[24446] | 1 | #!/usr/bin/perl -w
|
---|
| 2 |
|
---|
| 3 | ###########################################################################
|
---|
| 4 | #
|
---|
| 5 | # solr_passes.pl -- perl wrapper, akin to mgpp_passes, for Solr
|
---|
| 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 | # Heavily based on lucene_passes.pl, but does not need a SolrWrapper.jar
|
---|
| 29 | # style solution as Solr has its own XML syntax:
|
---|
| 30 | #
|
---|
| 31 | # http://wiki.apache.org/solr/UpdateXmlMessages
|
---|
| 32 | #
|
---|
| 33 | # This syntax is rather similar to what we already use, so the
|
---|
| 34 | # main task of monitor_xml() is to translate the XML syntax Greenstone uses
|
---|
| 35 | # into that needed by the solr server.
|
---|
| 36 |
|
---|
| 37 |
|
---|
| 38 | BEGIN {
|
---|
| 39 | die "GSDLHOME not set\n" unless defined $ENV{'GSDLHOME'};
|
---|
| 40 | die "GSDLOS not set\n" unless defined $ENV{'GSDLOS'};
|
---|
| 41 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib");
|
---|
| 42 | die "GEXT_SOLR not set\n" unless defined $ENV{'GEXT_SOLR'};
|
---|
[24453] | 43 |
|
---|
| 44 | my $solr_ext = $ENV{'GEXT_SOLR'};
|
---|
| 45 | unshift (@INC, "$solr_ext/perllib");
|
---|
[24446] | 46 | }
|
---|
| 47 |
|
---|
| 48 | use strict;
|
---|
| 49 | use util;
|
---|
[24453] | 50 | use solrutil;
|
---|
| 51 | use solrserver;
|
---|
[24446] | 52 |
|
---|
| 53 | # Not quite OO, but close enough for now
|
---|
| 54 | #
|
---|
[24453] | 55 | my $self = { 'solr_server' => undef };
|
---|
[24446] | 56 |
|
---|
| 57 | sub open_java_solr
|
---|
| 58 | {
|
---|
[24501] | 59 | my ($core,$full_builddir,$indexdir) = @_;
|
---|
[24446] | 60 |
|
---|
[24453] | 61 | # If the Solr/Jetty server is not already running, the following starts
|
---|
| 62 | # it up, and only returns when the server is "reading and listening"
|
---|
[24446] | 63 |
|
---|
[24501] | 64 | my $solr_server = new solrserver($full_builddir);
|
---|
[24453] | 65 | $solr_server->start();
|
---|
| 66 | $self->{'solr_server'} = $solr_server;
|
---|
[24446] | 67 |
|
---|
[24453] | 68 | # Now start up the solr-post command
|
---|
[24501] | 69 | &solrutil::open_post_pipe($core);
|
---|
[24446] | 70 | }
|
---|
| 71 |
|
---|
| 72 | sub close_java_solr
|
---|
| 73 | {
|
---|
[24453] | 74 | &solrutil::close_post_pipe();
|
---|
| 75 |
|
---|
| 76 | my $solr_server = $self->{'solr_server'};
|
---|
| 77 | if ($solr_server->explicitly_started()) {
|
---|
| 78 | $solr_server->stop();
|
---|
[24446] | 79 | }
|
---|
| 80 | }
|
---|
| 81 |
|
---|
[24447] | 82 | #----
|
---|
[24446] | 83 |
|
---|
| 84 | sub save_xml_doc
|
---|
| 85 | {
|
---|
| 86 | # This is identical to the one in lucene_passes.pl, and should be
|
---|
| 87 | # moved in to a package and shared ####
|
---|
| 88 |
|
---|
| 89 | my ($full_textdir,$output_filename,$doc_xml) = @_;
|
---|
| 90 |
|
---|
| 91 | my $dir_sep = &util::get_os_dirsep();
|
---|
| 92 |
|
---|
| 93 | my $full_output_filename = &util::filename_cat($full_textdir,$output_filename);
|
---|
