p, bufend=%p, offset=%lu\n", pdf, buffer, (unsigned long)bufsize, discard ? "true" : "false", (long)pdf->bufpos, pdf->buffer, pdf->bufptr, pdf->bufend, (unsigned long)(pdf->bufpos + (pdf->bufptr - pdf->buffer)));
 
   while (!eol)
   {
     // If there are characters ready in the buffer, use them...
-    while (!eol && pdf->bufptr < pdf->bufend && bufptr < bufend)
+    while (!eol && pdf->bufptr < pdf->bufend)
     {
       char ch = *(pdf->bufptr++);	// Next character in buffer
 
@@ -168,8 +169,10 @@ _pdfioFileGets(pdfio_file_t *pdf,	// I - PDF file
 	    pdf->bufptr ++;
 	}
       }
-      else
+      else if (bufptr < bufend)
         *bufptr++ = ch;
+      else if (!discard)
+        break;
     }
 
     // Fill the read buffer as needed...
@@ -424,7 +427,7 @@ off_t					// O - Offset from beginning of file
 _pdfioFileTell(pdfio_file_t *pdf)	// I - PDF file
 {
   if (pdf->bufptr)
-    return (pdf->bufpos + (pdf->bufptr - pdf->buffer));
+    return (pdf->bufpos + (off_t)(pdf->bufptr - pdf->buffer));
   else
     return (pdf->bufpos);
 }
@@ -452,7 +455,7 @@ _pdfioFileWrite(pdfio_file_t *pdf,	// I - PDF file
       if (!write_buffer(pdf, buffer, bytes))
         return (false);
 
-      pdf->bufpos += bytes;
+      pdf->bufpos += (off_t)bytes;
 
       return (true);
     }
@@ -478,7 +481,7 @@ fill_buffer(pdfio_file_t *pdf)		// I - PDF file
 
   // Advance current position in file as needed...
   if (pdf->bufend)
-    pdf->bufpos += pdf->bufend - pdf->buffer;
+    pdf->bufpos += (off_t)(pdf->bufend - pdf->buffer);
 
   // Try reading from the file...
   if ((bytes = read_buffer(pdf, pdf->buffer, sizeof(pdf->buffer))) <= 0)

pdfio-content.c

@@ -632,6 +632,8 @@ pdfioContentPathCurve23(
 //
 // 'pdfioContentPathEnd()' - Clear the current path.
 //
+// @since PDFio v1.1@
+//
 
 bool					// O - `true` on success, `false` on failure
 pdfioContentPathEnd(pdfio_stream_t *st)	// I - Stream
@@ -1111,6 +1113,8 @@ pdfioContentTextEnd(pdfio_stream_t *st)	// I - Stream
 // on "size". The text string must always use the UTF-8 (Unicode) encoding but
 // any control characters (such as newlines) are ignored.
 //
+// @since PDFio v1.2@
+//
 
 double					// O - Width
 pdfioContentTextMeasure(
@@ -1267,6 +1271,8 @@ pdfioContentTextMoveTo(
 //
 // 'pdfioContentTextNewLine()' - Move to the next line.
 //
+// @since PDFio v1.2@
+//
 
 bool					// O - `true` on success, `false` on failure
 pdfioContentTextNewLine(
@@ -1298,6 +1304,8 @@ pdfioContentTextNextLine(
 // specifies that the current font maps to full Unicode.  The "s" argument
 // specifies a UTF-8 encoded string.
 //
+// @since PDFio v1.2@
+//
 
 bool					// O - `true` on success, `false` on failure
 pdfioContentTextNewLineShow(
@@ -1346,6 +1354,8 @@ pdfioContentTextNewLineShow(
 // argument specifies that the current font maps to full Unicode.  The "format"
 // argument specifies a UTF-8 encoded `printf`-style format string.
 //
+// @since PDFio v1.2@
+//
 
 bool					// O - `true` on success, `false` on failure
 pdfioContentTextNewLineShowf(
@@ -1833,8 +1843,6 @@ pdfioFileCreateFontObjFromFile(
         {
           if ((w1 = ttfGetWidth(font, (int)i)) == w0 && i < 65530)
           {
-            size_t j;			// Look-ahead
-
             for (j = 1; j < 4; j ++)
             {
               if (ttfGetWidth(font, (int)(i + j)) != w0)
@@ -1937,7 +1945,8 @@ pdfioFileCreateFontObjFromFile(
   if (fd >= 0)
     close(fd);
 
-  _pdfioObjSetExtension(obj, font, (_pdfio_extfree_t)ttfDelete);
+  if (obj)
+    _pdfioObjSetExtension(obj, font, (_pdfio_extfree_t)ttfDelete);
 
   return (obj);
 }
@@ -2235,7 +2244,7 @@ pdfioFileCreateImageObjFromFile(
 // 'pdfioImageGetBytesPerLine()' - Get the number of bytes to read for each line.
 //
 
-size_t					// O - Number of bytes per line
+size_t					// O - Number of bytes per line or `0` on error
 pdfioImageGetBytesPerLine(
     pdfio_obj_t *obj)			// I - Image object
 {
@@ -2281,7 +2290,25 @@ pdfioImageGetBytesPerLine(
       colors = 1;
   }
 
-  return ((size_t)((width * colors * bpc + 7) / 8));
+  if (width < 0)
+  {
+    _pdfioFileError(obj->pdf, "Invalid image width %d.", width);
+    return (0);
+  }
+  else if (bpc != 1 && bpc != 2 && bpc != 4 && bpc != 8 && bpc != 16)
+  {
+    _pdfioFileError(obj->pdf, "Invalid image bits per component %d.", bpc);
+    return (0);
+  }
+  else if (colors < 1 || colors > 4)
+  {
+    _pdfioFileError(obj->pdf, "Invalid image number of colors %d.", colors);
+    return (0);
+  }
+  else
+  {
+    return ((size_t)((width * colors * bpc + 7) / 8));
+  }
 }
 
 

pdfio-crypto.c

@@ -98,7 +98,7 @@ static uint8_t pdf_passpad[32] =	// Padding for passwords
 
 static void	decrypt_user_key(pdfio_encryption_t encryption, const uint8_t *file_key, uint8_t user_key[32]);
 static void	encrypt_user_key(pdfio_encryption_t encryption, const uint8_t *file_key, uint8_t user_key[32]);
-static void	make_file_key(pdfio_encryption_t encryption, pdfio_permission_t permissions, const unsigned char *file_id, size_t file_idlen, const uint8_t *user_pad, const uint8_t *owner_key, uint8_t file_key[16]);
+static void	make_file_key(pdfio_encryption_t encryption, pdfio_permission_t permissions, const unsigned char *file_id, size_t file_idlen, const uint8_t *user_pad, const uint8_t *owner_key, bool encrypt_metadata, uint8_t file_key[16]);
 static void	make_owner_key(pdfio_encryption_t encryption, const uint8_t *owner_pad, const uint8_t *user_pad, uint8_t owner_key[32]);
 static void	make_user_key(const unsigned char *file_id, size_t file_idlen, uint8_t user_key[32]);
 static void	pad_password(const char *password, uint8_t pad[32]);
@@ -158,7 +158,7 @@ _pdfioCryptoLock(
         // Generate the encryption key
         file_id = pdfioArrayGetBinary(pdf->id_array, 0, &file_idlen);
 
-        make_file_key(encryption, permissions, file_id, file_idlen, user_pad, pdf->owner_key, pdf->file_key);
+        make_file_key(encryption, permissions, file_id, file_idlen, user_pad, pdf->owner_key, pdf->encrypt_metadata, pdf->file_key);
 	pdf->file_keylen = 16;
 
 	// Generate the user key...
@@ -409,13 +409,6 @@ _pdfioCryptoMakeReader(
   uint8_t	data[21];		// Key data
   _pdfio_md5_t	md5;			// MD5 state
   uint8_t	digest[16];		// MD5 digest value
-#if PDFIO_OBJ_CRYPT
-  pdfio_array_t	*id_array;		// Object ID array
-  unsigned char	*id_value;		// Object ID value
-  size_t	id_len;			// Length of object ID
-  uint8_t	temp_key[16];		// File key for object
-#endif // PDFIO_OBJ_CRYPT
-  uint8_t	*file_key;		// Computed file key to use
 
 
   PDFIO_DEBUG("_pdfioCryptoMakeReader(pdf=%p, obj=%p(%d), ctx=%p, iv=%p, ivlen=%p(%d))\n", pdf, obj, (int)obj->number, ctx, iv, ivlen, (int)*ivlen);
@@ -427,60 +420,6 @@ _pdfioCryptoMakeReader(
     return (NULL);
   }
 
-#if PDFIO_OBJ_CRYPT
-  if ((id_array = pdfioDictGetArray(pdfioObjGetDict(obj), "ID")) != NULL)
-  {
-    // Object has its own ID that will get used for encryption...
-    _pdfio_md5_t md5;			// MD5 context
-    uint8_t	file_digest[16];	// MD5 digest of file ID and pad
-    uint8_t	user_pad[32],		// Padded user password
-		own_user_key[32],	// Calculated user key
-		pdf_user_key[32];	// Decrypted user key
-
-    PDFIO_DEBUG("_pdfioCryptoMakeReader: Per-object file ID.\n");
-
-    if ((id_value = pdfioArrayGetBinary(id_array, 0, &id_len)) == NULL)
-    {
-      *ivlen = 0;
-      return (NULL);
-    }
-
-    _pdfioCryptoMD5Init(&md5);
-    _pdfioCryptoMD5Append(&md5, pdf_passpad, 32);
-    _pdfioCryptoMD5Append(&md5, id_value, id_len);
-    _pdfioCryptoMD5Finish(&md5, file_digest);
-
-    make_owner_key(pdf->encryption, pdf->password, pdf->owner_key, user_pad);
-    make_file_key(pdf->encryption, pdf->permissions, id_value, id_len, user_pad, pdf->owner_key, temp_key);
-    make_user_key(id_value, id_len, own_user_key);
-
-    if (memcmp(own_user_key, pdf->user_key, sizeof(own_user_key)))
-    {
-      PDFIO_DEBUG("_pdfioCryptoMakeReader: Not user password, trying owner password.\n");
-
-      make_file_key(pdf->encryption, pdf->permissions, id_value, id_len, pdf->password, pdf->owner_key, temp_key);
-      make_user_key(id_value, id_len, own_user_key);
-
-      memcpy(pdf_user_key, pdf->user_key, sizeof(pdf_user_key));
-      decrypt_user_key(pdf->encryption, temp_key, pdf_user_key);
-
-      if (memcmp(pdf->password, pdf_user_key, 32) && memcmp(own_user_key, pdf_user_key, 16))
-      {
-        _pdfioFileError(pdf, "Unable to unlock file.");
-	*ivlen = 0;
-	return (NULL);
-      }
-    }
-
-    file_key = temp_key;
-  }
-  else
-#endif // PDFIO_OBJ_CRYPT
-  {
-    // Use the default file key...
-    file_key = pdf->file_key;
-  }
-
   switch (pdf->encryption)
   {
     default :
@@ -490,20 +429,20 @@ _pdfioCryptoMakeReader(
 
     case PDFIO_ENCRYPTION_RC4_40 :
 	// Copy the key data for the MD5 hash.
-	memcpy(data, file_key, 16);
-	data[16] = (uint8_t)obj->number;
-	data[17] = (uint8_t)(obj->number >> 8);
-	data[18] = (uint8_t)(obj->number >> 16);
-	data[19] = (uint8_t)obj->generation;
-	data[20] = (uint8_t)(obj->generation >> 8);
+	memcpy(data, pdf->file_key, 5);
+	data[5] = (uint8_t)obj->number;
+	data[6] = (uint8_t)(obj->number >> 8);
+	data[7] = (uint8_t)(obj->number >> 16);
+	data[8] = (uint8_t)obj->generation;
+	data[9] = (uint8_t)(obj->generation >> 8);
 
         // Hash it...
         _pdfioCryptoMD5Init(&md5);
-	_pdfioCryptoMD5Append(&md5, data, sizeof(data));
+	_pdfioCryptoMD5Append(&md5, data, 10);
 	_pdfioCryptoMD5Finish(&md5, digest);
 
-        // Initialize the RC4 context using 40 bits of the digest...
-	_pdfioCryptoRC4Init(&ctx->rc4, digest, 5);
+        // Initialize the RC4 context using 80 bits of the digest...
+	_pdfioCryptoRC4Init(&ctx->rc4, digest, 10);
 	*ivlen = 0;
 	return ((_pdfio_crypto_cb_t)_pdfioCryptoRC4Crypt);
 
@@ -517,7 +456,7 @@ _pdfioCryptoMakeReader(
 
     case PDFIO_ENCRYPTION_RC4_128 :
 	// Copy the key data for the MD5 hash.
-	memcpy(data, file_key, 16);
+	memcpy(data, pdf->file_key, 16);
 	data[16] = (uint8_t)obj->number;
 	data[17] = (uint8_t)(obj->number >> 8);
 	data[18] = (uint8_t)(obj->number >> 16);
@@ -643,6 +582,8 @@ _pdfioCryptoUnlock(
 		file_idlen;		// Length of file ID
   _pdfio_md5_t	md5;			// MD5 context
   uint8_t	file_digest[16];	// MD5 digest of file ID and pad
+  double	p;			// Permissions value as a double
+  _pdfio_value_t *value;		// Encrypt dictionary value, if any
 
 
   // See if we support the type of encryption specified by the Encrypt object
@@ -658,7 +599,12 @@ _pdfioCryptoUnlock(
   revision = (int)pdfioDictGetNumber(encrypt_dict, "R");
   length   = (int)pdfioDictGetNumber(encrypt_dict, "Length");
 
