Updating PDF metadata with PDF::API2

I'm trying to run a perl script that uses metatagger to return metadata from a PDF and then use PDF::API2 to actually update the metadata in the PDF. However, I keep getting a seg fault after encountering certain pdfs. I've used eval statements around any of the metatagger or pdf::api2 commands but it still crashes.

If anyone could help, I would greatly appreaciate it.

Thanks.


#!/opt/Interwoven/TeamSite/iw-perl/bin/iwperl
use Data&colon;umper;
use File::Find;
use PDF::API2;
use XML:imple;
use XML::Twig;
my $logFile = "/tmp/pdf-MT-all-files.log";
my $dir = "/pdfdir";


if ( ! open (LOG,">$logFile") ){  
 print STDERR "Failed to write to debug log file: $logFile\\nError=$!";
}
else{
 *STDERR = *LOG;
 *STDOUT = *LOG;
 print LOG "\n\n####\n## START : " . localtime() . " ##\n######\n";   
}

open FILE, $skipList; 
my $skip_list_string = join("", );  

find(\&directory_processing, $dir);

#print LOG "Get files:@files\n"; 

sub directory_processing 
{
 #print LOG "count:$count\n";

 if($File::Find::name =~ m/.pdf$/is) {
  unless($skip_list_string=~m/$File::Find::name/is){
   print LOG "$File::Find::name\n";

   call_metagger($File::Find::name);
  }
 }
}

close(LOG);

sub call_metagger {
 my ($file)= @_; 
 my $xml = "";
 my $papi = "";
 my $found_meta_keyword_data = "";
 my $found_meta_title_data = "";
 my $found_meta_description_data = "";

 my $errFlag=0;
 my $hostname = `hostname`;
 $hostname =~ s/\s//g;
 my @data="";

 #my $mt_cmd = "/opt/Interwoven/TeamSite/bin/iwmtbatch -mthost utlp5055:9095 -suggest $file -nocontext -contextURL http://$hostname:80/iw-bin/iwmtcontext.cgi -dump";

 my $mt_cmd = "/opt/Interwoven/TeamSite/bin/iwmtbatch -mthost utlp5055:9095 -suggest $file -nocontext -contextURL http://$hostname:80/iw-bin/iwmtcontext.cgi -dump";

    print LOG "MT cmd: $mt_cmd\n";
    #print LOG "\nRunning Metatagger update...\n";
 eval{
  open(CMD, $mt_cmd." 2>&1 |");
@data = ;
  close(CMD);
 };
 unless($@){
  print LOG "checking for errors...\n";

  $tempdata = join("",@data); 
  $tempdata =~ s/.*<metadata>(.*)<\/metadata>.*/<metadata>$1<\/metadata>/gs;

  #print LOG "\n\nReturned data&colon;\n$tempdata\n";

  if(@data){
   foreach my $data (@data){   
    if($data =~ /\>ERROR\<\//){
     print LOG "\n\n failed: $data\n";
     $errFlag = 1;
    }
   }
  }else{
         $errFlag = 1;
         print LOG "\n\n failed \n";
  }
 }
 else{
  $errFlag=1;
 }
 if($tempdata && $errFlag != 1){ 

  my $tempMetaTaggerResultsXml = "\n" . $tempdata;
  undef @keywords;
  undef $keywords;
  undef @description;
  undef $description;
  undef @definitions;
  undef $definitions;

  print LOG "extracting data\n";
  my $t = XML::Twig->new();
  eval{
   $t->parse($tempMetaTaggerResultsXml);
   my $root = $t->root;
   my ($nameNode) = $root->get_xpath( '/metadata/attribute/name[string()="Keywords"]');
   my $attributeNode = $nameNode->parent();

   foreach my $valueNode ( $attributeNode->get_xpath( 'value' )){   
    my $tempKeyword=$valueNode->text;
    $tempKeyword=~s/[^a-zA-Z0-9 \t\r\n\v\f\-\\\/]/ /ig;
    push(@keywords, $tempKeyword);
    #print LOG "\nKeywords: " . $tempKeyword;
   }

   my ($nameNode) = $root->get_xpath( '/metadata/attribute/name[string()="Description"]');
   my $attributeNode = $nameNode->parent();
   foreach my $valueNode ( $attributeNode->get_xpath( 'value' )){   
    my $tempDescription=$valueNode->text;
    $tempDescription=~s/[^a-zA-Z0-9 \t\r\n\v\f\-\\\/]/ /ig;
    push(@description, $tempDescription);
    #print LOG "\nDescription: " . $tempDescription;
   }

   my @descriptorNodes = $root->get_xpath( '/metadata/facet/descriptor');

   foreach my $node ( @descriptorNodes){  
    my $labelNode = $node->get_xpath('label', 0);   
    my $tempDescriptor=$labelNode->text;
    $tempDescriptor=~s/[^a-zA-Z0-9 \t\r\n\v\f\-\\\/]/ /ig;
    push(@definitions, $tempDescriptor);
    #print LOG "\nDefinitions: " . $tempDescriptor;

