package C4::Search; # This file is part of Koha. # # Koha 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 3 of the License, or # (at your option) any later version. # # Koha 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 Koha; if not, see . use Modern::Perl; use C4::Context; use C4::Biblio qw( TransformMarcToKoha GetMarcFromKohaField GetFrameworkCode GetAuthorisedValueDesc GetBiblioData ); use C4::Koha qw( getFacets GetVariationsOfISBN GetNormalizedUPC GetNormalizedEAN GetNormalizedOCLCNumber GetNormalizedISBN getitemtypeimagelocation ); use Koha::DateUtils; use Koha::Libraries; use Koha::SearchEngine::QueryBuilder; use Lingua::Stem; use XML::Simple; use C4::XSLT qw( XSLTParse4Display ); use C4::Reserves qw( GetReserveStatus ); use C4::Charset qw( SetUTF8Flag ); use Koha::AuthorisedValues; use Koha::ItemTypes; use Koha::Libraries; use Koha::Logger; use Koha::Patrons; use Koha::Recalls; use Koha::RecordProcessor; use Koha::SearchFilters; use URI::Escape; use Business::ISBN; use MARC::Record; use MARC::Field; our (@ISA, @EXPORT_OK); BEGIN { require Exporter; @ISA = qw(Exporter); @EXPORT_OK = qw( FindDuplicate SimpleSearch searchResults getRecords buildQuery GetDistinctValues enabled_staff_search_views new_record_from_zebra z3950_search_args getIndexes ); } =head1 NAME C4::Search - Functions for searching the Koha catalog. =head1 SYNOPSIS See opac/opac-search.pl or catalogue/search.pl for example of usage =head1 DESCRIPTION This module provides searching functions for Koha's bibliographic databases =head1 FUNCTIONS =cut # make all your functions, whether exported or not; =head2 FindDuplicate ($biblionumber,$biblionumber,$title) = FindDuplicate($record); This function attempts to find duplicate records using a hard-coded, fairly simplistic algorithm =cut sub FindDuplicate { my ($record) = @_; my $dbh = C4::Context->dbh; my $result = TransformMarcToKoha({ record => $record }); my $sth; my $query; # search duplicate on ISBN, easy and fast.. # ... normalize first if ( $result->{isbn} ) { $result->{isbn} =~ s/\(.*$//; $result->{isbn} =~ s/\s+$//; $result->{isbn} =~ s/\|/OR/; $query = "isbn:$result->{isbn}"; } else { my $titleindex = 'ti,ext'; my $authorindex = 'au,ext'; my $op = 'AND'; $result->{title} =~ s /\\//g; $result->{title} =~ s /\"//g; $result->{title} =~ s /\(//g; $result->{title} =~ s /\)//g; $query = "$titleindex:\"$result->{title}\""; if ( $result->{author} ) { $result->{author} =~ s /\\//g; $result->{author} =~ s /\"//g; $result->{author} =~ s /\(//g; $result->{author} =~ s /\)//g; $query .= " $op $authorindex:\"$result->{author}\""; } } my $searcher = Koha::SearchEngine::Search->new({index => $Koha::SearchEngine::BIBLIOS_INDEX}); my ( $error, $searchresults, undef ) = $searcher->simple_search_compat($query,0,50); my @results; if (!defined $error) { foreach my $possible_duplicate_record (@{$searchresults}) { my $marcrecord = new_record_from_zebra( 'biblioserver', $possible_duplicate_record ); my $result = TransformMarcToKoha({ record => $marcrecord }); # FIXME :: why 2 $biblionumber ? if ($result) { push @results, $result->{'biblionumber'}; push @results, $result->{'title'}; } } } return @results; } =head2 SimpleSearch ( $error, $results, $total_hits ) = SimpleSearch( $query, $offset, $max_results, [@servers], [%options] ); This function provides a simple search API on the bibliographic catalog =over 2 =item C * $query can be a simple keyword or a complete CCL query * @servers is optional. Defaults to biblioserver as found in koha-conf.xml * $offset - If present, represents the number of records at the beginning to omit. Defaults to 0 * $max_results - if present, determines the maximum number of records to fetch. undef is All. defaults to undef. * %options is optional. (e.g. "skip_normalize" allows you to skip changing : to = ) =item C Returns an array consisting of three elements * $error is undefined unless an error is detected * $results is a reference to an array of records. * $total_hits is the number of hits that would have been returned with no limit If an error is returned the two other return elements are undefined. If error itself is undefined the other two elements are always defined =item C =back my ( $error, $marcresults, $total_hits ) = SimpleSearch($query); if (defined $error) { $template->param(query_error => $error); warn "error: ".$error; output_html_with_http_headers $input, $cookie, $template->output; exit; } my $hits = @{$marcresults}; my @results; for my $r ( @{$marcresults} ) { my $marcrecord = MARC::File::USMARC::decode($r); my $biblio = TransformMarcToKoha({ record => $marcrecord }); #build the iarray of hashs for the template. push @results, { title => $biblio->{'title'}, subtitle => $biblio->{'subtitle'}, biblionumber => $biblio->{'biblionumber'}, author => $biblio->{'author'}, publishercode => $biblio->{'publishercode'}, publicationyear => $biblio->{'publicationyear'}, }; } $template->param(result=>\@results); =cut sub SimpleSearch { my ( $query, $offset, $max_results, $servers, %options ) = @_; return ( 'No query entered', undef, undef ) unless $query; # FIXME hardcoded value. See catalog/search.pl & opac-search.pl too. my @servers = defined ( $servers ) ? @$servers : ( 'biblioserver' ); my @zoom_queries; my @tmpresults; my @zconns; my $results = []; my $total_hits = 0; # Initialize & Search Zebra for ( my $i = 0 ; $i < @servers ; $i++ ) { eval { $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 ); $query =~ s/:/=/g unless $options{skip_normalize}; $zoom_queries[$i] = ZOOM::Query::CCL2RPN->new( $query, $zconns[$i]); $tmpresults[$i] = $zconns[$i]->search( $zoom_queries[$i] ); # error handling my $error = $zconns[$i]->errmsg() . " (" . $zconns[$i]->errcode() . ") " . $zconns[$i]->addinfo() . " " . $zconns[$i]->diagset(); return ( $error, undef, undef ) if $zconns[$i]->errcode(); }; if ($@) { # caught a ZOOM::Exception my $error = $@->message() . " (" . $@->code() . ") " . $@->addinfo() . " " . $@->diagset(); warn $error." for query: $query"; return ( $error, undef, undef ); } } _ZOOM_event_loop( \@zconns, \@tmpresults, sub { my ($i, $size) = @_; my $first_record = defined($offset) ? $offset + 1 : 1; my $hits = $tmpresults[ $i - 1 ]->size(); $total_hits += $hits; my $last_record = $hits; if ( defined $max_results && $offset + $max_results < $hits ) { $last_record = $offset + $max_results; } for my $j ( $first_record .. $last_record ) { my $record = eval { $tmpresults[ $i - 1 ]->record( $j - 1 )->raw() ; # 0 indexed }; push @{$results}, $record if defined $record; } } ); foreach my $zoom_query (@zoom_queries) { $zoom_query->destroy(); } return ( undef, $results, $total_hits ); } =head2 getRecords ( undef, $results_hashref, \@facets_loop ) = getRecords ( $koha_query, $simple_query, $sort_by_ref, $servers_ref, $results_per_page, $offset, $branches, $itemtypes, $query_type, $scan, $opac ); The all singing, all dancing, multi-server, asynchronous, scanning, searching, record nabbing, facet-building See verbose embedded documentation. =cut sub getRecords { my ( $koha_query, $simple_query, $sort_by_ref, $servers_ref, $results_per_page, $offset, $branches, $itemtypes, $query_type, $scan, $opac ) = @_; my @servers = @$servers_ref; my @sort_by = @$sort_by_ref; $offset = 0 if $offset < 0; # Initialize variables for the ZOOM connection and results object my @zconns; my @results; my $results_hashref = (); # TODO simplify this structure ( { branchcode => $branchname } is enought) and remove this parameter $branches ||= { map { $_->branchcode => { branchname => $_->branchname } } Koha::Libraries->search->as_list }; # Initialize variables for the faceted results objects my $facets_counter = {}; my $facets_info = {}; my $facets = getFacets(); my @facets_loop; # stores the ref to array of hashes for template facets loop ### LOOP THROUGH THE SERVERS for ( my $i = 0 ; $i < @servers ; $i++ ) { $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 ); # perform the search, create the results objects # if this is a local search, use the $koha-query, if it's a federated one, use the federated-query my $query_to_use = ($servers[$i] =~ /biblioserver/) ? $koha_query : $simple_query; Koha::Logger->get->debug($simple_query) if $scan; # Check if we've got a query_type defined, if so, use it eval { if ($query_type) { if ($query_type =~ /^ccl/) { $query_to_use =~ s/\:/\=/g; # change : to = last minute (FIXME) $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i])); } elsif ($query_type =~ /^cql/) { $results[$i] = $zconns[$i]->search(ZOOM::Query::CQL->new($query_to_use, $zconns[$i])); } elsif ($query_type =~ /^pqf/) { $results[$i] = $zconns[$i]->search(ZOOM::Query::PQF->new($query_to_use, $zconns[$i])); } else { warn "Unknown query_type '$query_type'. Results undetermined."; } } elsif ($scan) { $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i])); } else { $results[$i] = $zconns[$i]->search(ZOOM::Query::CCL2RPN->new($query_to_use, $zconns[$i])); } }; if ($@) { warn "WARNING: query problem with $query_to_use " . $@; } # Concatenate the sort_by limits and pass them to the results object # Note: sort will override rank my $sort_by; foreach my $sort (@sort_by) { if ( $sort eq "author_az" || $sort eq "author_asc" ) { $sort_by .