Attachment 82148 Details for Bug 19893 – Bug 19893: Support for joined subfields in mappings

[patch] Bug 19893: Support for joined subfields in mappings

Bug-19893-Support-for-joined-subfields-in-mappings.patch (text/plain), 6.07 KB, created by Ere Maijala on 2018-11-09 10:06:19 UTC

(hide)

Description:

Filename:

MIME Type:

Creator: Ere Maijala

Created: 2018-11-09 10:06:19 UTC

Size: 6.07 KB

patch

obsolete

>From d695802ece5034988bf8e4bd8b7de050c3e7b301 Mon Sep 17 00:00:00 2001
>From: David Gustafsson <david.gustafsson@ub.gu.se>
>Date: Tue, 11 Sep 2018 17:56:44 +0200
>Subject: [PATCH] Bug 19893: Support for joined subfields in mappings
>
>Signed-off-by: Ere Maijala <ere.maijala@helsinki.fi>
>---
> Koha/SearchEngine/Elasticsearch.pm | 84 +++++++++++++++++++++++++++++++++-----
> 1 file changed, 74 insertions(+), 10 deletions(-)
>
>diff --git a/Koha/SearchEngine/Elasticsearch.pm b/Koha/SearchEngine/Elasticsearch.pm
>index 32ba39b..bd1c5e0 100644
>--- a/Koha/SearchEngine/Elasticsearch.pm
>+++ b/Koha/SearchEngine/Elasticsearch.pm
>@@ -80,7 +80,8 @@ sub get_elasticsearch {
>         $self->{elasticsearch} = Search::Elasticsearch->new(
>             client => "5_0::Direct",
>             nodes => $conf->{nodes},
>-            cxn_pool => 'Sniff'
>+            cxn_pool => 'Sniff',
>+            request_timeout => 60
>         );
>     }
>     return $self->{elasticsearch};
>@@ -348,8 +349,10 @@ sub marc_records_to_documents {
>                 }
>             }
>             else {
>-                my $subfields_mappings = $data_fields_rules->{$field->tag()};
>-                if ($subfields_mappings) {
>+                my $data_field_rules = $data_fields_rules->{$field->tag()};
>+
>+                if ($data_field_rules) {
>+                    my $subfields_mappings = $data_field_rules->{subfields};
>                     my $wildcard_mappings = $subfields_mappings->{'*'};
>                     foreach my $subfield ($field->subfields()) {
>                         my ($code, $data) = @{$subfield};
>@@ -361,6 +364,23 @@ sub marc_records_to_documents {
>                             _process_mappings($mappings, $data, $record_document);
>                         }
>                     }
>+
>+                    my $subfields_join_mappings = $data_field_rules->{subfields_join};
>+                    if ($subfields_join_mappings) {
>+                        foreach my $subfields_group (keys %{$subfields_join_mappings}) {
>+                            # Map each subfield to values, remove empty values, join with space
>+                            my $data = join(
>+                                ' ',
>+                                grep(
>+                                    $_,
>+                                    map { join(' ', $field->subfield($_)) } split(//, $subfields_group)
>+                                )
>+                            );
>+                            if ($data) {
>+                                _process_mappings($subfields_join_mappings->{$subfields_group}, $data, $record_document);
>+                            }
>+                        }
>+                    }
>                 }
>             }
>         }
>@@ -463,7 +483,7 @@ sub get_marc_mapping_rules {
>         }
>         return @mappings;
>     };
>-    my $field_spec_regexp = qr/^([0-9]{3})([0-9a-z]+)?(?:_\/(\d+(?:-\d+)?))?$/;
>+    my $field_spec_regexp = qr/^([0-9]{3})([()0-9a-z]+)?(?:_\/(\d+(?:-\d+)?))?$/;
>     my $leader_regexp = qr/^leader(?:_\/(\d+(?:-\d+)?))?$/;
>     my $rules = {
>         'leader' => [],
>@@ -488,19 +508,63 @@ sub get_marc_mapping_rules {
> 
>         if ($marc_field =~ $field_spec_regexp) {
>             my $field_tag = $1;
>-            my $subfields = defined $2 ? $2 : '*';
