| 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 | use strict;
|
|---|
| 36 | no strict 'refs'; # allow filehandles to be variables and viceversa
|
|---|
| 37 |
|
|---|
| 38 |
|
|---|
| 39 | use IncrementalBuildUtils;
|
|---|
| 40 |
|
|---|
| 41 | sub BEGIN {
|
|---|
| 42 | @lucenebuildproc::ISA = ('mgppbuildproc');
|
|---|
| 43 | }
|
|---|
| 44 |
|
|---|
| 45 |
|
|---|
| 46 | sub new {
|
|---|
| 47 | my $class = shift @_;
|
|---|
| 48 | my $self = new mgppbuildproc (@_);
|
|---|
| 49 |
|
|---|
| 50 | $self->{'numincdocs'} = 0;
|
|---|
| 51 |
|
|---|
| 52 | return bless $self, $class;
|
|---|
| 53 | }
|
|---|
| 54 |
|
|---|
| 55 |
|
|---|
| 56 | sub is_incremental_capable
|
|---|
| 57 | {
|
|---|
| 58 | my $self = shift (@_);
|
|---|
| 59 |
|
|---|
| 60 | # Unlike MG and MGPP, Lucene supports incremental building
|
|---|
| 61 | return 1;
|
|---|
| 62 | }
|
|---|
| 63 |
|
|---|
| 64 |
|
|---|
| 65 | sub text {
|
|---|
| 66 | my $self = shift (@_);
|
|---|
| 67 | my ($doc_obj,$file) = @_;
|
|---|
| 68 | my $handle = $self->{'output_handle'};
|
|---|
| 69 | my $outhandle = $self->{'outhandle'};
|
|---|
| 70 |
|
|---|
| 71 | # only output this document if it is one to be indexed
|
|---|
| 72 | return if ($doc_obj->get_doc_type() ne "indexed_doc");
|
|---|
| 73 |
|
|---|
| 74 | my $indexed_doc = $self->is_subcollection_doc($doc_obj);
|
|---|
| 75 |
|
|---|
| 76 | # this is another document
|
|---|
| 77 | $self->{'num_docs'} += 1;
|
|---|
| 78 |
|
|---|
| 79 | # get the parameters for the output
|
|---|
| 80 | # split on : just in case there is subcoll and lang stuff
|
|---|
| 81 | my ($fields) = split (/:/, $self->{'index'});
|
|---|
| 82 |
|
|---|
| 83 | my $doc_level = $mgppbuildproc::level_map{'document'};
|
|---|
| 84 |
|
|---|
| 85 | my $levels = $self->{'levels'};
|
|---|
| 86 | my $ldoc_level = $levels->{'document'};
|
|---|
| 87 | my $lsec_level = $levels->{'section'};
|
|---|
| 88 | my $lpar_level = $levels->{'paragraph'};
|
|---|
| 89 |
|
|---|
| 90 | my $gs2_id = "";
|
|---|
| 91 | if ($ldoc_level)
|
|---|
| 92 | {
|
|---|
| 93 | if ($self->{'db_level'} eq 'document')
|
|---|
| 94 | {
|
|---|
| 95 | $gs2_id = $self->{'num_docs'};
|
|---|
| 96 | }
|
|---|
| 97 | else
|
|---|
| 98 | {
|
|---|
| 99 | # default is section level
|
|---|
| 100 | $gs2_id = $self->{'num_sections'} + 1;
|
|---|
| 101 | }
|
|---|
| 102 | }
|
|---|
| 103 | my $gs2_docOID = $doc_obj->get_OID();
|
|---|
| 104 | my $documenttag = "<$doc_level xmlns:gs2=\"http://www.greenstone.org/gs2\" file=\"$file\" gs2:id=\"$gs2_id\" gs2:docOID=\"$gs2_docOID\">\n";
|
|---|
| 105 | my $documentendtag = "\n</$doc_level>\n";
|
|---|
| 106 |
|
|---|
| 107 | my ($sectiontag) = "";
|
|---|
| 108 | if ($lsec_level)
|
|---|
| 109 | {
|
|---|
| 110 | $sectiontag = $mgppbuildproc::level_map{'section'};
|
|---|
| 111 | }
|
|---|
| 112 | my ($parastarttag) = "";
|
|---|
| 113 | my ($paraendtag) = "";
|
|---|
| 114 | if ($self->{'levels'}->{'paragraph'})
|
|---|
| 115 | {
|
|---|
| 116 | if ($self->{'strip_html'})
|
|---|
| 117 | {
|
|---|
| 118 | $parastarttag = "<".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 119 | $paraendtag = "</".$mgppbuildproc::level_map{'paragraph'}.">";
