source: main/trunk/greenstone2/perllib/plugouts/GreenstoneXMLPlugout.pm@ 32512

Last change on this file since 32512 was 32512, checked in by ak19, 3 years ago

Refactoring some existing perl code as per Dr Bainbridge's ideas on this, so that MySQLPlugout, which inherits from GreenstoneXMLPlugout, can be structured better for reading.

  • Property svn:keywords set to Author Date Id Revision
File size: 7.7 KB
Line 
1###########################################################################
2#
3# GreenstoneXMLPlugout.pm -- the plugout module for Greenstone 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
26package GreenstoneXMLPlugout;
27
28use strict;
29no strict 'refs';
30no strict 'subs';
31
32eval {require bytes};
33use util;
34use FileUtils;
35use BasePlugout;
36use docprint;
37
38sub BEGIN {
39 @GreenstoneXMLPlugout::ISA = ('BasePlugout');
40}
41
42my $arguments = [
43 { 'name' => "group_size",
44 'desc' => "{BasePlugout.group_size}",
45 'type' => "int",
46 'deft' => "1",
47 'reqd' => "no",
48 'hiddengli' => "no"}
49 ];
50my $options = { 'name' => "GreenstoneXMLPlugout",
51 'desc' => "{GreenstoneXMLPlugout.desc}",
52 'abstract' => "no",
53 'inherits' => "yes",
54 'args' => $arguments };
55
56sub new {
57 my ($class) = shift (@_);
58 my ($plugoutlist, $inputargs,$hashArgOptLists) = @_;
59 push(@$plugoutlist, $class);
60
61 push(@{$hashArgOptLists->{"ArgList"}},@{$arguments});
62 push(@{$hashArgOptLists->{"OptList"}},$options);
63
64 my $self = new BasePlugout($plugoutlist,$inputargs,$hashArgOptLists);
65
66 if ($self->{'info_only'}) {
67 # don't worry about any options etc
68 return bless $self, $class;
69 }
70 return bless $self, $class;
71}
72
73sub is_group {
74 my $self = shift (@_);
75 return ($self->{'group_size'} > 1);
76}
77
78sub old_unused_saveas {
79 my $self = shift (@_);
80 my ($doc_obj, $doc_dir) = @_;
81 my $outhandler;
82 my $output_file;
83 if ($self->{'debug'}) {
84 $outhandler = STDOUT;
85 }
86 else {
87
88 $self->process_assoc_files($doc_obj, $doc_dir, '');
89 $self->process_metafiles_metadata ($doc_obj);
90
91 # open up the outhandler
92 if ($self->is_group() && !$self->{'new_doc_dir'}) {
93 # we already have a handle open ??
94 $outhandler = $self->{'group_outhandler'};
95 } else {
96 $output_file = &FileUtils::filenameConcatenate($self->{'output_dir'}, $doc_dir, "doc.xml");
97 # open the new handle
98 $self->open_xslt_pipe($output_file, $self->{'xslt_file'});
99
100 if (defined $self->{'xslt_writer'}){
101 $outhandler = $self->{'xslt_writer'};
102 }
103 else{
104 $outhandler = $self->get_output_handler($output_file);
105 }
106
107 if ($self->is_group()) {
108 $self->{'group_outhandler'} = $outhandler;
109 }
110 }
111 } # else not debug
112 binmode($outhandler,":utf8");
113
114 # only output the header if we have started a new doc
115 if (!$self->is_group() || $self->{'new_doc_dir'}) {
116 $self->output_xml_header($outhandler);
117 }
118
119 my $section_text = &docprint::get_section_xml($doc_obj);
120 print $outhandler $section_text;
121
122 # only output the footer if we are not doing group stuff. The group file will be finished in close_group_output
123 if (!$self->is_group()) {
124 $self->output_xml_footer($outhandler);
125 }
126
127 # close off the output - in a group process situation, this will be done by close_group_output
128 if (!$self->is_group() && !$self->{'debug'}) {
129 if (defined $self->{'xslt_writer'}){
130 $self->close_xslt_pipe();
131 }
132 else {
133 &FileUtils::closeFileHandle($output_file, \$outhandler) if defined $output_file;
134 }
135 }
136 $self->{'short_doc_file'} = &FileUtils::filenameConcatenate($doc_dir, "doc.xml");
137
138 $self->store_output_info_reference($doc_obj);
139
140}
141
142sub pre_saveas {
143 my $self = shift (@_);
144 my ($doc_obj, $doc_dir) = @_;
