1 package C4::External::Syndetics;
2 # Copyright (C) 2006 LibLime
3 # <jmf at liblime dot com>
5 # This file is part of Koha.
7 # Koha is free software; you can redistribute it and/or modify it
8 # under the terms of the GNU General Public License as published by
9 # the Free Software Foundation; either version 3 of the License, or
10 # (at your option) any later version.
12 # Koha is distributed in the hope that it will be useful, but
13 # WITHOUT ANY WARRANTY; without even the implied warranty of
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
15 # GNU General Public License for more details.
17 # You should have received a copy of the GNU General Public License
18 # along with Koha; if not, see <http://www.gnu.org/licenses>.
20 use XML::Simple qw( XMLout );
22 use LWP::Simple qw( $ua );
28 use vars qw(@ISA @EXPORT);
37 get_syndetics_editions
44 # package-level variable
45 my $parser = XML::LibXML->new();
49 C4::External::Syndetics - Functions for retrieving Syndetics content in Koha
53 This module provides facilities for retrieving Syndetics.com content in Koha
55 =head2 get_syndetics_summary
57 my $syndetics_summary= &get_syndetics_summary( $isbn );
59 Get Summary data from Syndetics
63 sub get_syndetics_index {
64 my ( $isbn, $upc, $oclc ) = @_;
66 return unless ( $isbn || $upc || $oclc );
68 my $response = _fetch_syndetics_content('INDEX.XML', $isbn, $upc, $oclc);
69 unless ($response->content_type =~ /xml/) {
73 my $content = $response->content;
74 my $xmlsimple = XML::Simple->new();
75 $response = $xmlsimple->XMLin(
79 my $syndetics_elements;
80 for my $available_type ('SUMMARY','TOC','FICTION','AWARDS1','SERIES1','SPSUMMARY','SPREVIEW', 'AVPROFILE', 'AVSUMMARY','DBCHAPTER','LJREVIEW','PWREVIEW','SLJREVIEW','CHREVIEW','BLREVIEW','HBREVIEW','KIREVIEW','CRITICASREVIEW','ANOTES') {
81 if (exists $response->{$available_type} && $response->{$available_type} =~ /$available_type/) {
82 $syndetics_elements->{$available_type} = $available_type;
86 return $syndetics_elements if $syndetics_elements;
89 sub get_syndetics_summary {
90 my ( $isbn, $upc, $oclc, $syndetics_elements ) = @_;
92 my $summary_type = exists($syndetics_elements->{'AVSUMMARY'}) ? 'AVSUMMARY.XML' : 'SUMMARY.XML';
93 my $response = _fetch_syndetics_content($summary_type, $isbn, $upc, $oclc);
94 unless ($response->content_type =~ /xml/) {
98 my $content = $response->content;
102 my $doc = $parser->parse_string($content);
103 $summary = $doc->findvalue('//Fld520');
106 warn "Error parsing Syndetics $summary_type";
108 return $summary if $summary;
111 sub get_syndetics_toc {
112 my ( $isbn,$upc,$oclc ) = @_;
114 my $response = _fetch_syndetics_content('TOC.XML', $isbn, $upc, $oclc);
115 unless ($response->content_type =~ /xml/) {
119 my $content = $response->content;
120 my $xmlsimple = XML::Simple->new();
121 $response = $xmlsimple->XMLin(
123 forcearray => [ qw(Fld970) ],
125 # manipulate response USMARC VarFlds VarDFlds Notes Fld520 a
127 $toc = \@{$response->{VarFlds}->{VarDFlds}->{SSIFlds}->{Fld970}} if $response;
131 sub get_syndetics_excerpt {
132 my ( $isbn,$upc,$oclc ) = @_;
134 my $response = _fetch_syndetics_content('DBCHAPTER.XML', $isbn, $upc, $oclc);
135 unless ($response->content_type =~ /xml/) {
139 my $content = $response->content;
140 my $xmlsimple = XML::Simple->new();
141 $response = $xmlsimple->XMLin(
143 forcearray => [ qw(Fld520) ],
145 # manipulate response USMARC VarFlds VarDFlds Notes Fld520 a
147 $excerpt = \@{$response->{VarFlds}->{VarDFlds}->{Notes}->{Fld520}} if $response;
148 return XMLout($excerpt, NoEscape => 1) if $excerpt;
151 sub get_syndetics_reviews {
152 my ( $isbn,$upc,$oclc,$syndetics_elements ) = @_;