   }

  };

  #print LOG "\nContent: \n" . $tempdata . "\n--End of content--\n";
  unless($@){
   foreach my $value (@description){
    $description .= ", $value";
   }

   my $count = 0;
   foreach my $value (@definitions){
    if($count < 10){
     $keywords .= ", $value";
    }
    $count++;
   }

   my $flag = 0;
   foreach my $value (@keywords){
    $flag = 0;
    foreach my $key (@definitions){
     if($value eq $key){
      $flag = 1;
     }
    }
    if($flag == 0){
     $keywords .= ", $value";
    }
   }


   #debug ("\nkeywords before:\n" . $keywords . "\n--End of keywords--\n");
   $keywords =~ s/\<(.*?)\>/ /g;
   $keywords =~ s/(\&)(.*?)(\/ /g; 
   $keywords =~ s/^\s+//g;
   $keywords =~ s/^\W+//g;


   #debug ("\nDescription before:\n" . $description . "\n--End of Description--\n");
   $description =~ s/(<)(.*?)(>)/ /g;
   $description =~ s/(\&)(.*?)(\/ /g; 
   $description =~ s/^\W+//g;


   chomp($keywords);
   chomp($description);


   $keywords =~ s/(\W)/\\$1/g;
   $description =~ s/(\W)/\\$1/g;





   if($keywords ne "" && $description ne ""){
    eval{update_pdf($keywords, $description, $file)};
    if($@){print LOG "Error calling update pdf";}
   }
  }
  else{
   print LOG "Errror parsing metadata\n";
  }
 }
}  

sub update_pdf {
 my ($found_meta_keyword_data, $found_meta_description_data, $file)= @_; 
 my $xml = "";
 my $pdf = "";
 print LOG "Description:$found_meta_description_data\n";
   print LOG "Keywords:$found_meta_keyword_data\n";

 print LOG "opening pdf\n";








 eval{$pdf=PDF::API2->open($file);}; 
 unless($@){
  eval{$xml=$pdf->xmpMetadata;};
  unless($@){

   #print LOG "PDFs Metadata reads: $xml\n";

   if($found_meta_keyword_data ne ""){
    unless($xml =~ m/.*?<\/pdf:Keywords>/is){
     unless($xml=~s/<\/pdfroducer>/<\/pdfroducer>$found_meta_keyword_data<\/pdf:Keywords>/is){
      $xml=~s/pdfroducer=\'.*?\'/pdfroducer=\'.*?\' pdf:Keywords=\'$found_meta_keyword_data\'/is; 
     }
    }
   }

   if($found_meta_description_data ne ""){
    unless($xml =~ m/.*?<\/dc:description>/is){
     unless($xml=~s/<\/dc:title>/<\/dc:title>$found_meta_description_data<\/rdf:li><\/rdf:Alt><\/dc:description>/is){
      $xml=~s/<\/dc:creator>/<\/dc:creator>$found_meta_description_data<\/rdf:li><\/rdf:Alt><\/dc:description>/is;
     }
    }
   }

   print LOG "about to update \n";
   $pdf->xmpMetadata($xml);
   $pdf->update;

   #print LOG "updated meta data&colon; $xml\n";
   $file=~s/\/iwmnt//i;
   my $lock_cmd = "/opt/Interwoven/TeamSite/bin/iwlock $file 'Metadata updated'";
   eval{my $lock = `$lock_cmd`;};
   if($@){print LOG "$file\t\tUnable to lock file\n\n";}

   print LOG "$file\t\tSuccessfully updated metadata\n\n";
  }
  else{
   print LOG "$file\t\tError in pdf metadata\n\n";
  }
  eval{$pdf->end;};
  if($@){print LOG "$file\t\tUnable to close pdf\n\n";}
 }
 else{
  print LOG "$file\t\tError opening File\n\n";
  print LOG "$@\n";
 } 
}
sub ok { printf STDERR "\n\n%s\n%s\n\n", '='x60, uc( "= @_" ) }

Find more posts tagged with

Comments

nipper

You very likely are on your own. Not too many people have MT loaded in a test environment where they can check this out.

However I would suggest running the commands from the Unix shell and see what happens. The only times I have seg-faulted Perl is too much data (typically endless loop caused).

rschmitt81

As far as I can tell, this fails after the script is unable to open a PDF.

PDF->open($file) returns something like Malformed XREF in PDF File. I would think the eval around this statement would prevent it from crashing anything else.

Migrateduser

Do the pdfs that it's failing on have internal Document Security on them that prevents editing of the pdf? I'm not talking about filesystem permissions, but the security inside the actual document.

Or it could be caused the the version of Adobe (Adobe 9 seems to give some people this error) that created it.

Otherwise, Google gave me these.

http://cpansearch.perl.org/src/LEOCHARRE/PDF-OCR2-1.20/lib/PDF/OCR2.pod