8 =head1 OAI-PMH for koha
10 This file is an implementation of the OAI-PMH protocol for koha. Its purpose
11 is to share metadata in Dublin core format with harvester like PKP-Harverster.
12 Presently, all the bibliographic records managed by the runing koha instance
13 are publicly shared (as the opac is).
15 =head1 Package MARC::Record::KOHADC
17 This package is a sub-class of the MARC::File::USMARC. It add methods and functions
18 to map the content of a marc record (of any flavor) to Dublin core.
19 As soon as it is possible, mapping between marc fields and there semantic
20 are got from ::GetMarcFromKohaField fonction from C4::Biblio (see also the "Koha
21 to MARC mapping" preferences).
25 package MARC::Record::KOHADC;
27 @ISA = qw(MARC::Record);
29 use MARC::File::USMARC;
31 sub new { # Get a MAR::Record as parameter and bless it as MARC::Record::KOHADC
34 bless $marc if( ref( $marc ) );
41 return $self->SUPER::subfield( @_ ) unless wantarray;
43 my @field = $self->field($t);
47 foreach $f ( @field ) {
48 push( @list, $f->subfield( $sf ) );
57 foreach my $kohafield ( @_ ) {
58 my ( $field, $subfield ) = ::GetMarcFromKohaField( $kohafield, '' );
59 push( @result, $field < 10 ? $marc->field( $field )->as_string() : $marc->subfield( $field, $subfield ) );
61 # @result>1 ? \@result : $result[0];
69 s/\&/\&/g; s/</</g;
81 &XMLescape( $self->getfields('biblio.title') );
86 &XMLescape( $self->getfields('biblio.author') );
91 &XMLescape( $self->getfields('bibliosubject.subject') );
96 my ($d,$h) = split( ' ', $self->{'biblio.timestamp'} );
102 my ($str) = @{$self->getfields('biblioitems.publicationyear')};
103 my ($y,$m,$d) = (substr($str,0,4), substr($str,4,2), substr($str,6,2));
105 $y=1970 unless($y>0); $m=1 unless($m>0); $d=1 unless($d>0);
107 sprintf( "%.4d-%.2d-%.2d", $y,$m,$d);
117 my $id = $self->getfields('biblio.biblionumber')->[0];
119 # get url of this script and assume that OAI server is in the same place as opac-detail script
120 # and build a direct link to the record.
121 my $uri = $ENV{'SCRIPT_URI'};
122 $uri= "http://" . $ENV{'HTTP_HOST'} . $ENV{'REQUEST_URI'} unless( $uri ); # SCRIPT_URI doesn't exist on all httpd server
125 C4::Context->preference("OAI-PMH:archiveID") .":" .$id,
126 "${uri}opac-detail.pl?bib=$id",
127 @{$self->getfields('biblioitems.isbn', 'biblioitems.issn')}
138 &XMLescape( $self->getfields('biblioitems.itemtype') );
143 &XMLescape( $self->getfields('biblioitems.publishercode') );
147 my $set = &OAI::KOHA::Set();
148 [ map( $_=$_->[0], @$set) ];
151 =head1 The OAI::KOHA package
153 This package is a subclass of the OAI::DC data provider. It overides needed methods
154 and provide the links between the OAI-PMH request and the koha application.
155 The data used in answers are from the koha table I<bibio>.
163 @ISA = ("C4::OAI::DC");
167 return the Set list to the I<verb=ListSets> query. Data are from the 'OAI-PMH:Set' preference.
173 # ['BRISE','Experimental unimarc set for BRISE network'],
174 # ['BRISE:EMSE','EMSE set in BRISE network']
177 # A blinder correctement
178 [ map( $_ = [ split(",", $_)], split( "\n",C4::Context->preference("OAI-PMH:Set") ) ) ];
183 The new method is the constructor for this class. It doesn't have any parameters and
184 get required data from koha preferences. Koha I<LibraryName> is used to identify the
185 OAI-PMH repository, I<OAI-PMH:MaxCount> is used to set the maximun number of records
186 returned at the same time in answers to I<verb=ListRecords> or I<verb=ListIdentifiers>
189 The method return a blessed reference.
196 my $classname = shift;
197 my $self = $classname->SUPER::new ();
200 $self->{'repositoryName'} = C4::Context->preference("LibraryName");
201 $self->{'MaxCount'} = C4::Context->preference("OAI-PMH:MaxCount");
202 $self->{'adminEmail'} = C4::Context->preference("KohaAdminEmailAddress");
204 bless $self, $classname;
210 The dispose method is used as a destructor. It call just the SUPER::dispose method.
218 $self->SUPER::dispose ();
223 my ($sec,$min,$hour,$mday,$mon,$year,$wday,$yday) = gmtime( time );
225 sprintf( "%.4d-%.2d-%.2d", $year+1900, $mon+1,$mday );
228 # build the resumptionTocken fom ($metadataPrefix,$offset,$from,$until)
230 =head2 buildResumptionToken and parseResumptionToken
232 Theses two functions are used to manage resumption tokens. The choosed syntax is simple as
233 possible, a token is only the metadata prefix, the offset in the full answer, the from and
234 the until date (in the yyyy-mm-dd format) joined by ':' caracter.