= "1=1003 sort( "yaz", $sort_by ) < 0 ) { warn "WARNING sort $sort_by failed"; } } } # finished looping through servers # The big moment: asynchronously retrieve results from all servers _ZOOM_event_loop( \@zconns, \@results, sub { my ( $i, $size ) = @_; my $results_hash; # loop through the results $results_hash->{'hits'} = $size; my $times; if ( $offset + $results_per_page <= $size ) { $times = $offset + $results_per_page; } else { $times = $size; } for ( my $j = $offset ; $j < $times ; $j++ ) { my $record; ## Check if it's an index scan if ($scan) { my ( $term, $occ ) = $results[ $i - 1 ]->display_term($j); # here we create a minimal MARC record and hand it off to the # template just like a normal result ... perhaps not ideal, but # it works for now my $tmprecord = MARC::Record->new(); $tmprecord->encoding('UTF-8'); my $tmptitle; my $tmpauthor; # the minimal record in author/title (depending on MARC flavour) if ( C4::Context->preference("marcflavour") eq "UNIMARC" ) { $tmptitle = MARC::Field->new( '200', ' ', ' ', a => $term, f => $occ ); $tmprecord->append_fields($tmptitle); } else { $tmptitle = MARC::Field->new( '245', ' ', ' ', a => $term, ); $tmpauthor = MARC::Field->new( '100', ' ', ' ', a => $occ, ); $tmprecord->append_fields($tmptitle); $tmprecord->append_fields($tmpauthor); } $results_hash->{'RECORDS'}[$j] = $tmprecord->as_usmarc(); } # not an index scan else { $record = $results[ $i - 1 ]->record($j)->raw(); # warn "RECORD $j:".$record; $results_hash->{'RECORDS'}[$j] = $record; } } $results_hashref->{ $servers[ $i - 1 ] } = $results_hash; # Fill the facets while we're looping, but only for the # biblioserver and not for a scan if ( !$scan && $servers[ $i - 1 ] =~ /biblioserver/ ) { $facets_counter = GetFacets( $results[ $i - 1 ] ); $facets_info = _get_facets_info( $facets ); } # BUILD FACETS if ( $servers[ $i - 1 ] =~ /biblioserver/ ) { for my $link_value ( sort { $a cmp $b } keys %$facets_counter ) { my @this_facets_array; for my $one_facet ( sort { $facets_counter->{$link_value} ->{$b} <=> $facets_counter->{$link_value} ->{$a} } keys %{ $facets_counter->{$link_value} } ) { # Sanitize the link value : parenthesis, question and exclamation mark will cause errors with CCL my $facet_link_value = $one_facet; $facet_link_value =~ s/[()!?¡¿؟]/ /g; # fix the length that will display in the label, my $facet_label_value = $one_facet; my $facet_max_length = C4::Context->preference( 'FacetLabelTruncationLength') || 20; $facet_label_value = substr( $one_facet, 0, $facet_max_length ) . "..." if length($facet_label_value) > $facet_max_length; # if it's a branch, label by the name, not the code, if ( $link_value =~ /branch/ ) { if ( defined $branches && ref($branches) eq "HASH" && defined $branches->{$one_facet} && ref( $branches->{$one_facet} ) eq "HASH" ) { $facet_label_value = $branches->{$one_facet} ->{'branchname'}; } else { $facet_label_value = "*"; } } # if it's a itemtype, label by the name, not the code, if ( $link_value =~ /itype/ ) { if ( defined $itemtypes && ref($itemtypes) eq "HASH" && defined $itemtypes->{$one_facet} && ref( $itemtypes->{$one_facet} ) eq "HASH" ) { $facet_label_value = $itemtypes->{$one_facet} ->{translated_description}; } } # also, if it's a location code, use the name instead of the code if ( $link_value =~ /location/ ) { # TODO Retrieve all authorised values at once, instead of 1 query per entry my $av = Koha::AuthorisedValues->search({ category => 'LOC', authorised_value => $one_facet }); $facet_label_value = $av->count ? $av->next->opac_description : ''; } # also, if it's a collection code, use the name instead of the code if ( $link_value =~ /ccode/ ) { # TODO Retrieve all authorised values at once, instead of 1 query per entry my $av = Koha::AuthorisedValues->search({ category => 'CCODE', authorised_value => $one_facet }); $facet_label_value = $av->count ? $av->next->opac_description : ''; } # but we're down with the whole label being in the link's title. push @this_facets_array, { facet_count => $facets_counter->{$link_value} ->{$one_facet}, facet_label_value => $facet_label_value, facet_title_value => $one_facet, facet_link_value => $facet_link_value, type_link_value => $link_value, } if ($facet_label_value); } push @facets_loop, { type_link_value => $link_value, type_id => $link_value . "_id", "type_label_" . $facets_info->{$link_value}->{'label_value'} => 1, label => $facets_info->{$link_value}->{'label_value'}, facets => \@this_facets_array, } unless ( ( $facets_info->{$link_value}->{'label_value'} =~ /Libraries/ ) and ( Koha::Libraries->search->count == 1 ) ); } } } ); # This sorts the facets into alphabetical order if (@facets_loop) { foreach my $f (@facets_loop) { if( C4::Context->preference('FacetOrder') eq 'Alphabetical' ){ $f->{facets} = [ sort { uc($a->{facet_label_value}) cmp uc($b->{facet_label_value}) } @{ $f->{facets} } ]; } } } return ( undef, $results_hashref, \@facets_loop ); } sub GetFacets { my $rs = shift; my $facets; my $use_zebra_facets = C4::Context->config('use_zebra_facets') // 0; if ( $use_zebra_facets ) { $facets = _get_facets_from_zebra( $rs ); } else { $facets = _get_facets_from_records( $rs ); } return $facets; } sub _get_facets_from_records { my $rs = shift; my $facets_maxrecs = C4::Context->preference('maxRecordsForFacets') // 20; my $facets_config = getFacets(); my $facets = {}; my $size = $rs->size(); my $jmax = $size > $facets_maxrecs ? $facets_maxrecs : $size; for ( my $j = 0 ; $j < $jmax ; $j++ ) { my $marc_record = new_record_from_zebra ( 'biblioserver', $rs->record( $j )->raw() ); if ( ! defined $marc_record ) { warn "ERROR DECODING RECORD - $@: " . $rs->record( $j )->raw(); next; } _get_facets_data_from_record( $marc_record, $facets_config, $facets ); } return $facets; } =head2 _get_facets_data_from_record C4::Search::_get_facets_data_from_record( $marc_record, $facets, $facets_counter ); Internal function that extracts facets information from a MARC::Record object and populates $facets_counter for using in getRecords. $facets is expected to be filled with C4::Koha::getFacets output (i.e. the configured facets for Zebra). =cut sub _get_facets_data_from_record { my ( $marc_record, $facets, $facets_counter ) = @_; for my $facet (@$facets) { my @used_datas = (); foreach my $tag ( @{ $facet->{ tags } } ) { # tag number is the first three digits my $tag_num = substr( $tag, 0, 3 ); # subfields are the remainder my $subfield_letters = substr( $tag, 3 ); my @fields = $marc_record->field( $tag_num ); foreach my $field (@fields) { # If $field->indicator(1) eq 'z', it means it is a 'see from' # field introduced because of IncludeSeeFromInSearches, so skip it next if $field->indicator(1) eq 'z'; my $data = $field->as_string( $subfield_letters, $facet->{ sep } ); $data =~ s/\s*(?