1 | ##########################################################################
|
---|
2 | #
|
---|
3 | # jenaTDBBuildproc.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) 1999 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 | # This document processor outputs a document for indexing (should be
|
---|
27 | # implemented by subclass) and storing in the database
|
---|
28 |
|
---|
29 | package jenaTDBBuildproc;
|
---|
30 |
|
---|
31 | use strict;
|
---|
32 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
---|
33 |
|
---|
34 | use docprint;
|
---|
35 | use util;
|
---|
36 | use FileUtils;
|
---|
37 |
|
---|
38 | use extrabuildproc;
|
---|
39 |
|
---|
40 |
|
---|
41 | BEGIN {
|
---|
42 | @jenaTDBBuildproc::ISA = ('extrabuildproc');
|
---|
43 | }
|
---|
44 |
|
---|
45 | sub new()
|
---|
46 | {
|
---|
47 | my $class = shift @_;
|
---|
48 |
|
---|
49 | my $self = new extrabuildproc (@_);
|
---|
50 |
|
---|
51 | # Do the following here so it doesn't keep checking (within the util.pm method)
|
---|
52 | # whether it needs to create the directory or not
|
---|
53 | my $tmp_dir = &util::get_collectlevel_tmp_dir();
|
---|
54 | $self->{'tmp_dir'} = $tmp_dir;
|
---|
55 |
|
---|
56 |
|
---|
57 | my $xslt_file_in = "gsdom2rdf.xsl";
|
---|
58 |
|
---|
59 | my $xslt_filename_in = &util::locate_config_file($xslt_file_in);
|
---|
60 | if (!defined $xslt_filename_in) {
|
---|
61 | print STDERR "Can not find $xslt_file_in, please make sure you have supplied the correct file path\n";
|
---|
62 | die "\n";
|
---|
63 | }
|
---|
64 |
|
---|
65 | my $xslt_filename_out = &FileUtils::filenameConcatenate($tmp_dir,$xslt_file_in);
|
---|
66 |
|
---|
67 | my $collection = $self->{'collection'};
|
---|
68 |
|
---|
69 | my $url_prefix = &util::get_full_greenstone_url_prefix();
|
---|
70 |
|
---|
71 | my $property_hashmap = { 'libraryurl' => $url_prefix,
|
---|
72 | 'collect' => $collection };
|
---|
73 |
|
---|
74 | file_copy_with_property_sub($xslt_filename_in,$xslt_filename_out,$property_hashmap);
|
---|
75 |
|
---|
76 | $self->{'xslt_file'} = $xslt_file_in;
|
---|
77 | $self->{'xslt_filename'} = $xslt_filename_out;
|
---|
78 |
|
---|
79 | return bless $self, $class;
|
---|
80 | }
|
---|
81 |
|
---|
82 |
|
---|
83 | sub property_lookup
|
---|
84 | {
|
---|
85 | my ($hashmap,$value) = @_;
|
---|
86 |
|
---|
87 | print STDERR "*** checking value = '$value'\n";
|
---|
88 |
|
---|
89 | print STDERR "*** lookup = ", $hashmap->{$value}, "\n";
|
---|
90 |
|
---|
91 | my $lookup = (defined $hashmap->{$value}) ? $hashmap->{$value} : "\@$value\@";
|
---|
92 |
|
---|
93 | return $lookup;
|
---|
94 | }
|
---|
95 |
|
---|
96 |
|
---|
97 | # Performs a text file copy, substituding substings of the form
|
---|
98 | # @xxx@ in the input file with the values set in hashmap
|
---|
99 | # passed in
|
---|
100 |
|
---|
101 | sub file_copy_with_property_sub
|
---|
102 | {
|
---|
103 | my ($filename_in,$filename_out,$property_hashmap) = @_;
|
---|
104 |
|
---|
105 | if (!open(FIN, "<$filename_in")) {
|
---|
106 | print STDERR "util::file_substitute_at_properteis failed to open $filename_in\n $!\n";
|
---|
107 | return;
|
---|
108 | }
|
---|
109 | binmode(FIN,":utf8");
|
---|
110 |
|
---|
111 | if (!open(FOUT, ">$filename_out")) {
|
---|
112 | print STDERR "util::file_substitute_at_properteis failed to open $filename_out\n $!\n";
|
---|
113 | return;
|
---|
114 | }
|
---|
115 | binmode(FOUT,":utf8");
|
---|
116 |
|
---|
117 | my $line;
|
---|
118 | while (defined($line = <FIN>)) {
|
---|
119 |
|
---|
120 | $line =~ s/\@([^@ ]+)\@/&property_lookup($property_hashmap,$1)/ige;
|
---|
121 |
|
---|
122 | print FOUT $line;
|
---|
123 | }
|
---|
124 |
|
---|
125 | close(FIN);
|
---|
126 | close(FOUT);
|
---|
127 | }
|
---|
128 |
|
---|
129 |
|
---|
130 | sub open_xslt_pipe
|
---|
131 | {
|
---|
132 | my $self = shift @_;
|
---|
133 | my ($output_file_name, $xslt_file)=@_;
|
---|
134 |
|
---|
135 | return unless defined $xslt_file and $xslt_file ne "" and &FileUtils::fileExists($xslt_file);
|
---|
136 |
|
---|
137 | my $apply_xslt_jar = &FileUtils::javaFilenameConcatenate($ENV{'GSDLHOME'},"bin","java","ApplyXSLT.jar");
