source: gsdl/trunk/perllib/plugouts/BasePlugout.pm@ 19180

###########################################################################
#
# BasePlugout.pm -- base class for all the plugout modules
# A component of the Greenstone digital library software
# from the New Zealand Digital Library Project at the 
# University of Waikato, New Zealand.
#
# Copyright (C) 2006 New Zealand Digital Library Project
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
###########################################################################

package BasePlugout;

eval {require bytes};

use strict;
no strict 'subs';
no strict 'refs';

use gsprintf 'gsprintf';
use printusage;
use parse2;
use GDBMUtils;


# suppress the annoying "subroutine redefined" warning that various
# gets cause under perl 5.6
$SIG{__WARN__} = sub {warn($_[0]) unless ($_[0] =~ /Subroutine\s+\S+\sredefined/)};

my $arguments = [ 
       { 'name' => "group_size", 
    'desc' => "{BasPlugout.group_size}",
    'type' => "int",
        'deft' =>  "1",
    'reqd' => "no",
    'hiddengli' => "no"},
       { 'name' => "output_info", 
    'desc' => "{BasPlugout.output_info}",
    'type' => "string",   
    'reqd' => "yes",
    'hiddengli' => "yes"},        
       { 'name' => "xslt_file", 
    'desc' => "{BasPlugout.xslt_file}",
    'type' => "string",
    'reqd' => "no",
    'hiddengli' => "no"},
       { 'name' => "output_handle", 
    'desc' => "{BasPlugout.output_handle}",
    'type' => "string",
        'deft' =>  'STDERR',
    'reqd' => "no",
    'hiddengli' => "yes"},
       { 'name' => "verbosity", 
    'desc' => "{BasPlugout.verbosity}",
    'type' => "int",
        'deft' =>  "0",
    'reqd' => "no",  
        'hiddengli' => "no"},
       { 'name' => "gzip_output", 
    'desc' => "{BasPlugout.gzip_output}",
    'type' => "flag",
    'reqd' => "no",  
        'hiddengli' => "no"},
       { 'name' => "debug",
     'desc' => "{BasPlugout.debug}",
     'type' => "flag",
     'reqd' => "no",
     'hiddengli' => "yes"}
];

my $options = { 'name'     => "BasePlugout",
        'desc'     => "{BasPlugout.desc}",
        'abstract' => "yes",
        'inherits' => "no",
        'args'     => $arguments};

sub new 
{
    my $class = shift (@_);

    my ($plugoutlist,$args,$hashArgOptLists) = @_;
    push(@$plugoutlist, $class);

    my $strPlugoutName = (defined $plugoutlist->[0]) ? $plugoutlist->[0] : $class;

    push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
    push(@{$hashArgOptLists->{"OptList"}},$options);

    my $self = {};
    $self->{'plugout_type'} = $class;
    $self->{'option_list'} = $hashArgOptLists->{"OptList"};
    $self->{"info_only"} = 0;

    # Check if gsdlinfo is in the argument list or not - if it is, don't parse 
    # the args, just return the object.  
    foreach my $strArg (@{$args})
    {
    if(defined $strArg && $strArg eq "-gsdlinfo")
    {
        $self->{"info_only"} = 1;
        return bless $self, $class;
    }
    }
    
    delete $self->{"info_only"};
    
    if(parse2::parse($args,$hashArgOptLists->{"ArgList"},$self) == -1)
    {
    my $classTempClass = bless $self, $class;
    print STDERR "<BadPlugout d=$self->{'plugout_name'}>\n";
    &gsprintf(STDERR, "\n{BasPlugout.bad_general_option}\n", $self->{'plugout_name'});
    $classTempClass->print_txt_usage("");  # Use default resource bundle
    die "\n";
    }

 
    if(defined $self->{'xslt_file'} &&  $self->{'xslt_file'} ne "")
    {
      ##$self->{'xslt_file'} =~ s/\"//g;##working on Windows???
    print STDERR "Can not find $self->{'xslt_file'}, please make sure you have supplied the correct file path\n" and  die "\n" unless (-e $self->{'xslt_file'});
    }