-  PDFIO_DEBUG("_pdfioCryptoUnlock: handler=%p(%s), version=%d, revision=%d, length=%d\n", (void *)handler, handler ? handler : "(null)", version, revision, length);
+  if ((value = _pdfioDictGetValue(encrypt_dict, "EncryptMetadata")) != NULL && value->type == PDFIO_VALTYPE_BOOLEAN)
+    pdf->encrypt_metadata = value->value.boolean;
+  else
+    pdf->encrypt_metadata = true;
+
+  PDFIO_DEBUG("_pdfioCryptoUnlock: handler=%p(%s), version=%d, revision=%d, length=%d, encrypt_metadata=%s\n", (void *)handler, handler ? handler : "(null)", version, revision, length, pdf->encrypt_metadata ? "true" : "false");
 
   if (!handler || strcmp(handler, "Standard"))
   {
@@ -750,8 +696,13 @@ _pdfioCryptoUnlock(
 
   // Grab the remaining values we need to unlock the PDF...
   pdf->file_keylen = (size_t)(length / 8);
-  pdf->permissions = (pdfio_permission_t)pdfioDictGetNumber(encrypt_dict, "P");
 
+  p = pdfioDictGetNumber(encrypt_dict, "P");
+  PDFIO_DEBUG("_pdfioCryptoUnlock: P=%.0f\n", p);
+  if (p < 0x7fffffff)			// Handle integers > 2^31-1
+    pdf->permissions = (pdfio_permission_t)p;
+  else
+    pdf->permissions = (pdfio_permission_t)(p - 4294967296.0);
   PDFIO_DEBUG("_pdfioCryptoUnlock: permissions=%d\n", pdf->permissions);
 
   owner_key = pdfioDictGetBinary(encrypt_dict, "O", &owner_keylen);
@@ -823,7 +774,7 @@ _pdfioCryptoUnlock(
       make_owner_key(pdf->encryption, pad, pdf->owner_key, user_pad);
       PDFIO_DEBUG("_pdfioCryptoUnlock: Upad=%02X%02X%02X%02X...%02X%02X%02X%02X\n", user_pad[0], user_pad[1], user_pad[2], user_pad[3], user_pad[28], user_pad[29], user_pad[30], user_pad[31]);
 
-      make_file_key(pdf->encryption, pdf->permissions, file_id, file_idlen, user_pad, pdf->owner_key, file_key);
+      make_file_key(pdf->encryption, pdf->permissions, file_id, file_idlen, user_pad, pdf->owner_key, pdf->encrypt_metadata, file_key);
       PDFIO_DEBUG("_pdfioCryptoUnlock: Fown=%02X%02X%02X%02X...%02X%02X%02X%02X\n", file_key[0], file_key[1], file_key[2], file_key[3], file_key[12], file_key[13], file_key[14], file_key[15]);
 
       make_user_key(file_id, file_idlen, own_user_key);
@@ -841,7 +792,7 @@ _pdfioCryptoUnlock(
       }
 
       // Not the owner password, try the user password...
-      make_file_key(pdf->encryption, pdf->permissions, file_id, file_idlen, pad, pdf->owner_key, file_key);
+      make_file_key(pdf->encryption, pdf->permissions, file_id, file_idlen, pad, pdf->owner_key, pdf->encrypt_metadata, file_key);
       PDFIO_DEBUG("_pdfioCryptoUnlock: Fuse=%02X%02X%02X%02X...%02X%02X%02X%02X\n", file_key[0], file_key[1], file_key[2], file_key[3], file_key[12], file_key[13], file_key[14], file_key[15]);
 
       make_user_key(file_id, file_idlen, own_user_key);
@@ -973,6 +924,8 @@ make_file_key(
     size_t              file_idlen,	// I - Length of file ID
     const uint8_t       *user_pad,	// I - Padded user password
     const uint8_t       *owner_key,	// I - Owner key
+    bool                encrypt_metadata,
+					// I - Encrypt metadata?
     uint8_t             file_key[16])	// O - Encryption key
 {
   size_t	i;			// Looping var
@@ -986,13 +939,25 @@ make_file_key(
   perm_bytes[2] = (uint8_t)(permissions >> 16);
   perm_bytes[3] = (uint8_t)(permissions >> 24);
 
+  PDFIO_DEBUG("make_file_key: user_pad[32]=<%02X%02X%02X%02X...%02X%02X%02X%02X>\n", user_pad[0], user_pad[1], user_pad[2], user_pad[3], user_pad[28], user_pad[29], user_pad[30], user_pad[31]);
+  PDFIO_DEBUG("make_file_key: owner_key[32]=<%02X%02X%02X%02X...%02X%02X%02X%02X>\n", owner_key[0], owner_key[1], owner_key[2], owner_key[3], owner_key[28], owner_key[29], owner_key[30], owner_key[31]);
+  PDFIO_DEBUG("make_file_key: permissions(%d)=<%02X%02X%02X%02X>\n", permissions, perm_bytes[0], perm_bytes[1], perm_bytes[2], perm_bytes[3]);
+
   _pdfioCryptoMD5Init(&md5);
   _pdfioCryptoMD5Append(&md5, user_pad, 32);
   _pdfioCryptoMD5Append(&md5, owner_key, 32);
   _pdfioCryptoMD5Append(&md5, perm_bytes, 4);
   _pdfioCryptoMD5Append(&md5, file_id, file_idlen);
+  if (!encrypt_metadata)
+  {
+    uint8_t meta_bytes[4] = { 0xff, 0xff, 0xff, 0xff };
+					// Metadata bytes
+    _pdfioCryptoMD5Append(&md5, meta_bytes, 4);
+  }
   _pdfioCryptoMD5Finish(&md5, digest);
 
+  PDFIO_DEBUG("make_file_key: first md5=<%02X%02X%02X%02X...%02X%02X%02X%02X>\n", digest[0], digest[1], digest[2], digest[3], digest[12], digest[13], digest[14], digest[15]);
+
   if (encryption != PDFIO_ENCRYPTION_RC4_40)
   {
     // MD5 the result 50 times..
@@ -1004,6 +969,8 @@ make_file_key(
     }
   }
 
+  PDFIO_DEBUG("make_file_key: file_key[16]=<%02X%02X%02X%02X...%02X%02X%02X%02X>\n", digest[0], digest[1], digest[2], digest[3], digest[12], digest[13], digest[14], digest[15]);
+
   memcpy(file_key, digest, 16);
 }
 
@@ -1054,9 +1021,11 @@ make_owner_key(
     // Encrypt 20 times...
     uint8_t encrypt_key[16];		// RC4 encryption key
 
-    for (i = 0; i < 20; i ++)
+    for (i = 20; i > 0;)
     {
       // XOR each byte in the digest with the loop counter to make a key...
+      i --;
+
       for (j = 0; j < sizeof(encrypt_key); j ++)
 	encrypt_key[j] = (uint8_t)(digest[j] ^ i);
 

pdfio-dict.c

@@ -20,6 +20,8 @@ static int	compare_pairs(_pdfio_pair_t *a, _pdfio_pair_t *b);
 //
 // 'pdfioDictClear()' - Remove a key/value pair from a dictionary.
 //
+// @since PDFio v1.4@
+//
 
 bool					// O - `true` if cleared, `false` otherwise
 pdfioDictClear(pdfio_dict_t *dict,	// I - Dictionary
@@ -345,6 +347,8 @@ pdfioDictGetDict(pdfio_dict_t *dict,	// I - Dictionary
 //
 // 'pdfioDictGetKey()' - Get the key for the specified pair.
 //
+// @since PDFio v1.4@
+//
 
 const char *				// O - Key for specified pair
 pdfioDictGetKey(pdfio_dict_t *dict,	// I - Dictionary
@@ -375,6 +379,8 @@ pdfioDictGetName(pdfio_dict_t *dict,	// I - Dictionary
 //
 // 'pdfioDictGetNumPairs()' - Get the number of key/value pairs in a dictionary.
 //
+// @since PDFio v1.4@
+//
 
 size_t					// O - Number of pairs
 pdfioDictGetNumPairs(pdfio_dict_t *dict)// I - Dictionary
@@ -465,127 +471,12 @@ pdfioDictGetString(pdfio_dict_t *dict,	// I - Dictionary
   else if (value && value->type == PDFIO_VALTYPE_BINARY && value->value.binary.datalen < 4096)
   {
     // Convert binary string to regular string...
-    char	temp[4096],		// Temporary string
-		*tempptr;		// Pointer into temporary string
-    unsigned char *dataptr;		// Pointer into the data string
+    char	temp[4096];		// Temporary UTF-8 string
 
-    if (!(value->value.binary.datalen & 1) && !memcmp(value->value.binary.data, "\376\377", 2))
+    if (!(value->value.binary.datalen & 1) && (!memcmp(value->value.binary.data, "\376\377", 2) || !memcmp(value->value.binary.data, "\377\376", 2)))
     {
-      // Copy UTF-16 BE
-      int	ch;			// Unicode character
-      size_t	remaining;		// Remaining bytes
-
-      for (dataptr = value->value.binary.data + 2, remaining = value->value.binary.datalen - 2, tempptr = temp; remaining > 1 && tempptr < (temp + sizeof(temp) - 5); dataptr += 2, remaining -= 2)
-      {
-        ch = (dataptr[0] << 8) | dataptr[1];
-
-        if (ch >= 0xd800 && ch <= 0xdbff && remaining > 3)
-	{
-	  // Multi-word UTF-16 char...
-          int lch;			// Lower bits
-
-	  lch = (dataptr[2] << 8) | dataptr[3];
-
-          if (lch < 0xdc00 || lch >= 0xdfff)
-	    break;
-
-          ch        = (((ch & 0x3ff) << 10) | (lch & 0x3ff)) + 0x10000;
-          dataptr   += 2;
-          remaining -= 2;
-	}
-	else if (ch >= 0xfffe)
-	{
-	  continue;
-	}
-
-        if (ch < 128)
-        {
-          // ASCII
-          *tempptr++ = (char)ch;
-        }
-        else if (ch < 4096)
-        {
-          // 2-byte UTF-8
-          *tempptr++ = (char)(0xc0 | (ch >> 6));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-        else if (ch < 65536)
-        {
-          // 3-byte UTF-8
-          *tempptr++ = (char)(0xe0 | (ch >> 12));
-          *tempptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-        else
-        {
-          // 4-byte UTF-8
-          *tempptr++ = (char)(0xe0 | (ch >> 18));
-          *tempptr++ = (char)(0x80 | ((ch >> 12) & 0x3f));
-          *tempptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-      }
-
-      *tempptr = '\0';
-    }
-    else if (!(value->value.binary.datalen & 1) && !memcmp(value->value.binary.data, "\377\376", 2))
-    {
-      // Copy UTF-16 LE
-      int	ch;			// Unicode character
-      size_t	remaining;		// Remaining bytes
-
-      for (dataptr = value->value.binary.data + 2, remaining = value->value.binary.datalen - 2, tempptr = temp; remaining > 1 && tempptr < (temp + sizeof(temp) - 5); dataptr += 2, remaining -= 2)
-      {
-        ch = (dataptr[1] << 8) | dataptr[0];
-
-        if (ch >= 0xd800 && ch <= 0xdbff && remaining > 3)
-	{
-	  // Multi-word UTF-16 char...
-          int lch;			// Lower bits
-
-	  lch = (dataptr[3] << 8) | dataptr[2];
-
-          if (lch < 0xdc00 || lch >= 0xdfff)
-	    break;
-
-          ch        = (((ch & 0x3ff) << 10) | (lch & 0x3ff)) + 0x10000;
-          dataptr   += 2;
-          remaining -= 2;
-	}
-	else if (ch >= 0xfffe)
-	{
-	  continue;
-	}
-
-        if (ch < 128)
-        {
-          // ASCII
-          *tempptr++ = (char)ch;
-        }
-        else if (ch < 4096)
-        {
-          // 2-byte UTF-8
-          *tempptr++ = (char)(0xc0 | (ch >> 6));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-        else if (ch < 65536)
-        {
-          // 3-byte UTF-8
-          *tempptr++ = (char)(0xe0 | (ch >> 12));
-          *tempptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-        else
-        {
-          // 4-byte UTF-8
-          *tempptr++ = (char)(0xe0 | (ch >> 18));
-          *tempptr++ = (char)(0x80 | ((ch >> 12) & 0x3f));
-          *tempptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
-          *tempptr++ = (char)(0x80 | (ch & 0x3f));
-        }
-      }
-
-      *tempptr = '\0';
+      // Copy UTF-16...
+      _pdfio_utf16cpy(temp, value->value.binary.data, value->value.binary.datalen, sizeof(temp));
     }
     else
     {
@@ -677,6 +568,8 @@ _pdfioDictGetValue(pdfio_dict_t *dict,	// I - Dictionary
 // The iteration continues as long as the callback returns `true` or all keys
 // have been iterated.
 //
+// @since PDFio v1.1@
+//
 
 void
 pdfioDictIterateKeys(
@@ -737,11 +630,6 @@ _pdfioDictRead(pdfio_file_t   *pdf,	// I - PDF file
       _pdfioFileError(pdf, "Invalid dictionary contents.");
       break;
     }
-    else if (_pdfioDictGetValue(dict, key + 1))
-    {
-      _pdfioFileError(pdf, "Duplicate dictionary key '%s'.", key + 1);
-      return (NULL);
-    }
 
