opac/unapi

   1 #!/usr/bin/perl
   2
   3 # Copyright 2008-2009 LibLime
   4 #
   5 # This file is part of Koha.
   6 #
   7 # Koha is free software; you can redistribute it and/or modify it under the
   8 # terms of the GNU General Public License as published by the Free Software
   9 # Foundation; either version 2 of the License, or (at your option) any later
  10 # version.
  11 #
  12 # Koha is distributed in the hope that it will be useful, but WITHOUT ANY
  13 # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR
  14 # A PARTICULAR PURPOSE.  See the GNU General Public License for more details.
  15 #
  16 # You should have received a copy of the GNU General Public License along
  17 # with Koha; if not, write to the Free Software Foundation, Inc.,
  18 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  19
  20 use strict;
  21 use warnings;
  22
  23 =head1 NAME
  24
  25 unapi - implement unAPI for the OPAC
  26
  27 =head1 SYNOPSIS
  28
  29 Retrieve http://library.example.org/cgi-bin/koha/unapi?id=koha:biblionumber:123&format=oai_dc
  30
  31 =head1 DESCRIPTION
  32
  33 Implements unAPI <http://unapi.info>, a small HTTP API for retrieving structured
  34 content from a web application.  The primary application of unAPI in Koha is to
  35 allow tools such as Zotero to identify and grab bibliographic record metadata in
  36 an XML format such as OAI DC, RSS2, MARCXML, or MODS.
  37
  38 =cut
  39
  40 use CGI;
  41 use C4::Context;
  42 use C4::Biblio;
  43 use XML::LibXML;
  44 use XML::LibXSLT;
  45
  46 my $cgi = CGI->new();
  47 binmode(STDOUT, ":encoding(UTF-8)"); #output as utf8
  48
  49 =head1 VARIABLES
  50
  51 =head2 $format_to_stylesheet_map
  52
  53 This hashref of hashrefs maps from a MARC flavour and unAPI format
  54 to the stylesheet that should be used to transform the bib MARCXML
  55 to the desired output format.  As new MARC XSLT stylesheets are added,
  56 (particularly for UNIMARC), this map should be updated.  Of course,
  57 if/when we add support for emitting a format that is not genreated
  58 by a stylesheet, the structure of this variable will have to be changed.
  59 At present, this doubles as the list of output formats supported by
  60 this unAPI implementation.
  61
  62 =cut
  63
  64 my $format_to_stylesheet_map = {
  65     'MARC21' => {
  66         'marcxml'      => 'identity.xsl',
  67         'marcxml-full' => 'identity.xsl',
  68         'mods'         => 'MARC21slim2MODS.xsl',
  69         'mods-full'    => 'MARC21slim2MODS.xsl',
  70         'mods3'        => 'MARC21slim2MODS3-1.xsl',
  71         'mods3-full'   => 'MARC21slim2MODS3-1.xsl',
  72         'oai_dc'       => 'MARC21slim2OAIDC.xsl',
  73         'rdfdc',       => 'MARC21slim2RDFDC.xsl',
  74         'rss2'         => 'MARC21slim2RSS2.xsl',
  75         'rss2-full'    => 'MARC21slim2RSS2.xsl',
  76         'srw_dc'       => 'MARC21slim2SRWDC.xsl',
  77     },
  78     'UNIMARC' => {
  79         'marcxml'      => 'identity.xsl',
  80         'marcxml-full' => 'identity.xsl',
  81         'oai_dc'       => 'UNIMARCslim2OAIDC.xsl',
  82     },
  83 };
  84
  85 =head2 $format_info
  86
  87 This hashref maps from unAPI output formats to the <format> elements
  88 used to describe them in an unAPI format request.