    $self->{'gs_count'} = 0;    

    $self->{'keep_import_structure'} = 0; 

    return bless $self, $class;

}

sub print_xml_usage
{
    my $self = shift(@_);
    my $header = shift(@_);
    my $high_level_information_only = shift(@_);
    
    # XML output is always in UTF-8
    gsprintf::output_strings_in_UTF8;

    if ($header) {
    &PrintUsage::print_xml_header("plugout");
    }
    $self->print_xml($high_level_information_only);
}


sub print_xml
{
    my $self = shift(@_);
    my $high_level_information_only = shift(@_);

    my $optionlistref = $self->{'option_list'};
    my @optionlist = @$optionlistref;
    my $plugoutoptions = shift(@$optionlistref);
    return if (!defined($plugoutoptions));

    gsprintf(STDERR, "<PlugoutInfo>\n");
    gsprintf(STDERR, "  <Name>$plugoutoptions->{'name'}</Name>\n");
    my $desc = gsprintf::lookup_string($plugoutoptions->{'desc'});
    $desc =~ s/</&amp;lt;/g; # doubly escaped
    $desc =~ s/>/&amp;gt;/g;
    gsprintf(STDERR, "  <Desc>$desc</Desc>\n");
    gsprintf(STDERR, "  <Abstract>$plugoutoptions->{'abstract'}</Abstract>\n");
    gsprintf(STDERR, "  <Inherits>$plugoutoptions->{'inherits'}</Inherits>\n");
    unless (defined($high_level_information_only)) {
    gsprintf(STDERR, "  <Arguments>\n");
    if (defined($plugoutoptions->{'args'})) {
        &PrintUsage::print_options_xml($plugoutoptions->{'args'});
    }
    gsprintf(STDERR, "  </Arguments>\n");

    # Recurse up the plugout hierarchy
    $self->print_xml();
    }
    gsprintf(STDERR, "</PlugoutInfo>\n");
}


sub print_txt_usage
{
    my $self = shift(@_);

    # Print the usage message for a plugout (recursively)
    my $descoffset = $self->determine_description_offset(0);
    $self->print_plugout_usage($descoffset, 1);
}

sub determine_description_offset
{
    my $self = shift(@_);
    my $maxoffset = shift(@_);

    my $optionlistref = $self->{'option_list'};
    my @optionlist = @$optionlistref;
    my $plugoutoptions = pop(@$optionlistref);
    return $maxoffset if (!defined($plugoutoptions));

    # Find the length of the longest option string of this download
    my $plugoutargs = $plugoutoptions->{'args'};
    if (defined($plugoutargs)) {
    my $longest = &PrintUsage::find_longest_option_string($plugoutargs);
    if ($longest > $maxoffset) {
        $maxoffset = $longest;
    }
    }

    # Recurse up the download hierarchy
    $maxoffset = $self->determine_description_offset($maxoffset);
    $self->{'option_list'} = \@optionlist;
    return $maxoffset;
}


sub print_plugout_usage
{
    my $self = shift(@_);
    my $descoffset = shift(@_);
    my $isleafclass = shift(@_);

    my $optionlistref = $self->{'option_list'};
    my @optionlist = @$optionlistref;
    my $plugoutoptions = shift(@$optionlistref);
    return if (!defined($plugoutoptions));

    my $plugoutname = $plugoutoptions->{'name'};
    my $plugoutargs = $plugoutoptions->{'args'};
    my $plugoutdesc = $plugoutoptions->{'desc'};

    # Produce the usage information using the data structure above
    if ($isleafclass) {
    if (defined($plugoutdesc)) {
        gsprintf(STDERR, "$plugoutdesc\n\n");
    }
    gsprintf(STDERR, " {common.usage}: plugout $plugoutname [{common.options}]\n\n");
    }

