| 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 | sub preprocess_text {
|
|---|
| 49 | my $self = shift (@_);
|
|---|
| 50 | my ($text, $strip_html, $para) = @_;
|
|---|
| 51 |
|
|---|
| 52 | my ($outtext) = "";
|
|---|
| 53 | if ($strip_html) {
|
|---|
| 54 | while ($text =~ /<([^>]*)>/ && $text ne "") {
|
|---|
| 55 |
|
|---|
| 56 | my $tag = $1;
|
|---|
| 57 | $outtext .= $`." "; #add everything before the matched tag
|
|---|
| 58 | $text = $'; #'everything after the matched tag
|
|---|
| 59 | if ($para && $tag =~ /^\s*p\s/i) {
|
|---|
| 60 | $outtext .= $para;
|
|---|
| 61 | }
|
|---|
| 62 | elsif ($tag =~ /^pre$/) { # a pre tag
|
|---|
| 63 | $text =~ /<\/pre>/; # find the closing pre tag
|
|---|
| 64 | my $tmp_text = $`; #everything before the closing pre tag
|
|---|
| 65 | $text = $'; #'everything after the </pre>
|
|---|
| 66 | $tmp_text =~ s/[<>]//g; # remove all < and >
|
|---|
| 67 | $outtext.= $tmp_text . " ";
|
|---|
| 68 | }
|
|---|
| 69 | }
|
|---|
| 70 |
|
|---|
| 71 | $outtext .= $text; # add any remaining text
|
|---|
| 72 | } #if strip_html
|
|---|
| 73 | else {
|
|---|
| 74 | $outtext = $text;
|
|---|
| 75 | }
|
|---|
| 76 | #if ($para) {
|
|---|
| 77 | #$text =~ s/(<p\b)/$para$1/gi;
|
|---|
| 78 | #return $text;
|
|---|
| 79 | #}
|
|---|
| 80 |
|
|---|
| 81 | # remove entities
|
|---|
| 82 | $outtext =~ s/&\w{1,10};//g;
|
|---|
| 83 |
|
|---|
| 84 | return $outtext;
|
|---|
| 85 | }
|
|---|
| 86 |
|
|---|
| 87 | sub text {
|
|---|
| 88 | my $self = shift (@_);
|
|---|
| 89 | my ($doc_obj,$file) = @_;
|
|---|
| 90 | my $handle = $self->{'output_handle'};
|
|---|
| 91 | my $outhandle = $self->{'outhandle'};
|
|---|
| 92 | my $indexed_doc = 1;
|
|---|
| 93 |
|
|---|
| 94 | # only output this document if it is one to be indexed
|
|---|
| 95 | return if ($doc_obj->get_doc_type() ne "indexed_doc");
|
|---|
| 96 |
|
|---|
| 97 | # see if this document belongs to this subcollection
|
|---|
| 98 | foreach my $indexexp (@{$self->{'indexexparr'}}) {
|
|---|
| 99 | $indexed_doc = 0;
|
|---|
| 100 | my ($field, $exp, $options) = split /\//, $indexexp;
|
|---|
| 101 | if (defined ($field) && defined ($exp)) {
|
|---|
| 102 | my ($bool) = $field =~ /^(.)/;
|
|---|
| 103 | $field =~ s/^.// if $bool eq '!';
|
|---|
| 104 | if ($field =~ /^filename$/i) {
|
|---|
| 105 | $field = $doc_obj->get_source_filename();
|
|---|
| 106 | } else {
|
|---|
| 107 | $field = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $field);
|
|---|
| 108 | }
|
|---|
| 109 | next unless defined $field;
|
|---|
| 110 | if ($bool eq '!') {
|
|---|
| 111 | if ($options =~ /^i$/i) {
|
|---|
| 112 | if ($field !~ /$exp/i) {$indexed_doc = 1; last;}
|
|---|
| 113 | } else {
|
|---|
| 114 | if ($field !~ /$exp/) {$indexed_doc = 1; last;}
|
|---|
| 115 | }
|
|---|
| 116 | } else {
|
|---|
| 117 | if ($options =~ /^i$/i) {
|
|---|
| 118 | if ($field =~ /$exp/i) {$indexed_doc = 1; last;}
|
|---|
| 119 | } else {
|
|---|
| 120 | if ($field =~ /$exp/) {$indexed_doc = 1; last;}
|
|---|
| 121 | }
|
|---|
| 122 | }
|
|---|
| 123 | }
|
|---|
| 124 | }
|
|---|
| 125 |
|
|---|
| 126 | # if this doc is so far in the sub collection, and we have lang info,
|
|---|
| 127 | # now we check the languages to see if it matches
|
|---|
| 128 | if($indexed_doc && defined $self->{'lang_meta'}) {
|
|---|