|
|---|
| 120 | }
|
|---|
| 121 | else
|
|---|
| 122 | {
|
|---|
| 123 | print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
|
|---|
| 124 | }
|
|---|
| 125 | }
|
|---|
| 126 |
|
|---|
| 127 | my $doc_section = 0; # just for this document
|
|---|
| 128 |
|
|---|
| 129 | my $text = "";
|
|---|
| 130 | $text .= $documenttag;
|
|---|
| 131 | # get the text for this document
|
|---|
| 132 | my $section = $doc_obj->get_top_section();
|
|---|
| 133 | while (defined $section)
|
|---|
| 134 | {
|
|---|
| 135 | # update a few statistics
|
|---|
| 136 | $doc_section++;
|
|---|
| 137 | $self->{'num_sections'}++;
|
|---|
| 138 |
|
|---|
| 139 | if ($sectiontag ne "")
|
|---|
| 140 | {
|
|---|
| 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 database
|
|---|
| 148 | my $indexed_section = $doc_obj->get_metadata_element($section, "gsdldoctype") || "indexed_section";
|
|---|
| 149 | if (($indexed_doc == 0) || ($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 | {
|
|---|
| 158 | # only deal with this field if it doesn't start with top or
|
|---|
| 159 | # this is the first section
|
|---|
| 160 | my $real_field = $field;
|
|---|
| 161 | next if (($real_field =~ s/^top//) && ($doc_section != 1));
|
|---|
| 162 |
|
|---|
| 163 | my $new_text = "";
|
|---|
| 164 | my $tmp_text = "";
|
|---|
| 165 |
|
|---|
| 166 | # If allfields is requested add all metadata fields and text as
|
|---|
| 167 | # belonging to the ZZ field
|
|---|
| 168 | if ($real_field eq "allfields") {
|
|---|
| 169 | # Text first - no html nor paragraph tags
|
|---|
| 170 | $new_text .= "$parastarttag<ZZ index=\"1\">\n";
|
|---|
| 171 | $tmp_text = $self->preprocess_text($doc_obj->get_text ($section), 1, "");
|
|---|
| 172 | &ghtml::htmlsafe($tmp_text);
|
|---|
| 173 | $new_text .= "$tmp_text</ZZ>$paraendtag\n";
|
|---|
| 174 | # Then Metadata
|
|---|
| 175 | my $metadata = $doc_obj->get_all_metadata ($section);
|
|---|
| 176 | foreach my $pair (@$metadata) {
|
|---|
| 177 | my ($mfield, $mvalue) = (@$pair);
|
|---|
| 178 | &ghtml::htmlsafe($mvalue);
|
|---|
| 179 | # check fields here, maybe others dont want - change to use dontindex!!
|
|---|
| 180 | if ($mfield ne "Identifier"
|
|---|
| 181 | && $mfield !~ /^gsdl/
|
|---|
| 182 | && $mfield ne "classifytype"
|
|---|
| 183 | && $mfield ne "assocfilepath"
|
|---|
| 184 | && defined $mvalue && $mvalue ne "") {
|
|---|
| 185 | $new_text .= "$parastarttag<ZZ index=\"1\">$mvalue</ZZ>$paraendtag\n";
|
|---|
| 186 | }
|
|---|
| 187 | if (!defined $self->{'indexfields'}->{$mfield}) {
|
|---|
| 188 | $self->{'indexfields'}->{$mfield} = 1;
|
|---|
| 189 | }
|
|---|
| 190 | }
|
|---|
| 191 | }
|
|---|
| 192 | # metadata - output all metadata we know about except gsdl stuff
|
|---|
| 193 | elsif ($real_field eq "metadata" || $real_field eq "allfields") {
|
|---|
| 194 | my $shortname = "";
|
|---|
| 195 | my $metadata = $doc_obj->get_all_metadata ($section);
|
|---|
| 196 | foreach my $pair (@$metadata) {
|
|---|
| 197 | my ($mfield, $mvalue) = (@$pair);
|
|---|
| 198 | &ghtml::htmlsafe($mvalue);
|
|---|
| 199 | # check fields here, maybe others dont want - change to use dontindex!!