    # Display the download options, if there are some
    if (defined($plugoutargs)) {
    # Calculate the column offset of the option descriptions
    my $optiondescoffset = $descoffset + 2;  # 2 spaces between options & descriptions

    if ($isleafclass) {
        gsprintf(STDERR, " {common.specific_options}:\n");
    }
    else {
        gsprintf(STDERR, " {common.general_options}:\n", $plugoutname);
    }

    # Display the download options
    &PrintUsage::print_options_txt($plugoutargs, $optiondescoffset);
    }

    # Recurse up the download hierarchy
    $self->print_plugout_usage($descoffset, 0);
    $self->{'option_list'} = \@optionlist;
}


sub error
{
      my ($strFunctionName,$strError) = @_;
    {
    print "Error occoured in BasePlugout.pm\n".
        "In Function: ".$strFunctionName."\n".
        "Error Message: ".$strError."\n";
    exit(-1);
    }  
}

# OIDtype may be "hash" or "incremental" or "dirname" or "assigned"
sub set_OIDtype {
    my $self = shift (@_);
    my ($type, $metadata) = @_;

    if ($type =~ /^(hash|incremental|dirname|assigned)$/) {
    $self->{'OIDtype'} = $type;
    } else {
    $self->{'OIDtype'} = "hash";
    }
    if ($type =~ /^assigned$/) {
    if (defined $metadata) {
        $self->{'OIDmetadata'} = $metadata;
    } else {
        $self->{'OIDmetadata'} = "dc.Identifier";
    }
    }
}

sub set_output_dir 
{
    my $self = shift @_;
    my ($output_dir) = @_;

   $self->{'output_dir'} = $output_dir;
}

sub setoutputdir 
{
    my $self = shift @_;
    my ($output_dir) = @_;

    $self->{'output_dir'} = $output_dir;
}

sub get_output_dir 
{
    my $self = shift (@_);

   return $self->{'output_dir'};
}

sub getoutputdir 
{
    my $self = shift (@_);

    return $self->{'output_dir'};
}

sub getoutputinfo 
{
    my $self = shift (@_);

    return $self->{'output_info'};
}


sub get_output_handler
{
    my $self = shift (@_);

    my ($output_file_name) = @_;
     
    open(*OUTPUT, ">$output_file_name") or die "Can not open a file handler for  $output_file_name\n";

    return *OUTPUT;            
}

sub release_output_handler
{
    my $self = shift (@_);
    my ($outhandler) = @_;

    close($outhandler);

}

sub output_xml_header {
    my $self = shift (@_);
    my ($handle,$docroot,$nondoctype) = @_;
    
    print $handle '<?xml version="1.0" encoding="UTF-8" standalone="no"?>' . "\n";
   
    if (!defined $nondoctype){
    print $handle '<!DOCTYPE Archive SYSTEM "http://greenstone.org/dtd/Archive/1.0/Archive.dtd">' . "\n";
    }

    print $handle "<$docroot>\n" if defined $docroot;
}

sub output_xml_footer {
    my $self = shift (@_);
    my ($handle,$docroot) = @_;
    print $handle "</$docroot>\n" if defined $docroot;
}

sub process {
    my $self = shift (@_);
    my ($doc_obj) = @_;
    
    $doc_obj->set_lastmodified();

     if ($self->{'group_size'} > 1) {
     $self->group_process ($doc_obj);
    return;
    }

    my $OID = $doc_obj->get_OID();
    $OID = "NULL" unless defined $OID;     

    my $top_section = $doc_obj->get_top_section();

    #get document's directory
    my $doc_dir = $self->get_doc_dir ($OID, $doc_obj->get_source_filename());
    
    my $output_info = $self->{'output_info'};
    return if (!defined $output_info);
     