     // Then get the next value...
     PDFIO_DEBUG("_pdfioDictRead: Reading value for '%s'.\n", key + 1);
@@ -751,8 +639,17 @@ _pdfioDictRead(pdfio_file_t   *pdf,	// I - PDF file
       _pdfioFileError(pdf, "Missing value for dictionary key '%s'.", key + 1);
       break;
     }
-
-    if (!_pdfioDictSetValue(dict, pdfioStringCreate(pdf, key + 1), &value))
+    else if (_pdfioDictGetValue(dict, key + 1))
+    {
+      // Issue 118: Discard duplicate key/value pairs, in the future this will
+      // be a warning message...
+      _pdfioValueDelete(&value);
+      if (_pdfioFileError(pdf, "WARNING: Discarding value for duplicate dictionary key '%s'.", key + 1))
+        continue;
+      else
+        break;
+    }
+    else if (!_pdfioDictSetValue(dict, pdfioStringCreate(pdf, key + 1), &value))
       break;
 
     PDFIO_DEBUG("_pdfioDictRead: Set %s.\n", key);

pdfio-file.c

@@ -61,9 +61,10 @@ _pdfioFileAddMappedObj(
   pdf->num_objmaps ++;
 
   map->obj        = dst_obj;
-  map->src_pdf    = src_obj->pdf;
   map->src_number = src_obj->number;
 
+  memcpy(map->src_id, src_obj->pdf->file_id, sizeof(map->src_id));
+
   // Sort as needed...
   if (pdf->num_objmaps > 1 && compare_objmaps(map, pdf->objmaps + pdf->num_objmaps - 2) < 0)
     qsort(pdf->objmaps, pdf->num_objmaps, sizeof(_pdfio_objmap_t), (int (*)(const void *, const void *))compare_objmaps);
@@ -110,6 +111,8 @@ pdfioFileClose(pdfio_file_t *pdf)	// I - PDF file
 {
   bool		ret = true;		// Return value
   size_t	i;			// Looping var
+  _pdfio_strbuf_t *current,		// Current string buffer
+		*next;			// Next string buffer
 
 
   // Range check input
@@ -152,6 +155,12 @@ pdfioFileClose(pdfio_file_t *pdf)	// I - PDF file
     free(pdf->strings[i]);
   free(pdf->strings);
 
+  for (current = pdf->strbuffers; current; current = next)
+  {
+    next = current->next;
+    free(current);
+  }
+
   free(pdf);
 
   return (ret);
@@ -173,8 +182,9 @@ pdfioFileClose(pdfio_file_t *pdf)	// I - PDF file
 // of 8.27x11in (the intersection of US Letter and ISO A4) is used.
 //
 // The "error_cb" and "error_cbdata" arguments specify an error handler callback
-// and its data pointer - if `NULL` the default error handler is used that
-// writes error messages to `stderr`.
+// and its data pointer - if `NULL` then the default error handler is used that
+// writes error messages to `stderr`.  The error handler callback should return
+// `true` to continue writing the PDF file or `false` to stop.
 //
 
 pdfio_file_t *				// O - PDF file or `NULL` on error
@@ -252,6 +262,8 @@ pdfioFileCreateArrayObj(
 // This function creates a new object with a name value in a PDF file.
 // You must call @link pdfioObjClose@ to write the object to the file.
 //
+// @since PDFio v1.4@
+//
 
 pdfio_obj_t *				// O - New object
 pdfioFileCreateNameObj(
@@ -281,6 +293,8 @@ pdfioFileCreateNameObj(
 // This function creates a new object with a number value in a PDF file.
 // You must call @link pdfioObjClose@ to write the object to the file.
 //
+// @since PDFio v1.2@
+//
 
 pdfio_obj_t *				// O - New object
 pdfioFileCreateNumberObj(
@@ -407,8 +421,9 @@ _pdfioFileCreateObj(
 // of 8.27x11in (the intersection of US Letter and ISO A4) is used.
 //
 // The "error_cb" and "error_cbdata" arguments specify an error handler callback
-// and its data pointer - if `NULL` the default error handler is used that
-// writes error messages to `stderr`.
+// and its data pointer - if `NULL` then the default error handler is used that
+// writes error messages to `stderr`.  The error handler callback should return
+// `true` to continue writing the PDF file or `false` to stop.
 //
 // > *Note*: Files created using this API are slightly larger than those
 // > created using the @link pdfioFileCreate@ function since stream lengths are
@@ -510,6 +525,8 @@ pdfioFileCreatePage(pdfio_file_t *pdf,	// I - PDF file
 // This function creates a new object with a string value in a PDF file.
 // You must call @link pdfioObjClose@ to write the object to the file.
 //
+// @since PDFio v1.2@
+//
 
 pdfio_obj_t *				// O - New object
 pdfioFileCreateStringObj(
@@ -661,7 +678,7 @@ _pdfioFileFindMappedObj(
     return (NULL);
 
   // Otherwise search for a match...
-  key.src_pdf    = src_pdf;
+  memcpy(key.src_id, src_pdf->file_id, sizeof(key.src_id));
   key.src_number = src_number;
 
   if ((match = (_pdfio_objmap_t *)bsearch(&key, pdf->objmaps, pdf->num_objmaps, sizeof(_pdfio_objmap_t), (int (*)(const void *, const void *))compare_objmaps)) != NULL)
@@ -979,8 +996,14 @@ pdfioFileGetVersion(
 // PDF file requires a password, the open will always fail.
 //
 // The "error_cb" and "error_cbdata" arguments specify an error handler callback
-// and its data pointer - if `NULL` the default error handler is used that
-// writes error messages to `stderr`.
+// and its data pointer - if `NULL` then the default error handler is used that
+// writes error messages to `stderr`.  The error handler callback should return
+// `true` to continue reading the PDF file or `false` to stop.
+//
+// > Note: Error messages starting with "WARNING:" are actually warning
+// > messages - the callback should normally return `true` to allow PDFio to
+// > try to resolve the issue.  In addition, some errors are unrecoverable and
+// > ignore the return value of the error callback.
 //
 
 pdfio_file_t *				// O - PDF file
@@ -998,6 +1021,10 @@ pdfioFileOpen(
 		*end;			// End of line
   ssize_t	bytes;			// Bytes read
   off_t		xref_offset;		// Offset to xref table
+  time_t	curtime;		// Creation date/time
+  unsigned char	*id_value;		// ID value
+  size_t	id_valuelen;		// Length of ID value
+  _pdfio_sha256_t ctx;			// Hashing context
 
 
   PDFIO_DEBUG("pdfioFileOpen(filename=\"%s\", password_cb=%p, password_cbdata=%p, error_cb=%p, error_cbdata=%p)\n", filename, (void *)password_cb, (void *)password_cbdata, (void *)error_cb, (void *)error_cbdata);
@@ -1041,7 +1068,7 @@ pdfioFileOpen(
   }
 
   // Read the header from the first line...
-  if (!_pdfioFileGets(pdf, line, sizeof(line)))
+  if (!_pdfioFileGets(pdf, line, sizeof(line), true))
     goto error;
 
   if ((strncmp(line, "%PDF-1.", 7) && strncmp(line, "%PDF-2.", 7)) || !isdigit(line[7] & 255))
@@ -1055,7 +1082,7 @@ pdfioFileOpen(
   pdf->version = strdup(line + 5);
 
   // Grab the last 1k of the file to find the start of the xref table...
-  if (_pdfioFileSeek(pdf, -1024, SEEK_END) < 0)
+  if (_pdfioFileSeek(pdf, 1 - sizeof(line), SEEK_END) < 0)
   {
     _pdfioFileError(pdf, "Unable to read startxref data.");
     goto error;
@@ -1067,28 +1094,48 @@ pdfioFileOpen(
     goto error;
   }
 
+  PDFIO_DEBUG("pdfioOpen: Read %d bytes at end of file.\n", (int)bytes);
+
   line[bytes] = '\0';
   end = line + bytes - 9;
 
   for (ptr = line; ptr < end; ptr ++)
   {
-    if (!memcmp(ptr, "startxref", 9))
+    if (!strncmp(ptr, "startxref", 9) && !strstr(ptr + 9, "startxref") && strtol(ptr + 9, NULL, 10) > 0)
       break;
   }
 
   if (ptr >= end)
   {
-    _pdfioFileError(pdf, "Unable to find start of xref table.");
-    goto error;
-  }
+    if (!_pdfioFileError(pdf, "WARNING: Unable to find start of cross-reference table, will attempt to rebuild."))
+      goto error;
 
-  xref_offset = (off_t)strtol(ptr + 9, NULL, 10);
-
-  if (!load_xref(pdf, xref_offset, password_cb, password_cbdata))
-  {
     if (!repair_xref(pdf, password_cb, password_cbdata))
       goto error;
   }
+  else
+  {
+    PDFIO_DEBUG("pdfioFileOpen: line=%p,ptr=%p(\"%s\")\n", line, ptr, ptr);
+
+    xref_offset = (off_t)strtol(ptr + 9, NULL, 10);
+
+    PDFIO_DEBUG("pdfioFileOpen: xref_offset=%lu\n", (unsigned long)xref_offset);
+
+    if (!load_xref(pdf, xref_offset, password_cb, password_cbdata))
+      goto error;
+  }
+
+  // Create the unique file identifier string for the object map...
+  curtime = pdfioFileGetCreationDate(pdf);
+
+  _pdfioCryptoSHA256Init(&ctx);
+  _pdfioCryptoSHA256Append(&ctx, (uint8_t *)pdf->filename, strlen(pdf->filename));
+  _pdfioCryptoSHA256Append(&ctx, (uint8_t *)&curtime, sizeof(curtime));
+  if ((id_value = pdfioArrayGetBinary(pdf->id_array, 0, &id_valuelen)) != NULL)
+    _pdfioCryptoSHA256Append(&ctx, id_value, id_valuelen);
+  if ((id_value = pdfioArrayGetBinary(pdf->id_array, 1, &id_valuelen)) != NULL)
+    _pdfioCryptoSHA256Append(&ctx, id_value, id_valuelen);
+  _pdfioCryptoSHA256Finish(&ctx, pdf->file_id);
 
   return (pdf);
 
@@ -1202,6 +1249,8 @@ pdfioFileSetPermissions(
   if (encryption == PDFIO_ENCRYPTION_NONE)
     return (true);
 
+  pdf->encrypt_metadata = true;
+
   return (_pdfioCryptoLock(pdf, permissions, encryption, owner_password, user_password));
 }
 
@@ -1342,10 +1391,11 @@ static int				// O - Result of comparison
 compare_objmaps(_pdfio_objmap_t *a,	// I - First object map
                 _pdfio_objmap_t *b)	// I - Second object map
 {
-  if (a->src_pdf < b->src_pdf)
-    return (-1);
-  else if (a->src_pdf > b->src_pdf)
-    return (1);
+  int	ret = memcmp(a->src_id, b->src_id, sizeof(a->src_id));
+					// Result of comparison
+
+  if (ret)
+    return (ret);
   else if (a->src_number < b->src_number)
     return (-1);
   else if (a->src_number > b->src_number)
@@ -1374,6 +1424,8 @@ create_common(
   pdfio_file_t	*pdf;			// New PDF file
   pdfio_dict_t	*dict;			// Dictionary
   unsigned char	id_value[16];		// File ID value
+  time_t	curtime;		// Creation date/time
+  _pdfio_sha256_t ctx;			// Hashing context
 
 
   PDFIO_DEBUG("create_common(filename=\"%s\", fd=%d, output_cb=%p, output_cbdata=%p, version=\"%s\", media_box=%p, crop_box=%p, error_cb=%p, error_cbdata=%p)\n", filename, fd, (void *)output_cb, (void *)output_cbdata, version, (void *)media_box, (void *)crop_box, (void *)error_cb, (void *)error_cbdata);
@@ -1464,7 +1516,9 @@ create_common(
   if ((dict = pdfioDictCreate(pdf)) == NULL)
     goto error;
 
-  pdfioDictSetDate(dict, "CreationDate", time(NULL));
+  curtime = time(NULL);
+
+  pdfioDictSetDate(dict, "CreationDate", curtime);
   pdfioDictSetString(dict, "Producer", "pdfio/" PDFIO_VERSION);
 
   if ((pdf->info_obj = pdfioFileCreateObj(pdf, dict)) == NULL)
@@ -1489,6 +1543,14 @@ create_common(
     pdfioArrayAppendBinary(pdf->id_array, id_value, sizeof(id_value));
   }
 
+  // Create the unique file identifier string for the object map...
+  _pdfioCryptoSHA256Init(&ctx);
+  _pdfioCryptoSHA256Append(&ctx, (uint8_t *)pdf->filename, strlen(pdf->filename));
+  _pdfioCryptoSHA256Append(&ctx, (uint8_t *)&curtime, sizeof(curtime));
+  _pdfioCryptoSHA256Append(&ctx, id_value, sizeof(id_value));
+  _pdfioCryptoSHA256Append(&ctx, id_value, sizeof(id_value));
+  _pdfioCryptoSHA256Finish(&ctx, pdf->file_id);
+
   return (pdf);
 
   // Common error handling code...
@@ -1667,7 +1729,10 @@ load_pages(pdfio_file_t *pdf,		// I - PDF file
   }
 
   if ((type = pdfioDictGetName(dict, "Type")) == NULL || (strcmp(type, "Pages") && strcmp(type, "Page")))
-    return (false);
+  {
+    if (!_pdfioFileError(pdf, "WARNING: No Type value for pages object."))
+      return (false);
+  }
 
   // If there is a Kids array, then this is a parent node and we have to look
   // at the child objects...
@@ -1734,31 +1799,32 @@ load_xref(
   int		generation;		// Generation number
   _pdfio_token_t tb;			// Token buffer/stack
   off_t		line_offset;		// Offset to start of line
+  pdfio_obj_t	*pages_obj;		// Pages object
 
 
   while (!done)
   {
     if (_pdfioFileSeek(pdf, xref_offset, SEEK_SET) != xref_offset)
     {
-      _pdfioFileError(pdf, "Unable to seek to start of xref table.");
-      return (false);
+      PDFIO_DEBUG("load_xref: Unable to seek to %lu.\n", (unsigned long)xref_offset);
+      goto repair;
     }
 
     do
     {
       line_offset = _pdfioFileTell(pdf);
 
