| 1 | ###########################################################################
|
|---|
| 2 | #
|
|---|
| 3 | # lucenebuildproc.pm -- perl wrapper for building index with Lucene
|
|---|
| 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 | package lucenebuildproc;
|
|---|
| 27 |
|
|---|
| 28 | # This document processor outputs a document
|
|---|
| 29 | # for lucene to process
|
|---|
| 30 |
|
|---|
| 31 | # Use same basic XML structure setup by mgppbuilder/mgppbuildproc
|
|---|
| 32 |
|
|---|
| 33 | use mgppbuildproc;
|
|---|
| 34 | use ghtml;
|
|---|
| 35 |
|
|---|
| 36 | sub BEGIN {
|
|---|
| 37 | @lucenebuildproc::ISA = ('mgppbuildproc');
|
|---|
| 38 | }
|
|---|
| 39 |
|
|---|
| 40 |
|
|---|
| 41 | sub new {
|
|---|
| 42 | my $class = shift @_;
|
|---|
| 43 | my $self = new mgppbuildproc (@_);
|
|---|
| 44 |
|
|---|
| 45 | return bless $self, $class;
|
|---|
| 46 | }
|
|---|
| 47 |
|
|---|
| 48 |
|
|---|
| 49 | sub is_incremental_capable
|
|---|
| 50 | {
|
|---|
| 51 | my $self = shift (@_);
|
|---|
| 52 |
|
|---|
| 53 | # Unlike MG and MGPP, Lucene supports incremental building
|
|---|
| 54 | return 1;
|
|---|
| 55 | }
|
|---|
| 56 |
|
|---|
| 57 |
|
|---|
| 58 | sub preprocess_text {
|
|---|
| 59 | my $self = shift (@_);
|
|---|
| 60 | my ($text, $strip_html, $para) = @_;
|
|---|
| 61 |
|
|---|
| 62 | # call the mgpp method first
|
|---|
| 63 | my ($new_text) = $self->SUPER::preprocess_text($text, $strip_html, $para);
|
|---|
| 64 |
|
|---|
| 65 | # remove entities
|
|---|
| 66 | $new_text =~ s/&\w{1,10};//g;
|
|---|
| 67 | # remove &
|
|---|
| 68 | $new_text =~ s/&//g;
|
|---|
| 69 |
|
|---|
| 70 | return $new_text;
|
|---|
| 71 | }
|
|---|
| 72 |
|
|---|
| 73 |
|
|---|
| 74 | sub text {
|
|---|
| 75 | my $self = shift (@_);
|
|---|
| 76 | my ($doc_obj,$file) = @_;
|
|---|
| 77 | my $handle = $self->{'output_handle'};
|
|---|
| 78 | my $outhandle = $self->{'outhandle'};
|
|---|
| 79 |
|
|---|
| 80 | # only output this document if it is one to be indexed
|
|---|
| 81 | return if ($doc_obj->get_doc_type() ne "indexed_doc");
|
|---|
| 82 |
|
|---|
| 83 | my $indexed_doc = $self->is_subcollection_doc($doc_obj);
|
|---|
| 84 |
|
|---|
| 85 | # this is another document
|
|---|
| 86 | $self->{'num_docs'} += 1;
|
|---|
| 87 |
|
|---|
| 88 | # get the parameters for the output
|
|---|
| 89 | # split on : just in case there is subcoll and lang stuff
|
|---|
| 90 | my ($fields) = split (/:/, $self->{'index'});
|
|---|
| 91 |
|
|---|
| 92 | my $doc_level = $mgppbuildproc::level_map{'document'};
|
|---|
| 93 | my $gs2ns = 'xmlns:gs2="http://www.greenstone.org/gs2"';
|
|---|
| 94 |
|
|---|
| 95 | my $levels = $self->{'levels'};
|
|---|
| 96 | my $ldoc_level = $levels->{'document'};
|
|---|
| 97 | my $lsec_level = $levels->{'section'};
|
|---|
| 98 | my $lpar_level = $levels->{'paragraph'};
|
|---|
| 99 |
|
|---|
| 100 | my $docid="";
|
|---|
| 101 | if ($ldoc_level) {
|
|---|
| 102 | if ($self->{'gdbm_level'} eq 'document') {
|
|---|
| 103 | my $doc_sec_num = $self->{'num_docs'};
|
|---|
| 104 | $docid = "gs2:id=\"$doc_sec_num\"";
|
|---|
| 105 | } else {
|
|---|
| 106 | # default is section level
|
|---|
| 107 | my $doc_sec_num = $self->{'num_sections'}+1;