|
|---|
| 200 | if ($mfield ne "Identifier"
|
|---|
| 201 | && $mfield !~ /^gsdl/
|
|---|
| 202 | && $mfield ne "classifytype"
|
|---|
| 203 | && $mfield ne "assocfilepath"
|
|---|
| 204 | && defined $mvalue && $mvalue ne "") {
|
|---|
| 205 |
|
|---|
| 206 | if (defined $self->{'indexfieldmap'}->{$mfield}) {
|
|---|
| 207 | $shortname = $self->{'indexfieldmap'}->{$mfield};
|
|---|
| 208 | }
|
|---|
| 209 | else {
|
|---|
| 210 | $shortname = $self->create_shortname($mfield);
|
|---|
| 211 | $self->{'indexfieldmap'}->{$mfield} = $shortname;
|
|---|
| 212 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 213 | }
|
|---|
| 214 | $new_text .= "$parastarttag<$shortname index=\"1\">$mvalue</$shortname>$paraendtag\n";
|
|---|
| 215 | if (!defined $self->{'indexfields'}->{$mfield}) {
|
|---|
| 216 | $self->{'indexfields'}->{$mfield} = 1;
|
|---|
| 217 | }
|
|---|
| 218 | }
|
|---|
| 219 | }
|
|---|
| 220 | }
|
|---|
| 221 | else {
|
|---|
| 222 | #individual metadata and or text specified - could be a comma separated list
|
|---|
| 223 | my $shortname="";
|
|---|
| 224 | if (defined $self->{'indexfieldmap'}->{$real_field}) {
|
|---|
| 225 | $shortname = $self->{'indexfieldmap'}->{$real_field};
|
|---|
| 226 | }
|
|---|
| 227 | else {
|
|---|
| 228 | $shortname = $self->create_shortname($real_field);
|
|---|
| 229 | $self->{'indexfieldmap'}->{$real_field} = $shortname;
|
|---|
| 230 | $self->{'indexfieldmap'}->{$shortname} = 1;
|
|---|
| 231 | }
|
|---|
| 232 |
|
|---|
| 233 | my @metadata_list = ();
|
|---|
| 234 | foreach my $submeta (split /,/, $real_field) {
|
|---|
| 235 | if ($submeta eq "text") {
|
|---|
| 236 | my $section_text = $doc_obj->get_text($section);
|
|---|
| 237 | if ($self->{'indexing_text'}) {
|
|---|
| 238 | # tag the text with <Text>...</Text>, add the <Paragraph> tags and always strip out HTML
|
|---|
| 239 | $new_text .= "$parastarttag<$shortname index=\"1\">\n";
|
|---|
| 240 | if ($parastarttag ne "") {
|
|---|
| 241 | $section_text = $self->preprocess_text($section_text, 1, "</$shortname>$paraendtag$parastarttag<$shortname index=\"1\">");
|
|---|
| 242 | }
|
|---|
| 243 | else {
|
|---|
| 244 | # we don't want to individually tag each paragraph if not doing para indexing
|
|---|
| 245 | $section_text = $self->preprocess_text($section_text, 1, "");
|
|---|
| 246 | }
|
|---|
| 247 | $new_text .= "$section_text</$shortname>$paraendtag\n";
|
|---|
| 248 | }
|
|---|
| 249 | else { # leave html stuff in, but escape the tags, and dont add Paragraph tags - never retrieve paras at the moment
|
|---|
| 250 | $tmp_text .= $doc_obj->get_text ($section);
|
|---|
| 251 | &ghtml::htmlsafe($tmp_text);
|
|---|
| 252 | $new_text .= $tmp_text;
|
|---|
| 253 | }
|
|---|
| 254 | }
|
|---|
| 255 | else {
|
|---|
| 256 | my @section_metadata = @{$doc_obj->get_metadata ($section, $submeta)};
|
|---|
| 257 | if ($section ne $doc_obj->get_top_section() && $self->{'indexing_text'} && defined ($self->{'sections_index_document_metadata'})) {
|
|---|