  89
  90 =cut
  91
  92 my $format_info = {
  93     'marcxml' => q(<format name="marcxml" type="application/xml" namespace_uri="http://www.loc.gov/MARC21/slim" docs="http://www.loc.gov/marcxml/" schema_location="http://www.loc.gov/standards/marcxml/schema/MARC21slim.xsd"/>),
  94     'marcxml-full' => q(<format name="marcxml-full" type="application/xml" namespace_uri="http://www.loc.gov/MARC21/slim" docs="http://www.loc.gov/marcxml/" schema_location="http://www.loc.gov/standards/marcxml/schema/MARC21slim.xsd"/>),
  95     'mods' => q(<format name="mods" type="application/xml" namespace_uri="http://www.loc.gov/mods/" docs="http://www.loc.gov/mods/" schema_location="http://www.loc.gov/standards/mods/mods.xsd"/>),
  96     'mods-full' => q(<format name="mods-full" type="application/xml" namespace_uri="http://www.loc.gov/mods/" docs="http://www.loc.gov/mods/" schema_location="http://www.loc.gov/standards/mods/mods.xsd"/>),
  97     'mods3' => q(<format name="mods3" type="application/xml" namespace_uri="http://www.loc.gov/mods/v3" docs="http://www.loc.gov/mods/" schema_location="http://www.loc.gov/standards/mods/v3/mods-3-1.xsd"/>),
  98     'mods3-full' => q(<format name="mods3-full" type="application/xml" namespace_uri="http://www.loc.gov/mods/v3" docs="http://www.loc.gov/mods/" schema_location="http://www.loc.gov/standards/mods/v3/mods-3-1.xsd"/>),
  99     'oai_dc' => q(<format name="oai_dc" type="application/xml" namespace_uri="http://www.openarchives.org/OAI/2.0/oai_dc/" schema_location="http://www.openarchives.org/OAI/2.0/oai_dc.xsd"/>),
 100     'rdfdc' => q(<format name="rdfdc" type="application/xml" namespace_uri="http://purl.org/dc/elements/1.1/" schema_location="http://purl.org/dc/elements/1.1/"/>),
 101     'rss2' => q(<format name="rss2" type="application/xml"/>),
 102     'rss2-full' => q(<format name="rss2-full" type="application/xml"/>),
 103     'srw_dc' => q(<format name="srw_dc" type="application/xml" namespace_uri="info:srw/schema/1/dc-schema" schema_location="http://www.loc.gov/z3950/agency/zing/srw/dc-schema.xsd"/>),
 104 };
 105
 106 my $id     = $cgi->param('id');
 107 my $format = $cgi->param('format');
 108
 109 if (not defined $format) {
 110     emit_formats($id);
 111 } elsif ($id) {
 112
 113     # koha:biblionumber:0152018484
 114     if ($id =~ /koha:biblionumber:(\d+)/) {
 115         my $biblionumber = $1;
 116
 117         my $content;
 118         eval {
 119             my $marcxml = GetXmlBiblio($biblionumber);
 120             unless (defined $marcxml) {
 121                 # no bib, so 404
 122                 print $cgi->header( -status => '404 record not found');
 123                 exit 0;
 124             }
 125
 126             my $transformer = get_transformer($format);
 127             unless (defined $transformer) {
 128                 print $cgi->header( -status => '406 invalid format requested' );
 129                 exit 0;
 130             }
 131             my $parser = XML::LibXML->new();
 132             my $record_dom = $parser->parse_string( $marcxml );
 133             $record_dom = $transformer->transform( $record_dom );
 134             $content = $record_dom->toString();
 135         };
 136         if ($@) {
 137             print $cgi->header( -status => '500 internal error ' . $@->code() . ": " . $@->message() );
 138             exit 0;
 139         }
 140
 141         print $cgi->header( -type =>'application/xml' );
 142         print $content;
 143     } else {
 144         # ID is obviously wrong, so 404
 145         print $cgi->header( -status => '404 record not found');
 146         exit 0;
 147     }
 148 } else {
 149     # supplied a format but no id - caller is doing it wrong
 150     print $cgi->header( -status => '400 bad request - if you specify format, must specify id');
 151     exit 0;
 152 }
 153
 154 exit 0;
 155
 156 sub emit_formats {
 157     my $id = shift;
 158
 159     if (defined $id) {
 160         print $cgi->header( -type =>'application/xml', -status => '300 multiple choices' );
 161     } else {
 162         print $cgi->header( -type =>'application/xml' );
 163     }
 164
 165     print "<?xml version='1.0' encoding='utf-8'  ?>\n";
 166     if (defined $id) {
 167         print qq(<formats id="$id">\n);
 168     } else {
 169         print "<formats>\n";
 170     }
 171
 172     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 173     foreach my $format (sort keys %{ $format_to_stylesheet_map->{$marcflavour} }) {
 174         print $format_info->{$format}, "\n";
 175     }
 176     print "</formats>\n";
 177     return;
 178 }
 179
 180
 181 sub get_transformer {
 182     my $format = lc shift;
 183
 184     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 185     return unless $format_to_stylesheet_map->{$marcflavour}->{$format};
 186
 187     my $xslt_file = C4::Context->config('intranetdir') .
 188                     "/koha-tmpl/intranet-tmpl/prog/en/xslt/" .
 189                     $format_to_stylesheet_map->{$marcflavour}->{$format};
 190
 191     my $parser = XML::LibXML->new();
 192     my $xslt = XML::LibXSLT->new();
 193     my $style_doc = $parser->parse_file( $xslt_file );
 194     my $stylesheet = $xslt->parse_stylesheet( $style_doc );
 195
 196     return $stylesheet;
 197 }
 198
 199 =head1 AUTHOR
 200
 201 Koha Development Team <http://koha-community.org/>
 202
 203 Originally written by Joshua Ferraro <jmf@liblime.com>
 204
 205 Improved by Galen Charlton <galen.charlton@liblime.com>
 206
 207 =cut