{ $facet->{ idx } }->{ $data }++; } } } } } =head2 _get_facets_from_zebra my $facets = _get_facets_from_zebra( $result_set ) Retrieves facets for a specified result set. It loops through the facets defined in C4::Koha::getFacets and returns a hash with the following structure: { facet_idx => { facet_value => count }, ... } =cut sub _get_facets_from_zebra { my $rs = shift; # save current elementSetName my $elementSetName = $rs->option( 'elementSetName' ); my $facets_loop = getFacets(); my $facets_data = {}; # loop through defined facets and fill the facets hashref foreach my $facet ( @$facets_loop ) { my $idx = $facet->{ idx }; my $sep = $facet->{ sep }; my $facet_values = _get_facet_from_result_set( $idx, $rs, $sep ); if ( $facet_values ) { # we've actually got a result $facets_data->{ $idx } = $facet_values; } } # set elementSetName to its previous value to avoid side effects $rs->option( elementSetName => $elementSetName ); return $facets_data; } =head2 _get_facet_from_result_set my $facet_values = C4::Search::_get_facet_from_result_set( $facet_idx, $result_set, $sep ) Internal function that extracts facet information for a specific index ($facet_idx) and returns a hash containing facet values and count: { $facet_value => $count , ... } Warning: this function has the side effect of changing the elementSetName for the result set. It is a helper function for the main loop, which takes care of backing it up for restoring. =cut sub _get_facet_from_result_set { my $facet_idx = shift; my $rs = shift; my $sep = shift; my $internal_sep = '<*>'; my $facetMaxCount = C4::Context->preference('FacetMaxCount') // 20; return if ( ! defined $facet_idx || ! defined $rs ); # zebra's facet element, untokenized index my $facet_element = 'zebra::facet::' . $facet_idx . ':0:' . $facetMaxCount; # configure zebra results for retrieving the desired facet $rs->option( elementSetName => $facet_element ); # get the facet record from result set my $facet = $rs->record( 0 )->raw; # if the facet has no restuls... return if !defined $facet; # TODO: benchmark DOM vs. SAX performance my $facet_dom = XML::LibXML->load_xml( string => ($facet) ); my @terms = $facet_dom->getElementsByTagName('term'); return if ! @terms; my $facets = {}; foreach my $term ( @terms ) { my $facet_value = $term->textContent; $facet_value =~ s/\s*(?{ $facet_value } += $term->getAttribute( 'occur' ); } return $facets; } =head2 _get_facets_info my $facets_info = C4::Search::_get_facets_info( $facets ) Internal function that extracts facets information and properly builds the data structure needed to render facet labels. =cut sub _get_facets_info { my $facets = shift; my $facets_info = {}; for my $facet ( @$facets ) { $facets_info->{ $facet->{ idx } }->{ label_value } = $facet->{ label }; } return $facets_info; } # TRUNCATION sub _detect_truncation { my ( $operand, $index ) = @_; my ( @nontruncated, @righttruncated, @lefttruncated, @rightlefttruncated, @regexpr ); $operand =~ s/^ //g; my @wordlist = split( /\s/, $operand ); foreach my $word (@wordlist) { if ( $word =~ s/^\*([^\*]+)\*$/$1/ ) { push @rightlefttruncated, $word; } elsif ( $word =~ s/^\*([^\*]+)$/$1/ ) { push @lefttruncated, $word; } elsif ( $word =~ s/^([^\*]+)\*$/$1/ ) { push @righttruncated, $word; } elsif ( index( $word, "*" ) < 0 ) { push @nontruncated, $word; } else { push @regexpr, $word; } } return ( \@nontruncated, \@righttruncated, \@lefttruncated, \@rightlefttruncated, \@regexpr ); } # STEMMING sub _build_stemmed_operand { my ($operand,$lang) = @_; require Lingua::Stem::Snowball ; my $stemmed_operand=q{}; # Stemmer needs language return $operand unless $lang; # If operand contains a digit, it is almost certainly an identifier, and should # not be stemmed. This is particularly relevant for ISBNs and ISSNs, which # can contain the letter "X" - for example, _build_stemmend_operand would reduce # "014100018X" to "x ", which for a MARC21 database would bring up irrelevant # results (e.g., "23 x 29 cm." from the 300$c). Bug 2098. return $operand if $operand =~ /\d/; # FIXME: the locale should be set based on the user's language and/or search choice #warn "$lang"; # Make sure we only use the first two letters from the language code $lang = lc(substr($lang, 0, 2)); # The language codes for the two variants of Norwegian will now be "nb" and "nn", # none of which Lingua::Stem::Snowball can use, so we need to "translate" them if ($lang eq 'nb' || $lang eq 'nn') { $lang = 'no'; } my $stemmer = Lingua::Stem::Snowball->new( lang => $lang, encoding => "UTF-8" ); my @words = split( / /, $operand ); my @stems = $stemmer->stem(\@words); for my $stem (@stems) { $stemmed_operand .= "$stem"; $stemmed_operand .= "?" unless ( $stem =~ /(and$|or$|not$)/ ) || ( length($stem) < 3 ); $stemmed_operand .= " "; } Koha::Logger->get->debug("STEMMED OPERAND: $stemmed_operand"); return $stemmed_operand; } # FIELD WEIGHTING sub _build_weighted_query { # FIELD WEIGHTING - This is largely experimental stuff. What I'm committing works # pretty well but could work much better if we had a smarter query parser my ( $operand, $stemmed_operand, $index ) = @_; my $stemming = C4::Context->preference("QueryStemming") || 0; my $weight_fields = C4::Context->preference("QueryWeightFields") || 0; my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0; $operand =~ s/"/ /g; # Bug 7518: searches with quotation marks don't work my $weighted_query = "(rk=("; # Specifies that we're applying rank # Keyword, or, no index specified if ( ( $index eq 'kw' ) || ( !$index ) ) { $weighted_query .= "Title-cover,ext,r1=\"$operand\""; # exact title-cover $weighted_query .= " or ti,ext,r2=\"$operand\""; # exact title $weighted_query .= " or Title-cover,phr,r3=\"$operand\""; # phrase title $weighted_query .= " or ti,wrdl,r4=\"$operand\""; # words in title #$weighted_query .= " or any,ext,r4=$operand"; # exact any #$weighted_query .=" or kw,wrdl,r5=\"$operand\""; # word list any $weighted_query .= " or wrdl,fuzzy,r8=\"$operand\"" if $fuzzy_enabled; # add fuzzy, word list $weighted_query .= " or wrdl,right-Truncation,r9=\"$stemmed_operand\"" if ( $stemming and $stemmed_operand ) ; # add stemming, right truncation $weighted_query .= " or wrdl,r9=\"$operand\""; # embedded sorting: 0 a-z; 1 z-a # $weighted_query .= ") or (sort1,aut=1"; } # Barcode searches should skip this process elsif ( $index eq 'bc' ) { $weighted_query .= "bc=\"$operand\""; } # Authority-number searches should skip this process elsif ( $index eq 'an' ) { $weighted_query .= "an=\"$operand\""; } # If the index is numeric, don't autoquote it. elsif ( $index =~ /,st-numeric$/ ) { $weighted_query .= " $index=$operand"; } # If the index already has more than one qualifier, wrap the operand # in quotes and pass it back (assumption is that the user knows what they # are doing and won't appreciate us mucking up their query elsif ( $index =~ ',' ) { $weighted_query .= " $index=\"$operand\""; } #TODO: build better cases based on specific search indexes else { $weighted_query .= " $index,ext,r1=\"$operand\""; # exact index #$weighted_query .= " or (title-sort-az=0 or $index,startswithnt,st-word,r3=$operand #)"; $weighted_query .= " or $index,phr,r3=\"$operand\""; # phrase index $weighted_query .= " or $index,wrdl,r6=\"$operand\""; # word list index $weighted_query .= " or $index,wrdl,fuzzy,r8=\"$operand\"" if $fuzzy_enabled; # add fuzzy, word list $weighted_query .= " or $index,wrdl,rt,r9=\"$stemmed_operand\"" if ( $stemming and $stemmed_operand ); # add stemming, right truncation } $weighted_query .