1 | #!/usr/bin/perl -w
|
---|
2 |
|
---|
3 | ###########################################################################
|
---|
4 | #
|
---|
5 | # lucene_passes.pl -- perl wrapper, akin to mgpp_passes, for Lucene
|
---|
6 | # A component of the Greenstone digital library software
|
---|
7 | # from the New Zealand Digital Library Project at the
|
---|
8 | # University of Waikato, New Zealand.
|
---|
9 | #
|
---|
10 | # Copyright (C) 1999 New Zealand Digital Library Project
|
---|
11 | #
|
---|
12 | # This program is free software; you can redistribute it and/or modify
|
---|
13 | # it under the terms of the GNU General Public License as published by
|
---|
14 | # the Free Software Foundation; either version 2 of the License, or
|
---|
15 | # (at your option) any later version.
|
---|
16 | #
|
---|
17 | # This program is distributed in the hope that it will be useful,
|
---|
18 | # but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
19 | # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
20 | # GNU General Public License for more details.
|
---|
21 | #
|
---|
22 | # You should have received a copy of the GNU General Public License
|
---|
23 | # along with this program; if not, write to the Free Software
|
---|
24 | # Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
|
---|
25 | #
|
---|
26 | ###########################################################################
|
---|
27 |
|
---|
28 |
|
---|
29 | BEGIN {
|
---|
30 | die "GSDLHOME not set\n" unless defined $ENV{'GSDLHOME'};
|
---|
31 | die "GSDLOS not set\n" unless defined $ENV{'GSDLOS'};
|
---|
32 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib");
|
---|
33 | }
|
---|
34 |
|
---|
35 |
|
---|
36 | use strict;
|
---|
37 | use util;
|
---|
38 |
|
---|
39 |
|
---|
40 | sub open_java_lucene
|
---|
41 | {
|
---|
42 | my ($doc_tag_level,$full_builddir,$indexdir,$java_lucene_options) = @_;
|
---|
43 |
|
---|
44 | # Is there a collection-specific bin/java/LuceneWrapper.jar file?
|
---|
45 | my $bin_java = &util::filename_cat($ENV{'GSDLCOLLECTDIR'},"bin","java");
|
---|
46 | my $classpath = &util::filename_cat($bin_java,"LuceneWrapper.jar");
|
---|
47 | if (!-f $classpath)
|
---|
48 | {
|
---|
49 | # No, so use the Greenstone one
|
---|
50 | $bin_java = &util::filename_cat($ENV{'GSDLHOME'},"bin","java");
|
---|
51 | $classpath = &util::filename_cat($bin_java,"LuceneWrapper.jar");
|
---|
52 | if(!-f $classpath) {
|
---|
53 | die "***** ERROR: $classpath does not exist\n";
|
---|
54 | }
|
---|
55 | }
|
---|
56 |
|
---|
57 | my $java_lucene = "java -classpath \"$classpath\" org.greenstone.LuceneWrapper.GS2LuceneIndexer";
|
---|
58 | my $java_cmd = "$java_lucene $java_lucene_options $doc_tag_level \"$full_builddir\" $indexdir";
|
---|
59 |
|
---|
60 | open (PIPEOUT, "| $java_cmd") or die "lucene_passes.pl - couldn't run $java_cmd\n";
|
---|
61 | }
|
---|
62 |
|
---|
63 |
|
---|
64 | sub close_java_lucene
|
---|
65 | {
|
---|
66 | close(PIPEOUT);
|
---|
67 | }
|
---|
68 |
|
---|
69 |
|
---|
70 | sub save_xml_doc
|
---|
71 | {
|
---|
72 | my ($full_textdir,$output_filename,$doc_xml) = @_;
|
---|
73 |
|
---|
74 | my $dir_sep = &util::get_os_dirsep();
|
---|
75 |
|
---|
76 | my $full_output_filename = &util::filename_cat($full_textdir,$output_filename);
|
---|
77 | my ($full_output_dir) = ($full_output_filename =~ m/^(.*$dir_sep)/x);
|
---|
78 | &util::mk_all_dir($full_output_dir);
|
---|
79 |
|
---|
80 | open(DOCOUT,">$full_output_filename")
|
---|
81 | || die "Unable to open $full_output_filename";
|
---|
82 |
|
---|
83 | print DOCOUT $doc_xml;
|
---|
84 | close(DOCOUT);
|
---|
85 |
|
---|
86 | my @secs = ($doc_xml =~ m/<Sec\s+gs2:id="\d+"\s*>.*?<\/Sec>/sg);
|
---|
87 | }
|
---|
88 |
|
---|
89 |
|
---|
90 | sub compress_xml_doc
|
---|
91 | {
|
---|
92 | my ($full_textdir,$output_filename) = @_;
|
---|
93 |
|
---|
94 | my $full_output_filename
|
---|
95 | = &util::filename_cat($full_textdir,$output_filename);
|
---|
96 |
|
---|
97 | `gzip $full_output_filename`;
|
---|
98 | }
|
---|
99 |
|
---|
100 |
|
---|
101 | # This appears to be the callback that gets the xml stream during the
|
---|
102 | # build process, so I need to intercept it here and call my XML RPC
|
---|
103 | # to insert into the Lucene database.
