Bug #86863 for CAM-PDF: CAM::PDF 1.58 "Expected object open tag"

Thu Jul 11 15:10:12 2013 p-fbsd-bugs [...] ziemba.us - Ticket created

Subject:	CAM::PDF 1.58 "Expected object open tag"
Date:	Thu, 11 Jul 2013 12:09:54 -0700
To:	bug-CAM-PDF [...] rt.cpan.org
From:	"G. Paul Ziemba" <p-fbsd-bugs [...] ziemba.us>

Thanks in advance for any suggestions: code: use CAM::PDF; # 1.58 $pdf = CAM::PDF->new($fn, undef, undef, {fault_tolerant => 1}); output: Expected object open tag 0 ^M xref^M2 15 ^M0000000016 00000 n^M 0000000... PDF file is attached. It was produced on Win2K using ABBYY FineReader (see meta-info in file). I'm running CAM::PDF 1.58 on FreeBSD 9.X. Acroread 8.1.7 and xpdf 3.03 display the file OK. For what it's worth, I tried fiddling with PDF.pm per item 3 of bug 78727 and found that if I added two blocks of testing for \n or \r, I stopped getting the "Expected object open tag" error message, but that the resulting value from $pdf->getPageText text had extra spaces embedded. I'm clearly missing something... sub _buildxref { . . . my $trailer; # BEGIN gpz - added per https://rt.cpan.org/Public/Bug/Display.html?id=78727 my $stuff=substr $self->{content}, $startxref, 1; if($stuff eq "\n" || $stuff eq "\r") { $startxref++; print __LINE__," $self->{pdfversion} - position was off a little\n"; } $stuff=substr $self->{content}, $startxref, 1; if($stuff eq "\n" || $stuff eq "\r") { $startxref++; print __LINE__," $self->{pdfversion} - position was off a little\n"; } # END - added per https://rt.cpan.org/Public/Bug/Display.html?id=78727 -- G. Paul Ziemba FreeBSD unix: 11:51AM up 85 days, 23:09, 1 user, load averages: 1.00, 0.81, 0.81

Download 2013_07_11_11_29_25_OCR.pdf
application/pdf 30.5k

Message body not shown because it is not plain text.

Tue Feb 21 10:56:06 2017 $_ = 'spro^^*%*^6ut# [...] &$%*c>#!^!#&!pan.org'; y/a-z.@//cd; print - Correspondence added

RT-Send-CC:

cdolan [...] cpan.org

On Thu Jul 11 15:10:12 2013, p-fbsd-bugs@ziemba.us wrote: Show quoted text

> Thanks in advance for any suggestions: > > code: > use CAM::PDF; # 1.58 > $pdf = CAM::PDF->new($fn, undef, undef, {fault_tolerant => 1}); > > output: > Expected object open tag > 0 ^M > xref^M2 15 ^M0000000016 00000 n^M > 0000000... > > PDF file is attached. It was produced on Win2K using ABBYY FineReader > (see meta-info in file). I'm running CAM::PDF 1.58 on FreeBSD 9.X. > Acroread 8.1.7 and xpdf 3.03 display the file OK. > > For what it's worth, I tried fiddling with PDF.pm per item 3 of bug > 78727 and found that if I added two blocks of testing for \n or \r, > I stopped getting the "Expected object open tag" error message, but > that the resulting value from $pdf->getPageText text had extra spaces > embedded. I'm clearly missing something... > > sub _buildxref > { > . > . > . > my $trailer; > # BEGIN gpz - added per > https://rt.cpan.org/Public/Bug/Display.html?id=78727 > my $stuff=substr $self->{content}, $startxref, 1; > if($stuff eq "\n" || $stuff eq "\r") > { > $startxref++; > print __LINE__," $self->{pdfversion} - position was off a > little\n"; > } > $stuff=substr $self->{content}, $startxref, 1; > if($stuff eq "\n" || $stuff eq "\r") > { > $startxref++; > print __LINE__," $self->{pdfversion} - position was off a > little\n"; > } > # END - added per > https://rt.cpan.org/Public/Bug/Display.html?id=78727

I don’t know why that did not work for you, but the attached simpler patch did work for me. I am also using ABBYY FineReader.

Subject:

open_M4pDttYw.txt

--- /Library/Perl/5.12/CAM/PDF.pm 2017-02-09 18:01:04.000000000 -0800 +++ CAM/PDF.pm 2017-02-21 07:51:25.000000000 -0800 @@ -572,9 +572,10 @@ my $objstreamrefs = shift; my $trailer; - if ('xref' eq substr $self->{content}, $startxref, 4) + if (substr $self->{content}, $startxref, 50, =~ /^(\s*)xref/) { - $trailer = $self->_buildxref_pdf14($startxref, $index, $versions); + $trailer = $self->_buildxref_pdf14($startxref + length $1, $index, + $versions); if ($trailer && exists $trailer->{XRefStm}) { if (!$self->_buildxref_pdf15($trailer->{XRefStm}->{value}, $index, $versions, $objstreamrefs))

Tue Feb 21 10:56:07 2017 The RT System itself - Status changed from 'new' to 'open'

Sun Feb 26 02:54:10 2017 $_ = 'spro^^*%*^6ut# [...] &$%*c>#!^!#&!pan.org'; y/a-z.@//cd; print - Correspondence added

On Thu Jul 11 15:10:12 2013, p-fbsd-bugs@ziemba.us wrote: Show quoted text

> Thanks in advance for any suggestions: > > code: > use CAM::PDF; # 1.58 > $pdf = CAM::PDF->new($fn, undef, undef, {fault_tolerant => 1}); > > output: > Expected object open tag > 0 ^M > xref^M2 15 ^M0000000016 00000 n^M > 0000000... > > PDF file is attached. It was produced on Win2K using ABBYY FineReader > (see meta-info in file). I'm running CAM::PDF 1.58 on FreeBSD 9.X. > Acroread 8.1.7 and xpdf 3.03 display the file OK.

At the risk of self-promotion, I will point out that reading and dumping the PDF file with PDF::Tiny makes it compatible with CAM::PDF: $ perl -MPDF::Tiny -e 'PDF::Tiny->new("old.pdf")->print(filename => "new.pdf")