[14927] | 1 | ###########################################################################
|
---|
| 2 | #
|
---|
| 3 | # GreenstoneMETSPlugout.pm -- the plugout module for METS archives
|
---|
| 4 | # A component of the Greenstone digital library software
|
---|
| 5 | # from the New Zealand Digital Library Project at the
|
---|
| 6 | # University of Waikato, New Zealand.
|
---|
| 7 | #
|
---|
| 8 | # Copyright (C) 2006 New Zealand Digital Library Project
|
---|
| 9 | #
|
---|
| 10 | # This program is free software; you can redistribute it and/or modify
|
---|
| 11 | # it under the terms of the GNU General Public License as published by
|
---|
| 12 | # the Free Software Foundation; either version 2 of the License, or
|
---|
| 13 | # (at your option) any later version.
|
---|
| 14 | #
|
---|
| 15 | # This program is distributed in the hope that it will be useful,
|
---|
| 16 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 17 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 18 | # GNU General Public License for more details.
|
---|
| 19 | #
|
---|
| 20 | # You should have received a copy of the GNU General Public License
|
---|
| 21 | # along with this program; if not, write to the Free Software
|
---|
| 22 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
| 23 | #
|
---|
| 24 | ###########################################################################
|
---|
| 25 |
|
---|
[16996] | 26 | package GreenstoneMETSPlugout;
|
---|
[14927] | 27 |
|
---|
| 28 | use strict;
|
---|
| 29 | no strict 'refs';
|
---|
| 30 |
|
---|
| 31 | #eval {require bytes};
|
---|
| 32 | #use util;
|
---|
| 33 | use METSPlugout;
|
---|
| 34 | #use docprint; # for escape_text
|
---|
| 35 |
|
---|
| 36 | sub BEGIN {
|
---|
| 37 | @GreenstoneMETSPlugout::ISA = ('METSPlugout');
|
---|
| 38 | }
|
---|
| 39 |
|
---|
| 40 | my $arguments = [
|
---|
| 41 | ];
|
---|
| 42 |
|
---|
| 43 | my $options = { 'name' => "GreenstoneMETSPlugout",
|
---|
| 44 | 'desc' => "{GreenstoneMETSPlugout.desc}",
|
---|
| 45 | 'abstract' => "no",
|
---|
| 46 | 'inherits' => "yes",
|
---|
| 47 | 'args' => $arguments
|
---|
| 48 | };
|
---|
| 49 |
|
---|
| 50 | sub new {
|
---|
| 51 | my ($class) = shift (@_);
|
---|
| 52 | my ($plugoutlist, $inputargs,$hashArgOptLists) = @_;
|
---|
| 53 | push(@$plugoutlist, $class);
|
---|
| 54 |
|
---|
[17203] | 55 | push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
|
---|
| 56 | push(@{$hashArgOptLists->{"OptList"}},$options);
|
---|
[14927] | 57 |
|
---|
[17203] | 58 | my $self = new METSPlugout($plugoutlist,$inputargs,$hashArgOptLists);
|
---|
| 59 |
|
---|
| 60 | return bless $self, $class;
|
---|
[14927] | 61 | }
|
---|
| 62 |
|
---|
| 63 |
|
---|
| 64 | sub output_mets_xml_header
|
---|
| 65 | {
|
---|
| 66 | my $self = shift(@_);
|
---|
| 67 | my ($handle, $OID, $doc_title) = @_;
|
---|
| 68 |
|
---|
| 69 | my $extra_attr = "OBJID=\"$OID:2\"";
|
---|
| 70 |
|
---|
| 71 | $self->output_mets_xml_header_extra_attribute($handle,$extra_attr);
|
---|
| 72 |
|
---|
| 73 | }
|
---|
| 74 |
|
---|
| 75 | #
|
---|
| 76 | # Print out docmets.xml file
|
---|
| 77 | #
|
---|
| 78 | sub output_mets_section
|
---|
| 79 | {
|
---|
| 80 | my $self = shift(@_);
|
---|
| 81 | my ($handle, $doc_obj, $section, $working_dir) = @_;
|
---|
| 82 |
|
---|
| 83 | # print out the dmdSection
|
---|
| 84 | print $handle $self->buffer_mets_dmdSection_section_xml($doc_obj,$section);
|
---|
| 85 |
|
---|