| 258 | if ($self->{'sections_index_document_metadata'} eq "always" || ( scalar(@section_metadata) == 0 && $self->{'sections_index_document_metadata'} eq "unless_section_metadata_exists")) {
|
|---|
| 259 | push (@section_metadata, @{$doc_obj->get_metadata ($doc_obj->get_top_section(), $submeta)});
|
|---|
| 260 | }
|
|---|
| 261 | }
|
|---|
| 262 | push (@metadata_list, @section_metadata);
|
|---|
| 263 | }
|
|---|
| 264 | }
|
|---|
| 265 | foreach my $item (@metadata_list) {
|
|---|
| 266 | &ghtml::htmlsafe($item);
|
|---|
| 267 | $new_text .= "$parastarttag<$shortname index=\"1\">$item</$shortname>$paraendtag\n";
|
|---|
| 268 | }
|
|---|
| 269 | }
|
|---|
| 270 | # filter the text
|
|---|
| 271 | $self->filter_text ($field, $new_text);
|
|---|
| 272 | $self->{'num_processed_bytes'} += length ($new_text);
|
|---|
| 273 |
|
|---|
| 274 | $text .= "$new_text";
|
|---|
| 275 | } # foreach field
|
|---|
| 276 |
|
|---|
| 277 | $text .= "\n</$sectiontag>\n" if ($sectiontag ne "");
|
|---|
| 278 |
|
|---|
| 279 | $section = $doc_obj->get_next_section($section);
|
|---|
| 280 | } #while defined section
|
|---|
| 281 | print $handle "$text\n$documentendtag";
|
|---|
| 282 | #print STDOUT "$text\n$documentendtag";
|
|---|
| 283 | }
|
|---|
| 284 |
|
|---|
| 285 | # /** We make this builder pretend to be a document processor so we can get
|
|---|
| 286 | # * information back from the plugins.
|
|---|
| 287 | # *
|
|---|
| 288 | # * @param $self A reference to this Lucene builder
|
|---|
| 289 | # * @param $doc_obj A reference to a document object representing what was
|
|---|
| 290 | # * parsed by the GAPlug
|
|---|
| 291 | # * @param $file The name of the file parsed as a string
|
|---|
| 292 | # *
|
|---|
| 293 | # * @author John Thompson, DL Consulting Ltd
|
|---|
| 294 | # */
|
|---|
| 295 | sub process()
|
|---|
| 296 | {
|
|---|
| 297 | my $self = shift (@_);
|
|---|
| 298 | my ($doc_obj, $file) = @_;
|
|---|
| 299 |
|
|---|
| 300 | # If this is called from any stage other than an incremental infodb we want
|
|---|
| 301 | # to pass through to the superclass of build
|
|---|
| 302 | if ($self->get_mode() eq "incinfodb")
|
|---|
| 303 | {
|
|---|
| 304 | print STDERR "*** Processing a document added using INCINFODB ***\n";
|
|---|
| 305 | my ($archivedir) = $file =~ /^(.*?)(?:\/|\\)[^\/\\]*$/;
|
|---|
| 306 | $archivedir = "" unless defined $archivedir;
|
|---|
| 307 | $archivedir =~ s/\\/\//g;
|
|---|
| 308 | $archivedir =~ s/^\/+//;
|
|---|
| 309 | $archivedir =~ s/\/+$//;
|
|---|
| 310 |
|
|---|
| 311 | # Number of files
|
|---|
| 312 | print STDERR "There are " . scalar($doc_obj->get_assoc_files()) . " associated documents...\n";
|
|---|
| 313 |
|
|---|
| 314 | # resolve the final filenames of the files associated with this document
|
|---|
| 315 | $self->assoc_files ($doc_obj, $archivedir);
|
|---|
| 316 |
|
|---|
| 317 | # is this a paged or a hierarchical document
|
|---|
| 318 | my ($thistype, $childtype) = $self->get_document_type ($doc_obj);
|
|---|
| 319 |
|
|---|
| 320 | # Determine the actual docnum by checking if we've processed any
|
|---|
| 321 | # previous incrementally added documents. If so, carry on from there.
