1 # $Id: Session.pm,v 1.19 2003-07-26 16:27:46 pop Exp $
3 # Zebra perl API header
4 # =============================================================================
5 package IDZebra::Session;
14 use IDZebra::Logger qw(:flags :calls);
15 use IDZebra::Resultset;
16 use IDZebra::ScanList;
17 use IDZebra::RetrievalRecord;
19 our $VERSION = do { my @r = (q$Revision: 1.19 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r };
20 our @ISA = qw(IDZebra::Logger Exporter);
21 our @EXPORT = qw (TRANS_RW TRANS_RO);
24 use constant TRANS_RW => 1;
25 use constant TRANS_RO => 0;
28 # -----------------------------------------------------------------------------
29 # Class constructors, destructor
30 # -----------------------------------------------------------------------------
32 my ($proto, %args) = @_;
33 my $class = ref($proto) || $proto;
35 $self->{args} = \%args;
37 bless ($self, $class);
38 $self->{cql_ct} = undef;
39 $self->{cql_mapfile} = "";
42 $self->{databases} = {};
46 my ($self, %args) = @_;
49 unless (defined($self->{zs})) {
50 if (defined($args{'configFile'})) {
51 $self->{zs} = IDZebra::start($args{'configFile'});
53 $self->{zs} = IDZebra::start("zebra.cfg");
60 if (defined($self->{zs})) {
61 IDZebra::stop($self->{zs}) if ($self->{zs});
68 my ($proto,%args) = @_;
71 if (ref($proto)) { $self = $proto; } else {
72 $self = $proto->new(%args);
76 %args = %{$self->{args}};
79 $self->start_service(%args);
81 unless (defined($self->{zs})) {
82 croak ("Falied to open zebra service");
85 unless (defined($self->{zh})) {
86 $self->{zh}=IDZebra::open($self->{zs});
89 # Reset result set counter
92 # This is needed in order to somehow initialize the service
93 $self->databases("Default");
95 # Load the default configuration
100 my $shadow = defined($args{shadow}) ? $args{shadow} : 0;
101 $self->shadow($shadow);
103 $self->{odr_input} = IDZebra::odr_createmem($IDZebra::ODR_DECODE);
104 $self->{odr_output} = IDZebra::odr_createmem($IDZebra::ODR_ENCODE);
111 unless (defined($self->{zh})) {
112 croak ("Zebra session is not opened");
122 # Delete all resulsets
123 my $r = IDZebra::deleteResultSet($self->{zh},
124 1, #Z_DeleteRequest_all,
128 while (IDZebra::trans_no($self->{zh}) > 0) {
129 logf (LOG_WARN,"Explicitly closing transaction with session");
133 IDZebra::close($self->{zh});
137 if ($self->{odr_input}) {
138 IDZebra::odr_reset($self->{odr_input});
139 IDZebra::odr_destroy($self->{odr_input});
140 $self->{odr_input} = undef;
143 if ($self->{odr_output}) {
144 IDZebra::odr_reset($self->{odr_output});
145 IDZebra::odr_destroy($self->{odr_output});
146 $self->{odr_output} = undef;
154 logf (LOG_LOG,"DESTROY $self");
157 if (defined ($self->{cql_ct})) {
158 IDZebra::cql_transform_close($self->{cql_ct});
162 # -----------------------------------------------------------------------------
163 # Record group selection This is a bit nasty... but used at many places
164 # -----------------------------------------------------------------------------
166 my ($self,%args) = @_;
169 $self->{rg} = $self->_makeRecordGroup(%args);
170 $self->_selectRecordGroup($self->{rg});
175 sub selectRecordGroup {
176 my ($self, $groupName) = @_;
178 $self->{rg} = $self->_getRecordGroup($groupName);
179 $self->_selectRecordGroup($self->{rg});
182 sub _displayRecordGroup {
183 my ($self, $rg) = @_;
184 print STDERR "-----\n";