    ##############################
    # call subclass' saveas method
    ##############################
    $self->saveas($doc_obj,$doc_dir);
    $self->archiveinf_gdbm($doc_obj,$doc_dir); 

}

sub store_output_info_reference {
    my $self = shift (@_);
    my ($doc_obj) = @_;

    my $output_info = $self->{'output_info'};
    my $metaname = $self->{'sortmeta'};
    if (!defined $metaname || $metaname !~ /\S/) {
    $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, "");
    return;
    }
    
    my $metadata = "";
    my $top_section = $doc_obj->get_top_section();
    
    my @commameta_list = split(/,/, $metaname);
    foreach my $cmn (@commameta_list) {
    my $meta = $doc_obj->get_metadata_element($top_section, $cmn);
    if ($meta) {
        # do remove prefix/suffix - this will apply to all values
        $meta =~ s/^$self->{'removeprefix'}// if defined $self->{'removeprefix'};          
        $meta =~ s/$self->{'removesuffix'}$// if defined $self->{'removesuffix'};
        $meta = &sorttools::format_metadata_for_sorting($cmn, $meta, $doc_obj);
        $metadata .= $meta if ($meta);
    }
    }

    # store reference in the output_info     
    $output_info->add_info($doc_obj->get_OID(),$self->{'short_doc_file'}, undef, $metadata);
    
}

sub group_process {

    my $self = shift (@_);
    my ($doc_obj) = @_;
    
    my $OID = $doc_obj->get_OID(); 
    $OID = "NULL" unless defined $OID;

    my $groupsize = $self->{'group_size'};
    my $gs_count = $self->{'gs_count'};
    my $open_new_file = (($gs_count % $groupsize)==0);
    my $outhandle = $self->{'output_handle'};

    # opening a new file, or document has assoicated files => directory needed
    if (($open_new_file) || (scalar(@{$doc_obj->get_assoc_files()})>0)) {
          
        # The directory the archive file (doc.xml) and all associated files
        # should end up in
        my $doc_dir;
        # If we've determined its time for a new file, open it now
        if ($open_new_file || !defined($self->{'gs_doc_dir'}))
          {
            $doc_dir = $self->get_doc_dir ($OID, $doc_obj->get_source_filename());
            # only if opening new file
        my $output_dir = $self->get_output_dir();
        &util::mk_all_dir ($output_dir) unless -e $output_dir;
        my $doc_file = &util::filename_cat ($output_dir, $doc_dir, "doc.xml");
        my $short_doc_file = &util::filename_cat ($doc_dir, "doc.xml");
        
        if ($gs_count>0)
        {
        return if (!$self->close_file_output());
        }

        open (GROUPPROCESS, ">$doc_file") or (print $outhandle "BasePlugout::group_process could not write to file $doc_file\n" and return);
            

        $self->{'gs_filename'} = $doc_file;
        $self->{'short_doc_file'} = $short_doc_file;
        $self->{'gs_OID'} = $OID;
            $self->{'gs_doc_dir'} = $doc_dir;

        $self->output_xml_header('BasePlugout::GROUPPROCESS','Archive');
    }
        # Otherwise load the same archive document directory used last time
        else
          {
            $doc_dir = $self->{'gs_doc_dir'};
          }

    # copy all the associated files, add this information as metadata
    # to the document
        print $outhandle "Writing associated files to $doc_dir\n";
    $self->process_assoc_files ($doc_obj, $doc_dir);
    }

    # save this document
    my $section_text = &docprint::get_section_xml($doc_obj,$doc_obj->get_top_section());
    print GROUPPROCESS $section_text;