-      if (!_pdfioFileGets(pdf, line, sizeof(line)))
+      if (!_pdfioFileGets(pdf, line, sizeof(line), true))
       {
-	_pdfioFileError(pdf, "Unable to read start of xref table.");
-	return (false);
+	PDFIO_DEBUG("load_xref: Unable to read line at offset %lu.\n", (unsigned long)line_offset);
+        goto repair;
       }
     }
     while (!line[0]);
 
     PDFIO_DEBUG("load_xref: line_offset=%lu, line='%s'\n", (unsigned long)line_offset, line);
 
-    if (isdigit(line[0] & 255) && strlen(line) > 4 && (!strcmp(line + strlen(line) - 4, " obj") || ((ptr = strstr(line, " obj")) != NULL && ptr[4] == '<')))
+    if (isdigit(line[0] & 255) && strlen(line) > 4 && (!strcmp(line + strlen(line) - 4, " obj") || ((ptr = strstr(line, " obj")) != NULL && (ptr[4] == '<' || isspace(ptr[4])))))
     {
       // Cross-reference stream
       pdfio_obj_t	*obj;		// Object
@@ -1780,14 +1846,14 @@ load_xref(
 
       if ((number = strtoimax(line, &ptr, 10)) < 1)
       {
-	_pdfioFileError(pdf, "Bad xref table header '%s'.", line);
-	return (false);
+        PDFIO_DEBUG("load_xref: Unable to scan object number.\n");
+        goto repair;
       }
 
       if ((generation = (int)strtol(ptr, &ptr, 10)) < 0 || (generation > 65535 && number != 0))
       {
-	_pdfioFileError(pdf, "Bad xref table header '%s'.", line);
-	return (false);
+        PDFIO_DEBUG("load_xref: Unable to scan generation number (%u).\n", (unsigned)generation);
+        goto repair;
       }
 
       while (isspace(*ptr & 255))
@@ -1795,14 +1861,14 @@ load_xref(
 
       if (strncmp(ptr, "obj", 3))
       {
-	_pdfioFileError(pdf, "Bad xref table header '%s'.", line);
-	return (false);
+        PDFIO_DEBUG("load_xref: No 'obj' after object number and generation (saw '%s').\n", ptr);
+        goto repair;
       }
 
       if (_pdfioFileSeek(pdf, line_offset + (off_t)(ptr + 3 - line), SEEK_SET) < 0)
       {
-        _pdfioFileError(pdf, "Unable to seek to xref object %lu %u.", (unsigned long)number, (unsigned)generation);
-        return (false);
+        PDFIO_DEBUG("load_xref: Unable to seek to start of cross-reference object dictionary.\n");
+        goto repair;
       }
 
       PDFIO_DEBUG("load_xref: Loading object %lu %u.\n", (unsigned long)number, (unsigned)generation);
@@ -1817,21 +1883,21 @@ load_xref(
 
       if (!_pdfioValueRead(pdf, obj, &tb, &trailer, 0))
       {
-        _pdfioFileError(pdf, "Unable to read cross-reference stream dictionary.");
-        return (false);
+        PDFIO_DEBUG("load_xref: Unable to read cross-reference object dictionary.\n");
+        goto repair;
       }
       else if (trailer.type != PDFIO_VALTYPE_DICT)
       {
-	_pdfioFileError(pdf, "Cross-reference stream does not have a dictionary.");
-	return (false);
+        PDFIO_DEBUG("load_xref: Expected dictionary for cross-reference object (type=%d).", trailer.type);
+        goto repair;
       }
 
       obj->value = trailer;
 
       if (!_pdfioTokenGet(&tb, line, sizeof(line)) || strcmp(line, "stream"))
       {
-        _pdfioFileError(pdf, "Unable to get stream after xref dictionary.");
-        return (false);
+        PDFIO_DEBUG("load_xref: No stream token after dictionary (got '%s').\n", line);
+        goto repair;
       }
 
       PDFIO_DEBUG("load_xref: tb.bufptr=%p, tb.bufend=%p, tb.bufptr[0]=0x%02x, tb.bufptr[0]=0x%02x\n", tb.bufptr, tb.bufend, tb.bufptr[0], tb.bufptr[1]);
@@ -1849,8 +1915,8 @@ load_xref(
 
       if ((w_array = pdfioDictGetArray(trailer.value.dict, "W")) == NULL)
       {
-	_pdfioFileError(pdf, "Cross-reference stream does not have required W key.");
-	return (false);
+        PDFIO_DEBUG("load_xref: Missing W array in cross-reference objection dictionary.\n");
+        goto repair;
       }
 
       w[0]    = (size_t)pdfioArrayGetNumber(w_array, 0);
@@ -1860,16 +1926,18 @@ load_xref(
       w_2     = w[0];
       w_3     = w[0] + w[1];
 
-      if (w[1] == 0 || w[2] > 4 || w[0] > sizeof(buffer) || w[1] > sizeof(buffer) || w[2] > sizeof(buffer) || w_total > sizeof(buffer))
+      PDFIO_DEBUG("W=[%u %u %u], w_total=%u\n", (unsigned)w[0], (unsigned)w[1], (unsigned)w[2], (unsigned)w_total);
+
+      if (pdfioArrayGetSize(w_array) > 3 || w[1] == 0 || w[2] > 4 || w[0] > sizeof(buffer) || w[1] > sizeof(buffer) || w[2] > sizeof(buffer) || w_total > sizeof(buffer))
       {
-	_pdfioFileError(pdf, "Cross-reference stream has invalid W key [%u %u %u].", (unsigned)w[0], (unsigned)w[1], (unsigned)w[2]);
-	return (false);
+        PDFIO_DEBUG("load_xref: Bad W array in cross-reference objection dictionary.\n");
+        goto repair;
       }
 
       if ((st = pdfioObjOpenStream(obj, true)) == NULL)
       {
-	_pdfioFileError(pdf, "Unable to open cross-reference stream.");
-	return (false);
+        PDFIO_DEBUG("load_xref: Unable to open cross-reference stream.\n");
+        goto repair;
       }
 
       for (index_n = 0; index_n < index_count; index_n += 2)
@@ -1889,7 +1957,20 @@ load_xref(
 	{
 	  count --;
 
-	  PDFIO_DEBUG("load_xref: number=%u %02X%02X%02X%02X%02X\n", (unsigned)number, buffer[0], buffer[1], buffer[2], buffer[3], buffer[4]);
+#ifdef DEBUG
+          if (w_total > 5)
+	    PDFIO_DEBUG("load_xref: number=%u %02X%02X%02X%02X%02X...\n", (unsigned)number, buffer[0], buffer[1], buffer[2], buffer[3], buffer[4]);
+	  else if (w_total == 5)
+	    PDFIO_DEBUG("load_xref: number=%u %02X%02X%02X%02X%02X\n", (unsigned)number, buffer[0], buffer[1], buffer[2], buffer[3], buffer[4]);
+	  else if (w_total == 4)
+	    PDFIO_DEBUG("load_xref: number=%u %02X%02X%02X%02X\n", (unsigned)number, buffer[0], buffer[1], buffer[2], buffer[3]);
+	  else if (w_total == 3)
+	    PDFIO_DEBUG("load_xref: number=%u %02X%02X%02X\n", (unsigned)number, buffer[0], buffer[1], buffer[2]);
+	  else if (w_total == 2)
+	    PDFIO_DEBUG("load_xref: number=%u %02X%02X\n", (unsigned)number, buffer[0], buffer[1]);
+	  else
+	    PDFIO_DEBUG("load_xref: number=%u %02X\n", (unsigned)number, buffer[0]);
+#endif // DEBUG
 
 	  // Check whether this is an object definition...
 	  if (w[0] > 0)
@@ -1971,6 +2052,7 @@ load_xref(
 	      else
 	      {
 		_pdfioFileError(pdf, "Too many object streams.");
+		pdfioStreamClose(st);
 	        return (false);
 	      }
 	    }
@@ -1979,7 +2061,10 @@ load_xref(
 	  {
 	    // Add this object...
 	    if (!add_obj(pdf, (size_t)number, (unsigned short)generation, (off_t)offset))
+	    {
+	      pdfioStreamClose(st);
 	      return (false);
+	    }
 	  }
 
 	  number ++;
@@ -2027,7 +2112,7 @@ load_xref(
 					// Offset of current line
 
       PDFIO_DEBUG("load_xref: Reading xref table starting at offset %lu\n", (unsigned long)trailer_offset);
-      while (_pdfioFileGets(pdf, line, sizeof(line)))
+      while (_pdfioFileGets(pdf, line, sizeof(line), false))
       {
         PDFIO_DEBUG("load_xref: '%s' at offset %lu\n", line, (unsigned long)trailer_offset);
 
@@ -2052,8 +2137,8 @@ load_xref(
 
 	if (sscanf(line, "%jd%jd", &number, &num_objects) != 2)
 	{
-	  _pdfioFileError(pdf, "Malformed xref table section '%s'.", line);
-	  return (false);
+	  PDFIO_DEBUG("load_xref: Unable to scan START COUNT from line.\n");
+	  goto repair;
 	}
 
 	// Read this group of objects...
@@ -2061,41 +2146,45 @@ load_xref(
 	{
 	  // Read a line from the file and validate it...
 	  if (_pdfioFileRead(pdf, line, 20) != 20)
-	    return (false);
+	  {
+	    PDFIO_DEBUG("load_xref: Unable to read 20 byte xref record.\n");
+	    goto repair;
+	  }
 
 	  line[20] = '\0';
 
-	  if (strcmp(line + 18, "\r\n") && strcmp(line + 18, " \n") && strcmp(line + 18, " \r"))
+	  if (strcmp(line + 18, "\r\n") && strcmp(line + 18, "\r\r") && strcmp(line + 18, " \n") && strcmp(line + 18, " \r"))
 	  {
-	    _pdfioFileError(pdf, "Malformed xref table entry '%s'.", line);
-	    return (false);
+	    PDFIO_DEBUG("load_xref: Bad end-of-line <%02X%02X>\n", line[18], line[19]);
+	    goto repair;
 	  }
+
 	  line[18] = '\0';
 
 	  // Parse the line
 	  if ((offset = strtoimax(line, &ptr, 10)) < 0)
 	  {
-	    _pdfioFileError(pdf, "Malformed xref table entry '%s'.", line);
-	    return (false);
+	    PDFIO_DEBUG("load_xref: Unable to scan offset.\n");
+	    goto repair;
 	  }
 
 	  if ((generation = (int)strtol(ptr, &ptr, 10)) < 0 || (generation > 65535 && offset != 0))
 	  {
-	    _pdfioFileError(pdf, "Malformed xref table entry '%s'.", line);
-	    return (false);
+	    PDFIO_DEBUG("load_xref: Unable to scan generation (%u).\n", (unsigned)generation);
+	    goto repair;
 	  }
 
 	  if (*ptr != ' ')
 	  {
-	    _pdfioFileError(pdf, "Malformed xref table entry '%s'.", line);
-	    return (false);
+	    PDFIO_DEBUG("load_xref: Missing space before type.\n");
+	    goto repair;
 	  }
 
 	  ptr ++;
 	  if (*ptr != 'f' && *ptr != 'n')
 	  {
-	    _pdfioFileError(pdf, "Malformed xref table entry '%s'.", line);
-	    return (false);
+	    PDFIO_DEBUG("load_xref: Bad type '%c'.\n", *ptr);
+	    goto repair;
 	  }
 
 	  if (*ptr == 'f')
@@ -2114,21 +2203,21 @@ load_xref(
 
       if (strncmp(line, "trailer", 7))
       {
-	_pdfioFileError(pdf, "Missing trailer.");
-	return (false);
+        PDFIO_DEBUG("load_xref: No trailer after xref table.\n");
+	goto repair;
       }
 
       _pdfioTokenInit(&tb, pdf, (_pdfio_tconsume_cb_t)_pdfioFileConsume, (_pdfio_tpeek_cb_t)_pdfioFilePeek, pdf);
 
       if (!_pdfioValueRead(pdf, NULL, &tb, &trailer, 0))
       {
-	_pdfioFileError(pdf, "Unable to read trailer dictionary.");
-	return (false);
+        PDFIO_DEBUG("load_xref: Unable to read trailer dictionary.\n");
+	goto repair;
       }
       else if (trailer.type != PDFIO_VALTYPE_DICT)
       {
-	_pdfioFileError(pdf, "Trailer is not a dictionary.");
-	return (false);
+        PDFIO_DEBUG("load_xref: Trailer not a dictionary (type=%d).\n", trailer.type);
+	goto repair;
       }
 
       PDFIO_DEBUG("load_xref: Got trailer dict.\n");
@@ -2150,8 +2239,7 @@ load_xref(
     }
     else
     {
-      _pdfioFileError(pdf, "Bad xref table header '%s'.", line);
-      return (false);
+      goto repair;
     }
 
     PDFIO_DEBUG("load_xref: Contents of trailer dictionary:\n");
@@ -2180,13 +2268,31 @@ load_xref(
 
   if ((pdf->root_obj = pdfioDictGetObj(pdf->trailer_dict, "Root")) == NULL)
   {
-    _pdfioFileError(pdf, "Missing Root object.");
-    return (false);
+    PDFIO_DEBUG("load_xref: Missing Root object.\n");
+    goto repair;
   }
 
   PDFIO_DEBUG("load_xref: Root=%p(%lu)\n", pdf->root_obj, (unsigned long)pdf->root_obj->number);
 