185 foreach my $key qw (groupName
196 print STDERR "$key:",$rg->{$key},"\n";
200 sub _cloneRecordGroup {
201 my ($self, $orig) = @_;
202 my $rg = IDZebra::recordGroup->new();
203 my $r = IDZebra::init_recordGroup($rg);
204 foreach my $key qw (groupName
216 $rg->{$key} = $orig->{$key} if ($orig->{$key});
221 sub _getRecordGroup {
222 my ($self, $groupName, $ext) = @_;
223 my $rg = IDZebra::recordGroup->new();
224 my $r = IDZebra::init_recordGroup($rg);
225 $rg->{groupName} = $groupName if ($groupName ne "");
226 $ext = "" unless ($ext);
227 $r = IDZebra::res_get_recordGroup($self->{zh}, $rg, $ext);
231 sub _makeRecordGroup {
232 my ($self, %args) = @_;
235 my @keys = keys(%args);
236 unless ($#keys >= 0) {
237 return ($self->{rg});
240 if ($args{groupName}) {
241 $rg = $self->_getRecordGroup($args{groupName});
243 $rg = $self->_cloneRecordGroup($self->{rg});
245 $self->_setRecordGroupOptions($rg, %args);
249 sub _setRecordGroupOptions {
250 my ($self, $rg, %args) = @_;
252 foreach my $key qw (databaseName
263 if (defined ($args{$key})) {
264 $rg->{$key} = $args{$key};
268 sub _selectRecordGroup {
269 my ($self, $rg) = @_;
271 my $r = IDZebra::set_group($self->{zh}, $rg);
273 unless ($dbName = $rg->{databaseName}) {
276 unless ($self->databases($dbName)) {
277 croak("Fatal error selecting database $dbName");
280 # -----------------------------------------------------------------------------
281 # Selecting databases for search (and also for updating - internally)
282 # -----------------------------------------------------------------------------
284 my ($self, @databases) = @_;
289 return (keys(%{$self->{databases}}));
294 foreach my $db (@databases) {
296 next if ($self->{databases}{$db});
300 foreach my $db (keys (%{$self->{databases}})) {
301 $changed++ unless ($tmp{$db});
306 delete ($self->{databases});
307 foreach my $db (@databases) {
308 $self->{databases}{$db}++;
311 if (IDZebra::select_databases($self->{zh},
315 "Could not select database(s) %s errCode=%d",
316 join(",",@databases),
320 logf(LOG_LOG,"Database(s) selected: %s",join(",",@databases));
323 return (keys(%{$self->{databases}}));
326 # -----------------------------------------------------------------------------
328 # -----------------------------------------------------------------------------
331 return(IDZebra::errCode($self->{zh}));
336 return(IDZebra::errString($self->{zh}));
341 return(IDZebra::errAdd($self->{zh}));
344 # -----------------------------------------------------------------------------
346 # -----------------------------------------------------------------------------
349 $m = TRANS_RW unless (defined ($m));
350 if (my $err = IDZebra::begin_trans($self->{zh},$m)) {
351 if ($self->errCode == 2) {
352 croak ("TRANS_RW not allowed within TRANS_RO");
354 croak("Error starting transaction; code:".
355 $self->errCode . " message: " . $self->errString);
363 my $stat = IDZebra::ZebraTransactionStatus->new();
364 IDZebra::end_trans($self->{zh}, $stat);
369 my ($self, $value) = @_;
372 $value = 0 unless (defined($value));
373 my $r =IDZebra::set_shadow_enable($self->{zh},$value);
375 return (IDZebra::get_shadow_enable($self->{zh}));
382 return(IDZebra::commit($self->{zh}));
386 # -----------------------------------------------------------------------------
387 # We don't really need that...