    $self->{'gs_count'}++;
}


sub saveas {
    my $self = shift (@_);
   
    die "Basplug::saveas function must be implemented in sub classes\n";
}

sub get_doc_dir {
    my $self = shift (@_);
    my ($OID, $source_filename) = @_;

    my $working_dir  = $self->get_output_dir();
    my $working_info = $self->{output_info}; 
    return if (!defined $working_info);

    my $doc_info = $working_info->get_info($OID);
    my $doc_dir = '';

    if (defined $doc_info && scalar(@$doc_info) >= 1)
    {
    # This OID already has an archives directory, so use it again
    $doc_dir = $doc_info->[0];
    $doc_dir =~ s/\/?((doc(mets)?)|(dublin_core))\.xml(\.gz)?$//;
    }
    elsif ($self->{'keep_import_structure'})
    {
    $source_filename = &File::Basename::dirname($source_filename);
    $source_filename =~ s/[\\\/]+/\//g;
    $source_filename =~ s/\/$//;

        $doc_dir = substr($source_filename, length($ENV{'GSDLIMPORTDIR'}) + 1);
    }

    # We have to use a new archives directory for this document
    if ($doc_dir eq "")
    {
    $doc_dir = $self->get_new_doc_dir ($working_info, $working_dir, $OID);
    }

    if (!defined $self->{'group'} || !$self->{'group'}){
    &util::mk_all_dir (&util::filename_cat ($working_dir, $doc_dir));
    }

    return $doc_dir;
}

sub get_new_doc_dir{
   my $self = shift (@_);  
   my($working_info,$working_dir,$OID) = @_;      
   
   
   my $doc_dir = "";
   my $doc_dir_rest = $OID;
   # remove any \ and / from the OID
   $doc_dir_rest =~ s/[\\\/]//g;
   my $doc_dir_num = 0;

   do {
       $doc_dir .= "/" if $doc_dir_num > 0;
       if ($doc_dir_rest =~ s/^(.{1,8})//) {
       $doc_dir .= $1;
       $doc_dir_num++;
       }
   } while ($doc_dir_rest ne "" && 
        ((-d &util::filename_cat ($working_dir, "$doc_dir.dir")) || 
         ($working_info->size() >= 1024 && $doc_dir_num < 2)));
   my $i = 1;
   my $doc_dir_base = $doc_dir;
   while (-d &util::filename_cat ($working_dir, "$doc_dir.dir")) {
       $doc_dir = "$doc_dir_base-$i";
       $i++;
   }
          
   return "$doc_dir.dir";
}

sub process_assoc_files {
    my $self = shift (@_);
    my ($doc_obj, $doc_dir, $handle) = @_;

    my $outhandle = $self->{'output_handle'};
    
    my $output_dir = $self->get_output_dir();
    return if (!defined $output_dir);

    &util::mk_all_dir ($output_dir) unless -e $output_dir;
      
    my $working_dir = &util::filename_cat($output_dir, $doc_dir);
    &util::mk_all_dir ($working_dir) unless -e $working_dir;

    my @assoc_files = ();
    my $filename;;

    my $source_filename = $doc_obj->get_source_filename();

    my $collect_dir = $ENV{'GSDLCOLLECTDIR'};

    if (defined $collect_dir) {
    my $dirsep_regexp = &util::get_os_dirsep();

    if ($collect_dir !~ /$dirsep_regexp$/) {
        $collect_dir .= &util::get_dirsep(); # ensure there is a slash at the end
    }

    # This test is never going to fail on Windows -- is this a problem?
     
    if ($source_filename !~ /^$dirsep_regexp/) {
        $source_filename = &util::filename_cat($collect_dir, $source_filename);
    }
    }


    # set the assocfile path (even if we have no assoc files - need this for lucene)
    $doc_obj->set_utf8_metadata_element ($doc_obj->get_top_section(),
                     "assocfilepath",
                     "$doc_dir");
    foreach my $assoc_file_rec (@{$doc_obj->get_assoc_files()}) {
    my ($dir, $afile) = $assoc_file_rec->[1] =~ /^(.*?)([^\/\\]+)$/;
    $dir = "" unless defined $dir;
        
    
    my $real_filename = $assoc_file_rec->[0];
    # for some reasons the image associate file has / before the full path
    $real_filename =~ s/^\\(.*)/$1/i;
    if (-e $real_filename) {