| 129 | $indexed_doc = 0;
|
|---|
| 130 | my $field = $doc_obj->get_metadata_element($doc_obj->get_top_section(), $self->{'lang_meta'});
|
|---|
| 131 | if (defined $field) {
|
|---|
| 132 | foreach my $lang (@{$self->{'langarr'}}) {
|
|---|
| 133 | my ($bool) = $lang =~ /^(.)/;
|
|---|
| 134 | if ($bool eq '!') {
|
|---|
| 135 | $lang =~ s/^.//;
|
|---|
| 136 | if ($field !~ /$lang/) {
|
|---|
| 137 | $indexed_doc = 1; last;
|
|---|
| 138 | }
|
|---|
| 139 | } else {
|
|---|
| 140 | if ($field =~ /$lang/) {
|
|---|
| 141 | $indexed_doc = 1; last;
|
|---|
| 142 | }
|
|---|
| 143 | }
|
|---|
| 144 | }
|
|---|
| 145 | }
|
|---|
| 146 | }
|
|---|
| 147 |
|
|---|
| 148 | # this is another document
|
|---|
| 149 | $self->{'num_docs'} += 1;
|
|---|
| 150 |
|
|---|
| 151 | # get the parameters for the output
|
|---|
| 152 | # split on : just in case there is subcoll and lang stuff
|
|---|
| 153 | my ($fields) = split (/:/, $self->{'index'});
|
|---|
| 154 |
|
|---|
| 155 | my $doc_level = $mgppbuildproc::level_map{'document'};
|
|---|
| 156 | my $gs2ns = 'xmlns:gs2="http://www.greenstone.org/gs2"';
|
|---|
| 157 |
|
|---|
| 158 | my $levels = $self->{'levels'};
|
|---|
| 159 | my $ldoc_level = $levels->{'document'};
|
|---|
| 160 | my $lsec_level = $levels->{'section'};
|
|---|
| 161 | my $lpar_level = $levels->{'paragraph'};
|
|---|
| 162 |
|
|---|
| 163 | my $docid="";
|
|---|
| 164 | if ($ldoc_level) {
|
|---|
| 165 | if ($self->{'gdbm_level'} eq 'document') {
|
|---|
| 166 | my $doc_sec_num = $self->{'num_docs'};
|
|---|
| 167 | $docid = "gs2:id=\"$doc_sec_num\"";
|
|---|
| 168 | } else {
|
|---|
| 169 | # default is section level
|
|---|
| 170 | my $doc_sec_num = $self->{'num_sections'}+1;
|
|---|
| 171 | $docid = "gs2:id=\"$doc_sec_num\"";
|
|---|
| 172 | }
|
|---|
| 173 | }
|
|---|
| 174 | my $documenttag = "<$doc_level $gs2ns file=\"$file\" $docid >\n";
|
|---|
| 175 | my $documentendtag = "\n</$doc_level>\n";
|
|---|
| 176 |
|
|---|
| 177 | my ($sectiontag) = "";
|
|---|
| 178 | if ($lsec_level) {
|
|---|
| 179 | $sectiontag = $mgppbuildproc::level_map{'section'};
|
|---|
| 180 | }
|
|---|
| 181 | my ($parastarttag) = "";
|
|---|
| 182 | my ($paraendtag) = "";
|
|---|
| 183 | if ($self->{'levels'}->{'paragraph'}) {
|
|---|
| 184 | if ($self->{'strip_html'}) {
|
|---|
| 185 | $parastarttag = "<".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 186 | $paraendtag = "</".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 187 | } else {
|
|---|
| 188 | print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
|
|---|
| 189 | }
|
|---|
| 190 | }
|
|---|
| 191 |
|
|---|
| 192 | my $doc_section = 0; # just for this document
|
|---|
| 193 |
|
|---|
| 194 | my $text = "";
|
|---|
| 195 | $text .= $documenttag;
|
|---|
| 196 | # get the text for this document
|
|---|
| 197 | my $section = $doc_obj->get_top_section();
|
|---|
| 198 | while (defined $section) {
|
|---|
| 199 | # update a few statistics
|
|---|
| 200 | $doc_section++;
|
|---|
| 201 | $self->{'num_sections'} += 1;
|
|---|
| 202 |
|
|---|
| 203 | if ($sectiontag ne "") {
|
|---|
| 204 | my $secid = "gs2:id=\"".$self->{'num_sections'}."\"";
|
|---|
| 205 | $text .= "\n<$sectiontag $secid >\n";
|
|---|
| 206 | }
|
|---|
| 207 |
|
|---|
| 208 | # if we are doing subcollections, then some docs shouldn't be indexed.