388 # -----------------------------------------------------------------------------
390 my ($self, $name) = @_;
391 if ($name !~/^(input|output)$/) {
392 croak("Undefined ODR '$name'");
394 IDZebra::odr_reset($self->{"odr_$name"});
397 # -----------------------------------------------------------------------------
399 # -----------------------------------------------------------------------------
403 return(IDZebra::init($self->{zh}));
409 return(IDZebra::compact($self->{zh}));
413 my ($self, %args) = @_;
415 my $rg = $self->_update_args(%args);
416 $self->_selectRecordGroup($rg);
418 IDZebra::repository_update($self->{zh});
419 $self->_selectRecordGroup($self->{rg});
424 my ($self, %args) = @_;
426 my $rg = $self->_update_args(%args);
427 $self->_selectRecordGroup($rg);
429 IDZebra::repository_delete($self->{zh});
430 $self->_selectRecordGroup($self->{rg});
435 my ($self, %args) = @_;
437 my $rg = $self->_update_args(%args);
438 $self->_selectRecordGroup($rg);
440 IDZebra::repository_show($self->{zh});
441 $self->_selectRecordGroup($self->{rg});
446 my ($self, %args) = @_;
447 my $rg = $self->_makeRecordGroup(%args);
448 $self->_selectRecordGroup($rg);
452 # -----------------------------------------------------------------------------
454 # -----------------------------------------------------------------------------
456 my ($self, %args) = @_;
458 my @args = $self->_record_update_args(%args);
459 my $stat = IDZebra::insert_record($self->{zh}, @args);
460 my $sysno = $args[2]; $stat = -1 * $stat if ($stat > 0);
461 return $stat ? $stat : $$sysno;
462 if ($stat) { return ($stat); } else { return $sysno};
466 my ($self, %args) = @_;
468 my @args = $self->_record_update_args(%args);
469 my $stat = IDZebra::update_record($self->{zh}, @args);
470 my $sysno = $args[2]; $stat = -1 * $stat if ($stat > 0);
471 return $stat ? $stat : $$sysno;
472 if ($stat) { return ($stat); } else { return $$sysno};
476 my ($self, %args) = @_;
478 my @args = $self->_record_update_args(%args);
479 my $stat = IDZebra::delete_record($self->{zh}, @args);
480 my $sysno = $args[2]; $stat = -1 * $stat if ($stat > 0);
481 return $stat ? $stat : $$sysno;
484 sub _record_update_args {
485 my ($self, %args) = @_;
487 my $sysno = $args{sysno} ? $args{sysno} : 0;
488 my $match = $args{match} ? $args{match} : "";
489 my $rectype = $args{recordType} ? $args{recordType} : "";
490 my $fname = $args{file} ? $args{file} : "<no file>";
491 my $force = $args{force} ? $args{force} : 0;
498 elsif ($args{file}) {
499 CORE::open (F, $args{file}) || warn ("Cannot open $args{file}");
500 $buff = join('',(<F>));
503 my $len = length($buff);
505 delete ($args{sysno});
506 delete ($args{match});
507 delete ($args{recordType});
508 delete ($args{file});
509 delete ($args{data});
510 delete ($args{force});
512 my $rg = $self->_makeRecordGroup(%args);
514 # If no record type is given, then try to find it out from the
517 if (my ($ext) = $fname =~ /\.(\w+)$/) {
518 my $rg2 = $self->_getRecordGroup($rg->{groupName},$ext);
519 $rectype = $rg2->{recordType};
523 $rg->{databaseName} = "Default" unless ($rg->{databaseName});
528 return ($rg, $rectype, \$sysno, $match, $fname, $buff, $len, $force);
531 # -----------------------------------------------------------------------------
534 my ($self,$mapfile) = @_;
536 if ($self->{cql_mapfile} ne $mapfile) {
537 unless (-f $mapfile) {
538 croak("Cannot find $mapfile");
540 if (defined ($self->{cql_ct})) {
541 IDZebra::cql_transform_close($self->{cql_ct});
543 $self->{cql_ct} = IDZebra::cql_transform_open_fname($mapfile);
544 $self->{cql_mapfile} = $mapfile;
547 return ($self->{cql_mapfile});
551 my ($self, $cqlquery) = @_;
552 unless (defined($self->{cql_ct})) {
553 croak("CQL map file is not specified yet.");
555 my $res = "\0" x 2048;
556 my $r = IDZebra::cql2pqf($self->{cql_ct}, $cqlquery, $res, 2048);
558 # carp ("Error transforming CQL query: '$cqlquery', status:$r");
565 # -----------------------------------------------------------------------------
567 # -----------------------------------------------------------------------------
569 my ($self, %args) = @_;
573 if ($args{cqlmap}) { $self->cqlmap($args{cqlmap}); }
581 ($query, $cqlstat) = $self->cql2pqf($args{cql});
583 croak ("Failed to transform query: '$args{cql}', ".