| 86 | print $handle "<mets:fileSec>\n";
|
---|
| 87 |
|
---|
| 88 | # print out the fileSection by sections
|
---|
| 89 | print $handle $self->buffer_mets_fileSection_section_xml($doc_obj,$section,$working_dir);
|
---|
| 90 |
|
---|
| 91 | # print out the whole fileSection
|
---|
| 92 | print $handle $self->buffer_mets_fileWhole_section_xml($doc_obj,$section,$working_dir);
|
---|
| 93 |
|
---|
| 94 | print $handle "</mets:fileSec>\n";
|
---|
| 95 |
|
---|
| 96 | # print out the StructMapSection by sections
|
---|
| 97 |
|
---|
| 98 | my $struct_type = "Section";
|
---|
| 99 |
|
---|
| 100 |
|
---|
| 101 | # consider making the following its own subroutine
|
---|
| 102 |
|
---|
| 103 | print $handle "<mets:structMap ID=\"Section\" TYPE=\"$struct_type\" LABEL=\"Section\">\n";
|
---|
| 104 | my $order_num=0;
|
---|
| 105 | print $handle $self->buffer_mets_StructMapSection_section_xml($doc_obj,$section, \$order_num);
|
---|
| 106 | print $handle "</mets:structMap>\n";
|
---|
| 107 |
|
---|
| 108 | print $handle '<mets:structMap ID="All" TYPE="Whole Document" LABEL="All">'."\n";
|
---|
| 109 | print $handle $self->buffer_mets_StructMapWhole_section_xml($doc_obj,$section);
|
---|
| 110 | print $handle "</mets:structMap>\n";
|
---|
| 111 |
|
---|
| 112 |
|
---|
| 113 | }
|
---|
| 114 |
|
---|
| 115 | sub buffer_mets_dmdSection_section_xml
|
---|
| 116 | {
|
---|
| 117 | my $self = shift(@_);
|
---|
| 118 | my ($doc_obj,$section) = @_;
|
---|
| 119 |
|
---|
| 120 | $section="" unless defined $section;
|
---|
| 121 |
|
---|
| 122 | my $section_ptr=$doc_obj->_lookup_section($section);
|
---|
| 123 | return "" unless defined $section_ptr;
|
---|
| 124 |
|
---|
| 125 | # convert section number
|
---|
| 126 | my $section_num ="1". $section;
|
---|
| 127 | my $dmd_num = $section_num;
|
---|
| 128 |
|
---|
| 129 | my $all_text = "";
|
---|
| 130 |
|
---|
| 131 | my $label_attr = "";
|
---|
| 132 | # TODO::
|
---|
| 133 | #print STDERR "***** Check that GROUPID in dmdSec is valid!!!\n";
|
---|
| 134 | #print STDERR "***** Check to see if <techMD> required\n";
|
---|
| 135 | # if it isn't allowed, go back and set $mdTag = dmdSec/amdSec
|
---|
| 136 |
|
---|
| 137 | $all_text .= "<mets:dmdSec ID=\"DM$dmd_num\" GROUPID=\"$section_num\">\n";
|
---|
| 138 |
|
---|
| 139 | $all_text .= " <mets:mdWrap $label_attr MDTYPE=\"OTHER\" OTHERMDTYPE=\"gsdl3\" ID=\"gsdl$section_num\">\n";
|
---|
| 140 | $all_text .= " <mets:xmlData>\n";
|
---|
| 141 |
|
---|
| 142 | foreach my $data (@{$section_ptr->{'metadata'}}){
|
---|
| 143 | my $escaped_value = &docprint::escape_text($data->[1]);
|
---|
| 144 | $all_text .= ' <gsdl3:Metadata name="'. $data->[0].'">'. $escaped_value. "</gsdl3:Metadata>\n";
|
---|
| 145 | if ($data->[0] eq "dc.Title") {
|
---|
| 146 | $all_text .= ' <gsdl3:Metadata name="Title">'. $escaped_value."</gsdl3:Metadata>\n";
|
---|
| 147 | }
|
---|
| 148 | }
|
---|
| 149 |
|
---|
| 150 | $all_text .= " </mets:xmlData>\n";
|
---|
| 151 | $all_text .= " </mets:mdWrap>\n";
|
---|
| 152 |
|
---|
| 153 | $all_text .= "</mets:dmdSec>\n";
|
---|
| 154 |
|
---|
| 155 |
|
---|
| 156 | foreach my $subsection (@{$section_ptr->{'subsection_order'}}){
|
---|
| 157 | $all_text .= $self->buffer_mets_dmdSection_section_xml($doc_obj,"$section.$subsection");
|
---|
| 158 | }
|
---|
| 159 |
|
---|
| 160 | $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;
|
---|
| 161 |
|
---|
| 162 | return $all_text;
|
---|
| 163 | }
|
---|
| 164 |
|
---|
