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 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 qw ( -utf8 );
  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     'NORMARC' => {
  79         'marcxml'      => 'identity.xsl',
  80         'marcxml-full' => 'identity.xsl',
  81         'mods'         => 'MARC21slim2MODS.xsl',
  82         'mods-full'    => 'MARC21slim2MODS.xsl',
  83         'mods3'        => 'MARC21slim2MODS3-1.xsl',
  84         'mods3-full'   => 'MARC21slim2MODS3-1.xsl',
  85         'oai_dc'       => 'MARC21slim2OAIDC.xsl',
  86         'rdfdc',       => 'MARC21slim2RDFDC.xsl',
  87         'rss2'         => 'MARC21slim2RSS2.xsl',
  88         'rss2-full'    => 'MARC21slim2RSS2.xsl',
  89         'srw_dc'       => 'MARC21slim2SRWDC.xsl',
  90     },
  91     'UNIMARC' => {
  92         'marcxml'      => 'identity.xsl',
  93         'marcxml-full' => 'identity.xsl',
  94         'oai_dc'       => 'UNIMARCslim2OAIDC.xsl',
  95         'rdfdc',       => 'UNIMARCslim2RDFDC.xsl',
  96         'srw_dc'       => 'UNIMARCslim2SRWDC.xsl',
  97     },
  98 };
  99
 100 =head2 $format_info
 101
 102 This hashref maps from unAPI output formats to the <format> elements
 103 used to describe them in an unAPI format request.
 104
 105 =cut
 106
 107 my $format_info = {
 108     '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"/>),
 109     '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"/>),
 110     '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"/>),
 111     '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"/>),
 112     '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"/>),
 113     '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"/>),
 114     '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"/>),
 115     '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/"/>),
 116     'rss2' => q(<format name="rss2" type="application/xml"/>),
 117     'rss2-full' => q(<format name="rss2-full" type="application/xml"/>),
 118     '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"/>),
 119 };
 120
 121 my $id     = $cgi->param('id');
 122 my $format = $cgi->param('format');
 123
 124 if (not defined $format) {
 125     emit_formats($id, $format_to_stylesheet_map, $format_info, $cgi);
 126 } elsif ($id) {
 127
 128     # koha:biblionumber:0152018484
 129     if ($id =~ /koha:biblionumber:(\d+)/) {
 130         my $biblionumber = $1;
 131
 132         my $content;
 133         eval {
 134             my $marcxml = GetXmlBiblio($biblionumber);
 135             unless (defined $marcxml) {
 136                 # no bib, so 404
 137                 print $cgi->header( -status => '404 record not found');
 138                 exit 0;
 139             }
 140
 141             my $transformer = get_transformer($format, $format_to_stylesheet_map, $format_info);
 142             unless (defined $transformer) {
 143                 print $cgi->header( -status => '406 invalid format requested' );
 144                 exit 0;
 145             }
 146             my $parser = XML::LibXML->new();
 147             my $record_dom = $parser->parse_string( $marcxml );
 148             $record_dom = $transformer->transform( $record_dom );
 149             $content = $transformer->output_as_chars( $record_dom );
 150         };
 151         if ($@) {
 152             print $cgi->header( -status => '500 internal error ' . $@->code() . ": " . $@->message() );
 153             exit 0;
 154         }
 155
 156         print $cgi->header( -type =>'application/xml', -charset => 'UTF-8' );
 157         print $content;
 158     } else {
 159         # ID is obviously wrong, so 404
 160         print $cgi->header( -status => '404 record not found');
 161         exit 0;
 162     }
 163 } else {
 164     # supplied a format but no id - caller is doing it wrong
 165     print $cgi->header( -status => '400 bad request - if you specify format, must specify id');
 166     exit 0;
 167 }
 168
 169 exit 0;
 170
 171 sub emit_formats {
 172     my ($id, $format_to_stylesheet_map, $format_info, $cgi) = @_;
 173
 174     if (defined $id) {
 175         print $cgi->header( -type =>'application/xml', -status => '300 multiple choices' );
 176     } else {
 177         print $cgi->header( -type =>'application/xml', -status => '200 Ok' );
 178     }
 179
 180     print "<?xml version='1.0' encoding='utf-8'  ?>\n";
 181     if (defined $id) {
 182         print qq(<formats id="$id">\n);
 183     } else {
 184         print "<formats>\n";
 185     }
 186
 187     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 188     foreach my $format (sort keys %{ $format_to_stylesheet_map->{$marcflavour} }) {
 189         print $format_info->{$format}, "\n";
 190     }
 191     print "</formats>\n";
 192     return;
 193 }
 194
 195
 196 sub get_transformer {
 197     my ($format, $format_to_stylesheet_map, $format_info) = @_;
 198     $format = lc $format;
 199
 200     my $marcflavour = uc(C4::Context->preference('marcflavour'));
 201     return unless $format_to_stylesheet_map->{$marcflavour}->{$format};
 202
 203     my $xslt_file = C4::Context->config('intrahtdocs') .
 204                     "/prog/en/xslt/" .
 205                     $format_to_stylesheet_map->{$marcflavour}->{$format};
 206
 207     my $parser = XML::LibXML->new();
 208     my $xslt = XML::LibXSLT->new();
 209     my $style_doc = $parser->parse_file( $xslt_file );
 210     my $stylesheet = $xslt->parse_stylesheet( $style_doc );
 211
 212     return $stylesheet;
 213 }
 214
 215 =head1 AUTHOR
 216
 217 Koha Development Team <http://koha-community.org/>
 218
 219 Originally written by Joshua Ferraro <jmf@liblime.com>
 220
 221 Improved by Galen Charlton <galen.charlton@liblime.com>
 222
 223 =cut