|
---|
104 | sub monitor_xml_stream
|
---|
105 | {
|
---|
106 | my ($mode, $full_textdir) = @_;
|
---|
107 |
|
---|
108 | my $doc_xml = "";
|
---|
109 | my $output_filename = "";
|
---|
110 |
|
---|
111 | my $line;
|
---|
112 | while (defined ($line = <STDIN>)) {
|
---|
113 | $doc_xml .= $line;
|
---|
114 | if ($line =~ m/^<Doc.+file=\"(.*?)\".*>$/) {
|
---|
115 | $output_filename = $1;
|
---|
116 |
|
---|
117 | }
|
---|
118 |
|
---|
119 | if ($line =~ m/^<\/Doc>$/) {
|
---|
120 | if ($mode eq "text") {
|
---|
121 | save_xml_doc($full_textdir,$output_filename,$doc_xml);
|
---|
122 | } elsif ($mode eq "index") {
|
---|
123 | # notify lucene indexer
|
---|
124 |
|
---|
125 | # SAX parser seems to be sensitive to blank lines
|
---|
126 | # => remove them
|
---|
127 | $doc_xml =~ s/\n+/\n/g;
|
---|
128 |
|
---|
129 | # print STDERR $doc_xml;
|
---|
130 |
|
---|
131 | ## print PIPEOUT "$output_filename\n";
|
---|
132 |
|
---|
133 | print PIPEOUT "$doc_xml";
|
---|
134 |
|
---|
135 |
|
---|
136 | #save_xml_doc($full_textdir, "$output_filename.txt", $doc_xml);
|
---|
137 | }
|
---|
138 | # compress file
|
---|
139 | ### compress_xml_doc($full_textdir,$output_filename);
|
---|
140 |
|
---|
141 | $doc_xml = "";
|
---|
142 | $output_filename = "";
|
---|
143 | }
|
---|
144 | }
|
---|
145 | }
|
---|
146 |
|
---|
147 |
|
---|
148 | # /** This checks the arguments on the command line, filters the
|
---|
149 | # * unknown command line arguments and then calls the open_java_lucene
|
---|
150 | # * function to begin processing. Most of the arguments are passed on
|
---|
151 | # * the command line of the java wrapper.
|
---|
152 | # *
|
---|
153 | # */
|
---|
154 | sub main
|
---|
155 | {
|
---|
156 | my (@argv) = @_;
|
---|
157 | my $argc = scalar(@argv);
|
---|
158 |
|
---|
159 | my $java_lucene_options = "";
|
---|
160 | my @filtered_argv = ();
|
---|
161 |
|
---|
162 | my $i = 0;
|
---|
163 | while ($i<$argc) {
|
---|
164 | if ($argv[$i] =~ m/^\-(.*)$/) {
|
---|
165 |
|
---|
166 | my $option = $1;
|
---|
167 |
|
---|
168 | # -removeold causes the existing index to be overwritten
|
---|
169 | if ($option eq "removeold") {
|
---|
170 | print STDERR "\n-removeold set\n";
|
---|
171 | $java_lucene_options .= "-removeold ";
|
---|
172 | }
|
---|
173 | # -verbosity <num>
|
---|
174 | elsif ($option eq "verbosity") {
|
---|
175 | $i++;
|
---|
176 | if ($i<$argc)
|
---|
177 | {
|
---|
178 | $java_lucene_options .= "-verbosity " . $argv[$i];
|
---|
179 | }
|
---|
180 | }
|
---|
181 | else {
|
---|
182 | print STDERR "Unrecognised minus option: -$option\n";
|
---|
183 | }
|
---|
184 | }
|
---|
185 | else {
|
---|
186 | push(@filtered_argv,$argv[$i]);
|
---|
187 | }
|
---|
188 | $i++;
|
---|
189 | }
|
---|
190 |
|
---|
191 | my $filtered_argc = scalar(@filtered_argv);
|
---|
192 |
|
---|
193 | if ($filtered_argc < 4) {
|
---|
194 | print STDERR "Usage: lucene_passes.pl [-removeold|-verbosity num] \"text\"|\"index\" doc-tag-level build-dir index-name\n";
|
---|
195 | exit 1;
|
---|
196 | }
|
---|
197 |
|
---|
198 | my $mode = $filtered_argv[0];
|
---|
199 | my $doc_tag_level = $filtered_argv[1];
|
---|
200 | my $full_builddir = $filtered_argv[2];
|
---|
201 | my $indexdir = $filtered_argv[3];
|
---|
202 | ### print STDERR "**** ARGS = ", join(" ", @argv), "\n";
|
---|
203 |
|
---|
204 | # We only need the Lucene handle opened if we are indexing the documents, not if we are just storing the text
|
---|
205 | if ($mode eq "index") {
|
---|
206 | open_java_lucene($doc_tag_level, $full_builddir, $indexdir, $java_lucene_options);
|
---|
207 | }
|
---|
208 |
|
---|
209 | print STDERR "Monitoring for input!\n";
|
---|
210 | my $full_textdir = &util::filename_cat($full_builddir,"text");
|
---|
211 | monitor_xml_stream($mode, $full_textdir);
|
---|
212 |
|
---|
213 | if ($mode eq "index") {
|
---|
214 | close_java_lucene();
|
---|
215 | }
|
---|
216 | }
|
---|
217 |
|
---|
218 |
|
---|
219 | &main(@ARGV);
|
---|