|
|---|
| 108 | $docid = "gs2:id=\"$doc_sec_num\"";
|
|---|
| 109 | }
|
|---|
| 110 | }
|
|---|
| 111 | my $documenttag = "<$doc_level $gs2ns file=\"$file\" $docid >\n";
|
|---|
| 112 | my $documentendtag = "\n</$doc_level>\n";
|
|---|
| 113 |
|
|---|
| 114 | my ($sectiontag) = "";
|
|---|
| 115 | if ($lsec_level) {
|
|---|
| 116 | $sectiontag = $mgppbuildproc::level_map{'section'};
|
|---|
| 117 | }
|
|---|
| 118 | my ($parastarttag) = "";
|
|---|
| 119 | my ($paraendtag) = "";
|
|---|
| 120 | if ($self->{'levels'}->{'paragraph'}) {
|
|---|
| 121 | if ($self->{'strip_html'}) {
|
|---|
| 122 | $parastarttag = "<".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 123 | $paraendtag = "</".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 124 | } else {
|
|---|
| 125 | print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
|
|---|
| 126 | }
|
|---|
| 127 | }
|
|---|
| 128 |
|
|---|
| 129 | my $doc_section = 0; # just for this document
|
|---|
| 130 |
|
|---|
| 131 | my $text = "";
|
|---|
| 132 | $text .= $documenttag;
|
|---|
| 133 | # get the text for this document
|
|---|
| 134 | my $section = $doc_obj->get_top_section();
|
|---|
| 135 | while (defined $section) {
|
|---|
| 136 | # update a few statistics
|
|---|
| 137 | $doc_section++;
|
|---|
| 138 | $self->{'num_sections'} += 1;
|
|---|
| 139 |
|
|---|
| 140 | if ($sectiontag ne "") {
|
|---|
| 141 | my $secid = "gs2:id=\"".$self->{'num_sections'}."\"";
|
|---|
| 142 | $text .= "\n<$sectiontag $secid >\n";
|
|---|
| 143 | }
|
|---|
| 144 |
|
|---|
| 145 | # if we are doing subcollections, then some docs shouldn't be indexed.
|
|---|
| 146 | # but we need to put the section tag placeholders in there so the
|
|---|
| 147 | # sections match up with gdbm db
|
|---|
| 148 | my $indexed_section = $doc_obj->get_metadata_element($section, "gsdldoctype") || "indexed_section";
|
|---|
| 149 | if (!$indexed_doc || ($indexed_section ne "indexed_section" && $indexed_section ne "indexed_doc")) {
|
|---|
| 150 | $text .= "\n</$sectiontag>\n" if ($sectiontag ne "");
|
|---|
| 151 | $section = $doc_obj->get_next_section($section);
|
|---|
| 152 | next;
|
|---|
| 153 | }
|
|---|
| 154 |
|
|---|
| 155 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
|---|
| 156 | foreach my $field (split (/;/, $fields)) {
|
|---|
| 157 | # only deal with this field if it doesn't start with top or
|
|---|
| 158 | # this is the first section
|
|---|
| 159 | my $real_field = $field;
|
|---|
| 160 | next if (($real_field =~ s/^top//) && ($doc_section != 1));
|
|---|
| 161 |
|
|---|
| 162 | my $new_text = "";
|
|---|
| 163 | my $tmp_text = "";
|
|---|
| 164 |
|
|---|
| 165 | # we get allfields by default - do nothing
|
|---|
| 166 | if ($real_field eq "allfields") {
|
|---|
| 167 |
|
|---|
| 168 | }
|
|---|
| 169 | # metadata - output all metadata we know about except gsdl stuff
|
|---|
| 170 | elsif ($real_field eq "metadata") {
|
|---|
| 171 | my $shortname = "";
|
|---|
| 172 | my $metadata = $doc_obj->get_all_metadata ($section);
|
|---|
| 173 | foreach $pair (@$metadata) {
|
|---|
| 174 | my ($mfield, $mvalue) = (@$pair);
|
|---|
| 175 | # check fields here, maybe others dont want - change to use dontindex!!
