1 | ###########################################################################
|
---|
2 | #
|
---|
3 | # GAPlug.pm
|
---|
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) 2001 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 |
|
---|
26 | # Processes GreenstoneArchive XML documents. Note that this plugin does no
|
---|
27 | # syntax checking (though the XML::Parser module tests for
|
---|
28 | # well-formedness). It's assumed that the GreenstoneArchive files conform
|
---|
29 | # to their DTD.
|
---|
30 |
|
---|
31 | # 12/05/02 Added usage datastructure - John Thompson
|
---|
32 |
|
---|
33 | package GAPlug;
|
---|
34 |
|
---|
35 | use XMLPlug;
|
---|
36 |
|
---|
37 | sub BEGIN {
|
---|
38 | @ISA = ('XMLPlug');
|
---|
39 | }
|
---|
40 |
|
---|
41 | my $options = { 'name' => "GAPlug",
|
---|
42 | 'desc' => "Processes GreenstoneArchive XML documents. Note that this plugin does no syntax checking (though the XML::Parser module tests for well-formedness). It's assumed that the GreenstoneArchive files conform to their DTD.",
|
---|
43 | 'inherits' => "yes" };
|
---|
44 |
|
---|
45 | sub new {
|
---|
46 | my $class = shift (@_);
|
---|
47 | my $self = new XMLPlug ($class, @_);
|
---|
48 |
|
---|
49 | # 14-05-02 To allow for proper inheritance of arguments - John Thompson
|
---|
50 | my $option_list = $self->{'option_list'};
|
---|
51 | push( @{$option_list}, $options );
|
---|
52 |
|
---|
53 | $self->{'section'} = "";
|
---|
54 | $self->{'section_level'} = 0;
|
---|
55 | $self->{'metadata_name'} = "";
|
---|
56 | $self->{'metadata_value'} = "";
|
---|
57 | $self->{'content'} = "";
|
---|
58 |
|
---|
59 | return bless $self, $class;
|
---|
60 | }
|
---|
61 |
|
---|
62 | sub xml_start_document {
|
---|
63 | }
|
---|
64 |
|
---|
65 | sub xml_end_document {
|
---|
66 | }
|
---|
67 |
|
---|
68 | sub xml_doctype {
|
---|
69 | my $self = shift(@_);
|
---|
70 | my ($expat, $name, $sysid, $pubid, $internal) = @_;
|
---|
71 |
|
---|
72 | # allow the short-lived and badly named "GreenstoneArchive" files to be processed
|
---|
73 | # as well as the "Archive" files which should now be created by import.pl
|
---|
74 | die "" if ($name !~ /^(Greenstone)?Archive$/);
|
---|
75 |
|
---|
76 | my $outhandle = $self->{'outhandle'};
|
---|
77 | print $outhandle "GAPLug: processing $self->{'file'}\n";
|
---|
78 | }
|
---|
79 |
|
---|
80 |
|
---|
81 | sub xml_start_tag {
|
---|
82 | my $self = shift(@_);
|
---|
83 | my ($expat, $element) = @_;
|
---|
84 |
|
---|
85 | $self->{'element'} = $element;
|
---|
86 | if ($element eq "Section") {
|
---|
87 | if ($self->{'section_level'} == 0) {
|
---|
88 | $self->open_document();
|
---|
89 | } else {
|
---|
90 | my $doc_obj = $self->{'doc_obj'};
|
---|
91 | $self->{'section'} =
|
---|
92 | $doc_obj->insert_section($doc_obj->get_end_child($self->{'section'}));
|
---|
93 | }
|
---|
94 | $self->{'section_level'} ++;
|
---|
95 | }
|
---|
96 | elsif ($element eq "Metadata") {
|
---|
97 | $self->{'metadata_name'} = $_{'name'};
|
---|
98 | }
|
---|
99 | }
|
---|
100 |
|
---|
101 | sub xml_end_tag {
|
---|
102 | my $self = shift(@_);
|
---|
103 | my ($expat, $element) = @_;