        $filename = &util::filename_cat($working_dir, $afile);

        &util::hard_link ($real_filename, $filename, $self->{'verbosity'});
        
        $doc_obj->add_utf8_metadata ($doc_obj->get_top_section(),
                     "gsdlassocfile",
                     "$afile:$assoc_file_rec->[2]:$dir");
    } elsif ($self->{'verbosity'} > 2) {
        print $outhandle "BasePlugout::process couldn't copy the associated file " .
        "$real_filename to $afile\n";
    }
    }
}


sub archiveinf_gdbm
{
    my $self = shift (@_);
    my ($doc_obj) = @_;

    my $verbosity = $self->{'verbosity'};

    my $collect_dir = $ENV{'GSDLCOLLECTDIR'};
    if (defined $collect_dir) {
    my $dirsep_regexp = &util::get_os_dirsep();

    if ($collect_dir !~ /$dirsep_regexp$/) {
        # ensure there is a slash at the end
        $collect_dir .= &util::get_dirsep(); 
    }
    }

    my $oid = $doc_obj->get_OID();
    my $source_filename = $doc_obj->get_source_filename();

    my $working_info = $self->{'output_info'}; 
    my $doc_info = $working_info->get_info($oid);
    my ($doc_file,$index_status) = @$doc_info;

    my $oid_files = { 'doc-file' => $doc_file,
              'index-status' => $index_status,
              'src-file' => $source_filename,
              'assoc-files' => [] };
    
    my %reverse_lookups = ( $source_filename => "1" );

    foreach my $assoc_file_rec (@{$doc_obj->get_assoc_files()}) {
    my $real_filename = $assoc_file_rec->[0];
    my $full_afile = $assoc_file_rec->[1];

    # for some reasons the image associate file has / before the full path
    $real_filename =~ s/^\\(.*)/$1/i;
    if (-e $real_filename) {

        if (defined $collect_dir) {
        my $collect_dir_re_safe = $collect_dir;
        $collect_dir_re_safe =~ s/\\/\\\\/g;
        $collect_dir_re_safe =~ s/\./\\./g;

        $real_filename =~ s/^$collect_dir_re_safe//;
        }

        $reverse_lookups{$real_filename} = 1;

        push(@{$oid_files->{'assoc-files'}},$full_afile);

    }
    else {
        print STDERR "Warning: archiveinf_gdbm()\n  $real_filename does not appear to be on the file system\n";
    }
    }

    # better not to commit to a particular db implementation, but
    # for simplicity, will use GDBM for now.

    my $output_dir = $self->{'output_dir'};

    my $doc_db = &util::filename_cat($output_dir,"archiveinf-doc.gdb");
    my $src_db = &util::filename_cat($output_dir,"archiveinf-src.gdb");

    my $doc_db_text = "";
    $doc_db_text .= "<doc-file>$oid_files->{'doc-file'}\n";
    $doc_db_text .= "<index-status>$oid_files->{'index-status'}\n";
    $doc_db_text .= "<src-file>$oid_files->{'src-file'}\n";
    foreach my $af (@{$oid_files->{'assoc-files'}}) {
    $doc_db_text .= "<assoc-file>$af\n";
    }
    chomp($doc_db_text); # remove trailing \n

    ##print STDERR "*** To set in db: \n\t$doc_db\n\t$oid\n\t$doc_db_text\n";
    &GDBMUtils::gdbmDatabaseSet($doc_db,$oid,$doc_db_text);

    foreach my $rl (keys %reverse_lookups) {
    &GDBMUtils::gdbmDatabaseAppend($src_db,$rl,"<oid>$oid\n");
    }
    