584 "status: ($cqlstat)");
588 croak ("No query given to search");
593 if ($args{databases}) {
594 @origdbs = $self->databases;
595 $self->databases(@{$args{databases}});
599 my $rsname = $args{rsname} ? $args{rsname} : $self->_new_setname;
601 my $rs = $self->_search_pqf($query, $rsname);
603 if ($args{databases}) {
604 $self->databases(@origdbs);
609 carp("Sort skipped due to search error: ".
612 $rs->sort($args{sort});
621 return ("set_".$self->{rscount}++);
625 my ($self, $query, $setname) = @_;
630 my $res = IDZebra::search_PQF($self->{zh},
635 my $rs = IDZebra::Resultset->new($self,
638 recordCount => $hits,
639 errCode => $self->errCode,
640 errString => $self->errString);
644 # -----------------------------------------------------------------------------
647 # Sorting of multiple result sets is not supported by zebra...
648 # -----------------------------------------------------------------------------
651 my ($self, $sortspec, $setname, @sets) = @_;
656 croak ("Sorting/merging of multiple resultsets is not supported now");
661 foreach my $rs (@sets) {
662 push (@setnames, $rs->{name});
663 $count += $rs->{recordCount}; # is this really sure ??? It doesn't
667 my $status = IDZebra::sort($self->{zh},
673 my $errCode = $self->errCode;
674 my $errString = $self->errString;
676 logf (LOG_LOG, "Sort status $setname: %d, errCode: %d, errString: %s",
677 $status, $errCode, $errString);
679 if ($status || $errCode) {$count = 0;}
681 my $rs = IDZebra::Resultset->new($self,
683 recordCount => $count,
685 errString => $errString);
689 # -----------------------------------------------------------------------------
691 # -----------------------------------------------------------------------------
693 my ($self, %args) = @_;
697 unless ($args{expression}) {
698 croak ("No scan expression given");
701 my $sl = IDZebra::ScanList->new($self,%args);
706 # ============================================================================
712 IDZebra::Session - A Zebra database server session for update and retrieval
716 $sess = IDZebra::Session->new(configFile => 'demo/zebra.cfg');
719 $sess = IDZebra::Session->open(configFile => 'demo/zebra.cfg',
720 groupName => 'demo1');
722 $sess->group(groupName => 'demo2');
728 $sess->update(path => 'lib');
730 my $s1=$sess->update_record(data => $rec1,
731 recordType => 'grs.perl.pod',
732 groupName => "demo1",
735 my $stat = $sess->end_trans;
737 $sess->databases('demo1','demo2');
739 my $rs1 = $sess->search(cqlmap => 'demo/cql.map',
740 cql => 'dc.title=IDZebra',
741 databases => [qw(demo1 demo2)]);
746 Zebra is a high-performance, general-purpose structured text indexing and retrieval engine. It reads structured records in a variety of input formats (eg. email, XML, MARC) and allows access to them through exact boolean search expressions and relevance-ranked free-text queries.
748 Zebra supports large databases (more than ten gigabytes of data, tens of millions of records). It supports incremental, safe database updates on live systems. You can access data stored in Zebra using a variety of Index Data tools (eg. YAZ and PHP/YAZ) as well as commercial and freeware Z39.50 clients and toolkits.
750 =head1 OPENING AND CLOSING A ZEBRA SESSIONS
752 For the time beeing only local database services are supported, the same way as calling zebraidx or zebrasrv from the command shell. In order to open a local Zebra database, with a specific configuration file, use
754 $sess = IDZebra::Session->new(configFile => 'demo/zebra.cfg');
759 $sess = IDZebra::Session->open(configFile => 'demo/zebra.cfg');
761 where $sess is going to be the object representing a Zebra Session. Whenever this variable gets out of scope, the session is closed, together with all active transactions, etc... Anyway, if you'd like to close the session, just say:
766 - close all transactions
767 - destroy all result sets and scan lists
770 Note, that if I<shadow registers> are enabled, the changes will not be committed automatically.
772 In the future different database access methods are going to be available,
775 $sess = IDZebra::Session->open(server => 'ostrich.technomat.hu:9999');
777 You can also use the B<record group> arguments described below directly when calling the constructor, or the open method:
779 $sess = IDZebra::Session->open(configFile => 'demo/zebra.cfg',
780 groupName => 'demo');
785 If you manage different sets of records that share common characteristics, you can organize the configuration settings for each type into "groups". See the Zebra manual on the configuration file (zebra.cfg).