-  return (load_pages(pdf, pdfioDictGetObj(pdfioObjGetDict(pdf->root_obj), "Pages"), 0));
+  if ((pages_obj = pdfioDictGetObj(pdfioObjGetDict(pdf->root_obj), "Pages")) == NULL)
+  {
+    PDFIO_DEBUG("load_xref: Missing Pages object.\n");
+    goto repair;
+  }
+
+  PDFIO_DEBUG("load_xref: Pages=%p(%lu)\n", pdf->root_obj, (unsigned long)pdf->root_obj->number);
+
+  return (load_pages(pdf, pages_obj, 0));
+
+  // If we get here the cross-reference table is busted - try repairing if the
+  // error callback says to proceed...
+
+  repair:
+
+  if (_pdfioFileError(pdf, "WARNING: Cross-reference is damaged, will attempt to rebuild."))
+    return (repair_xref(pdf, password_cb, password_data));
+  else
+    return (false);
 }
 
 
@@ -2200,7 +2306,7 @@ repair_xref(
     pdfio_password_cb_t password_cb,	// I - Password callback or `NULL` for none
     void                *password_data)	// I - Password callback data, if any
 {
-  char		line[16384],		// Line from file
+  char		line[1024],		// Line from file
 		*ptr;			// Pointer into line
   off_t		line_offset;		// Offset in file
   intmax_t	number;			// Object number
@@ -2208,13 +2314,22 @@ repair_xref(
   size_t	i;			// Looping var
   size_t	num_sobjs = 0;		// Number of object streams
   pdfio_obj_t	*sobjs[16384];		// Object streams to load
+  pdfio_dict_t	*backup_trailer = NULL;	// Backup trailer dictionary
+  pdfio_obj_t	*pages_obj;		// Pages object
 
 
-  // Read from the beginning of the file, looking for
+  // Clear trailer data...
+  pdf->trailer_dict = NULL;
+  pdf->root_obj     = NULL;
+  pdf->info_obj     = NULL;
+  pdf->pages_obj    = NULL;
+  pdf->encrypt_obj  = NULL;
+
+  // Read from the beginning of the file, looking for objects...
   if ((line_offset = _pdfioFileSeek(pdf, 0, SEEK_SET)) < 0)
     return (false);
 
-  while (_pdfioFileGets(pdf, line, sizeof(line)))
+  while (_pdfioFileGets(pdf, line, sizeof(line), true))
   {
     // See if this is the start of an object...
     if (line[0] >= '1' && line[0] <= '9')
@@ -2231,43 +2346,75 @@ repair_xref(
 	  pdfio_obj_t	*obj;		// Object
 	  _pdfio_token_t tb;		// Token buffer/stack
 
-          PDFIO_DEBUG("OBJECT %ld %d at offset %ld\n", (long)number, generation, (long)line_offset);
+          PDFIO_DEBUG("repair_xref: OBJECT %ld %d at offset %ld\n", (long)number, generation, (long)line_offset);
 
-	  if ((obj = add_obj(pdf, (size_t)number, (unsigned short)generation, line_offset)) == NULL)
+          if ((obj = pdfioFileFindObj(pdf, (size_t)number)) != NULL)
+          {
+            obj->offset = line_offset;
+          }
+          else if ((obj = add_obj(pdf, (size_t)number, (unsigned short)generation, line_offset)) == NULL)
 	  {
 	    _pdfioFileError(pdf, "Unable to allocate memory for object.");
 	    return (false);
 	  }
 
+          if (ptr[3])
+          {
+	    // Probably the start of the object dictionary, rewind the file so
+	    // we can read it...
+	    _pdfioFileSeek(pdf, line_offset + (ptr - line + 3), SEEK_SET);
+	  }
+
 	  _pdfioTokenInit(&tb, pdf, (_pdfio_tconsume_cb_t)_pdfioFileConsume, (_pdfio_tpeek_cb_t)_pdfioFilePeek, pdf);
 
 	  if (!_pdfioValueRead(pdf, obj, &tb, &obj->value, 0))
 	  {
-	    _pdfioFileError(pdf, "Unable to read cross-reference stream dictionary.");
-	    return (false);
+	    if (!_pdfioFileError(pdf, "WARNING: Unable to read object dictionary/value."))
+	      return (false);
+	    else
+	      continue;
 	  }
 
-	  if (_pdfioTokenGet(&tb, line, sizeof(line)) && strcmp(line, "stream"))
+	  if (_pdfioTokenGet(&tb, line, sizeof(line)))
 	  {
 	    const char *type = pdfioObjGetType(obj);
 					// Object type
 
 	    _pdfioTokenFlush(&tb);
-	    obj->stream_offset = _pdfioFileTell(pdf);
 
-	    if (type && !strcmp(type, "ObjStm") && num_sobjs < (sizeof(sobjs) / sizeof(sobjs[0])))
-	    {
-	      sobjs[num_sobjs] = obj;
-	      num_sobjs ++;
-	    }
-
-            if (type && !strcmp(type, "XRef") && !pdf->trailer_dict)
+            if (type && !strcmp(line, "stream"))
             {
-              // Save the trailer dictionary...
-	      pdf->trailer_dict = pdfioObjGetDict(obj);
-	      pdf->encrypt_obj  = pdfioDictGetObj(pdf->trailer_dict, "Encrypt");
-	      pdf->id_array     = pdfioDictGetArray(pdf->trailer_dict, "ID");
-            }
+              // Possible object or XRef stream...
+	      obj->stream_offset = _pdfioFileTell(pdf);
+
+	      if (!strcmp(type, "ObjStm") && num_sobjs < (sizeof(sobjs) / sizeof(sobjs[0])))
+	      {
+	        PDFIO_DEBUG("repair_xref: Object stream...\n");
+		sobjs[num_sobjs] = obj;
+		num_sobjs ++;
+	      }
+
+	      if (!strcmp(type, "XRef") && !pdf->trailer_dict)
+	      {
+		// Save the trailer dictionary...
+	        PDFIO_DEBUG("repair_xref: XRef stream...\n");
+		pdf->trailer_dict = pdfioObjGetDict(obj);
+		pdf->encrypt_obj  = pdfioDictGetObj(pdf->trailer_dict, "Encrypt");
+		pdf->id_array     = pdfioDictGetArray(pdf->trailer_dict, "ID");
+	      }
+	    }
+	    else if (type && !strcmp(line, "endobj"))
+	    {
+	      // Possible catalog or pages object...
+	      if (!strcmp(type, "Catalog"))
+	      {
+	        PDFIO_DEBUG("repair_xref: Catalog (root) object...\n");
+	        if (!backup_trailer)
+	          backup_trailer = pdfioDictCreate(pdf);
+
+	        pdfioDictSetObj(backup_trailer, "Root", obj);
+	      }
+	    }
 	  }
 	}
       }
@@ -2278,6 +2425,8 @@ repair_xref(
       _pdfio_token_t tb;		// Token buffer/stack
       _pdfio_value_t trailer;		// Trailer
 
+      PDFIO_DEBUG("repair_xref: line=\"%s\"\n", line);
+
       if (line[7])
       {
 	// Probably the start of the trailer dictionary, rewind the file so
@@ -2285,7 +2434,7 @@ repair_xref(
 	_pdfioFileSeek(pdf, line_offset + 7, SEEK_SET);
       }
 
-      PDFIO_DEBUG("TRAILER at offset %ld\n", (long)line_offset);
+      PDFIO_DEBUG("repair_xref: TRAILER at offset %ld\n", (long)line_offset);
 
       _pdfioTokenInit(&tb, pdf, (_pdfio_tconsume_cb_t)_pdfioFileConsume, (_pdfio_tpeek_cb_t)_pdfioFilePeek, pdf);
       if (!_pdfioValueRead(pdf, NULL, &tb, &trailer, 0))
@@ -2301,10 +2450,12 @@ repair_xref(
 
       _pdfioTokenFlush(&tb);
 
-      if (!pdf->trailer_dict)
+      if (_pdfioDictGetValue(trailer.value.dict, "Root"))
       {
 	// Save the trailer dictionary and grab the root (catalog) and info
 	// objects...
+	PDFIO_DEBUG("repair_xref: Using this trailer dictionary.\n");
+
 	pdf->trailer_dict = trailer.value.dict;
 	pdf->encrypt_obj  = pdfioDictGetObj(pdf->trailer_dict, "Encrypt");
 	pdf->id_array     = pdfioDictGetArray(pdf->trailer_dict, "ID");
@@ -2315,11 +2466,18 @@ repair_xref(
     line_offset = _pdfioFileTell(pdf);
   }
 
+  PDFIO_DEBUG("repair_xref: Stopped at line_offset=%lu\n", (unsigned long)line_offset);
+
+  if (!pdf->trailer_dict && backup_trailer)
+    pdf->trailer_dict = backup_trailer;
+
   // If the trailer contains an Encrypt key, try unlocking the file...
   if (pdf->encrypt_obj && !_pdfioCryptoUnlock(pdf, password_cb, password_data))
     return (false);
 
   // Load any stream objects...
+  PDFIO_DEBUG("repair_xref: Found %lu stream objects.\n", (unsigned long)num_sobjs);
+
   for (i = 0; i < num_sobjs; i ++)
   {
     if (!load_obj_stream(sobjs[i]))
@@ -2338,8 +2496,16 @@ repair_xref(
 
   PDFIO_DEBUG("repair_xref: Root=%p(%lu)\n", pdf->root_obj, (unsigned long)pdf->root_obj->number);
 
+  if ((pages_obj = pdfioDictGetObj(pdfioObjGetDict(pdf->root_obj), "Pages")) == NULL)
+  {
+    _pdfioFileError(pdf, "Missing Pages object.");
+    return (false);
+  }
+
+  PDFIO_DEBUG("repair_xref: Pages=%p(%lu)\n", pages_obj, (unsigned long)pages_obj->number);
+
   // Load pages...
-  return (load_pages(pdf, pdfioDictGetObj(pdfioObjGetDict(pdf->root_obj), "Pages"), 0));
+  return (load_pages(pdf, pages_obj, 0));
 }
 
 
@@ -2489,12 +2655,16 @@ write_trailer(pdfio_file_t *pdf)	// I - PDF file
             buffer[2] = (obj->offset >> 8) & 255;
             buffer[3] = obj->offset & 255;
             break;
+#ifdef _WIN32
+	default :
+#endif // _WIN32
         case 4 :
             buffer[1] = (obj->offset >> 24) & 255;
             buffer[2] = (obj->offset >> 16) & 255;
             buffer[3] = (obj->offset >> 8) & 255;
             buffer[4] = obj->offset & 255;
             break;
+#ifndef _WIN32 // Windows off_t is 32-bits?!?
         case 5 :
             buffer[1] = (obj->offset >> 32) & 255;
             buffer[2] = (obj->offset >> 24) & 255;
@@ -2529,6 +2699,7 @@ write_trailer(pdfio_file_t *pdf)	// I - PDF file
             buffer[7] = (obj->offset >> 8) & 255;
             buffer[8] = obj->offset & 255;
             break;
+#endif // !_WIN32
       }
 
       if (!pdfioStreamWrite(xref_st, buffer, offsize + 2))

pdfio-object.c

@@ -79,6 +79,10 @@ pdfioObjCopy(pdfio_file_t *pdf,		// I - PDF file
   if (srcobj->value.type == PDFIO_VALTYPE_NONE)
     _pdfioObjLoad(srcobj);
 
+  // See if we have already mapped this object...
+  if ((dstobj = _pdfioFileFindMappedObj(pdf, srcobj->pdf, srcobj->number)) != NULL)
+    return (dstobj);			// Yes, return that one...
+
   // Create the new object...
   if ((dstobj = _pdfioFileCreateObj(pdf, srcobj->pdf, NULL)) == NULL)
     return (NULL);
@@ -141,6 +145,7 @@ pdfioObjCreateStream(
     pdfio_obj_t    *obj,		// I - Object
     pdfio_filter_t filter)		// I - Type of compression to apply
 {
+  pdfio_stream_t *st;			// Stream
   pdfio_obj_t	*length_obj = NULL;	// Length object, if any
 
 
@@ -194,11 +199,13 @@ pdfioObjCreateStream(
   if (!_pdfioFilePuts(obj->pdf, "stream\n"))
     return (NULL);
 
-  obj->stream_offset    = _pdfioFileTell(obj->pdf);
-  obj->pdf->current_obj = obj;
+  obj->stream_offset = _pdfioFileTell(obj->pdf);
 
   // Return the new stream...
-  return (_pdfioStreamCreate(obj, length_obj, 0, filter));
+  if ((st = _pdfioStreamCreate(obj, length_obj, 0, filter)) != NULL)
+    obj->pdf->current_obj = obj;
+
+  return (st);
 }
 
 
@@ -330,6 +337,8 @@ pdfioObjGetLength(pdfio_obj_t *obj)	// I - Object
 //
 // 'pdfioObjGetName()' - Get the name value associated with an object.
 //
+// @since PDFio v1.4@
+//
 
 const char *				// O - Dictionary or `NULL` on error
 pdfioObjGetName(pdfio_obj_t *obj)	// I - Object
@@ -505,7 +514,7 @@ _pdfioObjLoad(pdfio_obj_t *obj)		// I - Object
   }
 
   // Decrypt as needed...
-  if (obj->pdf->encryption)
+  if (obj->pdf->encryption && obj->pdf->encrypt_metadata)
   {
     PDFIO_DEBUG("_pdfioObjLoad: Decrypting value...\n");
 
@@ -532,6 +541,9 @@ pdfio_stream_t *			// O - Stream or `NULL` on error
 pdfioObjOpenStream(pdfio_obj_t *obj,	// I - Object
                    bool        decode)	// I - Decode/decompress data?
 {
+  pdfio_stream_t	*st;		// Stream
+
+
   // Range check input...
   if (!obj)
     return (NULL);
@@ -554,9 +566,10 @@ pdfioObjOpenStream(pdfio_obj_t *obj,	// I - Object
     return (NULL);
 
