###########################################################################
#
# lucenebuildproc.pm -- perl wrapper for building index with Lucene
# A component of the Greenstone digital library software
# from the New Zealand Digital Library Project at the
# University of Waikato, New Zealand.
#
# Copyright (C) 1999 New Zealand Digital Library Project
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 2 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program; if not, write to the Free Software
# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
#
###########################################################################
package lucenebuildproc;
# This document processor outputs a document
# for lucene to process
# Use same basic XML structure setup by mgppbuilder/mgppbuildproc
use mgppbuildproc;
use ghtml;
sub BEGIN {
@lucenebuildproc::ISA = ('mgppbuildproc');
}
sub new {
my $class = shift @_;
my $self = new mgppbuildproc (@_);
return bless $self, $class;
}
sub is_incremental_capable
{
my $self = shift (@_);
# Unlike MG and MGPP, Lucene supports incremental building
return 1;
}
sub preprocess_text {
my $self = shift (@_);
my ($text, $strip_html, $para) = @_;
# call the mgpp method first
my ($new_text) = $self->SUPER::preprocess_text($text, $strip_html, $para);
# remove entities
$new_text =~ s/&\w{1,10};//g;
# remove &
$new_text =~ s/&//g;
return $new_text;
}
sub text {
my $self = shift (@_);
my ($doc_obj,$file) = @_;
my $handle = $self->{'output_handle'};
my $outhandle = $self->{'outhandle'};
# only output this document if it is one to be indexed
return if ($doc_obj->get_doc_type() ne "indexed_doc");
my $indexed_doc = $self->is_subcollection_doc($doc_obj);
# this is another document
$self->{'num_docs'} += 1;
# get the parameters for the output
# split on : just in case there is subcoll and lang stuff
my ($fields) = split (/:/, $self->{'index'});
my $doc_level = $mgppbuildproc::level_map{'document'};
my $gs2ns = 'xmlns:gs2="http://www.greenstone.org/gs2"';
my $levels = $self->{'levels'};
my $ldoc_level = $levels->{'document'};
my $lsec_level = $levels->{'section'};
my $lpar_level = $levels->{'paragraph'};
my $docid="";
if ($ldoc_level) {
if ($self->{'gdbm_level'} eq 'document') {
my $doc_sec_num = $self->{'num_docs'};
$docid = "gs2:id=\"$doc_sec_num\"";
} else {
# default is section level
my $doc_sec_num = $self->{'num_sections'}+1;
$docid = "gs2:id=\"$doc_sec_num\"";
}
}
my $documenttag = "<$doc_level $gs2ns file=\"$file\" $docid >\n";
my $documentendtag = "\n$doc_level>\n";
my ($sectiontag) = "";
if ($lsec_level) {
$sectiontag = $mgppbuildproc::level_map{'section'};
}
my ($parastarttag) = "";
my ($paraendtag) = "";
if ($self->{'levels'}->{'paragraph'}) {
if ($self->{'strip_html'}) {
$parastarttag = "<".$mgppbuildproc::level_map{'paragraph'}.">";
$paraendtag = "".$mgppbuildproc::level_map{'paragraph'}.">";
} else {
print $outhandle "Paragraph level can not be used with no_strip_html!. Not indexing Paragraphs.\n";
}
}
my $doc_section = 0; # just for this document
my $text = "";
$text .= $documenttag;
# get the text for this document
my $section = $doc_obj->get_top_section();
while (defined $section) {
# update a few statistics
$doc_section++;
$self->{'num_sections'} += 1;
if ($sectiontag ne "") {
my $secid = "gs2:id=\"".$self->{'num_sections'}."\"";
$text .= "\n<$sectiontag $secid >\n";
