use Koha::SearchEngine;
use Koha::SearchEngine::Search;
-use Koha::Util::Normalize qw/legacy_default remove_spaces upper_case lower_case/;
+use Koha::SearchEngine::QueryBuilder;
+use Koha::Util::Normalize qw/legacy_default remove_spaces upper_case lower_case ISBN/;
=head1 NAME
$matcher->add_matchpoint('isbn', 1000, [ { tag => '020', subfields => 'a', norms => [] } ]);
$matcher->add_simple_required_check('245', 'a', -1, 0, '', '245', 'a', -1, 0, '');
- $matcher->add_required_check([ { tag => '245', subfields => 'a', norms => [] } ],
+ $matcher->add_required_check([ { tag => '245', subfields => 'a', norms => [] } ],
[ { tag => '245', subfields => 'a', norms => [] } ]);
my @matches = $matcher->get_matches($marc_record, $max_matches);
$sth->execute($id);
my $row = $sth->fetchrow_hashref;
$sth->finish();
- return undef unless defined $row;
+ return unless defined $row;
my $self = {};
$self->{'id'} = $row->{'matcher_id'};
my $matcher_id = $self->{'id'};
$sth = $dbh->prepare_cached("INSERT INTO matchpoints (matcher_id, search_index, score)
VALUES (?, ?, ?)");
- $sth->execute($matcher_id, $matchpoint->{'index'}, $matchpoint->{'score'});
+ $sth->execute($matcher_id, $matchpoint->{'index'}, $matchpoint->{'score'}||0);
my $matchpoint_id = $dbh->{'mysql_insertid'};
my $seqnum = 0;
foreach my $component (@{ $matchpoint->{'components'} }) {
$sth->bind_param(2, $seqnum);
$sth->bind_param(3, $component->{'tag'});
$sth->bind_param(4, join "", sort keys %{ $component->{'subfields'} });
- $sth->bind_param(5, $component->{'offset'});
+ $sth->bind_param(5, $component->{'offset'}||0);
$sth->bind_param(6, $component->{'length'});
$sth->execute();
my $matchpoint_component_id = $dbh->{'mysql_insertid'};
my $self = shift;
my ($source_record, $max_matches) = @_;
- my %matches = ();
+ my $matches = {};
- my $QParser;
- $QParser = C4::Context->queryparser if (C4::Context->preference('UseQueryParser'));
foreach my $matchpoint ( @{ $self->{'matchpoints'} } ) {
my @source_keys = _get_match_keys( $source_record, $matchpoint );
next if scalar(@source_keys) == 0;
- # FIXME - because of a bug in QueryParser, an expression ofthe
- # format 'isbn:"isbn1" || isbn:"isbn2" || isbn"isbn3"...'
- # does not get parsed correctly, so we will not
- # do AggressiveMatchOnISBN if UseQueryParser is on
@source_keys = C4::Koha::GetVariationsOfISBNs(@source_keys)
if ( $matchpoint->{index} =~ /^isbn$/i
- && C4::Context->preference('AggressiveMatchOnISBN') )
- && !C4::Context->preference('UseQueryParser');
+ && C4::Context->preference('AggressiveMatchOnISBN') );
@source_keys = C4::Koha::GetVariationsOfISSNs(@source_keys)
if ( $matchpoint->{index} =~ /^issn$/i
- && C4::Context->preference('AggressiveMatchOnISSN') )
- && !C4::Context->preference('UseQueryParser');
+ && C4::Context->preference('AggressiveMatchOnISSN') );
# build query
my $query;
my $total_hits;
if ( $self->{'record_type'} eq 'biblio' ) {
- #NOTE: The QueryParser can't handle the CCL syntax of 'qualifier','qualifier', so fallback to non-QueryParser.
- #NOTE: You can see this in C4::Search::SimpleSearch() as well in a different way.
- if ($QParser && $matchpoint->{'index'} !~ m/\w,\w/) {
- $query = join( " || ",
- map { "$matchpoint->{'index'}:$_" } @source_keys );
+ my $phr = ( C4::Context->preference('AggressiveMatchOnISBN') || C4::Context->preference('AggressiveMatchOnISSN') ) ? ',phr' : q{};
+ $query = join( " OR ",
+ map { "$matchpoint->{'index'}$phr=\"$_\"" } @source_keys );
+ #NOTE: double-quote the values so you don't get a "Embedded truncation not supported" error when a term has a ? in it.
+
+ # Use state variables to avoid recreating the objects every time.
+ # With Elasticsearch this also avoids creating a massive amount of
+ # ES connectors that would eventually run out of file descriptors.
+ state $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
+ ( $error, $searchresults, $total_hits ) =
+ $searcher->simple_search_compat( $query, 0, $max_matches, undef, skip_normalize => 1 );
+
+ if ( defined $error ) {
+ warn "search failed ($query) $error";
}
else {
- my $phr = ( C4::Context->preference('AggressiveMatchOnISBN') || C4::Context->preference('AggressiveMatchOnISSN') ) ? ',phr' : q{};
- $query = join( " or ",
- map { "$matchpoint->{'index'}$phr=\"$_\"" } @source_keys );
- #NOTE: double-quote the values so you don't get a "Embedded truncation not supported" error when a term has a ? in it.