>+
>+            my @subfields;
>+            my @subfield_groups;
>+            # Parse and separate subfields form subfield groups
>+            if (defined $2) {
>+                my $subfield_group = '';
>+                my $open_group = 0;
>+
>+                foreach my $token (split //, $2) {
>+                    if ($token eq "(") {
>+                        if ($open_group) {
>+                            die("Unmatched opening parenthesis for $marc_field");
>+                        }
>+                        else {
>+                            $open_group = 1;
>+                        }
>+                    }
>+                    elsif ($token eq ")") {
>+                        if ($open_group) {
>+                            if ($subfield_group) {
>+                                push @subfield_groups, $subfield_group;
>+                                $subfield_group = '';
>+                            }
>+                            $open_group = 0;
>+                        }
>+                        else {
>+                            die("Unmatched closing parenthesis for $marc_field");
>+                        }
>+                    }
>+                    elsif ($open_group) {
>+                        $subfield_group .= $token;
>+                    }
>+                    else {
>+                        push @subfields, $token;
>+                    }
>+                }
>+            }
>+            else {
>+                push @subfields, '*';
>+            }
>+
>             my $range = defined $3 ? $3 : undef;
>+            my @mappings = _field_mappings($facet, $suggestible, $sort, $name, $type, $range);
>+
>             if ($field_tag < 10) {
>                 $rules->{control_fields}->{$field_tag} //= [];
>-                my @mappings = _field_mappings($facet, $suggestible, $sort, $name, $type, $range);
>                 push @{$rules->{control_fields}->{$field_tag}}, @mappings;
>             }
>             else {
>                 $rules->{data_fields}->{$field_tag} //= {};
>-                foreach my $subfield (split //, $subfields) {
>-                    $rules->{data_fields}->{$field_tag}->{$subfield} //= [];
>-                    my @mappings = _field_mappings($facet, $suggestible, $sort, $name, $type, $range);
>-                    push @{$rules->{data_fields}->{$field_tag}->{$subfield}}, @mappings;
>+                foreach my $subfield (@subfields) {
>+                    $rules->{data_fields}->{$field_tag}->{subfields}->{$subfield} //= [];
>+                    push @{$rules->{data_fields}->{$field_tag}->{subfields}->{$subfield}}, @mappings;
>+                }
>+                foreach my $subfield_group (@subfield_groups) {
>+                    $rules->{data_fields}->{$field_tag}->{subfields_join}->{$subfield_group} //= [];
>+                    push @{$rules->{data_fields}->{$field_tag}->{subfields_join}->{$subfield_group}}, @mappings;
>                 }
>             }
>         }
>-- 
>2.7.4

Actions: View | Diff | Splinter Review

Attachments on bug 19893: 70199 | 70200 | 70201 | 70202 | 70230 | 70325 | 73153 | 75109 | 75271 | 75272 | 75353 | 75428 | 75455 | 75588 | 75658 | 75660 | 76612 | 76613 | 76614 | 76629 | 78091 | 78092 | 78093 | 78094 | 78524 | 78561 | 78587 | 78588 | 78589 | 78590 | 78591 | 78592 | 78593 | 78594 | 78690 | 78691 | 78692 | 78693 | 78694 | 78695 | 78696 | 80969 | 80970 | 80971 | 80983 | 81014 | 81015 | 81284 | 81822 | 81901 | 81903 | 81955 | 81957 | 81958 | 81959 | 81960 | 81961 | 81962 | 81963 | 81964 | 81965 | 81966 | 82143 | 82144 | 82145 | 82146 | 82147 | 82148 | 82149 | 82150 | 82151 | 82152 | 82256 | 82257 | 82258 | 82259 | 82260 | 82261 | 82262 | 82263 | 82264 | 82265 | 82266 | 82267