[8520] | 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 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib/cpan");
|
---|
| 34 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib/plugins");
|
---|
| 35 | unshift (@INC, "$ENV{'GSDLHOME'}/perllib/classify");
|
---|
| 36 | }
|
---|
| 37 |
|
---|
| 38 | use util;
|
---|
[10163] | 39 | use ghtml;
|
---|
[8520] | 40 |
|
---|
| 41 |
|
---|
| 42 | sub open_java_lucene
|
---|
| 43 | {
|
---|
[10163] | 44 | my ($doc_tag_level,$full_builddir,$indexdir,$create,$verbosity) = @_;
|
---|
[8520] | 45 |
|
---|
| 46 | my $bin_java = &util::filename_cat($ENV{'GSDLHOME'},"bin","java");
|
---|
| 47 | my $classpath = &util::filename_cat($bin_java,"LuceneWrap.jar");
|
---|
| 48 |
|
---|
[9918] | 49 | my $java_lucene = "java -classpath \"$classpath\" GS2LuceneIndexer";
|
---|
[10163] | 50 | my $cmd_options = "$create -verbosity $verbosity";
|
---|
| 51 | my $java_cmd = "$java_lucene $cmd_options $doc_tag_level \"$full_builddir\" $indexdir";
|
---|
[8520] | 52 |
|
---|
| 53 | if (!open (PIPEOUT, "| $java_cmd")) {
|
---|
| 54 | die "$PROGNAME - couldn't run $java_cmd\n";
|
---|
| 55 | }
|
---|
| 56 | }
|
---|
| 57 |
|
---|
| 58 | sub close_java_lucene
|
---|
| 59 | {
|
---|
| 60 | close(PIPEOUT);
|
---|
| 61 | }
|
---|
| 62 |
|
---|
| 63 | sub save_xml_doc
|
---|
| 64 | {
|
---|
| 65 | my ($full_textdir,$output_filename,$doc_xml) = @_;
|
---|
[9198] | 66 | $dir_sep = &util::get_os_dirsep();
|
---|
[8520] | 67 |
|
---|
| 68 | my $full_output_filename
|
---|
| 69 | = &util::filename_cat($full_textdir,$output_filename);
|
---|
| 70 | my ($full_output_dir) = ($full_output_filename =~ m/^(.*$dir_sep)/x);
|
---|
| 71 | &util::mk_all_dir($full_output_dir);
|
---|
[9198] | 72 |
|
---|
[8520] | 73 | open(DOCOUT,">$full_output_filename")
|
---|
| 74 | || die "Unable to open $full_output_filename";
|
---|
| 75 |
|
---|
| 76 | print DOCOUT $doc_xml;
|
---|
| 77 | close(DOCOUT);
|
---|
[10163] | 78 |
|
---|
| 79 | my @secs = ($doc_xml =~ m/<Sec\s+gs2:id="\d+"\s*>.*?<\/Sec>/sg);
|
---|
| 80 |
|
---|
| 81 |
|
---|
[10165] | 82 | # Currently not used, but consult with DB before removing
|
---|
| 83 | # foreach my $sec (@secs) {
|
---|
| 84 | # my ($docnum,$sec_text) = ($sec =~ m/<Sec\s+gs2:id="(\d+)"\s*>(.*?)<\/Sec>/s);
|
---|
| 85 | # my $docnum_filename
|
---|
| 86 | # = &util::filename_cat($full_textdir,"$docnum.xml");
|
---|
[10163] | 87 |
|
---|
[10165] | 88 | #
|
---|
| 89 | # open(SECOUT,">$docnum_filename")
|
---|
| 90 | # || die "Unable to open $docnum_filename";
|
---|
| 91 |
|
---|
| 92 | # print SECOUT &ghtml::unescape_html($sec_text);
|
---|
| 93 | # close(SECOUT);
|
---|
| 94 | # }
|
---|
| 95 |
|
---|
[8520] | 96 | }
|
---|
| 97 |
|
---|
| 98 | sub compress_xml_doc
|
---|
| 99 | {
|
---|
| 100 | my ($full_textdir,$output_filename) = @_;
|
---|
| 101 |
|
---|
| 102 | my $full_output_filename
|
---|
| 103 | = &util::filename_cat($full_textdir,$output_filename);
|
---|
| 104 |
|
---|
| 105 | `gzip $full_output_filename`;
|
---|
| 106 | }
|
---|
| 107 |
|
---|
| 108 | sub monitor_xml_stream
|
---|
| 109 | {
|
---|
[9177] | 110 | my ($mode, $full_textdir) = @_;
|
---|
[8520] | 111 |
|
---|
| 112 | my $doc_xml = "";
|
---|
| 113 | my $output_filename = "";
|
---|
| 114 |
|
---|
| 115 | my $line;
|
---|
| 116 | while (defined ($line = <STDIN>)) {
|
