From 6249f039beb8024fc9cb7258fbe773f421a8b506 Mon Sep 17 00:00:00 2001 From: Tomas Cohen Arazi Date: Mon, 21 Oct 2013 16:17:57 -0700 Subject: [PATCH] Bug 11096: force XML syntax for Zebra in DOM This patch makes Koha <-> Zebra use MARCXML for the serialization. We currently use the USMARC, which remains for GRS-1. This functions are modified for that purpose: - C4::Context->Zconn - C4::Context-_new_Zconn - C4::Search->* - C4::AuthoritiesMarc->SearchAuthorities - C4::AuthoritiesMarc->merge - C4::AuthoritiesMarc->FindDuplicateAuthority C4::Context->config('zebra_bib_index_mode') and C4::Context->config('zebra_auth_index_mode') Calls to the unsafe ZOOM::Record->render()[2] method are removed. Due to this last change the code for building facets was rewritten. And for performance on the facets creation I pushed higher version dependencies for MARC::File::XML and MARC::Record (we rely on MARC::Field->as_string). Calls to MARC::Record->new_from_xml and MARC::Record->new_from_usmarc are wrapped with eval for catching problems [2]. To test: - prove t/db_dependent/Search.t should pass. - Searching should remain functional. - Indexing and searching for a big record should work (that's what the unit tests do). Regards To+ [1] http://search.cpan.org/~mirk/Net-Z3950-ZOOM/lib/ZOOM.pod#render() [2] a record that cannot be parsed by MARC::Record is simply skipped (bug 10684) Sponsored-by: Universidad Nacional de Cordoba Signed-off-by: Chris Cormack Signed-off-by: Kyle M Hall Passes koha-qa.pl, all unit test. Works as advertised. --- C4/AuthoritiesMarc.pm | 45 +++++++++++++------ C4/Context.pm | 42 ++++++++--------- C4/Installer/PerlDependencies.pm | 4 +- C4/Search.pm | 93 ++++++++++++++++++++++++------------- catalogue/search.pl | 2 +- circ/circulation.pl | 2 +- t/db_dependent/Search.t | 44 +++++++++++++++++- 7 files changed, 156 insertions(+), 76 deletions(-) diff --git a/C4/AuthoritiesMarc.pm b/C4/AuthoritiesMarc.pm index f7eb122..0fdea1e 100644 --- a/C4/AuthoritiesMarc.pm +++ b/C4/AuthoritiesMarc.pm @@ -266,10 +266,23 @@ sub SearchAuthorities { ##Here we have to extract MARC record and $authid from ZEBRA AUTHORITIES my $rec=$oAResult->record($counter); - my $marcdata=$rec->raw(); - my $authrecord; my $separator=C4::Context->preference('authoritysep'); - $authrecord = MARC::File::USMARC::decode($marcdata); + my $authrecord = eval { + if (C4::Context->config('zebra_auth_index_mode') eq 'dom') { + MARC::Record->new_from_xml( + $rec->raw(), + 'UTF-8' + ); + } else { + MARC::Record->new_from_usmarc($rec->raw()); + } + }; + + if ( !defined $authrecord or !defined $authrecord->field('001') ) { + $counter++; + next; + } + my $authid=$authrecord->field('001')->data(); my %newline; $newline{authid} = $authid; @@ -847,15 +860,17 @@ sub FindDuplicateAuthority { my $query='at:'.$authtypecode.' '; my $filtervalues=qr([\001-\040\!\'\"\`\#\$\%\&\*\+,\-\./:;<=>\?\@\(\)\{\[\]\}_\|\~]); if ($record->field($auth_tag_to_report)) { - foreach ($record->field($auth_tag_to_report)->subfields()) { - $_->[1]=~s/$filtervalues/ /g; $query.= " $op he:\"".$_->[1]."\"" if ($_->[0]=~/[A-z]/); - } + foreach ($record->field($auth_tag_to_report)->subfields()) { + $_->[1]=~s/$filtervalues/ /g; $query.= " $op he:\"".$_->[1]."