| 94 | my ($full_output_dir) = ($full_output_filename =~ m/^(.*$dir_sep)/x);
|
---|
[27780] | 95 | &FileUtils::makeAllDirectories($full_output_dir);
|
---|
[24446] | 96 |
|
---|
| 97 | open(DOCOUT,">$full_output_filename")
|
---|
| 98 | || die "Unable to open $full_output_filename";
|
---|
| 99 |
|
---|
| 100 | print DOCOUT $doc_xml;
|
---|
| 101 | close(DOCOUT);
|
---|
| 102 |
|
---|
| 103 | # What this the purpose of the following? ####
|
---|
| 104 | my @secs = ($doc_xml =~ m/<Sec\s+gs2:id="\d+"\s*>.*?<\/Sec>/sg);
|
---|
| 105 | }
|
---|
| 106 |
|
---|
| 107 |
|
---|
| 108 | sub compress_xml_doc
|
---|
| 109 | {
|
---|
| 110 | # This is identical to the one in lucene_passes.pl, and should be
|
---|
| 111 | # moved in to a package and shared ####
|
---|
| 112 |
|
---|
| 113 | my ($full_textdir,$output_filename) = @_;
|
---|
| 114 |
|
---|
| 115 | my $full_output_filename
|
---|
| 116 | = &util::filename_cat($full_textdir,$output_filename);
|
---|
| 117 |
|
---|
| 118 | # Greenstone ships with gzip for Windows
|
---|
| 119 | `gzip $full_output_filename`;
|
---|
| 120 | }
|
---|
| 121 |
|
---|
| 122 |
|
---|
| 123 | sub monitor_xml_stream
|
---|
| 124 | {
|
---|
| 125 | # based on lucene's monitor_xml_stream, but simplified
|
---|
| 126 | # as only now used when in "text" mode
|
---|
| 127 |
|
---|
| 128 | my ($full_textdir) = @_;
|
---|
| 129 |
|
---|
| 130 | my $doc_xml = "";
|
---|
| 131 | my $output_filename = "";
|
---|
| 132 |
|
---|
| 133 | my $line;
|
---|
| 134 | while (defined ($line = <STDIN>)) {
|
---|
| 135 |
|
---|
| 136 | $doc_xml .= $line;
|
---|
| 137 |
|
---|
| 138 | if ($line =~ m/^<Doc.+file=\"(.*?)\".*>$/) {
|
---|
| 139 | $output_filename = $1;
|
---|
| 140 | }
|
---|
| 141 |
|
---|
| 142 | if ($line =~ m/^<\/Doc>$/) {
|
---|
| 143 | save_xml_doc($full_textdir,$output_filename,$doc_xml);
|
---|
| 144 |
|
---|
| 145 | # Compress file
|
---|
| 146 | #
|
---|
| 147 | # The compress option was taken out for efficiency
|
---|
| 148 | # reasons. Consider putting it back in but making it a
|
---|
| 149 | # switch so a collection builder can decide for themselves on a
|
---|
| 150 | # case by case basis if they want to save on diskspace, but have
|
---|
| 151 | # the overhead of uncompressing at runtime
|
---|
| 152 |
|
---|
| 153 | ### compress_xml_doc($full_textdir,$output_filename);
|
---|
| 154 |
|
---|
| 155 | $doc_xml = "";
|
---|
| 156 | $output_filename = "";
|
---|
| 157 | }
|
---|
| 158 | }
|
---|
| 159 | }
|
---|
| 160 |
|
---|
| 161 |
|
---|
| 162 | sub pass_on_xml_stream
|
---|
| 163 | {
|
---|
| 164 | my $line;
|
---|
| 165 | while (defined ($line = <STDIN>)) {
|
---|
[24453] | 166 | &solrutil::print_to_post_pipe($line);
|
---|
[24446] | 167 | }
|
---|
| 168 | }
|
---|
| 169 |
|
---|
| 170 |
|
---|
| 171 |
|
---|
| 172 |
|
---|
| 173 | # /** This checks the arguments on the command line, filters the
|
---|
| 174 | # * unknown command line arguments and then calls the open_java_solr
|
---|
| 175 | # * function to begin processing.