= "))"; # close rank specification return $weighted_query; } =head2 getIndexes Return an array with available indexes. =cut sub getIndexes{ my @indexes = ( # biblio indexes 'ab', 'Abstract', 'acqdate', 'allrecords', 'an', 'Any', 'at', 'arl', 'arp', 'au', 'aub', 'aud', 'audience', 'auo', 'aut', 'Author', 'Author-in-order ', 'Author-personal-bibliography', 'Authority-Number', 'authtype', 'bc', 'Bib-level', 'biblionumber', 'bio', 'biography', 'callnum', 'cfn', 'Chronological-subdivision', 'cn-bib-source', 'cn-bib-sort', 'cn-class', 'cn-item', 'cn-prefix', 'cn-suffix', 'cpn', 'Code-institution', 'Conference-name', 'Conference-name-heading', 'Conference-name-see', 'Conference-name-seealso', 'Content-type', 'Control-number', 'cnum', 'Control-number-identifier', 'cni', 'copydate', 'Corporate-name', 'Corporate-name-heading', 'Corporate-name-see', 'Corporate-name-seealso', 'Country-publication', 'ctype', 'curriculum', 'date-entered-on-file', 'Date-of-acquisition', 'Date-of-publication', 'Date-time-last-modified', 'Dewey-classification', 'Dissertation-information', 'diss', 'dtlm', 'EAN', 'extent', 'fic', 'fiction', 'Form-subdivision', 'format', 'Geographic-subdivision', 'he', 'Heading', 'Heading-use-main-or-added-entry', 'Heading-use-series-added-entry ', 'Heading-use-subject-added-entry', 'Host-item', 'id-other', 'ident', 'Identifier-standard', 'Illustration-code', 'Index-term-genre', 'Index-term-uncontrolled', 'Interest-age-level', 'Interest-grade-level', 'ISBN', 'isbn', 'ISSN', 'issn', 'itemtype', 'kw', 'Koha-Auth-Number', 'l-format', 'language', 'language-original', 'lc-card', 'LC-card-number', 'lcn', 'lex', 'lexile-number', 'llength', 'ln', 'ln-audio', 'ln-subtitle', 'Local-classification', 'Local-number', 'Match-heading', 'Match-heading-see-from', 'Material-type', 'mc-itemtype', 'mc-rtype', 'mus', 'Multipart-resource-level', 'mrl', 'name', 'Music-number', 'Name-geographic', 'Name-geographic-heading', 'Name-geographic-see', 'Name-geographic-seealso', 'nb', 'Note', 'notes', 'ns', 'nt', 'Other-control-number', 'pb', 'Personal-name', 'Personal-name-heading', 'Personal-name-see', 'Personal-name-seealso', 'pl', 'Place-publication', 'pn', 'popularity', 'pubdate', 'Publisher', 'Provider', 'pv', 'Reading-grade-level', 'Record-control-number', 'rcn', 'Record-type', 'rtype', 'se', 'See', 'See-also', 'sn', 'Stock-number', 'su', 'Subject', 'Subject-heading-thesaurus', 'Subject-name-personal', 'Subject-subdivision', 'Summary', 'Suppress', 'su-geo', 'su-na', 'su-to', 'su-ut', 'ut', 'Term-genre-form', 'Term-genre-form-heading', 'Term-genre-form-see', 'Term-genre-form-seealso', 'ti', 'Title', 'Title-cover', 'Title-series', 'Title-uniform', 'Title-uniform-heading', 'Title-uniform-see', 'Title-uniform-seealso', 'totalissues', 'yr', # items indexes 'acqsource', 'barcode', 'bc', 'branch', 'ccode', 'classification-source', 'cn-sort', 'coded-location-qualifier', 'copynumber', 'damaged', 'datelastborrowed', 'datelastseen', 'holdingbranch', 'homebranch', 'issues', 'item', 'itemnumber', 'itype', 'Local-classification', 'location', 'lost', 'materials-specified', 'mc-ccode', 'mc-itype', 'mc-loc', 'notforloan', 'Number-local-acquisition', 'onloan', 'price', 'renewals', 'replacementprice', 'replacementpricedate', 'reserves', 'restricted', 'stack', 'stocknumber', 'inv', 'uri', 'withdrawn', # subject related ); return \@indexes; } =head2 buildQuery ( $error, $query, $simple_query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc, $query_type ) = buildQuery ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang); Build queries and limits in CCL, CGI, Human, handle truncation, stemming, field weighting, fuzziness, etc. See verbose embedded documentation. =cut sub buildQuery { my ( $operators, $operands, $indexes, $limits, $sort_by, $scan, $lang) = @_; my $query_desc; # dereference my @operators = $operators ? @$operators : (); my @indexes = $indexes ? @$indexes : (); my @operands = $operands ? @$operands : (); my @limits = $limits ? @$limits : (); my @sort_by = $sort_by ? @$sort_by : (); my $stemming = C4::Context->preference("QueryStemming") || 0; my $auto_truncation = C4::Context->preference("QueryAutoTruncate") || 0; my $weight_fields = C4::Context->preference("QueryWeightFields") || 0; my $fuzzy_enabled = C4::Context->preference("QueryFuzzy") || 0; my $query = $operands[0] // ""; my $simple_query = $operands[0]; # initialize the variables we're passing back my $query_cgi; my $query_type; my $limit = q{}; my $limit_cgi; my $limit_desc; my $cclq = 0; my $cclindexes = getIndexes(); if ( $query !~ /\s*(ccl=|pqf=|cql=)/ ) { while ( !$cclq && $query =~ /(?:^|\W)([\w-]+)(,[\w-]+)*[:=]/g ) { my $dx = lc($1); $cclq = grep { lc($_) eq $dx } @$cclindexes; } $query = "ccl=$query" if $cclq; } # add limits my %group_OR_limits; my $availability_limit; foreach my $this_limit (@limits) { next unless $this_limit; if ( $this_limit =~ /available/ ) { # ## 'available' is defined as (items.onloan is NULL) and (items.itemlost = 0) ## In English: ## all records not indexed in the onloan register (zebra) and all records with a value of lost equal to 0 $availability_limit .= "( (allrecords,AlwaysMatches='') and (not-onloan-count,st-numeric >= 1) and (lost,st-numeric=0) )"; $limit_cgi .= "&limit=available"; $limit_desc .= ""; } # group_OR_limits, prefixed by mc- # OR every member of the group elsif ( $this_limit =~ /mc/ ) { my ($k,$v) = split(/:/, $this_limit,2); if ( $k !~ /mc-i(tem)?type/ ) { # in case the mc-ccode value has complicating chars like ()'s inside it we wrap in quotes $this_limit =~ tr/"//d; $this_limit = $k.':"'.$v.'"'; } $group_OR_limits{$k} .= " or " if $group_OR_limits{$k}; $limit_desc .= " or " if $group_OR_limits{$k}; $group_OR_limits{$k} .= "$this_limit"; $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit); $limit_desc .= " $this_limit"; } elsif ( $this_limit =~ '^multibranchlimit:|^branch:' ) { $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit); $limit .= " and " if $limit || $query; my $branchfield = C4::Context->preference('SearchLimitLibrary'); my @branchcodes; if( $this_limit =~ '^multibranchlimit:' ){ my ($group_id) = ( $this_limit =~ /^multibranchlimit:(.*)$/ ); my $search_group = Koha::Library::Groups->find( $group_id ); @branchcodes = map { $_->branchcode } $search_group->all_libraries; @branchcodes = sort { $a cmp $b } @branchcodes; } else { @branchcodes = ( $this_limit =~ /^branch:(.*)$/ ); } if (@branchcodes) { if ( $branchfield eq "homebranch" ) { $this_limit = sprintf "(%s)", join " or ", map { 'homebranch: ' . $_ } @branchcodes; } elsif ( $branchfield eq "holdingbranch" ) { $this_limit = sprintf "(%s)", join " or ", map { 'holdingbranch: ' . $_ } @branchcodes; } else { $this_limit = sprintf "(%s or %s)", join( " or ", map { 'homebranch: ' . $_ } @branchcodes ), join( " or ", map { 'holdingbranch: ' . $_ } @branchcodes ); } } $limit .= "$this_limit"; $limit_desc .= " $this_limit"; } elsif ( $this_limit =~ '^search_filter:' ) { # Here we will get the query as a string, append to the limits, and pass through buildQuery # again to clean the terms and handle nested filters $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit); my ($filter_id) = ( $this_limit =~ /^search_filter:(.*)$/ ); my $search_filter = Koha::SearchFilters->find( $filter_id ); next unless $search_filter; my ($expanded_lim, $query_lim) = $search_filter->expand_filter; push @$expanded_lim, $query_lim; my ( $error, undef, undef, undef, undef, $fixed_limit, undef, undef, undef ) = buildQuery ( undef, undef, undef, $expanded_lim, undef, undef, $lang); $limit .= " and " if $limit || $query; $limit .= "$fixed_limit"; $limit_desc .= " $limit"; } # Regular old limits else { $limit .= " and " if $limit || $query; $limit .= "$this_limit"; $limit_cgi .= "&limit=" . uri_escape_utf8($this_limit); $limit_desc .= " $this_limit"; } } foreach my $k (keys (%group_OR_limits)) { $limit .= " and " if ( $query || $limit ); $limit .= "($group_OR_limits{$k})"; } if ($availability_limit) { $limit .= " and " if ( $query || $limit ); $limit .= "($availability_limit)"; } # for handling ccl, cql, pqf queries in diagnostic mode, skip the rest of the steps # DIAGNOSTIC ONLY!! if ( $query =~ /^ccl=/ ) { my $q=$'; # This is needed otherwise ccl= and &limit won't work together, and # this happens when selecting a subject on the opac-detail page my $original_q = $q; # without available part $q .