| 165 |
|
---|
| 166 |
|
---|
| 167 | sub doctxt_to_xlink
|
---|
| 168 | {
|
---|
| 169 | my $self = shift @_;
|
---|
| 170 | my ($fname,$working_dir) = @_;
|
---|
| 171 |
|
---|
| 172 | my $xlink_href = "file:$fname";
|
---|
| 173 |
|
---|
| 174 | return $xlink_href;
|
---|
| 175 |
|
---|
| 176 | }
|
---|
| 177 |
|
---|
| 178 |
|
---|
| 179 |
|
---|
| 180 | sub buffer_mets_fileSection_section_xml
|
---|
| 181 | {
|
---|
| 182 | my $self = shift(@_);
|
---|
| 183 | my ($doc_obj,$section,$working_dir,$is_recursive) = @_;
|
---|
| 184 |
|
---|
| 185 | my $is_txt_split = undef;
|
---|
| 186 |
|
---|
| 187 | my $all_text
|
---|
| 188 | = $self->SUPER::buffer_mets_fileSection_section_xml($doc_obj,$section,$working_dir,$is_txt_split);
|
---|
| 189 |
|
---|
| 190 | return $all_text;
|
---|
| 191 | }
|
---|
| 192 |
|
---|
| 193 | sub buffer_mets_fileWhole_section_xml
|
---|
| 194 | {
|
---|
| 195 | my $self = shift(@_);
|
---|
| 196 | my ($doc_obj,$section,$working_dir) = @_;
|
---|
| 197 |
|
---|
| 198 | my $section_ptr = $doc_obj-> _lookup_section($section);
|
---|
| 199 | return "" unless defined $section_ptr;
|
---|
| 200 |
|
---|
| 201 | my $all_text="";
|
---|
| 202 |
|
---|
| 203 | my $fileID=0;
|
---|
| 204 |
|
---|
| 205 | # Output the fileSection for the whole section
|
---|
| 206 | # => get the sourcefile and associative file
|
---|
| 207 |
|
---|
| 208 | my $id_root = "default";
|
---|
| 209 | my $opt_owner_id = "";
|
---|
| 210 |
|
---|
| 211 | $all_text .= " <mets:fileGrp ID=\"$id_root\">\n";
|
---|
| 212 |
|
---|
| 213 |
|
---|
| 214 | foreach my $data (@{$section_ptr->{'metadata'}}){
|
---|
| 215 | my $escaped_value = &docprint::escape_text($data->[1]);
|
---|
| 216 |
|
---|
| 217 | if ($data->[0] eq "gsdlsourcefilename") {
|
---|
| 218 | my ($dirPath) = $escaped_value =~ m/^(.*)[\/\\][^\/\\]*$/;
|
---|
| 219 |
|
---|
| 220 | ++$fileID;
|
---|
| 221 | $all_text .= " <mets:file MIMETYPE=\"text/xml\" ID=\"$id_root.$fileID\" $opt_owner_id >\n";
|
---|
| 222 |
|
---|
| 223 | $all_text .= ' <mets:FLocat LOCTYPE="URL" xlink:href="file:'.$data->[1].'" />'."\n";
|
---|
| 224 |
|
---|
| 225 | $all_text .= " </mets:file>\n";
|
---|
| 226 | }
|
---|
| 227 |
|
---|
| 228 | if ($data->[0] eq "gsdlassocfile"){
|
---|
| 229 |
|
---|
| 230 | $escaped_value =~ m/^(.*?):(.*):(.*)$/;
|
---|
| 231 | my $assoc_file = $1;
|
---|
| 232 | my $mime_type = $2;
|
---|
| 233 | my $assoc_dir = $3;
|
---|
| 234 |
|
---|
| 235 | my $assfilePath = ($assoc_dir eq "") ? $assoc_file : "$assoc_dir/$assoc_file";
|
---|
| 236 | ++$fileID;
|
---|
| 237 |
|
---|
| 238 | my $mime_attr = "MIMETYPE=\"$mime_type\"";
|
---|
| 239 | my $xlink_title = "xlink:title=\"$assoc_file\"";
|
---|
| 240 |
|
---|
| 241 | my $id_attr;
|
---|
| 242 | my $xlink_href;
|
---|
| 243 |
|
---|
| 244 | $id_attr = "ID=\"$id_root.$fileID\"";
|
---|
| 245 | $xlink_href = "xlink:href=\"$assfilePath\"";
|
---|
| 246 |
|
---|
| 247 | $all_text .= " <mets:file $mime_attr $id_attr $opt_owner_id >\n";
|
---|
| 248 | $all_text .= " <mets:FLocat LOCTYPE=\"URL\" $xlink_href $xlink_title />\n";
|
---|
| 249 |
|
---|
| 250 | $all_text .= " </mets:file>\n";
|
---|
| 251 |
|
---|
| 252 | }
|
---|
| 253 | }
|
---|
| 254 |
|
---|
| 255 | $all_text .= " </mets:fileGrp>\n";
|
---|
| 256 |
|
---|
| 257 | $all_text =~ s/[\x00-\x09\x0B\x0C\x0E-\x1F]//g;
|
---|
| 258 |
|
---|
| 259 | return $all_text;
|
---|
| 260 | }
|
---|
| 261 |
|
---|
| 262 |
|
---|
| 263 | 1;
|
---|