|
|---|
| 322 | # Otherwise we set the counter to be the same as the number of
|
|---|
| 323 | # sections encountered during the previous build
|
|---|
| 324 | if ($self->{'numincdocs'} == 0)
|
|---|
| 325 | {
|
|---|
| 326 | $self->{'numincdocs'} = $self->{'starting_num_sections'} + 1;
|
|---|
| 327 | }
|
|---|
| 328 |
|
|---|
| 329 | my $section = $doc_obj->get_top_section ();
|
|---|
| 330 | print STDERR "+ top section: '$section'\n";
|
|---|
| 331 | my $doc_OID = $doc_obj->get_OID();
|
|---|
| 332 | my $url = "";
|
|---|
| 333 | while (defined $section)
|
|---|
| 334 | {
|
|---|
| 335 | print STDERR "+ processing section: '$section'\n";
|
|---|
| 336 | # Attach all the other metadata to this document
|
|---|
| 337 | # output the fact that this document is a document (unless doctype
|
|---|
| 338 | # has been set to something else from within a plugin
|
|---|
| 339 | my $dtype = $doc_obj->get_metadata_element ($section, "doctype");
|
|---|
| 340 | if (!defined $dtype || $dtype !~ /\w/)
|
|---|
| 341 | {
|
|---|
| 342 | $doc_obj->add_utf8_metadata($section, "doctype", $dtype);
|
|---|
| 343 | }
|
|---|
| 344 | # output whether this node contains text
|
|---|
| 345 | if ($doc_obj->get_text_length($section) > 0)
|
|---|
| 346 | {
|
|---|
| 347 | $doc_obj->add_utf8_metadata($section, "hastxt", 1);
|
|---|
| 348 | }
|
|---|
| 349 | else
|
|---|
| 350 | {
|
|---|
| 351 | $doc_obj->add_utf8_metadata($section, "hastxt", 0);
|
|---|
| 352 | }
|
|---|
| 353 |
|
|---|
| 354 | # output archivedir if at top level
|
|---|
| 355 | if ($section eq $doc_obj->get_top_section())
|
|---|
| 356 | {
|
|---|
| 357 | $doc_obj->add_utf8_metadata($section, "archivedir", $archivedir);
|
|---|
| 358 | $doc_obj->add_utf8_metadata($section, "thistype", $thistype);
|
|---|
| 359 | }
|
|---|
| 360 |
|
|---|
| 361 | # output a list of children
|
|---|
| 362 | my $children = $doc_obj->get_children ($section);
|
|---|
| 363 | if (scalar(@$children) > 0)
|
|---|
| 364 | {
|
|---|
| 365 | $doc_obj->add_utf8_metadata($section, "childtype", $childtype);
|
|---|
| 366 | my @contains = ();
|
|---|
| 367 | foreach my $child (@$children)
|
|---|
| 368 | {
|
|---|
| 369 | if ($child =~ /^.*?\.(\d+)$/)
|
|---|
| 370 | {
|
|---|
| 371 | push (@contains, "\".$1");
|
|---|
| 372 | }
|
|---|
| 373 | else
|
|---|
| 374 | {
|
|---|
| 375 | push (@contains, "\".$child");
|
|---|
| 376 | }
|
|---|
| 377 | }
|
|---|
| 378 | $doc_obj->add_utf8_metadata($section, "contains", join(";", @contains));
|
|---|
| 379 | }
|
|---|
| 380 | #output the matching doc number
|
|---|
| 381 | print STDERR "+ docnum=" . $self->{'numincdocs'} . "\n";
|
|---|
| 382 | $doc_obj->add_utf8_metadata($section, "docnum", $self->{'numincdocs'});
|
|---|
| 383 |
|
|---|
| 384 | $self->{'numincdocs'}++;
|
|---|
| 385 | $section = $doc_obj->get_next_section($section);
|
|---|
| 386 | # if no sections wanted, only add the docs
|
|---|
| 387 | last if ($self->{'db_level'} eq "document");
|
|---|
| 388 | }
|
|---|
| 389 | print STDERR "\n*** incrementally add metadata from document at: " . $file . "\n";
|
|---|
| 390 | &IncrementalBuildUtils::addDocument($self->{'collection'}, $doc_obj, $doc_obj->get_top_section());