= $limit if $limit; return ( undef, $q, $q, "q=ccl=".uri_escape_utf8($q), $original_q, '', '', '', 'ccl' ); } if ( $query =~ /^cql=/ ) { return ( undef, $', $', "q=cql=".uri_escape_utf8($'), $', '', '', '', 'cql' ); } if ( $query =~ /^pqf=/ ) { $query_desc = $'; $query_cgi = "q=pqf=".uri_escape_utf8($'); return ( undef, $', $', $query_cgi, $query_desc, '', '', '', 'pqf' ); } # pass nested queries directly # FIXME: need better handling of some of these variables in this case # Nested queries aren't handled well and this implementation is flawed and causes users to be # unable to search for anything containing () commenting out, will be rewritten for 3.4.0 # if ( $query =~ /(\(|\))/ ) { # return ( # undef, $query, $simple_query, $query_cgi, # $query, $limit, $limit_cgi, $limit_desc, # 'ccl' # ); # } # Form-based queries are non-nested and fixed depth, so we can easily modify the incoming # query operands and indexes and add stemming, truncation, field weighting, etc. # Once we do so, we'll end up with a value in $query, just like if we had an # incoming $query from the user else { $query = "" ; # clear it out so we can populate properly with field-weighted, stemmed, etc. query my $previous_operand ; # a flag used to keep track if there was a previous query # if there was, we can apply the current operator # for every operand for ( my $i = 0 ; $i <= @operands ; $i++ ) { # COMBINE OPERANDS, INDEXES AND OPERATORS if ( ($operands[$i] // '') ne '' ) { $operands[$i]=~s/^\s+//; # A flag to determine whether or not to add the index to the query my $indexes_set; # If the user is sophisticated enough to specify an index, turn off field weighting, and stemming handling if ( $operands[$i] =~ /\w(:|=)/ || $scan ) { $weight_fields = 0; $stemming = 0; } else { $operands[$i] =~ s/\?/{?}/g; # need to escape question marks } my $operand = $operands[$i]; my $index = $indexes[$i] || 'kw'; # Add index-specific attributes #Afaik, this 'yr' condition will only ever be met in the staff interface advanced search #for "Publication date", since typing 'yr:YYYY' into the search box produces a CCL query, #which is processed higher up in this sub. Other than that, year searches are typically #handled as limits which are not processed her either. # Search ranges: Date of Publication, st-numeric if ( $index =~ /(yr|st-numeric)/ ) { #weight_fields/relevance search causes errors with date ranges #In the case of YYYY-, it will only return records with a 'yr' of YYYY (not the range) #In the case of YYYY-YYYY, it will return no results $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0; } # Date of Acquisition elsif ( $index =~ /acqdate/ ) { #stemming and auto_truncation would have zero impact since it already is YYYY-MM-DD format #Weight_fields probably SHOULD be turned OFF, otherwise you'll get records floating to the #top of the results just because they have lots of item records matching that date. #Fuzzy actually only applies during _build_weighted_query, and is reset there anyway, so #irrelevant here $stemming = $auto_truncation = $weight_fields = $fuzzy_enabled = 0; } # ISBN,ISSN,Standard Number, don't need special treatment elsif ( $index eq 'nb' || $index eq 'ns' || $index eq 'hi' ) { ( $stemming, $auto_truncation, $weight_fields, $fuzzy_enabled ) = ( 0, 0, 0, 0 ); if ( $index eq 'nb' ) { if ( C4::Context->preference("SearchWithISBNVariations") ) { my @isbns = C4::Koha::GetVariationsOfISBN( $operand ); $operands[$i] = $operand = '(' . join( ' OR ', map { 'nb=' . $_ } @isbns ) . ')'; $indexes[$i] = $index = 'kw'; } } if ( $index eq 'ns' ) { if ( C4::Context->preference("SearchWithISSNVariations") ) { my @issns = C4::Koha::GetVariationsOfISSN( $operand ); $operands[$i] = $operand = '(' . join( ' OR ', map { 'ns=' . $_ } @issns ) . ')'; $indexes[$i] = $index = 'kw'; } } } # Set default structure attribute (word list) my $struct_attr = q{}; unless ( $indexes_set || $index =~ /,(st-|phr|ext|wrdl)/ || $index =~ /^(nb|ns)$/ ) { $struct_attr = ",wrdl"; } # Some helpful index variants my $index_plus = $index . $struct_attr . ':'; my $index_plus_comma = $index . $struct_attr . ','; if ($auto_truncation){ unless ( $index =~ /,(st-|phr|ext)/ ) { #FIXME only valid with LTR scripts $operand=join(" ",map{ (index($_,"*")>0?"$_":"$_*") }split (/\s+/,$operand)); } } # Detect Truncation my $truncated_operand = q{}; my( $nontruncated, $righttruncated, $lefttruncated, $rightlefttruncated, $regexpr ) = _detect_truncation( $operand, $index ); Koha::Logger->get->debug( "TRUNCATION: NON:>@$nontruncated< RIGHT:>@$righttruncated< LEFT:>@$lefttruncated< RIGHTLEFT:>@$rightlefttruncated< REGEX:>@$regexpr<"); # Apply Truncation if ( scalar(@$righttruncated) + scalar(@$lefttruncated) + scalar(@$rightlefttruncated) > 0 ) { # Don't field weight or add the index to the query, we do it here $indexes_set = 1; undef $weight_fields; my $previous_truncation_operand; if (scalar @$nontruncated) { $truncated_operand .= "$index_plus @$nontruncated "; $previous_truncation_operand = 1; } if (scalar @$righttruncated) { $truncated_operand .= "and " if $previous_truncation_operand; $truncated_operand .= $index_plus_comma . "rtrn:@$righttruncated "; $previous_truncation_operand = 1; } if (scalar @$lefttruncated) { $truncated_operand .= "and " if $previous_truncation_operand; $truncated_operand .= $index_plus_comma . "ltrn:@$lefttruncated "; $previous_truncation_operand = 1; } if (scalar @$rightlefttruncated) { $truncated_operand .= "and " if $previous_truncation_operand; $truncated_operand .= $index_plus_comma . "rltrn:@$rightlefttruncated "; $previous_truncation_operand = 1; } } $operand = $truncated_operand if $truncated_operand; Koha::Logger->get->debug("TRUNCATED OPERAND: >$truncated_operand<"); # Handle Stemming my $stemmed_operand = q{}; $stemmed_operand = _build_stemmed_operand($operand, $lang) if $stemming; Koha::Logger->get->debug("STEMMED OPERAND: >$stemmed_operand<"); # Handle Field Weighting my $weighted_operand = q{}; if ($weight_fields) { $weighted_operand = _build_weighted_query( $operand, $stemmed_operand, $index ); $operand = $weighted_operand; $indexes_set = 1; } Koha::Logger->get->debug("FIELD WEIGHTED OPERAND: >$weighted_operand<"); #Use relevance ranking when not using a weighted query (which adds relevance ranking of its own) #N.B. Truncation is mutually exclusive with Weighted Queries, #so even if QueryWeightFields is turned on, QueryAutoTruncate will turn it off, thus #the need for this relevance wrapper. $operand = "(rk=($operand))" unless $weight_fields; ($query,$query_cgi,$query_desc,$previous_operand) = _build_initial_query({ query => $query, query_cgi => $query_cgi, query_desc => $query_desc, operator => ($operators[ $i - 1 ]) ? $operators[ $i - 1 ] : '', parsed_operand => $operand, original_operand => $operands[$i] // '', index => $index, index_plus => $index_plus, indexes_set => $indexes_set, previous_operand => $previous_operand, }); } #/if $operands } # /for } Koha::Logger->get->debug("QUERY BEFORE LIMITS: >$query<"); # Normalize the query and limit strings # This is flawed , means we can't search anything with : in it # if user wants to do ccl or cql, start the query with that # $query =~ s/:/=/g; #NOTE: We use several several different regexps here as you can't have variable length lookback assertions $query =~ s/(?<=(ti|au|pb|su|an|kw|mc|nb|ns)):/=/g; $query =~ s/(?<=(wrdl)):/=/g; $query =~ s/(?<=(trn|phr)):/=/g; $query =~ s/(?<=(st-numeric)):/=/g; $query =~ s/(?<=(st-year)):/=/g; $query =~ s/(?<=(st-date-normalized)):/=/g; # Removing warnings for later substitutions $query //= q{}; $query_desc //= q{}; $query_cgi //= q{}; $limit //= q{}; $limit_desc //= q{}; $limit_cgi //= q{}; $simple_query //= q{}; $limit =~ s/:/=/g; for ( $query, $query_desc, $limit, $limit_desc ) { s/ +/ /g; # remove extra spaces s/^ //g; # remove any beginning spaces s/ $//g; # remove any ending spaces s/==/=/g; # remove double == from query } $query_cgi =~ s/^&//; # remove unnecessary & from beginning of the query cgi for ($query_cgi,$simple_query) { s/"//g; } # append the limit to the query $query .