---|
| 117 | $line =~ s/ /&nbsp;/g;
|
---|
| 118 | $doc_xml .= $line;
|
---|
| 119 | if ($line =~ m/^<Doc.+file=\"(.*?)\".*>$/) {
|
---|
| 120 | $output_filename = $1;
|
---|
| 121 |
|
---|
| 122 | }
|
---|
[9177] | 123 |
|
---|
[8520] | 124 | if ($line =~ m/^<\/Doc>$/) {
|
---|
[9177] | 125 | if ($mode eq "text") {
|
---|
| 126 | save_xml_doc($full_textdir,$output_filename,$doc_xml);
|
---|
| 127 | } elsif ($mode eq "index") {
|
---|
| 128 | # notify lucene indexer
|
---|
| 129 | # print STDERR $doc_xml;
|
---|
[8520] | 130 | ## print PIPEOUT "$output_filename\n";
|
---|
[9177] | 131 | print PIPEOUT "$doc_xml";
|
---|
[9918] | 132 | #save_xml_doc($full_textdir, "$output_filename.txt", $doc_xml);
|
---|
[9177] | 133 | }
|
---|
[8520] | 134 | # compress file
|
---|
| 135 | ### compress_xml_doc($full_textdir,$output_filename);
|
---|
| 136 |
|
---|
| 137 | $doc_xml = "";
|
---|
| 138 | $output_filename = "";
|
---|
| 139 | }
|
---|
| 140 | }
|
---|
| 141 | }
|
---|
| 142 |
|
---|
| 143 | sub main
|
---|
| 144 | {
|
---|
| 145 | my (@argv) = @_;
|
---|
| 146 | my $argc = scalar(@argv);
|
---|
| 147 |
|
---|
[10163] | 148 | my $create = "";
|
---|
| 149 | my $verbosity = 1;
|
---|
| 150 |
|
---|
| 151 | my @filtered_argv = ();
|
---|
| 152 |
|
---|
| 153 | my $i = 0;
|
---|
| 154 | while ($i<$argc) {
|
---|
| 155 | if ($argv[$i] =~ m/^-(.*)$/) {
|
---|
| 156 |
|
---|
| 157 | my $option = $1;
|
---|
| 158 |
|
---|
| 159 | # -create causes build to be incremental
|
---|
| 160 | if ($option eq ("create")) {
|
---|
| 161 | $create = "-create";
|
---|
| 162 | }
|
---|
| 163 |
|
---|
| 164 | # -verbosity num
|
---|
| 165 | elsif ($option eq "verbosity") {
|
---|
| 166 | $i++;
|
---|
| 167 | if ($i<$argc) {
|
---|
| 168 | $verbosity = $argv[$i];
|
---|
| 169 | }
|
---|
| 170 | }
|
---|
| 171 | else {
|
---|
| 172 | print STDERR "Unrecognised minus option: -$option\n";
|
---|
| 173 | }
|
---|
| 174 | }
|
---|
| 175 | else {
|
---|
| 176 | push(@filtered_argv,$argv[$i]);
|
---|
| 177 | }
|
---|
| 178 | $i++;
|
---|
| 179 | }
|
---|
| 180 |
|
---|
| 181 | my $filtered_argc = scalar(@filtered_argv);
|
---|
| 182 |
|
---|
| 183 | if ($filtered_argc < 4) {
|
---|
| 184 | print STDERR "Usage: $PROGNAME [-create|-verbosity num] \"text\"|\"index\" doc-tag-level build-dir index-name\n";
|
---|
[8520] | 185 | exit 1;
|
---|
| 186 | }
|
---|
| 187 |
|
---|
[10163] | 188 | my $mode = $filtered_argv[0];
|
---|
| 189 | my $doc_tag_level = $filtered_argv[1];
|
---|
| 190 | my $full_builddir = $filtered_argv[2];
|
---|
| 191 | my $indexdir = $filtered_argv[3];
|
---|
| 192 | ### print STDERR "**** ARGS = ", join(" ", @argv), "\n";
|
---|
[8520] | 193 |
|
---|
| 194 | my $full_textdir = &util::filename_cat($full_builddir,"text");
|
---|
| 195 |
|
---|
[9177] | 196 | if ($mode eq "index") {
|
---|
| 197 | # don't need the lucene stuff if we are just storing the docs
|
---|
[10163] | 198 | open_java_lucene($doc_tag_level,$full_builddir,$indexdir,$create,$verbosity);
|
---|
[9177] | 199 | }
|
---|
| 200 | monitor_xml_stream($mode, $full_textdir);
|
---|
| 201 | if ($mode eq "index") {
|
---|
| 202 | close_java_lucene();
|
---|
| 203 | }
|
---|
[8520] | 204 | }
|
---|
| 205 |
|
---|
| 206 | $PROGNAME = $0;
|
---|
| 207 | $PROGNAME =~ s/^.*\/(.*)$/$1/;
|
---|
| 208 |
|
---|
| 209 | &main(@ARGV);
|
---|
| 210 |
|
---|