\"" if ($_->[0]=~/[A-z]/); + } } my ($error, $results, $total_hits) = C4::Search::SimpleSearch( $query, 0, 1, [ "authorityserver" ] ); # there is at least 1 result => return the 1st one if (!defined $error && @{$results} ) { - my $marcrecord = MARC::File::USMARC::decode($results->[0]); - return $marcrecord->field('001')->data,BuildSummary($marcrecord,$marcrecord->field('001')->data,$authtypecode); + my $marcrecord = MARC::Record->new_from_xml( + $results->[0] + ); + return $marcrecord->field('001')->data,BuildSummary($marcrecord,$marcrecord->field('001')->data,$authtypecode); } # no result, returns nothing return; @@ -1452,13 +1467,15 @@ sub merge { } my $z=0; while ( $z<$count ) { - my $rec; - $rec=$oResult->record($z); - my $marcdata = $rec->raw(); - my $marcrecordzebra= MARC::Record->new_from_usmarc($marcdata); + my $marcrecordzebra= MARC::Record->new_from_xml( + $oResult->record($z)->raw(), + 'UTF-8' + ); my ( $biblionumbertagfield, $biblionumbertagsubfield ) = &GetMarcFromKohaField( "biblio.biblionumber", '' ); - my $i = ($biblionumbertagfield < 10) ? $marcrecordzebra->field($biblionumbertagfield)->data : $marcrecordzebra->subfield($biblionumbertagfield, $biblionumbertagsubfield); - my $marcrecorddb=GetMarcBiblio($i); + my $i = ($biblionumbertagfield < 10) + ? $marcrecordzebra->field( $biblionumbertagfield )->data + : $marcrecordzebra->subfield( $biblionumbertagfield, $biblionumbertagsubfield ); + my $marcrecorddb = GetMarcBiblio($i); push @reccache, $marcrecorddb; $z++; } diff --git a/C4/Context.pm b/C4/Context.pm index 3914420..e96d399 100644 --- a/C4/Context.pm +++ b/C4/Context.pm @@ -706,8 +706,6 @@ sub Zconn { $context->{"Zconn"}->{$server}->destroy() if defined($context->{"Zconn"}->{$server}); $context->{"Zconn"}->{$server} = &_new_Zconn($server,$async,$auth,$piggyback,$syntax); - $context->{ Zconn }->{ $server }->option( - preferredRecordSyntax => C4::Context->preference("marcflavour") ); return $context->{"Zconn"}->{$server}; } } @@ -731,14 +729,28 @@ sub _new_Zconn { my $tried=0; # first attempt my $Zconn; # connection object - $server = "biblioserver" unless $server; - $syntax = "usmarc" unless $syntax; + my $elementSetName; + + $server //= "biblioserver"; + $syntax //= "XML"; + + if ( ($server eq 'biblioserver' and $context->{'config'}->{'zebra_bib_index_mode'} eq 'grs1') + or ($server eq 'authorityserver' and $context->{'config'}->{'zebra_auth_index_mode'} eq 'grs1')) { + + $elementSetName = 'F'; + $syntax = 'usmarc'; + + } else { + + $elementSetName = 'marcxml'; + $syntax = 'XML'; + } my $host = $context->{'listen'}->{$server}->{'content'}; my $servername = $context->{"config"}->{$server}; my $user = $context->{"serverinfo"}->{$server}->{"user"}; my $password = $context->{"serverinfo"}->{$server}->{"password"}; - $auth = 1 if($user && $password); + $auth = 1 if($user && $password); retry: eval { # set options @@ -750,7 +762,7 @@ sub _new_Zconn { $o->option(cqlfile=> $context->{"server"}->{$server}->{"cql2rpn"}); $o->option(cclfile=> $context->{"serverinfo"}->{$server}->{"ccl2rpn"}); $o->option(preferredRecordSyntax => $syntax); - $o->option(elementSetName => "F"); # F for 'full' as opposed to B for 'brief' + $o->option(elementSetName => $elementSetName); $o->option(databaseName => ($servername?$servername:"biblios")); # create a new connection object @@ -765,23 +777,7 @@ sub _new_Zconn { } }; -# if ($@) { -# # Koha manages the Zebra server -- this doesn't work currently for me because of permissions issues -# # Also, I'm skeptical about whether it's the best approach -# warn "problem with Zebra"; -# if ( C4::Context->preference("ManageZebra") ) { -# if ($@->code==10000 && $tried==0) { ##No connection try restarting Zebra -# $tried=1; -# warn "trying to restart Zebra"; -# my $res=system("zebrasrv -f $ENV{'KOHA_CONF'} >/koha/log/zebra-error.log"); -# goto "retry"; -# } else { -# warn "Error ", $@->code(), ": ", $@->message(), "\n"; -# $Zconn="error"; -# return $Zconn; -# } -# } -# } + return $Zconn; } diff --git a/C4/Installer/PerlDependencies.pm b/C4/Installer/PerlDependencies.pm index 9b4b3b6..59a91e8 100644 --- a/C4/Installer/PerlDependencies.pm +++ b/C4/Installer/PerlDependencies.pm @@ -462,7 +462,7 @@ our $PERL_DEPS = { 'MARC::Record' => { 'usage' => 'Core', 'required' => '1', - 'min_ver' => '2' + 'min_ver' => '2.0.6' }, 'Locale::Currency::Format' => { 'usage' => 'Core', @@ -487,7 +487,7 @@ our $PERL_DEPS = { 'MARC::File::XML' => { 'usage' => 'Core', 'required' => '1', - 'min_ver' => '0.88' + 'min_ver' => '1.0.1' }, 'XML::SAX::Writer' => { 'usage' => 'Core', diff --git a/C4/Search.pm b/C4/Search.pm index 18a0f5f..c108323 100644 --- a/C4/Search.pm +++ b/C4/Search.pm @@ -144,8 +144,19 @@ sub FindDuplicate { my @results; if (!defined $error) { foreach my $possible_duplicate_record (@{$searchresults}) { - my $marcrecord = - MARC::Record->new_from_usmarc($possible_duplicate_record); + my $marcrecord = eval { + if ( C4::Context->config('zebra_bib_index_mode') eq 'dom') { + MARC::Record->new_from_xml( + $possible_duplicate_record, + 'UTF-8' + ); + } else { + MARC::Record->new_from_usmarc( + $possible_duplicate_record + ); + } + }; + my $result = TransformMarcToKoha( $dbh, $marcrecord, '' ); # FIXME :: why 2 $biblionumber ? @@ -289,10 +300,11 @@ sub SimpleSearch { } for my $j ( $first_record .. $last_record ) { - my $record = + my $record = eval { $tmpresults[ $i - 1 ]->record( $j - 1 )->raw() ; # 0 indexed - push @{$results}, $record; + }; + push @{$results}, $record if defined $record; } } ); @@ -446,6 +458,7 @@ sub getRecords { else { $times = $size; } + for ( my $j = $offset ; $j < $times ; $j++ ) { my $records_hash; my $record; @@ -455,7 +468,7 @@ sub getRecords { my ( $term, $occ ) = $results[ $i - 1 ]->term($j); # here we create a minimal MARC record and hand it off to the - # template just like a normal result ... perhaps not ideal, but + # template just like a normal result ... serhaps not ideal, but # it works for now my $tmprecord = MARC::Record->new(); $tmprecord->encoding('UTF-8'); @@ -488,7 +501,6 @@ sub getRecords { # not an index scan else { $record = $results[ $i - 1 ]->record($j)->raw(); - # warn "RECORD $j:".$record; $results_hash->{'RECORDS'}[$j] = $record; } @@ -503,39 +515,47 @@ sub getRecords { $size > $facets_maxrecs ? $facets_maxrecs : $size; for my $facet (@$facets) { for ( my $j = 0 ; $j < $jmax ; $j++ ) { - my $render_record = - $results[ $i - 1 ]->record($j)->render(); + + my $marc_record = eval { + if ( C4::Context->config('zebra_bib_index_mode') eq 'dom' ) { + MARC::Record->new_from_xml( + $results[ $i - 1 ]->record($j)->raw(), + 'UTF-8' + ); + } else { + MARC::Record->new_from_usmarc( + $results[ $i - 1 ]->record($j)->raw() + ); + } + }; + + if ( $@ ) { + warn "ERROR DECODING RECORD - $@: " . + $results[ $i - 1 ]->record($j)->raw(); + next; + } + my @used_datas = (); + foreach my $tag ( @{ $facet->{tags} } ) { # avoid first line my $tag_num = substr( $tag, 0, 3 ); - my $letters = substr( $tag, 3 ); - my $field_pattern = - '\n' . $tag_num . ' ([^z][^\n]+)'; - $field_pattern = '\n' . $tag_num . ' ([^\n]+)' - if ( int($tag_num) < 10 ); - my @field_tokens = - ( $render_record =~ /$field_pattern/g ); - foreach my $field_token (@field_tokens) { - my @subf = ( $field_token =~ - /\$([a-zA-Z0-9]) ([^\$]+)/g ); - my @values; - for ( my $i = 0 ; $i < @subf ; $i += 2 ) { - if ( $letters =~ $subf[$i] ) { - my $value = $subf[ $i + 1 ]; - $value =~ s/^ *//; - $value =~ s/ *$//; - push @values, $value; - } - } - my $data = join( $facet->{sep}, @values ); + my $subfield_letters = substr( $tag, 3 ); + # Removed when as_string fixed + my @subfields = $subfield_letters =~ /./sg; + + my @fields = $marc_record->field($tag_num); + foreach my $field (@fields) { + my $data = $field->as_string( $subfield_letters, $facet->{sep} ); + unless ( $data ~~ @used_datas ) { - $facets_counter->{ $facet->{idx} } - ->{$data}++; push @used_datas, $data; } - } # fields + + $facets_counter->{ $facet->{idx} }->{$data}++; + + } # fields } # field codes } # records $facets_info->{ $facet->{idx} }->{label_value} = @@ -1700,13 +1720,20 @@ sub searchResults { $times = $hits; # FIXME: if $hits is undefined, why do we want