}
# if we are doing subcollections, then some docs shouldn't be indexed.
# but we need to put the section tag placeholders in there so the
# sections match up with gdbm db
if (!$indexed_doc) {
$text .= "\n$sectiontag>\n" if ($sectiontag ne "");
$section = $doc_obj->get_next_section($section);
next;
}
$self->{'num_bytes'} += $doc_obj->get_text_length ($section);
foreach my $field (split (/;/, $fields)) {
# only deal with this field if it doesn't start with top or
# this is the first section
my $real_field = $field;
next if (($real_field =~ s/^top//) && ($doc_section != 1));
my $new_text = "";
my $tmp_text = "";
# we get allfields by default - do nothing
if ($real_field eq "allfields") {
}
# metadata - output all metadata we know about except gsdl stuff
elsif ($real_field eq "metadata") {
my $shortname = "";
my $metadata = $doc_obj->get_all_metadata ($section);
foreach $pair (@$metadata) {
my ($mfield, $mvalue) = (@$pair);
# check fields here, maybe others dont want - change to use dontindex!!
if ($mfield ne "Identifier"
&& $mfield !~ /^gsdl/
&& $mfield ne "classifytype"
&& $mfield ne "assocfilepath"
&& defined $mvalue && $mvalue ne "") {
if (defined $self->{'indexfieldmap'}->{$mfield}) {
$shortname = $self->{'indexfieldmap'}->{$mfield};
}
else {
$shortname = $self->create_shortname($mfield);
$self->{'indexfieldmap'}->{$mfield} = $shortname;
$self->{'indexfieldmap'}->{$shortname} = 1;
}
$new_text .= "$parastarttag<$shortname index=\"1\">$mvalue$shortname>$paraendtag\n";
if (!defined $self->{'indexfields'}->{$mfield}) {
$self->{'indexfields'}->{$mfield} = 1;
}
}
}
}
else {
#individual metadata and or text specified - could be a comma separated list
my $shortname="";
if (defined $self->{'indexfieldmap'}->{$real_field}) {
$shortname = $self->{'indexfieldmap'}->{$real_field};
}
else {
$shortname = $self->create_shortname($real_field);
$self->{'indexfieldmap'}->{$real_field} = $shortname;
$self->{'indexfieldmap'}->{$shortname} = 1;
}
my @metadata_list = ();
foreach $submeta (split /,/, $real_field) {
if ($submeta eq "text") {
if ($self->{'indexing_text'}) { #tag the text with ..., add the tags and strip out html if needed
$new_text .= "$parastarttag<$shortname index=\"1\">\n";
$tmp_text .= $doc_obj->get_text ($section);
if ($parastarttag ne "") {
$tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "$shortname>$paraendtag$parastarttag<$shortname index=\"1\">");
} else {
# we don't want to individually tag each paragraph if not doing para indexing
$tmp_text = $self->preprocess_text($tmp_text, $self->{'strip_html'}, "");
}
$new_text .= "$tmp_text$shortname>$paraendtag\n";
}
else { # leave html stuff in, but escape the tags, and dont add Paragraph tags - never retrieve paras at the moment
$tmp_text .= $doc_obj->get_text ($section);
&ghtml::htmlsafe($tmp_text);
$new_text .= $tmp_text;
}
}
else {
my @section_metadata = @{$doc_obj->get_metadata ($section, $submeta)};
if ($self->{'indexing_text'} && defined ($self->{'sections_index_document_metadata'})) {
if ($self->{'sections_index_document_metadata'} eq "always" || ( scalar(@section_metadata) == 0 && $self->{'sections_index_document_metadata'} eq "unless_section_metadata_exists")) {
push (@section_metadata, @{$doc_obj->get_metadata ($doc_obj->get_top_section(), $submeta)});
}
}
push (@metadata_list, @section_metadata);
}
}
foreach my $item (@metadata_list) {
$new_text .= "$parastarttag<$shortname index=\"1\">$item$shortname>$paraendtag\n";
# remove entities
$new_text =~ s/&\w{1,10};//g;
# remove &
$new_text =~ s/&//g;
}
}
# filter the text
$self->filter_text ($field, $new_text);
$self->{'num_processed_bytes'} += length ($new_text);
$text .= "$new_text";
} # foreach field
$text .= "\n$sectiontag>\n" if ($sectiontag ne "");
$section = $doc_obj->get_next_section($section);
} #while defined section
print $handle "$text\n$documentendtag";
#print STDOUT "$text\n$documentendtag";
}
1;