= " " . $limit; Koha::Logger->get->debug( sprintf "buildQuery returns\nQUERY:%s\nQUERY CGI:%s\nQUERY DESC:%s\nLIMIT:%s\nLIMIT CGI:%s\nLIMIT DESC:%s", $query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc ); return ( undef, $query, $simple_query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc, $query_type ); } =head2 _build_initial_query ($query, $query_cgi, $query_desc, $previous_operand) = _build_initial_query($initial_query_params); Build a section of the initial query containing indexes, operators, and operands. =cut sub _build_initial_query { my ($params) = @_; my $operator = ""; if ($params->{previous_operand}){ #If there is a previous operand, add a supplied operator or the default 'and' $operator = ($params->{operator}) ? ($params->{operator}) : 'AND'; } #NOTE: indexes_set is typically set when doing truncation or field weighting my $operand = ($params->{indexes_set}) ? $params->{parsed_operand} : $params->{index_plus}.$params->{parsed_operand}; #e.g. "kw,wrdl:test" #e.g. " and kw,wrdl:test" $params->{query} .= " " . $operator . " " . $operand; $params->{query_cgi} .= "&op=".uri_escape_utf8($operator) if $operator; $params->{query_cgi} .= "&idx=".uri_escape_utf8($params->{index}) if $params->{index}; $params->{query_cgi} .= "&q=".uri_escape_utf8($params->{original_operand}) if ( $params->{original_operand} ne '' ); #e.g. " and kw,wrdl: test" $params->{query_desc} .= " " . $operator . " " . ( $params->{index_plus} // q{} ) . " " . ( $params->{original_operand} // q{} ); $params->{previous_operand} = 1 unless $params->{previous_operand}; #If there is no previous operand, mark this as one return ($params->{query}, $params->{query_cgi}, $params->{query_desc}, $params->{previous_operand}); } =head2 searchResults my @search_results = searchResults($search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, @marcresults); Format results in a form suitable for passing to the template =cut # IMO this subroutine is pretty messy still -- it's responsible for # building the HTML output for the template sub searchResults { my ( $search_context, $searchdesc, $hits, $results_per_page, $offset, $scan, $marcresults, $xslt_variables ) = @_; my $dbh = C4::Context->dbh; my @newresults; require C4::Items; $search_context->{'interface'} = 'opac' if !$search_context->{'interface'} || $search_context->{'interface'} ne 'intranet'; my ($is_opac, $hidelostitems); if ($search_context->{'interface'} eq 'opac') { $hidelostitems = C4::Context->preference('hidelostitems'); $is_opac = 1; } my $record_processor = Koha::RecordProcessor->new({ filters => 'ViewPolicy' }); #Build branchnames hash my %branches = map { $_->branchcode => $_->branchname } Koha::Libraries->search({}, { order_by => 'branchname' })->as_list; # FIXME - We build an authorised values hash here, using the default framework # though it is possible to have different authvals for different fws. my $shelflocations = { map { $_->{authorised_value} => $_->{lib} } Koha::AuthorisedValues->get_descriptions_by_koha_field( { frameworkcode => '', kohafield => 'items.location' } ) }; # get notforloan authorised value list (see $shelflocations FIXME) my $av = Koha::MarcSubfieldStructures->search({ frameworkcode => '', kohafield => 'items.notforloan', authorised_value => [ -and => {'!=' => undef }, {'!=' => ''}] }); my $notforloan_authorised_value = $av->count ? $av->next->authorised_value : undef; #Get itemtype hash my $itemtypes = Koha::ItemTypes->search_with_localization; my %itemtypes = map { $_->{itemtype} => $_ } @{ $itemtypes->unblessed }; #search item field code my ($itemtag, undef) = &GetMarcFromKohaField( "items.itemnumber" ); ## find column names of items related to MARC my %subfieldstosearch; my @columns = Koha::Database->new()->schema()->resultset('Item')->result_source->columns; for my $column ( @columns ) { my ( $tagfield, $tagsubfield ) = &GetMarcFromKohaField( "items." . $column ); if ( defined $tagsubfield ) { $subfieldstosearch{$column} = $tagsubfield; } } # handle which records to actually retrieve my $times; # Times is which record to process up to if ( $hits && $offset + $results_per_page <= $hits ) { $times = $offset + $results_per_page; } else { $times = $hits // 0; # If less hits than results_per_page+offset we go to the end } my $marcflavour = C4::Context->preference("marcflavour"); # We get the biblionumber position in MARC my ($bibliotag,$bibliosubf)=GetMarcFromKohaField( 'biblio.biblionumber' ); # set stuff for XSLT processing here once, not later again for every record we retrieved my $userenv = C4::Context->userenv; my $logged_in_user = ( defined $userenv and $userenv->{number} ) ? Koha::Patrons->find( $userenv->{number} ) : undef; my $patron_category_hide_lost_items = ($logged_in_user) ? $logged_in_user->category->hidelostitems : 0; # loop through all of the records we've retrieved for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) { my $marcrecord; if ($scan) { # For Scan searches we built USMARC data $marcrecord = MARC::Record->new_from_usmarc( $marcresults->[$i]); } else { # Normal search, render from Zebra's output $marcrecord = new_record_from_zebra( 'biblioserver', $marcresults->[$i] ); if ( ! defined $marcrecord ) { warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i]; next; } } my $fw = $scan ? undef : $bibliotag < 10 ? GetFrameworkCode($marcrecord->field($bibliotag)->data) : GetFrameworkCode($marcrecord->subfield($bibliotag,$bibliosubf)); SetUTF8Flag($marcrecord); my $oldbiblio = TransformMarcToKoha({ record => $marcrecord, limit_table => 'no_items' }); $oldbiblio->{result_number} = $i + 1; $oldbiblio->{normalized_upc} = GetNormalizedUPC( $marcrecord,$marcflavour); $oldbiblio->{normalized_ean} = GetNormalizedEAN( $marcrecord,$marcflavour); $oldbiblio->{normalized_oclc} = GetNormalizedOCLCNumber($marcrecord,$marcflavour); $oldbiblio->{normalized_isbn} = GetNormalizedISBN($oldbiblio->{isbn},$marcrecord,$marcflavour); # Use existing ISBN from record if we got one $oldbiblio->{content_identifier_exists} = 1 if ($oldbiblio->{normalized_isbn} or $oldbiblio->{normalized_oclc} or $oldbiblio->{normalized_ean} or $oldbiblio->{normalized_upc}); # edition information, if any $oldbiblio->{edition} = $oldbiblio->{editionstatement}; my $itemtype = $oldbiblio->{itemtype} ? $itemtypes{$oldbiblio->{itemtype}} : undef; # add imageurl to itemtype if there is one $oldbiblio->{imageurl} = $itemtype ? getitemtypeimagelocation( $search_context->{'interface'}, $itemtype->{imageurl} ) : q{}; # Build summary if there is one (the summary is defined in the itemtypes table) $oldbiblio->{description} = $itemtype ? $itemtype->{translated_description} : q{}; # Pull out the items fields my @fields = $marcrecord->field($itemtag); $marcrecord->delete_fields( @fields ) unless C4::Context->preference('PassItemMarcToXSLT'); my $marcflavor = C4::Context->preference("marcflavour"); # adding linked items that belong to host records if ( C4::Context->preference('EasyAnalyticalRecords') ) { my $analyticsfield = '773'; if ($marcflavor eq 'MARC21') { $analyticsfield = '773'; } elsif ($marcflavor eq 'UNIMARC') { $analyticsfield = '461'; } foreach my $hostfield ( $marcrecord->field($analyticsfield)) { my $hostbiblionumber = $hostfield->subfield("0"); my $linkeditemnumber = $hostfield->subfield("9"); if( $hostbiblionumber ) { my $linkeditemmarc = C4::Items::GetMarcItem( $hostbiblionumber, $linkeditemnumber ); if ($linkeditemmarc) { my $linkeditemfield = $linkeditemmarc->field($itemtag); if ($linkeditemfield) { push( @fields, $linkeditemfield ); } } } } } # Setting item statuses for display my @available_items_loop; my @onloan_items_loop; my @other_items_loop; my $available_items; my $onloan_items; my $other_items; my $ordered_count = 0; my $available_count = 0; my $onloan_count = 0; my $longoverdue_count = 0; my $other_count = 0; my $withdrawn_count = 0; my $itemlost_count = 0; my $hideatopac_count = 0; my $itembinding_count = 0; my $itemdamaged_count = 0; my $item_in_transit_count = 0; my $item_onhold_count = 0; my $notforloan_count = 0; my $item_recalled_count = 0; my $items_count = scalar(@fields); my $maxitems_pref = C4::Context->preference('maxItemsinSearchResults'); my $maxitems = $maxitems_pref ? $maxitems_pref - 1 : 1; my @hiddenitems; # hidden itemnumbers based on OpacHiddenItems syspref # loop through every item foreach my $field (@fields) { my $item; # populate the items hash foreach my $code ( keys %subfieldstosearch ) { $item->{$code} = $field->subfield( $subfieldstosearch{$code} ); } unless ( $item->{itemnumber} ) { warn "MARC item without itemnumber retrieved for biblio ($oldbiblio->{biblionumber})"; next; } $item->{description} = $itemtypes{ $item->{itype} }{translated_description} if $item->{itype}; # OPAC hidden items if ($is_opac) { # hidden based on OpacHiddenItems syspref or because lost my $hi = Koha::Items->search( { itemnumber => $item->{itemnumber} } ) ->filter_by_visible_in_opac({ patron => $search_context->{patron} }); unless ( $hi->count ) { push @hiddenitems, $item->{itemnumber}; $hideatopac_count++; next; } } my $hbranch = C4::Context->preference('StaffSearchResultsDisplayBranch'); my $otherbranch = $hbranch eq 'homebranch' ? 