}


sub set_sortmeta {
    my $self = shift (@_);
    my ($sortmeta, $removeprefix, $removesuffix) = @_;
    
    $self->{'sortmeta'} = $sortmeta;
    if (defined ($removeprefix) && $removeprefix ) {
    $removeprefix =~ s/^\^//; # don't need a leading ^
    $self->{'removeprefix'} = $removeprefix;
    }
    if (defined ($removesuffix) && $removesuffix) {
    $removesuffix =~ s/\$$//; # don't need a trailing $
    $self->{'removesuffix'} = $removesuffix;
    }
}

sub open_xslt_pipe
{
    my $self = shift @_;
    my ($output_file_name, $xslt_file)=@_;

   return unless defined $xslt_file and $xslt_file ne "" and -e $xslt_file;

    my $java_class_path =  &util::filename_cat ($ENV{'GSDLHOME'},"bin","java");

    my $mapping_file_path = "";

    if ($ENV{'GSDLOS'} eq "windows"){
    $java_class_path .=";".&util::filename_cat ($ENV{'GSDLHOME'},"bin","java","xalan.jar");
    $xslt_file = "\"file:///".$xslt_file."\"";
    $mapping_file_path = "\"file:///";
    }
    else{
    $java_class_path .=":".&util::filename_cat ($ENV{'GSDLHOME'},"bin","java","xalan.jar");
    }


    $java_class_path = "\"".$java_class_path."\"";

    my $cmd = "| java -cp $java_class_path org.nzdl.gsdl.ApplyXSLT -t $xslt_file "; 

    if (defined $self->{'mapping_file'} and $self->{'mapping_file'} ne ""){
    my $mapping_file_path = "\"".$self->{'mapping_file'}."\""; 
    $cmd .= "-m $mapping_file_path";
    }

    open(*XMLWRITER, $cmd)
    or die "can't open pipe to xslt: $!";

    
    $self->{'xslt_writer'} = *XMLWRITER;

    print XMLWRITER "<?DocStart?>\n";       
    print XMLWRITER "$output_file_name\n";

 
  }
  

sub close_xslt_pipe
{
  my $self = shift @_;

  
  return unless defined $self->{'xslt_writer'} ;
    
  my $xsltwriter = $self->{'xslt_writer'};
  
  print $xsltwriter "<?DocEnd?>\n";
  close($xsltwriter);

  undef $self->{'xslt_writer'};

}

sub close_file_output
{
    my ($self) = @_;
 
    # make sure that the handle has been opened - it won't be if we failed
    # to import any documents...
    if (defined(fileno(GROUPPROCESS))) {
    $self->output_xml_footer('GROUPPROCESS','Archive');    
    close GROUPPROCESS;
    }

    my $OID = $self->{'gs_OID'};
    my $short_doc_file = $self->{'short_doc_file'};
   
    if ($self->{'gzip'}) {
    my $doc_file = $self->{'gs_filename'};
    `gzip $doc_file`;
    $doc_file .= ".gz";
    $short_doc_file .= ".gz";
    if (!-e $doc_file) {
         my $outhandle = $self->{'output_handle'};
        print $outhandle "error while gzipping: $doc_file doesn't exist\n";
        return 0;
    }
    }

    # store reference in output_info
    my $output_info = $self->{'output_info'};
    return 0 if (!defined $output_info);
    $output_info->add_info($OID, $short_doc_file, undef, undef);
    return 1;
}


#the subclass should implement this method if is_group method could return 1.
sub close_group_output{
   my $self = shift (@_);        
}

sub is_group {
    my $self = shift (@_);
    return 0;        
}

my $dc_set = { Title => 1,       
           Creator => 1, 
           Subject => 1, 
           Description => 1, 
           Publisher => 1, 
           Contributor => 1, 
           Date => 1, 
           Type => 1, 
           Format => 1, 
           Identifier => 1, 
           Source => 1, 
           Language => 1, 
           Relation => 1, 
           Coverage => 1, 
           Rights => 1};