|
|---|
| 209 | # but we need to put the section tag placeholders in there so the
|
|---|
| 210 | # sections match up with gdbm db
|
|---|
| 211 | if ($indexed_doc) {
|
|---|
| 212 | #if ($self->{'indexing_text'}) {
|
|---|
| 213 | # $text .= "$parastarttag"; # only add para tags for indexing
|
|---|
| 214 | # note that we assume that metadata will not be asked for for the compressed text, so we add para tags without checking for indexing_text
|
|---|
| 215 | # }
|
|---|
| 216 | $self->{'num_bytes'} += $doc_obj->get_text_length ($section);
|
|---|
| 217 | foreach my $field (split (/,/, $fields)) {
|
|---|
| 218 | # only deal with this field if it doesn't start with top or
|
|---|
| 219 | # this is the first section
|
|---|
| 220 | my $real_field = $field;
|
|---|
| 221 | if (!($real_field =~ s/^top//) || ($doc_section == 1)) {
|
|---|
| 222 | my $new_text = "";
|
|---|
| 223 | my $tmp_text = "";
|
|---|
| 224 | if ($real_field eq "text") {
|
|---|
| 225 | if ($self->{'indexing_text'}) { #tag the text with <Text>...</Text>, add the <Paragraph> tags and strip out html if needed
|
|---|
| 226 | $new_text .= "$parastarttag<TX index=\"1\">\n";
|
|---|
| 227 | $tmp_text .= $doc_obj->get_text ($section);
|
|---|
| 228 | if ($parastarttag =~ "") {
|
|---|
| 229 | # we don't want to individually tag each paragraph if not doing para indexing
|
|---|
| 230 | $tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "");
|
|---|
| 231 | } else {
|
|---|
| 232 | $tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "</TX>$paraendtag$parastarttag<TX index=\"1\">");
|
|---|
| 233 | }
|
|---|
| 234 |
|
|---|
| 235 | $new_text .= "$tmp_text</TX>$paraendtag\n";
|
|---|
| 236 | #if (!defined $self->{'indexfields'}->{'TextOnly'}) {
|
|---|
| 237 | #$self->{'indexfields'}->{'TextOnly'} = 1;
|
|---|
| 238 | #}
|
|---|
| 239 | }
|
|---|
| 240 | else { # leave html stuff in, but escape the tags, and dont add Paragraph tags - never retrieve paras at the moment
|
|---|
| 241 | if ($self->{'store_text'}) {
|
|---|
| 242 | $tmp_text .= $doc_obj->get_text ($section);
|
|---|
| 243 | &ghtml::htmlsafe($tmp_text);
|
|---|
| 244 | $new_text .= $tmp_text;
|
|---|
| 245 | }
|
|---|
| 246 | }
|
|---|
| 247 | } else { # metadata field
|
|---|
| 248 | if ($real_field eq "allfields") { #ignore
|
|---|
| 249 | }
|
|---|
| 250 | elsif ($real_field eq "metadata") { # insert all metadata
|
|---|
| 251 | #except gsdl stuff
|
|---|
| 252 | my $shortname = "";
|
|---|
| 253 | my $metadata = $doc_obj->get_all_metadata ($section);
|
|---|
| 254 | foreach $pair (@$metadata) {
|
|---|
| 255 | my ($mfield, $mvalue) = (@$pair);
|
|---|
| 256 | # check fields here, maybe others dont want - change to use dontindex!!