to equal it? } - my $marcflavour = C4::Context->preference("marcflavour"); + my $marcflavour = C4::Context->preference("marcflavour"); # We get the biblionumber position in MARC my ($bibliotag,$bibliosubf)=GetMarcFromKohaField('biblio.biblionumber',''); # loop through all of the records we've retrieved for ( my $i = $offset ; $i <= $times - 1 ; $i++ ) { - my $marcrecord = eval { MARC::File::USMARC::decode( $marcresults->[$i] ); }; + my $marcrecord = eval { + if ( C4::Context->config('zebra_bib_index_mode') eq 'dom') { + MARC::Record->new_from_xml( $marcresults->[$i], 'UTF-8' ); + } else { + MARC::Record->new_from_usmarc( $marcresults->[$i] ); + } + }; + if ( $@ ) { warn "ERROR DECODING RECORD - $@: " . $marcresults->[$i]; next; diff --git a/catalogue/search.pl b/catalogue/search.pl index 892ff15..4ee0387 100755 --- a/catalogue/search.pl +++ b/catalogue/search.pl @@ -660,7 +660,7 @@ for (my $i=0;$i<@servers;$i++) { elsif ($server =~/authorityserver/) { # this is the local authority server my @inner_sup_results_array; for my $sup_record ( @{$results_hashref->{$server}->{"RECORDS"}} ) { - my $marc_record_object = MARC::Record->new_from_usmarc($sup_record); + my $marc_record_object = MARC::Record->new_from_xml($sup_record,'UTF-8'); # warn "Authority Found: ".$marc_record_object->as_formatted(); push @inner_sup_results_array, { 'title' => $marc_record_object->field(100)->subfield('a'), diff --git a/circ/circulation.pl b/circ/circulation.pl index 9769731..d587a00 100755 --- a/circ/circulation.pl +++ b/circ/circulation.pl @@ -309,7 +309,7 @@ if ($barcode) { foreach my $hit ( @{$results} ) { my $chosen = TransformMarcToKoha( C4::Context->dbh, - MARC::Record->new_from_usmarc($hit) ); + MARC::Record->new_from_xml($hit,'UTF-8') ); # offer all barcodes individually foreach my $barcode ( sort split(/\s*\|\s*/, $chosen->{barcode}) ) { diff --git a/t/db_dependent/Search.t b/t/db_dependent/Search.t index cdc87dc..887bf0f 100644 --- a/t/db_dependent/Search.t +++ b/t/db_dependent/Search.t @@ -359,6 +359,43 @@ sub run_marc21_search_tests { getRecords('au:Lessig', 'au:Lessig', [], [ 'biblioserver' ], '20', 0, undef, \%branches, \%itemtypes, 'ccl', undef); is($results_hashref->{biblioserver}->{hits}, 4, "getRecords title search for 'Australia' matched right number of records"); +if ( $indexing_mode eq 'dom' ) { + ( undef, $results_hashref, $facets_loop ) = + getRecords('salud', 'salud', [], [ 'biblioserver' ], '19', 0, undef, \%branches, \%itemtypes, 'ccl', undef); + ok(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper() =~ m/^Efectos del ambiente/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[7],'UTF-8')->title_proper() eq 'Salud y seguridad de los trabajadores del sector salud: manual para gerentes y administradores^ies' && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[18],'UTF-8')->title_proper() =~ m/^Indicadores de resultados identificados/ + , "Simple relevance sorting in getRecords matches old behavior"); + + ( undef, $results_hashref, $facets_loop ) = + getRecords('salud', 'salud', [ 'author_az' ], [ 'biblioserver' ], '38', 0, undef, \%branches, \%itemtypes, 'ccl', undef); + ok(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper() =~ m/la enfermedad laboral\^ies$/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[6],'UTF-8')->title_proper() =~ m/^Indicadores de resultados identificados/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[18],'UTF-8')->title_proper() eq 'World health statistics 2009^ien' + , "Simple ascending author sorting in getRecords matches old behavior"); + + ( undef, $results_hashref, $facets_loop ) = + getRecords('salud', 'salud', [ 'author_za' ], [ 'biblioserver' ], '38', 0, undef, \%branches, \%itemtypes, 'ccl', undef); + ok(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper() eq 