'holdingbranch' : 'homebranch'; # set item's branch name, use HomeOrHoldingBranch syspref first, fall back to the other one if ($item->{$hbranch}) { $item->{'branchname'} = $branches{$item->{$hbranch}}; $item->{'branchcode'} = $item->{$hbranch}; } elsif ($item->{$otherbranch}) { # Last resort $item->{'branchname'} = $branches{$item->{$otherbranch}}; $item->{'branchcode'} = $item->{$otherbranch}; } my $prefix = ( $item->{$hbranch} ? $item->{$hbranch} . '--' : q{} ) . ( $item->{location} ? $item->{location} : q{} ) . ( $item->{itype} ? $item->{itype} : q{} ) . ( $item->{ccode} ? $item->{ccode} : q{} ) . ( $item->{itemcallnumber} ? $item->{itemcallnumber} : q{} ); # For each grouping of items (onloan, available, unavailable), we build a key to store relevant info about that item if ( $item->{onloan} and $logged_in_user and !( $patron_category_hide_lost_items and $item->{itemlost} ) ) { $onloan_count++; my $key = $prefix . $item->{onloan} . $item->{barcode}; $onloan_items->{$key}->{due_date} = $item->{onloan}; $onloan_items->{$key}->{count}++ if $item->{$hbranch}; $onloan_items->{$key}->{branchname} = $item->{branchname}; $onloan_items->{$key}->{branchcode} = $item->{branchcode}; $onloan_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location}; $onloan_items->{$key}->{itemcallnumber} = $item->{itemcallnumber}; $onloan_items->{$key}->{description} = $item->{description}; $onloan_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype} }->{imageurl} ); $onloan_items->{$key}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE'); # if something's checked out and lost, mark it as 'long overdue' if ( $item->{itemlost} ) { $onloan_items->{$key}->{longoverdue}++; $longoverdue_count++; } } # items not on loan, but still unavailable ( lost, withdrawn, damaged ) else { my $itemtype = C4::Context->preference("item-level_itypes")? $item->{itype}: $oldbiblio->{itemtype}; $item->{notforloan} = 1 if !$item->{notforloan} && $itemtype && $itemtypes{ $itemtype }->{notforloan}; # item is on order if ( $item->{notforloan} < 0 ) { $ordered_count++; } elsif ( $item->{notforloan} > 0 ) { $notforloan_count++; } # is item in transit? my $transfertwhen = ''; my ($transfertfrom, $transfertto); # is item on the reserve shelf? my $reservestatus = ''; # is item a waiting recall? my $recallstatus = ''; unless ($item->{withdrawn} || $item->{itemlost} || $item->{damaged} || $item->{notforloan} || ( C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') && $items_count > C4::Context->preference('MaxSearchResultsItemsPerRecordStatusCheck') ) ) { # A couple heuristics to limit how many times # we query the database for item transfer information, sacrificing # accuracy in some cases for speed; # # 1. don't query if item has one of the other statuses # 2. don't check transit status if the bib has # more than 20 items # # FIXME: to avoid having the query the database like this, and to make # the in transit status count as unavailable for search limiting, # should map transit status to record indexed in Zebra. my $item_object = Koha::Items->find($item->{itemnumber}); my $transfer = defined($item_object) ? $item_object->get_transfer : undef; ( $transfertwhen, $transfertfrom, $transfertto ) = defined($transfer) ? ( $transfer->datesent, $transfer->frombranch, $transfer->tobranch ) : ( '', '', '' ); $reservestatus = C4::Reserves::GetReserveStatus( $item->{itemnumber} ); if ( C4::Context->preference('UseRecalls') ) { if ( Koha::Recalls->search({ item_id => $item->{itemnumber}, status => 'waiting' })->count ) { $recallstatus = 'Waiting'; } } } # item is withdrawn, lost, damaged, not for loan, reserved or in transit if ( $item->{withdrawn} || $item->{itemlost} || $item->{damaged} || $item->{notforloan} || $reservestatus eq 'Waiting' || $recallstatus eq 'Waiting' || ($transfertwhen && $transfertwhen ne '')) { $withdrawn_count++ if $item->{withdrawn}; $itemlost_count++ if $item->{itemlost}; $itemdamaged_count++ if $item->{damaged}; $item_in_transit_count++ if $transfertwhen && $transfertwhen ne ''; $item_onhold_count++ if $reservestatus eq 'Waiting'; $item_recalled_count++ if $recallstatus eq 'Waiting'; $item->{status} = ($item->{withdrawn}//q{}) . "-" . ($item->{itemlost}//q{}) . "-" . ($item->{damaged}//q{}) . "-" . ($item->{notforloan}//q{}); $other_count++; my $key = $prefix . $item->{status}; foreach (qw(withdrawn itemlost damaged branchname itemcallnumber)) { $other_items->{$key}->{$_} = $item->{$_}; } $other_items->{$key}->{branchcode} = $item->{branchcode}; $other_items->{$key}->{intransit} = ( $transfertwhen ne '' ) ? 1 : 0; $other_items->{$key}->{recalled} = ($recallstatus) ? 1 : 0; $other_items->{$key}->{onhold} = ($reservestatus) ? 1 : 0; $other_items->{$key}->{notforloan} = GetAuthorisedValueDesc('','',$item->{notforloan},'','',$notforloan_authorised_value) if $notforloan_authorised_value and $item->{notforloan}; $other_items->{$key}->{count}++ if $item->{$hbranch}; $other_items->{$key}->{location} = $shelflocations->{ $item->{location} } if $item->{location}; $other_items->{$key}->{description} = $item->{description}; $other_items->{$key}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} ); $other_items->{$key}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE'); } # item is available else { $available_count++; $available_items->{$prefix}->{count}++ if $item->{$hbranch}; foreach (qw(branchname itemcallnumber description)) { $available_items->{$prefix}->{$_} = $item->{$_}; } $available_items->{$prefix}->{branchcode} = $item->{branchcode}; $available_items->{$prefix}->{location} = $shelflocations->{ $item->{location} } if $item->{location}; $available_items->{$prefix}->{imageurl} = getitemtypeimagelocation( $search_context->{'interface'}, $itemtypes{ $item->{itype}//q{} }->{imageurl} ); $available_items->{$prefix}->{collectioncode} = GetAuthorisedValueDesc('','',$item->{ccode},'','','CCODE'); } } } # notforloan, item level and biblioitem level # if all items are hidden, do not show the record if ( C4::Context->preference('OpacHiddenItemsHidesRecord') && $items_count > 0 && $hideatopac_count == $items_count) { next; } my ( $availableitemscount, $onloanitemscount, $otheritemscount ); for my $key ( sort keys %$onloan_items ) { (++$onloanitemscount > $maxitems) and last; push @onloan_items_loop, $onloan_items->{$key}; } for my $key ( sort keys %$other_items ) { (++$otheritemscount > $maxitems) and last; push @other_items_loop, $other_items->{$key}; } for my $key ( sort keys %$available_items ) { (++$availableitemscount > $maxitems) and last; push @available_items_loop, $available_items->{$key} } # XSLT processing of some stuff # we fetched the sysprefs already before the loop through all retrieved record! if (!$scan) { $record_processor->options({ frameworkcode => $fw, interface => $search_context->{'interface'} }); $record_processor->process($marcrecord); $oldbiblio->{XSLTResultsRecord} = XSLTParse4Display( { biblionumber => $oldbiblio->{biblionumber}, record => $marcrecord, xsl_syspref => ( $is_opac ? 