   // Open the stream...
-  obj->pdf->current_obj = obj;
+  if ((st = _pdfioStreamOpen(obj, decode)) != NULL)
+    obj->pdf->current_obj = obj;
 
-  return (_pdfioStreamOpen(obj, decode));
+  return (st);
 }
 
 

pdfio-private.h

@@ -10,7 +10,7 @@
 #ifndef PDFIO_PRIVATE_H
 #  define PDFIO_PRIVATE_H
 #  ifdef _WIN32
-#    define _CRT_SECURE_NO_WARNINGS	// Disable bogus VS warnings/errors...
+#    define _CRT_SECURE_NO_WARNINGS 1	// Disable bogus VS warnings/errors...
 #  endif // _WIN32
 #  include "pdfio.h"
 #  include <stdarg.h>
@@ -28,16 +28,16 @@
 #    define access	_access		// Map standard POSIX/C99 names
 #    define close	_close
 #    define fileno	_fileno
-#    define lseek	_lseek
+#    define lseek(f,o,w) (off_t)_lseek((f),(long)(o),(w))
 #    define mkdir(d,p)	_mkdir(d)
 #    define open	_open
-#    define read	_read
+#    define read(f,b,s)	_read((f),(b),(unsigned)(s))
 #    define rmdir	_rmdir
 #    define snprintf	_snprintf
 #    define strdup	_strdup
 #    define unlink	_unlink
 #    define vsnprintf	_vsnprintf
-#    define write	_write
+#    define write(f,b,s) _write((f),(b),(unsigned)(s))
 #    ifndef F_OK
 #      define F_OK	00		// POSIX parameters/flags
 #      define W_OK	02
@@ -94,6 +94,7 @@
 //
 
 #  define PDFIO_MAX_DEPTH	32	// Maximum nesting depth for values
+#  define PDFIO_MAX_STRING	65536	// Maximum length of string
 
 typedef void (*_pdfio_extfree_t)(void *);
 					// Extension data free function
@@ -220,13 +221,22 @@ struct _pdfio_dict_s			// Dictionary
 typedef struct _pdfio_objmap_s		// PDF object map
 {
   pdfio_obj_t	*obj;			// Object for this file
-  pdfio_file_t	*src_pdf;		// Source PDF file
+  unsigned char	src_id[32];		// Source PDF file file identifier
   size_t	src_number;		// Source object number
 } _pdfio_objmap_t;
 
+typedef struct _pdfio_strbuf_s		// PDF string buffer
+{
+  struct _pdfio_strbuf_s *next;		// Next string buffer
+  bool		bufused;		// Is this string buffer being used?
+  char		buffer[PDFIO_MAX_STRING + 32];
+					// String buffer
+} _pdfio_strbuf_t;
+
 struct _pdfio_file_s			// PDF file structure
 {
   char		*filename;		// Filename
+  unsigned char	file_id[32];		// File identifier bytes
   struct lconv	*loc;			// Locale data
   char		*version;		// Version number
   pdfio_rect_t	media_box,		// Default MediaBox value
@@ -261,6 +271,7 @@ struct _pdfio_file_s			// PDF file structure
   pdfio_obj_t	*cp1252_obj,		// CP1252 font encoding object
 		*unicode_obj;		// Unicode font encoding object
   pdfio_array_t	*id_array;		// ID array
+  bool		encrypt_metadata;	// Encrypt metadata?
 
   // Allocated data elements
   size_t	num_arrays,		// Number of arrays
@@ -283,6 +294,7 @@ struct _pdfio_file_s			// PDF file structure
   size_t	num_strings,		// Number of strings
 		alloc_strings;		// Allocated strings
   char		**strings;		// Nul-terminated strings
+  _pdfio_strbuf_t *strbuffers;		// String buffers
 };
 
 struct _pdfio_obj_s			// Object
@@ -329,6 +341,7 @@ struct _pdfio_stream_s			// Stream
 
 extern size_t		_pdfio_strlcpy(char *dst, const char *src, size_t dstsize) _PDFIO_INTERNAL;
 extern double		_pdfio_strtod(pdfio_file_t *pdf, const char *s) _PDFIO_INTERNAL;
+extern void		_pdfio_utf16cpy(char *dst, const unsigned char *src, size_t srclen, size_t dstsize) _PDFIO_INTERNAL;
 extern ssize_t		_pdfio_vsnprintf(pdfio_file_t *pdf, char *buffer, size_t bufsize, const char *format, va_list ap) _PDFIO_INTERNAL;
 
 extern bool		_pdfioArrayDecrypt(pdfio_file_t *pdf, pdfio_obj_t *obj, pdfio_array_t *a, size_t depth) _PDFIO_INTERNAL;
@@ -372,7 +385,7 @@ extern bool		_pdfioFileError(pdfio_file_t *pdf, const char *format, ...) _PDFIO_
 extern pdfio_obj_t	*_pdfioFileFindMappedObj(pdfio_file_t *pdf, pdfio_file_t *src_pdf, size_t src_number) _PDFIO_INTERNAL;
 extern bool		_pdfioFileFlush(pdfio_file_t *pdf) _PDFIO_INTERNAL;
 extern int		_pdfioFileGetChar(pdfio_file_t *pdf) _PDFIO_INTERNAL;
-extern bool		_pdfioFileGets(pdfio_file_t *pdf, char *buffer, size_t bufsize) _PDFIO_INTERNAL;
+extern bool		_pdfioFileGets(pdfio_file_t *pdf, char *buffer, size_t bufsize, bool discard) _PDFIO_INTERNAL;
 extern ssize_t		_pdfioFilePeek(pdfio_file_t *pdf, void *buffer, size_t bytes) _PDFIO_INTERNAL;
 extern bool		_pdfioFilePrintf(pdfio_file_t *pdf, const char *format, ...) _PDFIO_INTERNAL;
 extern bool		_pdfioFilePuts(pdfio_file_t *pdf, const char *s) _PDFIO_INTERNAL;
@@ -390,6 +403,8 @@ extern bool		_pdfioObjWriteHeader(pdfio_obj_t *obj) _PDFIO_INTERNAL;
 extern pdfio_stream_t	*_pdfioStreamCreate(pdfio_obj_t *obj, pdfio_obj_t *length_obj, size_t cbsize, pdfio_filter_t compression) _PDFIO_INTERNAL;
 extern pdfio_stream_t	*_pdfioStreamOpen(pdfio_obj_t *obj, bool decode) _PDFIO_INTERNAL;
 
+extern char		*_pdfioStringAllocBuffer(pdfio_file_t *pdf);
+extern void		_pdfioStringFreeBuffer(pdfio_file_t *pdf, char *buffer);
 extern bool		_pdfioStringIsAllocated(pdfio_file_t *pdf, const char *s) _PDFIO_INTERNAL;
 
 extern void		_pdfioTokenClear(_pdfio_token_t *tb) _PDFIO_INTERNAL;

pdfio-stream.c

@@ -259,7 +259,7 @@ _pdfioStreamCreate(
     {
       colors = 1;
     }
-    else if (colors < 0 || colors > 4)
+    else if (colors < 0 || colors > 32)
     {
       _pdfioFileError(st->pdf, "Unsupported Colors value %d.", colors);
       free(st);
@@ -270,7 +270,7 @@ _pdfioStreamCreate(
     {
       columns = 1;
     }
-    else if (columns < 0)
+    else if (columns < 0 || columns > 65536)
     {
       _pdfioFileError(st->pdf, "Unsupported Columns value %d.", columns);
       free(st);
@@ -532,7 +532,7 @@ _pdfioStreamOpen(pdfio_obj_t *obj,	// I - Object
       {
         colors = 1;
       }
-      else if (colors < 0 || colors > 4)
+      else if (colors < 0 || colors > 32)
       {
         _pdfioFileError(st->pdf, "Unsupported Colors value %d.", colors);
 	goto error;
@@ -542,7 +542,7 @@ _pdfioStreamOpen(pdfio_obj_t *obj,	// I - Object
       {
         columns = 1;
       }
-      else if (columns < 0)
+      else if (columns < 0 || columns > 65536)
       {
         _pdfioFileError(st->pdf, "Unsupported Columns value %d.", columns);
 	goto error;
@@ -562,6 +562,13 @@ _pdfioStreamOpen(pdfio_obj_t *obj,	// I - Object
         if (predictor >= 10)
           st->pbsize ++;		// Add PNG predictor byte
 
+        if (st->pbsize < 2)
+        {
+          _pdfioFileError(st->pdf, "Bad Predictor buffer size %lu.", (unsigned long)st->pbsize);
+	  goto error;
+        }
+
+        PDFIO_DEBUG("_pdfioStreamOpen: st->predictor=%d, st->pbpixel=%u, st->pbsize=%lu\n", st->predictor, (unsigned)st->pbpixel, (unsigned long)st->pbsize);
         if ((st->prbuffer = calloc(1, st->pbsize - 1)) == NULL || (st->psbuffer = calloc(1, st->pbsize)) == NULL)
         {
           _pdfioFileError(st->pdf, "Unable to allocate %lu bytes for Predictor buffers.", (unsigned long)st->pbsize);
@@ -690,8 +697,9 @@ pdfioStreamPeek(pdfio_stream_t *st,	// I - Stream
 // 'pdfioStreamPrintf()' - Write a formatted string to a stream.
 //
 // This function writes a formatted string to a stream.  In addition to the
-// standard `printf` format characters, you can use "%N" to format a PDF name
-// value ("/Name") and "%S" to format a PDF string ("(String)") value.
+// standard `printf` format characters, you can use "%H" to format a HTML/XML
+// string value, "%N" to format a PDF name value ("/Name"), and "%S" to format
+// a PDF string ("(String)") value.
 //
 
 bool					// O - `true` on success, `false` on failure
@@ -1227,7 +1235,18 @@ stream_read(pdfio_stream_t *st,		// I - Stream
       }
 
       // Apply predictor for this line
-      PDFIO_DEBUG("stream_read: Line %02X %02X %02X %02X %02X.\n", sptr[-1], sptr[0], sptr[0], sptr[2], sptr[3]);
+#ifdef DEBUG
+      if (remaining > 4)
+	PDFIO_DEBUG("stream_read: Line %02X %02X %02X %02X %02X ...\n", sptr[-1], sptr[0], sptr[1], sptr[2], sptr[3]);
+      else if (remaining > 3)
+	PDFIO_DEBUG("stream_read: Line %02X %02X %02X %02X %02X.\n", sptr[-1], sptr[0], sptr[1], sptr[2], sptr[3]);
+      else if (remaining > 2)
+	PDFIO_DEBUG("stream_read: Line %02X %02X %02X %02X.\n", sptr[-1], sptr[0], sptr[1], sptr[2]);
+      else if (remaining > 1)
+	PDFIO_DEBUG("stream_read: Line %02X %02X %02X.\n", sptr[-1], sptr[0], sptr[1]);
+      else
+	PDFIO_DEBUG("stream_read: Line %02X %02X.\n", sptr[-1], sptr[0]);
+#endif // DEBUG
 
       switch (sptr[-1])
       {

pdfio-string.c

@@ -158,6 +158,89 @@ _pdfio_strtod(pdfio_file_t *pdf,	// I - PDF file
 }
 
 
+//
+// '_pdfio_utf16cpy()' - Convert UTF-16 to UTF-8.
+//
+
+void
+_pdfio_utf16cpy(
+    char                *dst,		// I - Destination buffer for UTF-8
+    const unsigned char *src,		// I - Source UTF-16
+    size_t              srclen,		// I - Length of UTF-16
+    size_t              dstsize)	// I - Destination buffer size
+{
+  char	*dstptr = dst,			// Pointer into buffer
+	*dstend = dst + dstsize - 5;	// End of buffer
+  int	ch;				// Unicode character
+  bool	is_be = !memcmp(src, "\376\377", 2);
+					// Big-endian strings?
+
+
+  // Loop through the UTF-16 string, converting to Unicode then UTF-8...
+  for (src += 2, srclen -= 2; srclen > 1 && dstptr < dstend; src += 2, srclen -= 2)
+  {
+    // Initial character...
+    if (is_be)
+      ch = (src[0] << 8) | src[1];
+    else
+      ch = (src[1] << 8) | src[0];
+
+    if (ch >= 0xd800 && ch <= 0xdbff && srclen > 3)
+    {
+      // Multi-word UTF-16 char...
+      int lch;			// Lower bits
+
+      if (is_be)
+	lch = (src[2] << 8) | src[3];
+      else
+	lch = (src[3] << 8) | src[2];
+
+      if (lch < 0xdc00 || lch >= 0xdfff)
+	break;
+
+      ch     = (((ch & 0x3ff) << 10) | (lch & 0x3ff)) + 0x10000;
+      src    += 2;
+      srclen -= 2;
+    }
+    else if (ch >= 0xfffe)
+    {
+      continue;
+    }
+
+    // Convert Unicode to UTF-8...
+    if (ch < 128)
+    {
+      // ASCII
+      *dstptr++ = (char)ch;
+    }
+    else if (ch < 4096)
+    {
+      // 2-byte UTF-8
+      *dstptr++ = (char)(0xc0 | (ch >> 6));
+      *dstptr++ = (char)(0x80 | (ch & 0x3f));
+    }
+    else if (ch < 65536)
+    {
+      // 3-byte UTF-8
+      *dstptr++ = (char)(0xe0 | (ch >> 12));
+      *dstptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
+      *dstptr++ = (char)(0x80 | (ch & 0x3f));
+    }
+    else
+    {
+      // 4-byte UTF-8
+      *dstptr++ = (char)(0xe0 | (ch >> 18));
+      *dstptr++ = (char)(0x80 | ((ch >> 12) & 0x3f));
+      *dstptr++ = (char)(0x80 | ((ch >> 6) & 0x3f));
+      *dstptr++ = (char)(0x80 | (ch & 0x3f));
+    }
+  }
+
+  // Nul-terminate the UTF-8 string...
+  *dstptr = '\0';
+}
+
+
 //
 // '_pdfio_vsnprintf()' - Format a string.
 //
@@ -403,6 +486,59 @@ _pdfio_vsnprintf(pdfio_file_t *pdf,	// I - PDF file
 	    }
 	    break;
 