|
|---|
| 176 | if ($mfield ne "Identifier"
|
|---|
| 177 | && $mfield !~ /^gsdl/
|
|---|
| 178 | && $mfield ne "classifytype"
|
|---|
| 179 | && $mfield ne "assocfilepath"
|
|---|
| 180 | && defined $mvalue && $mvalue ne "") {
|
|---|
| 181 |
|
|---|
| 182 | if (defined $self->{'indexfieldmap'}->{$mfield}) {
|
|---|
| 183 | $shortname = $self->{'indexfieldmap'}->{$mfield};
|
|---|
| 184 | }
|
|---|
| 185 | else {
|
|---|
| 186 | $shortname = $self->create_shortname($mfield);
|
|---|
| 187 | $self->{'indexfieldmap'}->{$mfield} = $shortname;
|
|---|
| 188 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 189 | }
|
|---|
| 190 | $new_text .= "$parastarttag<$shortname index=\"1\">$mvalue</$shortname>$paraendtag\n";
|
|---|
| 191 | if (!defined $self->{'indexfields'}->{$mfield}) {
|
|---|
| 192 | $self->{'indexfields'}->{$mfield} = 1;
|
|---|
| 193 | }
|
|---|
| 194 | }
|
|---|
| 195 | }
|
|---|
| 196 |
|
|---|
| 197 | }
|
|---|
| 198 | else {
|
|---|
| 199 | #individual metadata and or text specified - could be a comma separated list
|
|---|
| 200 | my $shortname="";
|
|---|
| 201 | if (defined $self->{'indexfieldmap'}->{$real_field}) {
|
|---|
| 202 | $shortname = $self->{'indexfieldmap'}->{$real_field};
|
|---|
| 203 | }
|
|---|
| 204 | else {
|
|---|
| 205 | $shortname = $self->create_shortname($real_field);
|
|---|
| 206 | $self->{'indexfieldmap'}->{$real_field} = $shortname;
|
|---|
| 207 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 208 | }
|
|---|
| 209 |
|
|---|
| 210 | my @metadata_list = ();
|
|---|
| 211 | foreach $submeta (split /,/, $real_field) {
|
|---|
| 212 | if ($submeta eq "text") {
|
|---|
| 213 | if ($self->{'indexing_text'}) { #tag the text with <Text>...</Text>, add the <Paragraph> tags and strip out html if needed
|
|---|
| 214 | $new_text .= "$parastarttag<$shortname index=\"1\">\n";
|
|---|
| 215 | $tmp_text .= $doc_obj->get_text ($section);
|
|---|
| 216 | if ($parastarttag ne "") {
|
|---|
| 217 | $tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "</$shortname>$paraendtag$parastarttag<$shortname index=\"1\">");
|
|---|
| 218 | } else {
|
|---|
| 219 | # we don't want to individually tag each paragraph if not doing para indexing
|
|---|
| 220 | $tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "");
|
|---|
| 221 | }
|
|---|
| 222 | $new_text .= "$tmp_text</$shortname>$paraendtag\n";
|
|---|
| 223 | }
|
|---|
| 224 | else { # leave html stuff in, but escape the tags, and dont add Paragraph tags - never retrieve paras at the moment
|
|---|
| 225 | $tmp_text .= $doc_obj->get_text ($section);
|
|---|
| 226 | &ghtml::htmlsafe($tmp_text);
|
|---|
| 227 | $new_text .= $tmp_text;
|
|---|
| 228 |
|
|---|
| 229 | }
|
|---|
| 230 | }
|
|---|
| 231 | else {
|
|---|
| 232 | my @section_metadata = @{$doc_obj->get_metadata ($section, $submeta)};
|
|---|
| 233 | if ($section ne $doc_obj->get_top_section() && $self->{'indexing_text'} && defined ($self->{'sections_index_document_metadata'})) {
|
|---|
| 234 | if ($self->{'sections_index_document_metadata'} eq "always" || ( scalar(@section_metadata) == 0 && $self->{'sections_index_document_metadata'} eq "unless_section_metadata_exists")) {
|
|---|
| 235 | push (@section_metadata, @{$doc_obj->get_metadata ($doc_obj->get_top_section(), $submeta)});
|
|---|
| 236 | }
|
|---|
| 237 | }
|
|---|
| 238 | push (@metadata_list, @section_metadata);
|
|---|
| 239 | }
|
|---|
| 240 | }
|
|---|
| 241 | foreach my $item (@metadata_list) {
|
|---|
| 242 | $new_text .= "$parastarttag<$shortname index=\"1\">$item</$shortname>$paraendtag\n";
|
|---|
| 243 | }
|
|---|
| 244 |
|
|---|
| 245 | # remove entities
|
|---|
| 246 | $new_text =~ s/&\w{1,10};//g;
|
|---|
| 247 | # remove &
|
|---|
| 248 | $new_text =~ s/&//g;
|
|---|
| 249 | }
|
|---|
| 250 |
|
|---|
| 251 | # filter the text
|
|---|
| 252 | $self->filter_text ($field, $new_text);
|
|---|
| 253 | $self->{'num_processed_bytes'} += length ($new_text);
|
|---|
| 254 | $text .= "$new_text";
|
|---|
| 255 | } # foreach field
|
|---|
| 256 |
|
|---|
| 257 | $text .= "\n</$sectiontag>\n" if ($sectiontag ne "");
|
|---|
| 258 |
|
|---|
| 259 | $section = $doc_obj->get_next_section($section);
|
|---|
| 260 | } #while defined section
|
|---|
| 261 | print $handle "$text\n$documentendtag";
|
|---|
| 262 | #print STDOUT "$text\n$documentendtag";
|
|---|
| 263 | }
|
|---|
| 264 |
|
|---|
| 265 | 1;
|
|---|
| 266 |
|
|---|