155 my $review_sources = [
156 {title => 'Library Journal Review', file => 'LJREVIEW.XML', element => 'LJREVIEW'},
157 {title => 'Publishers Weekly Review', file => 'PWREVIEW.XML', element => 'PWREVIEW'},
158 {title => 'School Library Journal Review', file => 'SLJREVIEW.XML', element => 'SLJREVIEW'},
159 {title => 'CHOICE Review', file => 'CHREVIEW.XML', element => 'CHREVIEW'},
160 {title => 'Booklist Review', file => 'BLREVIEW.XML', element => 'BLREVIEW'},
161 {title => 'Horn Book Review', file => 'HBREVIEW.XML', element => 'HBREVIEW'},
162 {title => 'Kirkus Book Review', file => 'KIREVIEW.XML', element => 'KIREVIEW'},
163 {title => 'Criticas Review', file => 'CRITICASREVIEW.XML', element => 'CRITICASREVIEW'},
164 {title => 'Spanish Review', file => 'SPREVIEW.XML', element => 'SPREVIEW'},
167 for my $source (@$review_sources) {
168 if ($syndetics_elements->{$source->{element}} and $source->{element} =~ $syndetics_elements->{$source->{element}}) {
171 #warn "Skipping $source->{element} doesn't match $syndetics_elements->{$source->{element}} \n";
174 my $response = _fetch_syndetics_content($source->{file}, $isbn, $upc, $oclc);
175 unless ($response->content_type =~ /xml/) {
179 my $content = $response->content;
182 my $doc = $parser->parse_string($content);
184 # note that using findvalue strips any HTML elements embedded
185 # in that review. That helps us handle slight differences
186 # in the output provided by Syndetics 'old' and 'new' versions
187 # of their service and cleans any questionable HTML that
188 # may be present in the reviews, but does mean that any
189 # <B> and <I> tags used to format the review are also gone.
190 my $result = $doc->findvalue('//Fld520');
191 push @reviews, {title => $source->{title}, reviews => [ { content => $result } ]} if $result;
194 warn "Error parsing Syndetics $source->{title} review";
200 sub get_syndetics_editions {
201 my ( $isbn,$upc,$oclc ) = @_;
203 my $response = _fetch_syndetics_content('FICTION.XML', $isbn, $upc, $oclc);
204 unless ($response->content_type =~ /xml/) {
208 my $content = $response->content;
210 my $xmlsimple = XML::Simple->new();
211 $response = $xmlsimple->XMLin(
213 forcearray => [ qw(Fld020) ],
215 # manipulate response USMARC VarFlds VarDFlds Notes Fld520 a
217 $similar_items = \@{$response->{VarFlds}->{VarDFlds}->{NumbCode}->{Fld020}} if $response;
218 return $similar_items if $similar_items;
221 sub get_syndetics_anotes {
222 my ( $isbn,$upc,$oclc) = @_;
224 my $response = _fetch_syndetics_content('ANOTES.XML', $isbn, $upc, $oclc);
225 unless ($response->content_type =~ /xml/) {
229 my $content = $response->content;
231 my $xmlsimple = XML::Simple->new();
232 $response = $xmlsimple->XMLin(
234 forcearray => [ qw(Fld980) ],
238 for my $fld980 (@{$response->{VarFlds}->{VarDFlds}->{SSIFlds}->{Fld980}}) {
239 # this is absurd, but sometimes this data serializes differently
240 if(ref($fld980->{a}->{content}) eq 'ARRAY') {
241 for my $content (@{$fld980->{a}->{content}}) {
242 push @anotes, {content => $content};
247 push @anotes, {content => $fld980->{a}->{content}};
253 sub _fetch_syndetics_content {
254 my ( $element, $isbn, $upc, $oclc ) = @_;
256 $isbn = '' unless defined $isbn;
257 $upc = '' unless defined $upc;
258 $oclc = '' unless defined $oclc;
260 my $syndetics_client_code = C4::Context->preference('SyndeticsClientCode');
262 my $url = "http://www.syndetics.com/index.aspx?isbn=$isbn/$element&client=$syndetics_client_code&type=xw10&upc=$upc&oclc=$oclc";
263 my $ua = LWP::UserAgent->new;
266 my $response = $ua->get($url);
268 warn "could not retrieve $url" unless $response->content;
281 Joshua Ferraro <jmf@liblime.com>