+	case 'H' : // XML/HTML string
+	    if ((s = va_arg(ap, char *)) == NULL)
+	      s = "(null)";
+
+            // Loop through the literal string...
+            while (*s)
+            {
+              // Escape special characters
+              if (*s == '&')
+              {
+                // &amp;
+                if (bufptr < bufend)
+                {
+                  _pdfio_strlcpy(bufptr, "&amp;", (size_t)(bufend - bufptr + 1));
+                  bufptr += strlen(bufptr);
+                }
+
+                bytes += 5;
+              }
+              else if (*s == '<')
+              {
+                // &lt;
+                if (bufptr < bufend)
+                {
+                  _pdfio_strlcpy(bufptr, "&lt;", (size_t)(bufend - bufptr + 1));
+                  bufptr += strlen(bufptr);
+                }
+
+                bytes += 4;
+              }
+              else if (*s == '>')
+              {
+                // &gt;
+                if (bufptr < bufend)
+                {
+                  _pdfio_strlcpy(bufptr, "&gt;", (size_t)(bufend - bufptr + 1));
+                  bufptr += strlen(bufptr);
+                }
+
+                bytes += 4;
+              }
+              else
+              {
+                // Literal character...
+                if (bufptr < bufend)
+                  *bufptr++ = *s;
+                bytes ++;
+              }
+
+              s ++;
+            }
+	    break;
+
 	case 'S' : // PDF string
 	    if ((s = va_arg(ap, char *)) == NULL)
 	      s = "(null)";
@@ -532,6 +668,41 @@ _pdfio_vsnprintf(pdfio_file_t *pdf,	// I - PDF file
 }
 
 
+//
+// '_pdfioStringAllocBuffer()' - Allocate a string buffer.
+//
+
+char *					// O - Buffer or `NULL` on error
+_pdfioStringAllocBuffer(
+    pdfio_file_t *pdf)			// I - PDF file
+{
+  _pdfio_strbuf_t	*current;	// Current string buffer
+
+
+  // See if we have an available string buffer...
+  for (current = pdf->strbuffers; current; current = current->next)
+  {
+    if (!current->bufused)
+    {
+      current->bufused = true;
+      return (current->buffer);
+    }
+  }
+
+  // Didn't find one, allocate a new one...
+  if ((current = calloc(1, sizeof(_pdfio_strbuf_t))) == NULL)
+    return (NULL);
+
+  // Add to the linked list of string buffers...
+  current->next    = pdf->strbuffers;
+  current->bufused = true;
+
+  pdf->strbuffers = current;
+
+  return (current->buffer);
+}
+
+
 //
 // 'pdfioStringCreate()' - Create a durable literal string.
 //
@@ -642,6 +813,29 @@ pdfioStringCreatef(
 }
 
 
+//
+// '_pdfioStringFreeBuffer()' - Free a string buffer.
+//
+
+void
+_pdfioStringFreeBuffer(
+    pdfio_file_t *pdf,			// I - PDF file
+    char         *buffer)		// I - String buffer
+{
+  _pdfio_strbuf_t	*current;	// Current string buffer
+
+
+  for (current = pdf->strbuffers; current; current = current->next)
+  {
+    if (current->buffer == buffer)
+    {
+      current->bufused = false;
+      break;
+    }
+  }
+}
+
+
 //
 // '_pdfioStringIsAllocated()' - Check whether a string has been allocated.
 //

pdfio-token.c

@@ -1,7 +1,7 @@
 //
 // PDF token parsing functions for PDFio.
 //
-// Copyright © 2021-2023 by Michael R Sweet.
+// Copyright © 2021-2025 by Michael R Sweet.
 //
 // Licensed under Apache License v2.0.  See the file "LICENSE" for more
 // information.
@@ -528,13 +528,6 @@ _pdfioTokenRead(_pdfio_token_t *tb,	// I - Token buffer/stack
 	    return (false);
 	  }
 	}
-
-	if (bufptr == (buffer + 1))
-	{
-	  _pdfioFileError(tb->pdf, "Empty name.");
-	  *bufptr = '\0';
-	  return (false);
-	}
 	break;
 
     case '<' : // Potential hex string

pdfio-value.c

@@ -125,7 +125,7 @@ _pdfioValueDecrypt(pdfio_file_t   *pdf,	// I - PDF file
   _pdfio_crypto_ctx_t	ctx;		// Decryption context
   _pdfio_crypto_cb_t	cb;		// Decryption callback
   size_t		ivlen;		// Number of initialization vector bytes
-  uint8_t		temp[32768];	// Temporary buffer for decryption
+  uint8_t		*temp = NULL;	// Temporary buffer for decryption
   size_t		templen;	// Number of actual data bytes
   time_t		timeval;	// Date/time value
 
@@ -152,11 +152,16 @@ _pdfioValueDecrypt(pdfio_file_t   *pdf,	// I - PDF file
 
     case PDFIO_VALTYPE_BINARY :
 	// Decrypt the binary string...
-	if (v->value.binary.datalen > (sizeof(temp) - 32))
+	if (v->value.binary.datalen > PDFIO_MAX_STRING)
 	{
 	  _pdfioFileError(pdf, "Unable to read encrypted binary string - too long.");
 	  return (false);
 	}
+	else if ((temp = (uint8_t *)_pdfioStringAllocBuffer(pdf)) == NULL)
+	{
+	  _pdfioFileError(pdf, "Unable to read encrypted binary string - out of memory.");
+	  return (false);
+	}
 
 	ivlen = v->value.binary.datalen;
 	if ((cb = _pdfioCryptoMakeReader(pdf, obj, &ctx, v->value.binary.data, &ivlen)) == NULL)
@@ -167,29 +172,59 @@ _pdfioValueDecrypt(pdfio_file_t   *pdf,	// I - PDF file
 	// Copy the decrypted string back to the value and adjust the length...
 	memcpy(v->value.binary.data, temp, templen);
 
-	if (pdf->encryption >= PDFIO_ENCRYPTION_AES_128)
+	if (pdf->encryption >= PDFIO_ENCRYPTION_AES_128 && temp[templen - 1] <= templen)
 	  v->value.binary.datalen = templen - temp[templen - 1];
 	else
 	  v->value.binary.datalen = templen;
+
+        _pdfioStringFreeBuffer(pdf, (char *)temp);
 	break;
 
     case PDFIO_VALTYPE_STRING :
         // Decrypt regular string...
         templen = strlen(v->value.string);
-	if (templen > (sizeof(temp) - 33))
+	if (templen > (PDFIO_MAX_STRING - 1))
 	{
 	  _pdfioFileError(pdf, "Unable to read encrypted string - too long.");
 	  return (false);
 	}
+	else if ((temp = (uint8_t *)_pdfioStringAllocBuffer(pdf)) == NULL)
+	{
+	  _pdfioFileError(pdf, "Unable to read encrypted binary string - out of memory.");
+	  return (false);
+	}
 
 	ivlen = templen;
 	if ((cb = _pdfioCryptoMakeReader(pdf, obj, &ctx, (uint8_t *)v->value.string, &ivlen)) == NULL)
 	  return (false);
 
 	templen = (cb)(&ctx, temp, (uint8_t *)v->value.string + ivlen, templen - ivlen);
+
+	if (pdf->encryption >= PDFIO_ENCRYPTION_AES_128 && temp[templen - 1] <= templen)
+	  templen -= temp[templen - 1];
+
 	temp[templen] = '\0';
 
-        if ((timeval = get_date_time((char *)temp)) != 0)
+        if ((templen & 1) == 0 && (!memcmp(temp, "\376\377", 2) || !memcmp(temp, "\377\376", 2)))
+        {
+          // Convert UTF-16 to UTF-8...
+          char	utf8[4096];		// Temporary string
+
+          _pdfio_utf16cpy(utf8, temp, templen, sizeof(utf8));
+
+	  if ((timeval = get_date_time((char *)utf8)) != 0)
+	  {
+	    // Change the type to date...
+	    v->type       = PDFIO_VALTYPE_DATE;
+	    v->value.date = timeval;
+	  }
+	  else
+	  {
+	    // Copy the decrypted string back to the value...
+	    v->value.string = pdfioStringCreate(pdf, utf8);
+	  }
+        }
+        else if ((timeval = get_date_time((char *)temp)) != 0)
         {
           // Change the type to date...
           v->type       = PDFIO_VALTYPE_DATE;
@@ -200,6 +235,8 @@ _pdfioValueDecrypt(pdfio_file_t   *pdf,	// I - PDF file
           // Copy the decrypted string back to the value...
 	  v->value.string = pdfioStringCreate(pdf, (char *)temp);
 	}
+
+        _pdfioStringFreeBuffer(pdf, (char *)temp);
         break;
   }
 
@@ -300,7 +337,9 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
                 _pdfio_value_t *v,	// I - Value
                 size_t         depth)	// I - Depth of value
 {
-  char		token[32768];		// Token buffer
+  _pdfio_value_t *ret = NULL;		// Return value
+  char		*token = _pdfioStringAllocBuffer(pdf);
+					// Token buffer
   time_t	timeval;		// Date/time value
 #ifdef DEBUG
   static const char * const valtypes[] =
@@ -322,8 +361,11 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
 
   PDFIO_DEBUG("_pdfioValueRead(pdf=%p, obj=%p, v=%p)\n", pdf, obj, v);
 
-  if (!_pdfioTokenGet(tb, token, sizeof(token)))
-    return (NULL);
+  if (!token)
+    goto done;
+
+  if (!_pdfioTokenGet(tb, token, PDFIO_MAX_STRING))
+    goto done;
 
   if (!strcmp(token, "["))
   {
@@ -331,12 +373,14 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
     if (depth >= PDFIO_MAX_DEPTH)
     {
       _pdfioFileError(pdf, "Too many nested arrays.");
-      return (NULL);
+      goto done;
     }
 
     v->type = PDFIO_VALTYPE_ARRAY;
     if ((v->value.array = _pdfioArrayRead(pdf, obj, tb, depth + 1)) == NULL)
-      return (NULL);
+      goto done;
+
+    ret = v;
   }
   else if (!strcmp(token, "<<"))
   {
@@ -344,29 +388,38 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
     if (depth >= PDFIO_MAX_DEPTH)
     {
       _pdfioFileError(pdf, "Too many nested dictionaries.");
-      return (NULL);
+      goto done;
     }
 
     v->type = PDFIO_VALTYPE_DICT;
     if ((v->value.dict = _pdfioDictRead(pdf, obj, tb, depth + 1)) == NULL)
-      return (NULL);
-  }
-  else if (!strncmp(token, "(D:", 3) && (timeval = get_date_time(token + 1)) != 0)
-  {
-    v->type       = PDFIO_VALTYPE_DATE;
-    v->value.date = timeval;
+      goto done;
+
+    ret = v;
   }
   else if (token[0] == '(')
   {
-    // String
-    v->type         = PDFIO_VALTYPE_STRING;
-    v->value.string = pdfioStringCreate(pdf, token + 1);
+    if ((timeval = get_date_time(token + 1)) != 0)
+    {
+      // Date
+      v->type       = PDFIO_VALTYPE_DATE;
+      v->value.date = timeval;
+      ret           = v;
+    }
+    else
+    {
+      // String
+      v->type         = PDFIO_VALTYPE_STRING;
+      v->value.string = pdfioStringCreate(pdf, token + 1);
+      ret           = v;
+    }
   }
   else if (token[0] == '/')
   {
     // Name
     v->type       = PDFIO_VALTYPE_NAME;
     v->value.name = pdfioStringCreate(pdf, token + 1);
+    ret           = v;
   }
   else if (token[0] == '<')
   {
@@ -379,7 +432,7 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
     if ((v->value.binary.data = (unsigned char *)malloc(v->value.binary.datalen)) == NULL)
     {
       _pdfioFileError(pdf, "Out of memory for hex string.");
-      return (NULL);
+      goto done;
     }
 
     // Convert hex to binary...
@@ -406,6 +459,8 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
 
       *dataptr++ = (unsigned char)d;
     }
+
+    ret = v;
   }
   else if (strchr("0123456789-+.", token[0]) != NULL)
   {
@@ -493,7 +548,8 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
 
 	  PDFIO_DEBUG("_pdfioValueRead: Returning indirect value %lu %u R.\n", (unsigned long)v->value.indirect.number, v->value.indirect.generation);
 
-	  return (v);
+	  ret = v;
+	  goto done;
 	}
       }
     }
@@ -501,27 +557,41 @@ _pdfioValueRead(pdfio_file_t   *pdf,	// I - PDF file
     // If we get here, we have a number...
     v->type         = PDFIO_VALTYPE_NUMBER;
     v->value.number = _pdfio_strtod(pdf, token);
+    ret             = v;
   }
   else if (!strcmp(token, "true") || !strcmp(token, "false"))
   {
     // Boolean value
     v->type          = PDFIO_VALTYPE_BOOLEAN;
     v->value.boolean = !strcmp(token, "true");
+    ret              = v;
   }
   else if (!strcmp(token, "null"))
   {
     // null value
     v->type = PDFIO_VALTYPE_NULL;
+    ret     = v;
   }
   else
   {
     _pdfioFileError(pdf, "Unexpected '%s' token seen.", token);
-    return (NULL);
   }
 
-  PDFIO_DEBUG("_pdfioValueRead: Returning %s value.\n", valtypes[v->type]);
+  done:
 