236 I<buildResumptionToken> get the four elements as parameters and return the ':' separated
239 I<parseResumptionToken> is used to set the default values to the from and until date, the
240 metadata prefix using the resumption tocken if necessary. This function have four parameters
241 (from,until,metadata prefix and resumption tocken) which can be undefined and return every
242 time this list of values correctly set. The missing values are set with defaults: offset=0,
243 from= 1970-01-01 and until is set to current date.
247 sub buildResumptionToken {
251 # parse the resumptionTocken
252 sub parseResumptionToken {
253 my ($from, $until, $metadataPrefix, $resumptionToken) = @_;
256 if( $resumptionToken ) {
257 ($metadataPrefix,$offset,$from,$until) = split( ':', $resumptionToken );
260 $from = "1970-01-01" unless( $from );
261 $until = &now unless( $until );
262 ($metadataPrefix, $offset, $from, $until );
265 =head2 Archive_ListSets
267 return the full list Set to the I<verb=ListSets> query. Data are from the 'OAI-PMH:Set' preference.
271 # get full list of sets from the archive
277 =head2 Archive_GetRecord
279 This method select the record specified as its first parameter from the koha I<biblio>
280 table and return a reference to a MARC::Record::KOHADC object.
284 # get a single record from the archive
285 sub Archive_GetRecord
287 my ($self, $identifier, $metadataFormat) = @_;
288 my $dbh = C4::Context->dbh;
289 my $sth = $dbh->prepare("SELECT biblionumber,timestamp FROM biblio WHERE biblionumber=?");
290 my $prefixID = C4::Context->preference("OAI-PMH:archiveID"); $prefixID=qr{$prefixID:};
292 $identifier =~ s/^$prefixID//;
294 $sth->execute( $identifier );
296 if( my $r = $sth->fetchrow_hashref() ) {
297 my $marc = new MARC::Record::KOHADC( ::GetMarcBiblio( $identifier ) );
299 $marc->{'biblio.timestamp'} = $r->{'timestamp'};
303 warn("Archive_GetRecord : no MARC record for " . C4::Context->preference("OAI-PMH:archiveID") . ":" . $identifier);
307 $self->AddError ('idDoesNotExist', 'The value of the identifier argument is unknown or illegal in this repository');
311 =head2 Archive_ListRecords
313 This method return a list of 'MaxCount' references to MARC::Record::KOHADC object build from the
314 koha I<biblio> table according to its parameters : set, from and until date, metadata prefix
315 and resumption token.
319 # list metadata records from the archive
320 sub Archive_ListRecords
322 my ($self, $set, $from, $until, $metadataPrefix, $resumptionToken) = @_;
328 my $dbh = C4::Context->dbh;
329 my $sth = $dbh->prepare("SELECT biblionumber,timestamp FROM biblio WHERE DATE(timestamp) >= ? and DATE(timestamp) <= ? LIMIT ? OFFSET ?");
332 ($metadataPrefix, $offset, $from, $until ) = &parseResumptionToken($from, $until, $metadataPrefix, $resumptionToken);
334 #warn( "Archive_ListRecords : $set, $from, $until, $metadataPrefix, $resumptionToken\n");
335 $sth->execute( $from,$until,$self->{'MaxCount'}?$self->{'MaxCount'}:100000, $offset );
337 while( my $r = $sth->fetchrow_hashref() ) {
338 my $marc = new MARC::Record::KOHADC( ::GetMarcBiblio( $r->{'biblionumber'} ) );
339 unless( $marc ) { # somme time there is problems within koha, and we can't get valid marc record
340 warn("Archive_ListRecords : no MARC record for " . C4::Context->preference("OAI-PMH:archiveID") .":" . $r->{'biblionumber'} );
343 $marc->{'biblio.timestamp'} = $r->{'timestamp'};
344 push( @allrows, $marc );
347 $sth = $dbh->prepare("SELECT count(*) FROM biblioitems WHERE DATE(timestamp) >= ? and DATE(timestamp) <= ?");
348 $sth->execute($from, $until);
349 ( $count ) = $sth->fetchrow_array();
352 $self->AddError ('noRecordsMatch', 'The combination of the values of arguments results in an empty set');
355 if( $offset + $self->{'MaxCount'} < $count ) { # Not at the end
356 $offset = $offset + $self->{'MaxCount'};
357 $resumptionToken = &buildResumptionToken($metadataPrefix,$offset,$from,$until);
358 $tokenInfo = { 'completeListSize' => $count, 'cursor' => $offset };
361 $resumptionToken = '';
364 ( \@allrows, $resumptionToken, $metadataPrefix, $tokenInfo );
371 The I<main> function is the starting point of the service. The first step is
372 to verify if the service is enable using the 'OAI-PMH' preference value
373 (See Koha systeme preferences).
375 If the service is enable, it create a new instance of the OAI::KOHA data
376 provider (see before) and run the service.
381 print "Status:404 OAI-PMH service is disabled\n";
382 print "Content-type: text/plain\n\n";
384 print "OAI-PMH service is disable.\n";
389 return &disable() unless( C4::Context->preference('OAI-PMH') );
391 my $OAI = new OAI::KOHA();