'World health statistics 2009^ien' && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[12],'UTF-8')->title_proper() =~ m/^Indicadores de resultados identificados/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[18],'UTF-8')->title_proper() =~ m/la enfermedad laboral\^ies$/ + , "Simple descending author sorting in getRecords matches old behavior"); + + ( undef, $results_hashref, $facets_loop ) = + getRecords('salud', 'salud', [ 'pubdate_asc' ], [ 'biblioserver' ], '38', 0, undef, \%branches, \%itemtypes, 'ccl', undef); + ok(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper() eq 'Manual de higiene industrial^ies' && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[7],'UTF-8')->title_proper() =~ m/seguridad e higiene del trabajo\^ies$/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[18],'UTF-8')->title_proper() =~ m/^Indicadores de resultados identificados/ + , "Simple ascending publication date sorting in getRecords matches old behavior"); + + ( undef, $results_hashref, $facets_loop ) = + getRecords('salud', 'salud', [ 'pubdate_dsc' ], [ 'biblioserver' ], '38', 0, undef, \%branches, \%itemtypes, 'ccl', undef); + ok(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper() =~ m/^Estado de salud/ && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[7],'UTF-8')->title_proper() eq 'World health statistics 2009^ien' && + MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[18],'UTF-8')->title_proper() eq 'Manual de higiene industrial^ies' + , "Simple descending publication date sorting in getRecords matches old behavior"); + +} elsif ( $indexing_mode eq 'grs1' ){ ( undef, $results_hashref, $facets_loop ) = getRecords('salud', 'salud', [], [ 'biblioserver' ], '19', 0, undef, \%branches, \%itemtypes, 'ccl', undef); ok(MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[0])->title_proper() =~ m/^Efectos del ambiente/ && @@ -393,14 +430,17 @@ sub run_marc21_search_tests { MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[7])->title_proper() eq 'World health statistics 2009^ien' && MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[18])->title_proper() eq 'Manual de higiene industrial^ies' , "Simple descending publication date sorting in getRecords matches old behavior"); +} +TODO: { + local $TODO = "Switch relevance search to MARCXML too"; ( undef, $results_hashref, $facets_loop ) = getRecords('books', 'books', [ 'relevance' ], [ 'biblioserver' ], '20', 0, undef, \%branches, \%itemtypes, undef, 1); $record = MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[0]); is($record->title_proper(), 'books', "Scan returned requested item"); is($record->subfield('100', 'a'), 2, "Scan returned correct number of records matching term"); - # Time to test buildQuery and searchResults too. +} my ( $query, $simple_query, $query_cgi, $query_desc, $limit, $limit_cgi, $limit_desc, @@ -544,7 +584,7 @@ sub run_marc21_search_tests { is(MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[0])->title_proper(), 'Salud y seguridad de los trabajadores del sector salud: manual para gerentes y administradores^ies', "Weighted query returns best match first"); } else { local $TODO = "Query weighting does not behave exactly the same in DOM vs. GRS"; - is(MARC::Record::new_from_usmarc($results_hashref->{biblioserver}->{RECORDS}->[0])->title_proper(), 'Salud y seguridad de los trabajadores del sector salud: manual para gerentes y administradores^ies', "Weighted query returns best match first"); + is(MARC::Record::new_from_xml($results_hashref->{biblioserver}->{RECORDS}->[0],'UTF-8')->title_proper(), 'Salud y seguridad de los trabajadores del sector salud: manual para gerentes y administradores^ies', "Weighted query returns best match first"); } $QueryStemming = $QueryWeightFields = $QueryFuzzy = $QueryRemoveStopwords = 0; -- 1.7.2.5