787 For each open session a default record group is assigned. You can configure it in the constructor, or by the B<group> method:
789 $sess->group(groupName => ..., ...)
791 The following options are available:
797 This will select the named record group, and load the corresponding settings from the configuration file. All subsequent values will overwrite those...
799 =item B<databaseName>
801 The name of the (logical) database the updated records will belong to.
805 This path is used for directory updates (B<update>, B<delete> methods);
809 This option determines how to identify your records. See I<Zebra manual: Locating Records>
813 The record type used for indexing.
815 =item B<flagStoreData>
817 Specifies whether the records should be stored internally in the Zebra system files. If you want to maintain the raw records yourself, this option should be false (0). If you want Zebra to take care of the records for you, it should be true(1).
819 =item B<flagStoreKeys>
821 Specifies whether key information should be saved for a given group of records. If you plan to update/delete this type of records later this should be specified as 1; otherwise it should be 0 (default), to save register space.
827 =item B<fileVerboseLimit>
829 Skip log messages, when doing a directory update, and the specified number of files are processed...
831 =item B<databaseNamePath>
835 =item B<explainDatabase>
837 The name of the explain database to be used
841 Follow links when doing directory update.
845 You can use the same parameters calling all update methods.
847 =head1 TRANSACTIONS (READ / WRITE LOCKS)
849 A transaction is a block of record update (insert / modify / delete) or retrieval procedures. So, all call to such function will implicitly start a transaction, unless one is already started by
855 $sess->begin_trans(TRANS_RW)
857 (these two are equivalents). The effect of this call is a kind of lock: if you call is a write lock is put on the registers, so other processes trying to update the database will be blocked. If there is already an RW (Read-Write) transaction opened by another process, the I<begin_trans> call will be blocked.
861 $sess->begin_trans(TRANS_RO),
863 if you would like to put on a "read lock". This one is B<deprecated>, as while you have explicitly opened a transaction for read, you can't open another one for update. For example:
865 $sess->begin_trans(TRANS_RO);
866 $sess->begin_tran(TRANS_RW); # invalid, die here
872 $sess->begin_tran(TRANS_RW);
873 $sess->begin_trans(TRANS_RO);
877 is valid, but probably useless. Note again, that for each retrieval call, an RO transaction is opened. I<TRANS_RW> and I<TRANS_RO> are exported by default by IDZebra::Session.pm.
879 For multiple per-record I<updates> it's efficient to start transactions explicitly: otherwise registers (system files, vocabularies, etc..) are updated one by one. After finishing all requested updates, use
881 $stat = $sess->end_trans;
883 The return value is a ZebraTransactionStatus object, containing the following members as a hash reference:
885 $stat->{processed} # Number of records processed
886 $stat->{updated} # Number of records processed
887 $stat->{deleted} # Number of records processed
888 $stat->{inserted} # Number of records processed
889 $stat->{stime} # System time used
890 $stat->{utime} # User time used
892 Normally, if the perl code dies due to some runtime error, or the session is closed, then the API attempts to close all pending transactions.
894 =head1 THE SHADOW REGISTERS
896 The Zebra server supports updating of the index structures. That is, you can add, modify, or remove records from databases managed by Zebra without rebuilding the entire index. Since this process involves modifying structured files with various references between blocks of data in the files, the update process is inherently sensitive to system crashes, or to process interruptions: Anything but a successfully completed update process will leave the register files in an unknown state, and you will essentially have no recourse but to re-index everything, or to restore the register files from a backup medium. Further, while the update process is active, users cannot be allowed to access the system, as the contents of the register files may change unpredictably.