145 my $outhandler;
146 my $output_file;
147 if ($self->{'debug'}) {
148 $outhandler = STDOUT;
149 }
150 else {
151
152 $self->process_assoc_files($doc_obj, $doc_dir, '');
153 $self->process_metafiles_metadata ($doc_obj);
154
155 # open up the outhandler
156 if ($self->is_group() && !$self->{'new_doc_dir'}) {
157 # we already have a handle open ??
158 $outhandler = $self->{'group_outhandler'};
159 } else {
160 $output_file = &FileUtils::filenameConcatenate($self->{'output_dir'}, $doc_dir, "doc.xml");
161 # open the new handle
162 $self->open_xslt_pipe($output_file, $self->{'xslt_file'});
163
164 if (defined $self->{'xslt_writer'}){
165 $outhandler = $self->{'xslt_writer'};
166 }
167 else{
168 $outhandler = $self->get_output_handler($output_file);
169 }
170
171 if ($self->is_group()) {
172 $self->{'group_outhandler'} = $outhandler;
173 }
174 }
175 } # else not debug
176 binmode($outhandler,":utf8");
177
178 # only output the header if we have started a new doc
179 if (!$self->is_group() || $self->{'new_doc_dir'}) {
180 $self->output_xml_header($outhandler);
181 }
182
183 return ($outhandler, $output_file);
184}
185
186sub saveas {
187 my $self = shift (@_);
188 my ($doc_obj, $doc_dir) = @_;
189
190 # pre
191 my ($outhandler, $output_file) = $self->pre_saveas(@_);
192 push(@_, $outhandler, $output_file);
193
194 # write out the doc xml file for the current document
195 my $section_text = &docprint::get_section_xml($doc_obj);
196 print $outhandler $section_text;
197
198 # post
199 $self->post_saveas(@_);
200}
201
202sub post_saveas {
203 my $self = shift (@_);
204 my ($doc_obj, $doc_dir, $outhandler, $output_file) = @_;
205
206 # only output the footer if we are not doing group stuff. The group file will be finished in close_group_output
207 if (!$self->is_group()) {
208 $self->output_xml_footer($outhandler);
209 }
210
211 # close off the output - in a group process situation, this will be done by close_group_output
212 if (!$self->is_group() && !$self->{'debug'}) {
213 if (defined $self->{'xslt_writer'}){
214 $self->close_xslt_pipe();
215 }
216 else {
217 &FileUtils::closeFileHandle($output_file, \$outhandler) if defined $output_file;
218 }
219 }
220 $self->{'short_doc_file'} = &FileUtils::filenameConcatenate($doc_dir, "doc.xml");
221
222 $self->store_output_info_reference($doc_obj);
223}
224
225sub output_xml_header {
226 my $self = shift (@_);
227 my ($outhandle) = @_;
228
229 print $outhandle '<?xml version="1.0" encoding="utf-8" standalone="no"?>' . "\n";
230 print $outhandle "<!DOCTYPE Archive SYSTEM \"http://greenstone.org/dtd/Archive/1.0/Archive.dtd\">\n";
231 print $outhandle "<Archive>\n";
232}
233
234sub output_xml_footer {
235 my $self = shift (@_);
236 my ($outhandle) = @_;
237
238 print $outhandle "</Archive>\n";
239}
240
241sub close_group_output
242{
243 my $self = shift(@_);
244
245 # make sure that the handle has been opened - it won't be if we failed
246 # to import any documents...
247 my $outhandle = $self->{'group_outhandler'};
248 if (defined(fileno($outhandle))) {
249 $self->output_xml_footer($outhandle);
250 &FileUtils::closeFileHandle("", \$outhandle);
251 undef $self->{'group_outhandler'}
252 }
253
254 my $OID = $self->{'gs_OID'};
255 my $short_doc_file = $self->{'short_doc_file'};
256
257 ### TODO - from here is old code. check that it is still valid.
258 if ($self->{'gzip'}) {
259 my $doc_file = $self->{'gs_filename'};
260 `gzip $doc_file`;
261 $doc_file .= ".gz";
262 $short_doc_file .= ".gz";
263 if (!&FileUtils::fileExists($doc_file)) {
264 my $outhandle = $self->{'output_handle'};
265 print $outhandle "error while gzipping: $doc_file doesn't exist\n";
266 return 0;
267 }
268 }
269
270 return 1;
271}
272
273
2741;
275
Note: See TracBrowser for help on using the repository browser.