PDF::API2 fails combining pdf 1.3 and 1.4

[Anonymous Monk]

use strict;
use PDF::API2;
use Carp;

my $pdfile = "out.pdf";
my $pdf = PDF::API2->new( -file => $pdfile) or croak "can't open $pdfi
+le as output";
   #open a blank page
   my $page = $pdf->page;
   $page->mediabox ('A4');
   my %font = (
      Helvetica => {
         Roman  => $pdf->corefont( 'Helvetica',         -encoding => '
+latin1' ),
      },
   );

# put some text
   my $header = $page->text;
   $header->font( $font{'Helvetica'}{'Roman'}, 24 );
   $header->fillcolor('darkblue');
   $header->translate( 50, 750);
   my $ht = "Hello, this is the first page"; 
   $header->text($ht);

# add pages from other .pdf files;
   my $include;
   foreach $include ("1_3.PDF", "1_4.PDF") 
   {
      $page = $pdf->page;
      $page->mediabox ('A4');
      my $include_object = PDF::API2->open($include);
      my $sourcepage = $include_object->openpage(1);
      my $xo = $pdf->importPageIntoForm($include_object, 1);
      my $gfx = $page->gfx;
      $gfx->formimage($xo, 0, 0, 1.0);
   }
$pdf->save; 
$pdf->end();

The above code produces a .pdf file that's perferctly readable by ghostview, but causes an error in Reader XI. I need to produce a .pdf files that collates pages from different sources, inserting text pages that include computed information. The report will be used in a Windows 7 office environment in which avoiding acrobat is not an option.

The problem arises, I believe, when combining .pdf of the 1.3 and 1.4 standard, but the reason could be another. The input files and result are available for download at https://www.dropbox.com/sh/12jofjihcs67ph7/aPnL4y5xZb .

I tried both formimage and importpage, sprinkled some save/restore in the code, the 'pageencaps' option, but it didn't cut it.

My environment is strawberry perl v5.16.2 on Windows 7 32 bits. My version of PDF::API2 is 2.020 MSWin32

Thank you for your help!

[Anonymous Monk]

You could try CAM::PDF ;) appendpdf.pl - Append one PDF to another

[Anonymous Monk]

I could, but I need also to include also jpeg and tiff images. After a quick look, I had the impression that CAM::PDF does not support jpeg and tiff, but I may be in error, of course.

Besides, I am not sure that this is a bug of PDF::API2. It could be an Acrobat's problem or the result of a malformed .pdf file. Unfortunately, I have no control on my source files. Interestingly, I tried to combine the two source .pdf with Acrobat pro 9 and it also fails, but in a different way. I tried also pdftk and found that it behaves in the same way as Acrobat pro.

I may have to convert the .pdf into bitmaps and include them as such, but it seems a pity.

Thanks,

Giovanni

[Anonymous Monk]

I learned that you should not combine pdf files with different pdf versions. They can be converted by reinterpreting them in ghostscript, as below.

gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4
            -dNOPAUSE -dQUIET -dBATCH
            -sOutputFile=clean.pdf original.pdf
# In windows, replace 'gs' with 'gswin32c.exe'

that made the trick for me. It's trivial to call ghostscrip with 'exec' or 'system' from perl.

It wasn't a perl problem and I should not have posted it in perlmonks, probably, but I wanted to share the solution anyway.

Cheers

Giovanni