+ foreach my $matched ( @{$searchresults} ) {
+ my $target_record = C4::Search::new_record_from_zebra( 'biblioserver', $matched );
+ my ( $biblionumber_tag, $biblionumber_subfield ) = C4::Biblio::GetMarcFromKohaField( "biblio.biblionumber" );
+ my $id = ( $biblionumber_tag > 10 ) ?
+ $target_record->field($biblionumber_tag)->subfield($biblionumber_subfield) :
+ $target_record->field($biblionumber_tag)->data();
+ $matches->{$id}->{score} += $matchpoint->{score};
+ $matches->{$id}->{record} = $target_record;
+ }
}
- my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX});
- ( $error, $searchresults, $total_hits ) =
- $searcher->simple_search_compat( $query, 0, $max_matches, undef, skip_normalize => 1 );
}
elsif ( $self->{'record_type'} eq 'authority' ) {
- my $authresults;
my @marclist;
my @and_or;
my @excluding = [];
push @operator, 'exact';
push @value, $key;
}
- require C4::AuthoritiesMarc;
- ( $authresults, $total_hits ) =
- C4::AuthoritiesMarc::SearchAuthorities(
- \@marclist, \@and_or, \@excluding, \@operator,
- \@value, 0, 20, undef,
- 'AuthidAsc', 1
- );
- foreach my $result (@$authresults) {
- push @$searchresults, $result->{'authid'};
- }
- }
+ # Use state variables to avoid recreating the objects every time.
+ # With Elasticsearch this also avoids creating a massive amount of
+ # ES connectors that would eventually run out of file descriptors.
+ state $builder = Koha::SearchEngine::QueryBuilder->new({index => $Koha::SearchEngine::AUTHORITIES_INDEX});
+ state $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::AUTHORITIES_INDEX});
+ my $search_query = $builder->build_authorities_query_compat(
+ \@marclist, \@and_or, \@excluding, \@operator,
+ \@value, undef, 'AuthidAsc'
+ );
+ my ( $authresults, $total ) = $searcher->search_auth_compat( $search_query, 0, 20 );
- if ( defined $error ) {
- warn "search failed ($query) $error";
- }
- else {
- foreach my $matched ( @{$searchresults} ) {
- $matches{$matched} += $matchpoint->{'score'};
+ foreach my $result (@$authresults) {
+ my $id = $result->{authid};
+ $matches->{$id}->{score} += $matchpoint->{'score'};
+ $matches->{$id}->{record} = $id;
}
}
}
# get rid of any that don't meet the threshold
- %matches = map { ($matches{$_} >= $self->{'threshold'}) ? ($_ => $matches{$_}) : () } keys %matches;
-
- # get rid of any that don't meet the required checks
- %matches = map { _passes_required_checks($source_record, $_, $self->{'required_checks'}) ? ($_ => $matches{$_}) : () }
- keys %matches unless ($self->{'record_type'} eq 'auth');
+ $matches = { map { ($matches->{$_}->{score} >= $self->{'threshold'}) ? ($_ => $matches->{$_}) : () } keys %$matches };
my @results = ();
if ($self->{'record_type'} eq 'biblio') {
require C4::Biblio;
- foreach my $marcblob (keys %matches) {
- my $target_record = C4::Search::new_record_from_zebra('biblioserver',$marcblob);
- my $record_number;
- my $result = C4::Biblio::TransformMarcToKoha($target_record, '');
- $record_number = $result->{'biblionumber'};
- push @results, { 'record_id' => $record_number, 'score' => $matches{$marcblob} };
+ # get rid of any that don't meet the required checks
+ $matches = {
+ map {
+ _passes_required_checks( $source_record, $matches->{$_}->{'record'}, $self->{'required_checks'} )
+ ? ( $_ => $matches->{$_} )
+ : ()
+ } keys %$matches
+ };
+
+ foreach my $id ( keys %$matches ) {
+ push @results, {
+ record_id => $id,
+ score => $matches->{$id}->{score}
+ };
}
} elsif ($self->{'record_type'} eq 'authority') {
require C4::AuthoritiesMarc;
- foreach my $authid (keys %matches) {
- push @results, { 'record_id' => $authid, 'score' => $matches{$authid} };
+ foreach my $id (keys %$matches) {
+ push @results, {
+ record_id => $id,
+ score => $matches->{$id}->{score}
+ };
}
}
@results = sort {
@results = @results[0..$max_matches-1];
}
return @results;
-
}
=head2 dump
}
sub _passes_required_checks {
- my ($source_record, $target_blob, $matchchecks) = @_;
- my $target_record = MARC::Record->new_from_usmarc($target_blob); # FIXME -- need to avoid parsing record twice
+ my ($source_record, $target_record, $matchchecks) = @_;
# no checks supplied == automatic pass
return 1 if $#{ $matchchecks } == -1;
elsif ( $norm eq 'legacy_default' ) {
$key = legacy_default($key);
}
+ elsif ( $norm eq 'ISBN' ) {
+ $key = ISBN($key);
+ }
} else {
warn "Invalid normalization routine required ($norm)"
unless $norm eq 'none';
'remove_spaces',
'upper_case',
'lower_case',
- 'legacy_default'
+ 'legacy_default',
+ 'ISBN'
);
}
projects / srvgit / blobdiff