'OPACXSLTResultsDisplay' : 'XSLTResultsDisplay' ), fix_amps => 1, hidden_items => \@hiddenitems, xslt_variables => $xslt_variables, } ); } my $biblio_object = Koha::Biblios->find( $oldbiblio->{biblionumber} ); $oldbiblio->{biblio_object} = $biblio_object; $oldbiblio->{coins} = eval { $biblio_object->get_coins } if $biblio_object && C4::Context->preference('COinSinOPACResults') && $is_opac; my $can_place_holds = 1; # if biblio level itypes are used and itemtype is notforloan, it can't be reserved either if (!C4::Context->preference("item-level_itypes")) { if ($itemtype && $itemtype->{notforloan}) { $can_place_holds = 0; } } else { $can_place_holds = $biblio_object->items->filter_by_for_hold()->count if $biblio_object; } $oldbiblio->{norequests} = 1 unless $can_place_holds; $oldbiblio->{items_count} = $items_count; $oldbiblio->{available_items_loop} = \@available_items_loop; $oldbiblio->{onloan_items_loop} = \@onloan_items_loop; $oldbiblio->{other_items_loop} = \@other_items_loop; $oldbiblio->{availablecount} = $available_count; $oldbiblio->{availableplural} = 1 if $available_count > 1; $oldbiblio->{onloancount} = $onloan_count; $oldbiblio->{onloanplural} = 1 if $onloan_count > 1; $oldbiblio->{othercount} = $other_count; $oldbiblio->{otherplural} = 1 if $other_count > 1; $oldbiblio->{withdrawncount} = $withdrawn_count; $oldbiblio->{itemlostcount} = $itemlost_count; $oldbiblio->{damagedcount} = $itemdamaged_count; $oldbiblio->{intransitcount} = $item_in_transit_count; $oldbiblio->{onholdcount} = $item_onhold_count; $oldbiblio->{recalledcount} = $item_recalled_count; $oldbiblio->{orderedcount} = $ordered_count; $oldbiblio->{notforloancount} = $notforloan_count; if (C4::Context->preference("AlternateHoldingsField") && $items_count == 0) { my $fieldspec = C4::Context->preference("AlternateHoldingsField"); my $subfields = substr $fieldspec, 3; my $holdingsep = C4::Context->preference("AlternateHoldingsSeparator") || ' '; my @alternateholdingsinfo = (); my @holdingsfields = $marcrecord->field(substr $fieldspec, 0, 3); for my $field (@holdingsfields) { my %holding = ( holding => '' ); my $havesubfield = 0; for my $subfield ($field->subfields()) { if ((index $subfields, $$subfield[0]) >= 0) { $holding{'holding'} .= $holdingsep if (length $holding{'holding'} > 0); $holding{'holding'} .= $$subfield[1]; $havesubfield++; } } if ($havesubfield) { push(@alternateholdingsinfo, \%holding); } } $oldbiblio->{'ALTERNATEHOLDINGS'} = \@alternateholdingsinfo; } push( @newresults, $oldbiblio ); } return @newresults; } =head2 enabled_staff_search_views %hash = enabled_staff_search_views() This function returns a hash that contains three flags obtained from the system preferences, used to determine whether a particular staff search results view is enabled. =over 2 =item C * $hash{can_view_MARC} is true only if the MARC view is enabled * $hash{can_view_ISBD} is true only if the ISBD view is enabled * $hash{can_view_labeledMARC} is true only if the Labeled MARC view is enabled =item C =back $template->param ( C4::Search::enabled_staff_search_views ); =cut sub enabled_staff_search_views { return ( can_view_MARC => C4::Context->preference('viewMARC'), # 1 if the staff search allows the MARC view can_view_ISBD => C4::Context->preference('viewISBD'), # 1 if the staff search allows the ISBD view can_view_labeledMARC => C4::Context->preference('viewLabeledMARC'), # 1 if the staff search allows the Labeled MARC view ); } =head2 z3950_search_args $arrayref = z3950_search_args($matchpoints) This function returns an array reference that contains the search parameters to be passed to the Z39.50 search script (z3950_search.pl). The array elements are hash refs whose keys are name and value, and whose values are the name of a search parameter, the value of that search parameter and the URL encoded value of that parameter. The search parameter names are lccn, isbn, issn, title, author, dewey and subject. The search parameter values are obtained from the bibliographic record whose data is in a hash reference in $matchpoints, as returned by Biblio::GetBiblioData(). If $matchpoints is a scalar, it is assumed to be an unnamed query descriptor, e.g. a general purpose search argument. In this case, the returned array contains only entry: the key is 'title' and the value is derived from $matchpoints. If a search parameter value is undefined or empty, it is not included in the returned array. The returned array reference may be passed directly to the template parameters. =over 2 =item C * $array containing hash refs as described above =item C =back $data = Biblio::GetBiblioData($bibno); $template->param ( MYLOOP => C4::Search::z3950_search_args($data) ) *OR* $template->param ( MYLOOP => C4::Search::z3950_search_args($searchscalar) ) =cut sub z3950_search_args { my $bibrec = shift; my $isbn_string = ref( $bibrec ) ? $bibrec->{title} : $bibrec; my $isbn = Business::ISBN->new( $isbn_string ); if (defined $isbn && $isbn->is_valid) { if ( ref($bibrec) ) { $bibrec->{isbn} = $isbn_string; $bibrec->{title} = undef; } else { $bibrec = { isbn => $isbn_string }; } } else { $bibrec = { title => $bibrec } if !ref $bibrec; } my $array = []; for my $field (qw/ lccn isbn issn title author dewey subject /) { push @$array, { name => $field, value => $bibrec->{$field} } if defined $bibrec->{$field}; } return $array; } =head2 GetDistinctValues($field); C<$field> is a reference to the fields array =cut sub GetDistinctValues { my ($fieldname,$string)=@_; # returns a reference to a hash of references to branches... if ($fieldname=~/\./){ my ($table,$column)=split /\./, $fieldname; my $dbh = C4::Context->dbh; my $sth = $dbh->prepare("select DISTINCT($column) as value, count(*) as cnt from $table ".($string?" where $column like \"$string%\"":"")."group by value order by $column "); $sth->execute; my $elements=$sth->fetchall_arrayref({}); return $elements; } else { $string||= qq(""); my @servers=qw; my (@zconns,@results); for ( my $i = 0 ; $i < @servers ; $i++ ) { $zconns[$i] = C4::Context->Zconn( $servers[$i], 1 ); $results[$i] = $zconns[$i]->scan( ZOOM::Query::CCL2RPN->new( qq"$fieldname $string", $zconns[$i]) ); } # The big moment: asynchronously retrieve results from all servers my @elements; _ZOOM_event_loop( \@zconns, \@results, sub { my ( $i, $size ) = @_; for ( my $j = 0 ; $j < $size ; $j++ ) { my %hashscan; @hashscan{qw(value cnt)} = $results[ $i - 1 ]->display_term($j); push @elements, \%hashscan; } } ); return \@elements; } } =head2 _ZOOM_event_loop _ZOOM_event_loop(\@zconns, \@results, sub { my ( $i, $size ) = @_; .... } ); Processes a ZOOM event loop and passes control to a closure for processing the results, and destroying the resultsets. =cut sub _ZOOM_event_loop { my ($zconns, $results, $callback) = @_; while ( ( my $i = ZOOM::event( $zconns ) ) != 0 ) { my $ev = $zconns->[ $i - 1 ]->last_event(); if ( $ev == ZOOM::Event::ZEND ) { next unless $results->[ $i - 1 ]; my $size = $results->[ $i - 1 ]->size(); if ( $size > 0 ) { $callback->($i, $size); } } } foreach my $result (@$results) { $result->destroy(); } } =head2 new_record_from_zebra Given raw data from a searchengine result set, return a MARC::Record object This helper function is needed to take into account all the involved system preferences and configuration variables to properly create the MARC::Record object. If we are using GRS-1, then the raw data we get from Zebra should be USMARC data. If we are using DOM, then it has to be MARCXML. If we are using elasticsearch, it'll already be a MARC::Record and this function needs a new name. =cut sub new_record_from_zebra { my $server = shift; my $raw_data = shift; # Set the default indexing modes my $search_engine = C4::Context->preference("SearchEngine"); if ($search_engine eq 'Elasticsearch') { return ref $raw_data eq 'MARC::Record' ? $raw_data : MARC::Record->new_from_xml( $raw_data, 'UTF-8' ); } my $index_mode = ( $server eq 'biblioserver' ) ? C4::Context->config('zebra_bib_index_mode') // 'dom' : C4::Context->config('zebra_auth_index_mode') // 'dom'; my $marc_record = eval { if ( $index_mode eq 'dom' ) { MARC::Record->new_from_xml( $raw_data, 'UTF-8' ); } else { MARC::Record->new_from_usmarc( $raw_data ); } }; if ($@) { return; } else { return $marc_record; } } END { } # module clean-up code here (global destructor) 1; __END__ =head1 AUTHOR Koha Development Team =cut