|
|---|
| 391 | }
|
|---|
| 392 | else
|
|---|
| 393 | {
|
|---|
| 394 | $self->mgppbuildproc::process(@_);
|
|---|
| 395 | }
|
|---|
| 396 | }
|
|---|
| 397 | # /** process() **/
|
|---|
| 398 |
|
|---|
| 399 |
|
|---|
| 400 | # Following methods seem to be no different to those defined in basebuildproc.pm
|
|---|
| 401 | # From inspection, it looks like these ones can be removed
|
|---|
| 402 |
|
|---|
| 403 |
|
|---|
| 404 | sub get_num_docs {
|
|---|
| 405 | my $self = shift (@_);
|
|---|
| 406 | #rint STDERR "get_num_docs(): $self->{'num_docs'}\n";
|
|---|
| 407 | return $self->{'num_docs'};
|
|---|
| 408 | }
|
|---|
| 409 |
|
|---|
| 410 | sub get_num_sections {
|
|---|
| 411 | my $self = shift (@_);
|
|---|
| 412 | #rint STDERR "get_num_sections(): $self->{'num_sections'}\n";
|
|---|
| 413 | return $self->{'num_sections'};
|
|---|
| 414 | }
|
|---|
| 415 |
|
|---|
| 416 | # num_bytes is the actual number of bytes in the collection
|
|---|
| 417 | # this is normally the same as what's processed during text compression
|
|---|
| 418 | sub get_num_bytes {
|
|---|
| 419 | my $self = shift (@_);
|
|---|
| 420 | #rint STDERR "get_num_bytes(): $self->{'num_bytes'}\n";
|
|---|
| 421 | return $self->{'num_bytes'};
|
|---|
| 422 | }
|
|---|
| 423 |
|
|---|
| 424 |
|
|---|
| 425 | # This is similar to mgppbuildproc's preprocess_text but adds extra spaces
|
|---|
| 426 | # Otherwise the removal of tags below might lead to Lucene turning
|
|---|
| 427 | # "...farming</p>\n<p>EDWARD.." into "farmingedward"
|
|---|
| 428 | # (example from demo collection b20cre)
|
|---|
| 429 | # Many thanks to John Thompson, DL Consulting Ltd. (www.dlconsulting.com)
|
|---|
| 430 | sub preprocess_text
|
|---|
| 431 | {
|
|---|
| 432 | my $self = shift (@_);
|
|---|
| 433 | my ($text, $strip_html, $para) = @_;
|
|---|
| 434 | # at this stage, we do not do paragraph tags unless have strip_html -
|
|---|
| 435 | # it will result in a huge mess of non-xml
|
|---|
| 436 | return unless $strip_html;
|
|---|
| 437 |
|
|---|
| 438 | my $new_text = $text;
|
|---|
| 439 |
|
|---|
| 440 | # if we have <pre> tags, we can have < > inside them, need to delete
|
|---|
| 441 | # the <> before stripping tags
|
|---|
| 442 | $new_text =~ s/<pre>(.*?)<\/pre>/$self->remove_gtlt($1,$para)/gse;
|
|---|
| 443 |
|
|---|
| 444 | if ($para eq "") {
|
|---|
| 445 | # just remove all tags
|
|---|
| 446 | $new_text =~ s/<[^>]*>/ /gs;
|
|---|
| 447 | } else {
|
|---|
| 448 | # strip all tags except <p> tags which get turned into $para
|
|---|
| 449 | $new_text =~ s/<([^>]*)>/$self->process_tags($1, $para)/gse;
|
|---|
| 450 | }
|
|---|
| 451 |
|
|---|
| 452 | # It's important that we remove name entities because otherwise the text passed to Lucene for indexing
|
|---|
| 453 | # may not be valid XML (eg. if HTML-only entities like are used)
|
|---|
| 454 | $new_text =~ s/&\w{1,10};//g;
|
|---|
| 455 | # Remove stray '&' characters, except in &#nnnn; or &#xhhhh; entities (which are valid XML)
|
|---|
| 456 | $new_text =~ s/&([^\#])/ $1/g;
|
|---|
| 457 |
|
|---|
| 458 | return $new_text;
|
|---|
| 459 | }
|
|---|
| 460 |
|
|---|
| 461 |
|
|---|
| 462 | 1;
|
|---|
| 463 |
|
|---|