|
---|
138 | my $xalan_jar = &FileUtils::javaFilenameConcatenate($ENV{'GSDLHOME'},"bin","java","xalan.jar");
|
---|
139 |
|
---|
140 | my $java_class_path = &util::javapathname_cat($apply_xslt_jar,$xalan_jar);
|
---|
141 |
|
---|
142 | $xslt_file = &util::makeFilenameJavaCygwinCompatible($xslt_file);
|
---|
143 |
|
---|
144 | my $mapping_file_path = "";
|
---|
145 |
|
---|
146 | my $cmd = "| java -cp \"$java_class_path\" org.nzdl.gsdl.ApplyXSLT -t \"$xslt_file\" ";
|
---|
147 |
|
---|
148 |
|
---|
149 | if (defined $self->{'mapping_file'} and $self->{'mapping_file'} ne ""){
|
---|
150 | my $mapping_file_path = "\"".$self->{'mapping_file'}."\"";
|
---|
151 | $cmd .= "-m $mapping_file_path";
|
---|
152 | }
|
---|
153 |
|
---|
154 | open(*XMLWRITER, $cmd)
|
---|
155 | or die "can't open pipe to xslt: $!";
|
---|
156 |
|
---|
157 | $self->{'xslt_writer'} = *XMLWRITER;
|
---|
158 |
|
---|
159 | print XMLWRITER "<?DocStart?>\n";
|
---|
160 | print XMLWRITER "$output_file_name\n";
|
---|
161 |
|
---|
162 | }
|
---|
163 |
|
---|
164 |
|
---|
165 | sub close_xslt_pipe
|
---|
166 | {
|
---|
167 | my $self = shift @_;
|
---|
168 |
|
---|
169 |
|
---|
170 | return unless defined $self->{'xslt_writer'} ;
|
---|
171 |
|
---|
172 | my $xsltwriter = $self->{'xslt_writer'};
|
---|
173 |
|
---|
174 | print $xsltwriter "<?DocEnd?>\n";
|
---|
175 | close($xsltwriter);
|
---|
176 |
|
---|
177 | undef $self->{'xslt_writer'};
|
---|
178 |
|
---|
179 | }
|
---|
180 |
|
---|
181 |
|
---|
182 | sub textedit {
|
---|
183 | my $self = shift (@_);
|
---|
184 | my ($doc_obj) = @_;
|
---|
185 | my $handle = $self->{'output_handle'};
|
---|
186 |
|
---|
187 | my $doc_oid = $doc_obj->get_OID();
|
---|
188 |
|
---|
189 | my $tmp_dir = $self->{'tmp_dir'};
|
---|
190 | my $tmp_doc_filename = &FileUtils::filenameConcatenate($tmp_dir,"doc-$doc_oid.ttl");
|
---|
191 | my $tmp_doc_filename_cc = &util::makeFilenameJavaCygwinCompatible($tmp_doc_filename);
|
---|
192 |
|
---|
193 | my $xslt_filename = $self->{'xslt_filename'};
|
---|
194 | $self->open_xslt_pipe($tmp_doc_filename_cc, $xslt_filename); # stops with error if not able to open pipe
|
---|
195 |
|
---|
196 | my $outhandler = $self->{'xslt_writer'};
|
---|
197 | binmode($outhandler,":utf8");
|
---|
198 |
|
---|
199 | my $section_text = &docprint::get_section_xml($doc_obj,$doc_obj->get_top_section());
|
---|
200 | print $outhandler $section_text;
|
---|
201 |
|
---|
202 | $self->close_xslt_pipe();
|
---|
203 |
|
---|
204 | # now feed to generated file to jena's (TDB) tripple store
|
---|
205 |
|
---|
206 | my $outhandle = $self->{'outhandle'};
|
---|
207 | print $outhandle " Inserting tripples for $doc_oid\n";
|
---|
208 |
|
---|
209 | my $collection = $self->{'collection'};
|
---|
210 |
|
---|
211 | if (-f $tmp_doc_filename) {
|
---|
212 |
|
---|
213 | my $cmd = "gs-triplestore-add $collection \"$tmp_doc_filename\"";
|
---|
214 |
|
---|
215 | my $status = system($cmd);
|
---|
216 | if ($status != 0) {
|
---|
217 | print STDERR "Error: failed to run:\n $cmd\n$!\n";
|
---|
218 | }
|
---|
219 |
|
---|
220 | unlink $tmp_doc_filename;
|
---|
221 | }
|
---|
222 | else {
|
---|
223 | print STDERR "*** Failed to generate: $tmp_doc_filename\n";
|
---|
224 | }
|
---|
225 |
|
---|
226 | }
|
---|
227 |
|
---|
228 |
|
---|
229 | sub text {
|
---|
230 | my $self = shift (@_);
|
---|
231 | my ($doc_obj,$file) = @_;
|
---|
232 |
|
---|
233 | $self->textedit($doc_obj,$file,"add");
|
---|
234 | }
|
---|
235 |
|
---|
236 | sub textreindex
|
---|
237 | {
|
---|
238 | my $self = shift @_;
|
---|
239 | my ($doc_obj,$file) = @_;
|
---|
240 |
|
---|
241 | $self->textedit($doc_obj,$file,"update");
|
---|
242 | }
|
---|
243 |
|
---|
244 | sub textdelete
|
---|
245 | {
|
---|
246 | my $self = shift @_;
|
---|
247 |
|
---|
248 | my ($doc_obj,$file) = @_;
|
---|
249 |
|
---|
250 | print STDERR "Warning: jenaTDB command-line does not currently support delete operation\n";
|
---|
251 |
|
---|
252 | # $self->textedit($doc_obj,$file,"delete");
|
---|
253 | }
|
---|
254 |
|
---|
255 |
|
---|
256 |
|
---|
257 |
|
---|
258 |
|
---|
259 | 1;
|
---|