-  return (v);
+  if (token)
+    _pdfioStringFreeBuffer(pdf, token);
+
+  if (ret)
+  {
+    PDFIO_DEBUG("_pdfioValueRead: Returning %s value.\n", valtypes[ret->type]);
+    return (ret);
+  }
+  else
+  {
+    PDFIO_DEBUG("_pdfioValueRead: Returning NULL.\n");
+    return (NULL);
+  }
 }
 
 
@@ -546,8 +616,10 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
     case PDFIO_VALTYPE_BINARY :
         {
           size_t	databytes;	// Bytes to write
-          uint8_t	temp[32768],	// Temporary buffer for encryption
+          uint8_t	*temp = NULL,	// Temporary buffer for encryption
 			*dataptr;	// Pointer into data
+          bool          ret = false;	// Return value
+
 
           if (obj && pdf->encryption)
           {
@@ -556,11 +628,16 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
 	    _pdfio_crypto_cb_t cb;	// Encryption callback
 	    size_t	ivlen;		// Number of initialization vector bytes
 
-            if (v->value.binary.datalen > (sizeof(temp) - 32))
+            if (v->value.binary.datalen > PDFIO_MAX_STRING)
             {
 	      _pdfioFileError(pdf, "Unable to write encrypted binary string - too long.");
 	      return (false);
             }
+            else if ((temp = (uint8_t *)_pdfioStringAllocBuffer(pdf)) == NULL)
+            {
+	      _pdfioFileError(pdf, "Unable to write encrypted binary string - out of memory.");
+	      return (false);
+            }
 
 	    cb        = _pdfioCryptoMakeWriter(pdf, obj, &ctx, temp, &ivlen);
 	    databytes = (cb)(&ctx, temp + ivlen, v->value.binary.data, v->value.binary.datalen) + ivlen;
@@ -573,18 +650,25 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
           }
 
           if (!_pdfioFilePuts(pdf, "<"))
-            return (false);
+            goto bindone;
 
           for (; databytes > 1; databytes -= 2, dataptr += 2)
           {
             if (!_pdfioFilePrintf(pdf, "%02X%02X", dataptr[0], dataptr[1]))
-              return (false);
+              goto bindone;
           }
 
-          if (databytes > 0)
-            return (_pdfioFilePrintf(pdf, "%02X>", dataptr[0]));
-          else
-            return (_pdfioFilePuts(pdf, ">"));
+          if (databytes > 0 && !_pdfioFilePrintf(pdf, "%02X", dataptr[0]))
+            goto bindone;
+
+	  ret = _pdfioFilePuts(pdf, ">");
+
+          bindone:
+
+          if (temp)
+            _pdfioStringFreeBuffer(pdf, (char *)temp);
+
+          return (ret);
         }
 
     case PDFIO_VALTYPE_BOOLEAN :
@@ -609,7 +693,7 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
 	  if (obj && pdf->encryption)
 	  {
 	    // Write encrypted string...
-	    uint8_t	temp[32768],	// Encrypted bytes
+	    uint8_t	temp[64],	// Encrypted bytes
 			*tempptr;	// Pointer into encrypted bytes
 	    _pdfio_crypto_ctx_t ctx;	// Encryption context
 	    _pdfio_crypto_cb_t cb;	// Encryption callback
@@ -660,7 +744,7 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
         if (obj && pdf->encryption)
         {
           // Write encrypted string...
-          uint8_t	temp[32768],	// Encrypted bytes
+          uint8_t	*temp = NULL,	// Encrypted bytes
 			*tempptr;	// Pointer into encrypted bytes
           _pdfio_crypto_ctx_t ctx;	// Encryption context
           _pdfio_crypto_cb_t cb;	// Encryption callback
@@ -668,29 +752,41 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
 					// Length of value
 			ivlen,		// Number of initialization vector bytes
 			tempbytes;	// Number of output bytes
+	  bool		ret = false;	// Return value
 
-          if (len > (sizeof(temp) - 32))
+          if (len > PDFIO_MAX_STRING)
           {
             _pdfioFileError(pdf, "Unable to write encrypted string - too long.");
             return (false);
           }
+          else if ((temp = (uint8_t *)_pdfioStringAllocBuffer(pdf)) == NULL)
+          {
+            _pdfioFileError(pdf, "Unable to write encrypted string - out of memory.");
+            return (false);
+          }
 
           cb        = _pdfioCryptoMakeWriter(pdf, obj, &ctx, temp, &ivlen);
           tempbytes = (cb)(&ctx, temp + ivlen, (const uint8_t *)v->value.string, len) + ivlen;
 
           if (!_pdfioFilePuts(pdf, "<"))
-            return (false);
+            goto strdone;
 
           for (tempptr = temp; tempbytes > 1; tempbytes -= 2, tempptr += 2)
           {
             if (!_pdfioFilePrintf(pdf, "%02X%02X", tempptr[0], tempptr[1]))
-              return (false);
+              goto strdone;
           }
 
-          if (tempbytes > 0)
-            return (_pdfioFilePrintf(pdf, "%02X>", *tempptr));
-          else
-	    return (_pdfioFilePuts(pdf, ">"));
+          if (tempbytes > 0 && !_pdfioFilePrintf(pdf, "%02X", *tempptr))
+            goto strdone;
+
+	  ret = _pdfioFilePuts(pdf, ">");
+
+          strdone :
+
+          _pdfioStringFreeBuffer(pdf, (char *)temp);
+
+          return (ret);
         }
         else
         {
@@ -707,31 +803,59 @@ _pdfioValueWrite(pdfio_file_t   *pdf,	// I - PDF file
 // 'get_date_time()' - Convert PDF date/time value to time_t.
 //
 
-static time_t				// O - Time in seconds
+static time_t				// O - Time in seconds or `0` for none
 get_date_time(const char *s)		// I - PDF date/time value
 {
   int		i;			// Looping var
   struct tm	dateval;		// Date value
-  int		offset;			// Date offset
+  int		offset = 0;		// Date offset in seconds
+  time_t	t;			// Time value
 
 
   PDFIO_DEBUG("get_date_time(s=\"%s\")\n", s);
 
   // Possible date value of the form:
   //
-  //   (D:YYYYMMDDhhmmssZ)
-  //   (D:YYYYMMDDhhmmss+HH'mm)
-  //   (D:YYYYMMDDhhmmss-HH'mm)
+  //   D:YYYYMMDDhhmmssZ
+  //   D:YYYYMMDDhhmmss+HH'mm
+  //   D:YYYYMMDDhhmmss-HH'mm
   //
 
+  if (strncmp(s, "D:", 2))
+    return (0);
+
   for (i = 2; i < 16; i ++)
   {
+    // Look for date/time digits...
     if (!isdigit(s[i] & 255) || !s[i])
       break;
   }
 
-  if (i >= 16)
+  if (i < 6 || (i & 1))
   {
+    // Short year or missing digit...
+    return (0);
+  }
+
+  memset(&dateval, 0, sizeof(dateval));
+
+  dateval.tm_year = (s[2] - '0') * 1000 + (s[3] - '0') * 100 + (s[4] - '0') * 10 + s[5] - '0' - 1900;
+  if (i > 6)
+    dateval.tm_mon  = (s[6] - '0') * 10 + s[7] - '0' - 1;
+  if (i > 8)
+    dateval.tm_mday = (s[8] - '0') * 10 + s[9] - '0';
+  else
+    dateval.tm_mday = 1;
+  if (i > 10)
+    dateval.tm_hour = (s[10] - '0') * 10 + s[11] - '0';
+  if (i > 12)
+    dateval.tm_min  = (s[12] - '0') * 10 + s[13] - '0';
+  if (i > 14)
+    dateval.tm_sec  = (s[14] - '0') * 10 + s[15] - '0';
+
+  if (i >= 16 && s[i])
+  {
+    // Get zone info...
     if (s[i] == 'Z')
     {
       // UTC...
@@ -742,14 +866,20 @@ get_date_time(const char *s)		// I - PDF date/time value
       // Timezone offset from UTC...
       if (isdigit(s[i + 1] & 255) && isdigit(s[i + 2] & 255) && s[i + 3] == '\'' && isdigit(s[i + 4] & 255) && isdigit(s[i + 5] & 255))
       {
+	offset = (s[i + 1] - '0') * 36000 + (s[i + 2] - '0') * 3600 + (s[i + 4] - '0') * 600 + (s[i + 5] - '0') * 60;
+	if (s[i] == '-')
+	  offset = -offset;
+
 	i += 6;
+
+	// Accept trailing quote, per PDF spec...
 	if (s[i] == '\'')
 	  i ++;
       }
     }
-    else if (!s[i])
+    else
     {
-      // Missing zone info, invalid date string...
+      // Random zone info, invalid date string...
       return (0);
     }
   }
@@ -760,26 +890,31 @@ get_date_time(const char *s)		// I - PDF date/time value
     return (0);
   }
 
-  // Date value...
-  memset(&dateval, 0, sizeof(dateval));
+  // Convert date value to time_t...
+#if _WIN32
+  if ((t = _mkgmtime(&dateval)) <= 0)
+    return (0);
 
-  dateval.tm_year = (s[2] - '0') * 1000 + (s[3] - '0') * 100 + (s[4] - '0') * 10 + s[5] - '0' - 1900;
-  dateval.tm_mon  = (s[6] - '0') * 10 + s[7] - '0' - 1;
-  dateval.tm_mday = (s[8] - '0') * 10 + s[9] - '0';
-  dateval.tm_hour = (s[10] - '0') * 10 + s[11] - '0';
-  dateval.tm_min  = (s[12] - '0') * 10 + s[13] - '0';
-  dateval.tm_sec  = (s[14] - '0') * 10 + s[15] - '0';
+#elif defined(HAVE_TIMEGM)
+  if ((t = timegm(&dateval)) <= 0)
+    return (0);
 
-  if (s[16] == 'Z')
-  {
-    offset = 0;
-  }
-  else
-  {
-    offset = (s[17] - '0') * 600 + (s[18] - '0') * 60 + (s[19] - '0') * 10 + s[20] - '0';
-    if (s[16] == '-')
-      offset = -offset;
-  }
+#else
+  if ((t = mktime(&dateval)) <= 0)
+    return (0);
 
-  return (mktime(&dateval) + offset);
+#  if defined(HAVE_TM_GMTOFF)
+  // Adjust the time value using the "tm_gmtoff" and "tm_isdst" members.  As
+  // noted by M-HT on Github, this DST hack will fail in timezones where the
+  // DST offset is not one hour, such as Australia/Lord_Howe.  Fortunately,
+  // this is unusual and most systems support the "timegm" function...
+  t += dateval.tm_gmtoff - 3600 * dateval.tm_isdst;
+#  else
+  // Adjust the time value using the even more legacy "timezone" variable,
+  // which also reflects any DST offset...
+  t += timezone;
+#  endif // HAVE_TM_GMTOFF
+#endif // _WIN32
+
+  return (t - offset);
 }

pdfio.h

@@ -23,7 +23,7 @@ extern "C" {
 // Version numbers...
 //
 
-#  define PDFIO_VERSION		"1.5.1"
+#  define PDFIO_VERSION		"1.5.3"
 #  define PDFIO_VERSION_MAJOR	1
 #  define PDFIO_VERSION_MINOR	5
 
@@ -46,7 +46,7 @@ extern "C" {
 //
 
 #  if _WIN32
-typedef __int64 ssize_t;		// POSIX type not present on Windows... @private@
+typedef __int64 ssize_t;		// POSIX type not present on Windows @private@
 #  endif // _WIN32
 
 typedef struct _pdfio_array_s pdfio_array_t;
@@ -62,7 +62,7 @@ typedef bool (*pdfio_error_cb_t)(pdfio_file_t *pdf, const char *message, void *d
 typedef enum pdfio_encryption_e		// PDF encryption modes
 {
   PDFIO_ENCRYPTION_NONE = 0,		// No encryption
-  PDFIO_ENCRYPTION_RC4_40,		// 40-bit RC4 encryption (PDF 1.3)
+  PDFIO_ENCRYPTION_RC4_40,		// 40-bit RC4 encryption (PDF 1.3, reading only)
   PDFIO_ENCRYPTION_RC4_128,		// 128-bit RC4 encryption (PDF 1.4)
   PDFIO_ENCRYPTION_AES_128,		// 128-bit AES encryption (PDF 1.6)
   PDFIO_ENCRYPTION_AES_256		// 256-bit AES encryption (PDF 2.0) @exclude all@

pdfio_native.nuspec

@@ -3,7 +3,7 @@
     <metadata>
         <id>pdfio_native</id>
         <title>PDFio Library for VS2019+</title>
-        <version>1.5.1</version>
+        <version>1.5.3</version>
         <authors>Michael R Sweet</authors>
         <owners>michaelrsweet</owners>
         <projectUrl>https://github.com/michaelrsweet/pappl</projectUrl>
@@ -16,7 +16,7 @@
         <copyright>Copyright © 2019-2025 by Michael R Sweet</copyright>
         <tags>pdf file native</tags>
 	<dependencies>
-	    <dependency id="pdfio_native.redist" version="1.5.1" />
+	    <dependency id="pdfio_native.redist" version="1.5.3" />
 	    <dependency id="libpng_native.redist" version="1.6.30" />
 	    <dependency id="zlib_native.redist" version="1.2.11" />
 	</dependencies>

pdfio_native.redist.nuspec

@@ -3,7 +3,7 @@
     <metadata>
         <id>pdfio_native.redist</id>
         <title>PDFio Library for VS2019+</title>
-        <version>1.5.1</version>
+        <version>1.5.3</version>
         <authors>Michael R Sweet</authors>
         <owners>michaelrsweet</owners>
         <projectUrl>https://github.com/michaelrsweet/pappl</projectUrl>