|
|---|
| 257 | if ($mfield ne "Identifier"
|
|---|
| 258 | && $mfield !~ /^gsdl/
|
|---|
| 259 | && $mfield ne "classifytype"
|
|---|
| 260 | && $mfield ne "assocfilepath"
|
|---|
| 261 | && defined $mvalue && $mvalue ne "") {
|
|---|
| 262 |
|
|---|
| 263 | if (defined $self->{'indexfieldmap'}->{$mfield}) {
|
|---|
| 264 | $shortname = $self->{'indexfieldmap'}->{$mfield};
|
|---|
| 265 | }
|
|---|
| 266 | else {
|
|---|
| 267 | $shortname = $self->create_shortname($mfield);
|
|---|
| 268 | $self->{'indexfieldmap'}->{$mfield} = $shortname;
|
|---|
| 269 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 270 | }
|
|---|
| 271 | $new_text .= "$parastarttag<$shortname index=\"1\">$mvalue</$shortname>$paraendtag\n";
|
|---|
| 272 | if (!defined $self->{'indexfields'}->{$mfield}) {
|
|---|
| 273 | $self->{'indexfields'}->{$mfield} = 1;
|
|---|
| 274 | }
|
|---|
| 275 | }
|
|---|
| 276 | }
|
|---|
| 277 |
|
|---|
| 278 | }
|
|---|
| 279 | else { #individual metadata specified
|
|---|
| 280 | my $shortname="";
|
|---|
| 281 | #if (!defined $self->{'indexfields'}->{$real_field}) {
|
|---|
| 282 | #$self->{'indexfields'}->{$real_field} = 1;
|
|---|
| 283 | #}
|
|---|
| 284 | if (defined $self->{'indexfieldmap'}->{$real_field}) {
|
|---|
| 285 | $shortname = $self->{'indexfieldmap'}->{$real_field};
|
|---|
| 286 | }
|
|---|
| 287 | else {
|
|---|
| 288 | $shortname = $self->create_shortname($real_field);
|
|---|
| 289 | $self->{'indexfieldmap'}->{$real_field} = $shortname;
|
|---|
| 290 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 291 | }
|
|---|
| 292 | foreach $item (@{$doc_obj->get_metadata ($section, $real_field)}) {
|
|---|
| 293 | $new_text .= "$parastarttag<$shortname index=\"1\">$item</$shortname>$paraendtag\n";
|
|---|
| 294 | # remove entities
|
|---|
| 295 | $new_text =~ s/&\w{1,10};//g;
|
|---|
| 296 | }
|
|---|
| 297 | }
|
|---|
| 298 |
|
|---|
| 299 | }
|
|---|
| 300 |
|
|---|
| 301 | # filter the text
|
|---|
| 302 | $self->filter_text ($field, $new_text);
|
|---|
| 303 | $self->{'num_processed_bytes'} += length ($new_text);
|
|---|
| 304 | $text .= "$new_text";
|
|---|
| 305 | }
|
|---|
| 306 | }
|
|---|
| 307 | } # if (indexed_doc)
|
|---|
| 308 |
|
|---|
| 309 | $text .= "\n</$sectiontag>\n" if ($sectiontag ne "");
|
|---|
| 310 |
|
|---|
| 311 | $section = $doc_obj->get_next_section($section);
|
|---|
| 312 | } #while defined section
|
|---|
| 313 | print $handle "$text\n$documentendtag";
|
|---|
| 314 | }
|
|---|
| 315 |
|
|---|
| 316 | 1;
|
|---|
| 317 |
|
|---|
