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
   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.
  11 #
  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.
  16 #
  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>.
  19
  20 use Modern::Perl;
  21
  22 =head1 NAME
  23
  24 unapi - implement unAPI for the OPAC
  25
  26 =head1 SYNOPSIS
  27
  28 Retrieve http://library.example.org/cgi-bin/koha/unapi?id=koha:biblionumber:123&format=oai_dc
  29
  30 =head1 DESCRIPTION
  31
  32 Implements unAPI <http://unapi.info>, a small HTTP API for retrieving structured
  33 content from a web application.  The primary application of unAPI in Koha is to
  34 allow tools such as Zotero to identify and grab bibliographic record metadata in
  35 an XML format such as OAI DC, RSS2, MARCXML, or MODS.
  36
  37 =cut
  38
  39 use CGI qw ( -utf8 );
  40 use C4::Context;
  41 use C4::Biblio;
  42 use Koha::XSLT_Handler;
  43
  44 my $cgi = CGI->new();
  45 binmode(STDOUT, ":encoding(UTF-8)"); #output as utf8
  46
  47 =head1 VARIABLES
  48
  49 =head2 $format_to_stylesheet_map
  50
  51 This hashref of hashrefs maps from a MARC flavour and unAPI format
  52 to the stylesheet that should be used to transform the bib MARCXML
  53 to the desired output format.  As new MARC XSLT stylesheets are added,
  54 (particularly for UNIMARC), this map should be updated.  Of course,
  55 if/when we add support for emitting a format that is not genreated
  56 by a stylesheet, the structure of this variable will have to be changed.
  57 At present, this doubles as the list of output formats supported by
  58 this unAPI implementation.
  59
  60 =cut
  61
  62 my $format_to_stylesheet_map = {
  63     'MARC21' => {
  64         'marcxml'      => 'identity.xsl',
  65         'marcxml-full' => 'identity.xsl',
  66         'mods'         => 'MARC21slim2MODS.xsl',
  67         'mods-full'    => 'MARC21slim2MODS.xsl',
  68         'mods3'        => 'MARC21slim2MODS3-1.xsl',
  69         'mods3-full'   => 'MARC21slim2MODS3-1.xsl',
  70         'oai_dc'       => 'MARC21slim2OAIDC.xsl',
  71         'rdfdc',       => 'MARC21slim2RDFDC.xsl',
  72         'rss2'         => 'MARC21slim2RSS2.xsl',
  73         'rss2-full'    => 'MARC21slim2RSS2.xsl',
  74         'srw_dc'       => 'MARC21slim2SRWDC.xsl',
  75     },
  76     'NORMARC' => {
  77         'marcxml'      => 'identity.xsl',
  78         'marcxml-full' => 'identity.xsl',
  79         'mods'         => 'MARC21slim2MODS.xsl',
  80         'mods-full'    => 'MARC21slim2MODS.xsl',
  81         'mods3'        => 'MARC21slim2MODS3-1.xsl',
  82         'mods3-full'   => 'MARC21slim2MODS3-1.xsl',
  83         'oai_dc'       => 'MARC21slim2OAIDC.xsl',
  84         'rdfdc',       => 'MARC21slim2RDFDC.xsl',
  85         'rss2'         => 'MARC21slim2RSS2.xsl',
  86         'rss2-full'    => 'MARC21slim2RSS2.xsl',
  87         'srw_dc'       => 'MARC21slim2SRWDC.xsl',
  88     },
  89     'UNIMARC' => {
  90         'marcxml'      => 'identity.xsl',
  91         'marcxml-full' => 'identity.xsl',
  92         'oai_dc'       => 'UNIMARCslim2OAIDC.xsl',
  93         'rdfdc',       => 'UNIMARCslim2RDFDC.xsl',
  94         'srw_dc'       => 'UNIMARCslim2SRWDC.xsl',
  95     },
  96 };
  97
  98 =head2 $format_info
  99
 100 This hashref maps from unAPI output formats to the <format> elements
 101 used to describe them in an unAPI format request.