|
---|
104 |
|
---|
105 | if ($element eq "Section") {
|
---|
106 | $self->{'section_level'} --;
|
---|
107 | $self->{'section'} = $self->{'doc_obj'}->get_parent_section ($self->{'section'});
|
---|
108 | $self->close_document() if $self->{'section_level'} == 0;
|
---|
109 | }
|
---|
110 | elsif ($element eq "Metadata") {
|
---|
111 | $self->{'doc_obj'}->add_utf8_metadata($self->{'section'}, $self->{'metadata_name'},
|
---|
112 | $self->{'metadata_value'});
|
---|
113 | $self->{'metadata_name'} = "";
|
---|
114 | $self->{'metadata_value'} = "";
|
---|
115 | }
|
---|
116 | elsif ($element eq "Content" && $self->{'content'} ne "") {
|
---|
117 | $self->{'doc_obj'}->add_utf8_text($self->{'section'}, $self->{'content'});
|
---|
118 | $self->{'content'} = "";
|
---|
119 | }
|
---|
120 |
|
---|
121 | $self->{'element'} = "";
|
---|
122 | }
|
---|
123 |
|
---|
124 | sub xml_text {
|
---|
125 | my $self = shift(@_);
|
---|
126 | my ($expat) = @_;
|
---|
127 |
|
---|
128 | if ($self->{'element'} eq "Metadata") {
|
---|
129 | $self->{'metadata_value'} .= $_;
|
---|
130 | }
|
---|
131 | elsif ($self->{'element'} eq "Content") {
|
---|
132 | $self->{'content'} .= $_;
|
---|
133 | }
|
---|
134 | }
|
---|
135 |
|
---|
136 | sub open_document {
|
---|
137 | my $self = shift(@_);
|
---|
138 |
|
---|
139 | # create a new document
|
---|
140 | $self->{'doc_obj'} = new doc ();
|
---|
141 | $self->{'section'} = "";
|
---|
142 | }
|
---|
143 |
|
---|
144 | sub close_document {
|
---|
145 | my $self = shift(@_);
|
---|
146 |
|
---|
147 | # add the associated files
|
---|
148 | my $assoc_files =
|
---|
149 | $self->{'doc_obj'}->get_metadata($self->{'doc_obj'}->get_top_section(), "gsdlassocfile");
|
---|
150 |
|
---|
151 | # for when "assocfilepath" isn't the same directory that doc.xml is in...
|
---|
152 | my $assoc_filepath_list= $self->{'doc_obj'}->get_metadata($self->{'doc_obj'}->get_top_section(), "assocfilepath");
|
---|
153 |
|
---|
154 | my $assoc_filepath=shift (@$assoc_filepath_list);
|
---|
155 | if (defined ($assoc_filepath)) {
|
---|
156 | # make absolute rather than relative...
|
---|
157 | $self->{'filename'} =~ m@^(.*[\\/]archives)@;
|
---|
158 | $assoc_filepath = "$1/$assoc_filepath/";
|
---|
159 | } else {
|
---|
160 | $assoc_filepath = $self->{'filename'};
|
---|
161 | $assoc_filepath =~ s/[^\\\/]*$//;
|
---|
162 | }
|
---|
163 |
|
---|
164 | foreach my $assoc_file_info (@$assoc_files) {
|
---|
165 | my ($assoc_file, $mime_type, $dir) = split (":", $assoc_file_info);
|
---|
166 | my $real_dir = &util::filename_cat($assoc_filepath, $assoc_file),
|
---|
167 | my $assoc_dir = (defined $dir && $dir ne "")
|
---|
168 | ? &util::filename_cat($dir, $assoc_file) : $assoc_file;
|
---|
169 | $self->{'doc_obj'}->associate_file($real_dir, $assoc_dir, $mime_type);
|
---|
170 | }
|
---|
171 | $self->{'doc_obj'}->delete_metadata($self->{'doc_obj'}->get_top_section(), "gsdlassocfile");
|
---|
172 |
|
---|
173 | # process the document
|
---|
174 | $self->{'processor'}->process($self->{'doc_obj'}, $self->{'file'});
|
---|
175 | }
|
---|
176 |
|
---|
177 |
|
---|
178 | 1;
|
---|
179 |
|
---|
180 |
|
---|