|
---|
| 176 | # */
|
---|
| 177 | sub main
|
---|
| 178 | {
|
---|
| 179 | my (@argv) = @_;
|
---|
| 180 | my $argc = scalar(@argv);
|
---|
| 181 |
|
---|
| 182 | my @filtered_argv = ();
|
---|
| 183 |
|
---|
| 184 | my $i = 0;
|
---|
| 185 | while ($i<$argc) {
|
---|
| 186 | if ($argv[$i] =~ m/^\-(.*)$/) {
|
---|
| 187 |
|
---|
| 188 | my $option = $1;
|
---|
| 189 |
|
---|
| 190 | # -verbosity <num>
|
---|
[24501] | 191 | if ($option eq "verbosity") {
|
---|
[24446] | 192 | $i++;
|
---|
| 193 | if ($i<$argc)
|
---|
| 194 | {
|
---|
| 195 | # solr indexing has no support for verbosity
|
---|
| 196 | # => parse to be compatible with calling program, but supress it
|
---|
| 197 | # for solr-post.jar
|
---|
| 198 | }
|
---|
| 199 | }
|
---|
| 200 | else {
|
---|
| 201 | print STDERR "Unrecognised minus option: -$option\n";
|
---|
| 202 | }
|
---|
| 203 | }
|
---|
| 204 | else {
|
---|
| 205 | push(@filtered_argv,$argv[$i]);
|
---|
| 206 | }
|
---|
| 207 | $i++;
|
---|
| 208 | }
|
---|
| 209 |
|
---|
| 210 | my $filtered_argc = scalar(@filtered_argv);
|
---|
| 211 |
|
---|
[24501] | 212 | if ($filtered_argc < 4) {
|
---|
| 213 | print STDERR "Usage: solr_passes.pl [-verbosity num] core \"text\"|\"index\" build-dir index-name\n";
|
---|
[24446] | 214 | exit 1;
|
---|
| 215 | }
|
---|
| 216 |
|
---|
[24501] | 217 | my $core = $filtered_argv[0];
|
---|
[24446] | 218 | my $mode = $filtered_argv[1];
|
---|
[24501] | 219 | my $full_builddir = $filtered_argv[2];
|
---|
| 220 | my $indexdir = $filtered_argv[3];
|
---|
[24446] | 221 |
|
---|
| 222 | # We only need the Solr handle opened if we are indexing the
|
---|
| 223 | # documents, not if we are just storing the text
|
---|
| 224 | if ($mode eq "index") {
|
---|
[24501] | 225 | open_java_solr($core, $full_builddir, $indexdir);
|
---|
[24446] | 226 | }
|
---|
| 227 |
|
---|
| 228 | if ($mode eq "text") {
|
---|
| 229 | print STDERR "Monitoring for input!\n";
|
---|
| 230 | my $full_textdir = &util::filename_cat($full_builddir,"text");
|
---|
| 231 | monitor_xml_stream($full_textdir);
|
---|
| 232 | }
|
---|
| 233 | else {
|
---|
[24447] | 234 | print STDERR "Streaming document input onto Solr server!\n";
|
---|
[24446] | 235 | pass_on_xml_stream();
|
---|
| 236 | }
|
---|
| 237 |
|
---|
| 238 |
|
---|
| 239 | if ($mode eq "index") {
|
---|
| 240 | close_java_solr();
|
---|
| 241 | }
|
---|
| 242 | }
|
---|
| 243 |
|
---|
| 244 |
|
---|
| 245 | &main(@ARGV);
|
---|