 102
 103 =cut
 104
 105 my $format_info = {
 106     '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"/>),
 107     '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"/>),
 108     '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"/>),
 109     '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"/>),
 110     '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"/>),
 111     '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"/>),
 112     '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"/>),
 113     '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/"/>),
 114     'rss2' => q(<format name="rss2" type="application/xml"/>),
 115     'rss2-full' => q(<format name="rss2-full" type="application/xml"/>),
 116     '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"/>),
 117 };
 118
 119 my $id     = $cgi->param('id');
 120 my $format = $cgi->param('format');
 121
 122 if (not defined $format) {
 123     emit_formats($id, $format_to_stylesheet_map, $format_info, $cgi);
 124 } elsif ($id) {
 125
 126     # koha:biblionumber:0152018484
 127     if ($id =~ /koha:biblionumber:(\d+)/) {
 128         my $biblionumber = $1;
 129
 130         my $content;
 131
 132         my $marcxml = GetXmlBiblio($biblionumber);
 133         unless (defined $marcxml) {
 134             # no bib, so 404
 135             print $cgi->header( -status => '404 record not found');
 136             exit 0;
 137         }
 138
 139         my $xslt_file = get_xslt_file( $format, $format_to_stylesheet_map, $format_info );
 140         unless( defined $xslt_file ) {
 141             print $cgi->header( -status => '406 invalid format requested' );
 142             exit 0;
 143         }
 144         my $xslt_engine = Koha::XSLT_Handler->new;
 145         $content = $xslt_engine->transform({
 146             xml => $marcxml,
 147             file => $xslt_file,
 148         });
 149
 150         if( !defined $content || $xslt_engine->err ) {
 151             print $cgi->header( -status => '500 internal error' );
 152             exit 0;
 153         }
 154
 155         print $cgi->header( -type =>'application/xml', -charset => 'UTF-8' );
 156         print $content;
 157     } else {
 158         # ID is obviously wrong, so 404
 159         print $cgi->header( -status => '404 record not found');
 160         exit 0;
 161     }
 162 } else {
 163     # supplied a format but no id - caller is doing it wrong
 164     print $cgi->header( -status => '400 bad request - if you specify format, must specify id');
 165     exit 0;
 166 }
 167
 168 exit 0;
 169
 170 sub emit_formats {
 171     my ($id, $format_to_stylesheet_map, $format_info, $cgi) = @_;
 172
 173     if (defined $id) {
 174         print $cgi->header( -type =>'application/xml', -status => '300 multiple choices' );
 175     } else {
 176         print $cgi->header( -type =>'application/xml', -status => '200 Ok' );
 177     }
 178
 179     print "<?xml version='1.0' encoding='utf-8'  ?>\n";
 180     if (defined $id) {
 181         print qq(<formats id="$id">\n);
 182     } else {
 183         print "<formats>\n";
 184     }
 185
 186     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 187     foreach my $format (sort keys %{ $format_to_stylesheet_map->{$marcflavour} }) {
 188         print $format_info->{$format}, "\n";
 189     }
 190     print "</formats>\n";
 191     return;
 192 }
 193
 194
 195 sub get_xslt_file {
 196     my ($format, $format_to_stylesheet_map, $format_info) = @_;
 197     $format = lc $format;
 198
 199     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 200     return unless $format_to_stylesheet_map->{$marcflavour}->{$format};
 201
 202     my $xslt_file = C4::Context->config('intrahtdocs') .
 203                     "/prog/en/xslt/" .
 204                     $format_to_stylesheet_map->{$marcflavour}->{$format};
 205
 206     return $xslt_file;
 207 }
 208
 209 =head1 AUTHOR
 210
 211 Koha Development Team <http://koha-community.org/>
 212
 213 Originally written by Joshua Ferraro <jmf@liblime.com>
 214
 215 Improved by Galen Charlton <galen.charlton@liblime.com>
 216
 217 =cut