# returns an XML representation of the dublin core metadata
# if dc meta is not found, try ex mete
sub get_dc_metadata {
    my $self = shift(@_);
    my ($doc_obj, $section, $version) = @_;
    
    # build up string of dublin core metadata
    $section="" unless defined $section;
    
    my $section_ptr = $doc_obj->_lookup_section($section);
    return "" unless defined $section_ptr;


    my $explicit_dc = {};
    my $explicit_ex = {};

    my $all_text="";
    foreach my $data (@{$section_ptr->{'metadata'}}){
    my $escaped_value = &docprint::escape_text($data->[1]);
    if ($data->[0]=~ m/^dc\./) {
        $data->[0] =~ tr/[A-Z]/[a-z]/;

        $data->[0] =~ m/^dc\.(.*)/;
        my $dc_element =  $1;

        if (!defined $explicit_dc->{$dc_element}) {
        $explicit_dc->{$dc_element} = [];
        }
        push(@{$explicit_dc->{$dc_element}},$escaped_value);

        if (defined $version && ($version eq "oai_dc")) {
        $all_text .= "   <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
        }
        else {
        # qualifier???
        $all_text .= '   <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
        }

    } 
    elsif (($data->[0] =~ m/^ex\./) || ($data->[0] !~ m/\./)) {
        $data->[0] =~ m/^(ex\.)?(.*)/;
        my $ex_element =  $2;
        my $lc_ex_element = lc($ex_element);

        if (defined $dc_set->{$ex_element}) {
        if (!defined $explicit_ex->{$lc_ex_element}) {
            $explicit_ex->{$lc_ex_element} = [];
        }
        push(@{$explicit_ex->{$lc_ex_element}},$escaped_value);
        }
    }
    }

    # go through dc_set and for any element *not* defined in explicit_dc
    # that does exist in explicit_ex, add it in as metadata
    foreach my $k ( keys %$dc_set ) {
    my $lc_k = lc($k);

    if (!defined $explicit_dc->{$lc_k}) {
        if (defined $explicit_ex->{$lc_k}) {

        foreach my $v (@{$explicit_ex->{$lc_k}}) {
            my $dc_element    = $lc_k;
            my $escaped_value = $v;

            if (defined $version && ($version eq "oai_dc")) {
            $all_text .= "   <dc:$dc_element>$escaped_value</dc:$dc_element>\n";
            }
            else {
            $all_text .= '   <dcvalue element="'. $dc_element.'">'. $escaped_value. "</dcvalue>\n";
            }
            
        }
        }
    }
    }

    if ($all_text eq "") {
    $all_text .= "   There is no Dublin Core metatdata in this document\n";
    }   
    $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;

    return $all_text;
}

# Build up dublin_core metadata.  Priority given to dc.* over ex.*
# This method was apparently added by Jeffrey and committed by Shaoqun.
# But we don't know why it was added, so not using it anymore.
sub new_get_dc_metadata {
    
    my $self = shift(@_);
    my ($doc_obj, $section, $version) = @_;

    # build up string of dublin core metadata
    $section="" unless defined $section;
    
    my $section_ptr=$doc_obj->_lookup_section($section);
    return "" unless defined $section_ptr;

    my $all_text = "";
    foreach my $data (@{$section_ptr->{'metadata'}}){
    my $escaped_value = &docprint::escape_text($data->[1]);
    my $dc_element =  $data->[0];
    
    my @array = split('\.',$dc_element);
    my ($type,$name);

    if(defined $array[1])
    {
        $type = $array[0];
        $name = $array[1];
    }
    else
    {
        $type = "ex";
        $name = $array[0];
    }
    
    $all_text .= '   <Metadata Type="'. $type.'" Name="'.$name.'">'. $escaped_value. "</Metadata>\n";
    }
    return $all_text;
}


1;