898 You can solve these problems by enabling the shadow register system in Zebra. During the updating procedure, zebraidx will temporarily write changes to the involved files in a set of "shadow files", without modifying the files that are accessed by the active server processes. If the update procedure is interrupted by a system crash or a signal, you simply repeat the procedure - the register files have not been changed or damaged, and the partially written shadow files are automatically deleted before the new updating procedure commences.
900 At the end of the updating procedure (or in a separate operation, if you so desire), the system enters a "commit mode". First, any active server processes are forced to access those blocks that have been changed from the shadow files rather than from the main register files; the unmodified blocks are still accessed at their normal location (the shadow files are not a complete copy of the register files - they only contain those parts that have actually been modified). If the commit process is interrupted at any point during the commit process, the server processes will continue to access the shadow files until you can repeat the commit procedure and complete the writing of data to the main register files. You can perform multiple update operations to the registers before you commit the changes to the system files, or you can execute the commit operation at the end of each update operation. When the commit phase has completed successfully, any running server processes are instructed to switch their operations to the new, operational register, and the temporary shadow files are deleted.
902 By default, (in the API !) the use of shadow registers is disabled. If zebra is configured that way (there is a "shadow" entry in zebra.cfg), then the shadow system can be enabled by calling:
910 If shadow system is enabled, then you have to commit changes you did, by calling:
914 Note, that you can also determine shadow usage in the session constructor:
916 $sess = IDZebra::Session->open(configFile => 'demo/zebra.cfg',
919 Changes to I<shadow> will not have effect, within a I<transaction> (ie.: a transaction is started either with shadow enabled or disabled). For more details, read Zebra documentation: I<Safe Updating - Using Shadow Registers>.
923 There are two ways to update data in a Zebra database using the perl API. You can update an entire directory structure just the way it's done by zebraidx:
925 $sess->update(path => 'lib');
927 This will update the database with the files in directory "lib", according to the current record group settings.
931 This will update the database with the files, specified by the default record group setting. I<path> has to be specified there...
933 $sess->update(groupName => 'demo1',
936 Update the database with files in "lib" according to the settings of group "demo1"
938 $sess->delete(groupName => 'demo1',
941 Delete the records derived from the files in directory "lib", according to the "demo1" group settings. Sounds complex? Read zebra documentation about identifying records.
943 You can also update records one by one, even directly from the memory:
945 $sysno = $sess->update_record(data => $rec1,
946 recordType => 'grs.perl.pod',
947 groupName => "demo1");
949 This will update the database with the given record buffer. Note, that in this case recordType is explicitly specified, as there is no filename given, and for the demo1 group, no default record type is specified. The return value is the system assigned id of the record.
951 You can also index a single file:
953 $sysno = $sess->update_record(file => "lib/IDZebra/Data1.pm");
955 Or, provide a buffer, and a filename (where filename will only be used to identify the record, if configured that way, and possibly to find out it's record type):
957 $sysno = $sess->update_record(data => $rec1,
958 file => "lib/IDZebra/Data1.pm");
960 And some crazy stuff:
962 $sysno = $sess->delete_record(sysno => $sysno);
964 where sysno in itself is sufficient to identify the record
966 $sysno = $sess->delete_record(data => $rec1,
967 recordType => 'grs.perl.pod',
968 groupName => "demo1");
970 This case the record is extracted, and if already exists, located in the database, then deleted...
972 $sysno = $sess->update_record(data => $rec1,
974 recordType => 'grs.perl.pod',
975 groupName => "demo1");
977 Don't try this at home! This case, the record identifier string (which is normally generated according to the rules set in I<recordId> member of the record group, or in the I<recordId> parameter) is provided directly.... Looks much better this way:
979 $sysno = $sess->update_record(data => $rec1,
980 databaseName => 'books',
981 recordId => '(bib1,ISBN)',
982 recordType => 'grs.perl.pod',
986 You can notice, that it's not necessary to define a record group in zebra.cfg: you can do it "on the fly" in your code.
988 B<Important:> Note, that one record can be updated only once within a transaction - all subsequent updates are skipped. If you'd like to override this feature, use the I<force=E<gt>1> flag:
990 $sysno = $sess->update_record(data => $rec1,
991 recordType => 'grs.perl.pod',
992 groupName => "demo1",
995 If you don't like to update the record, if it alerady exists, use the I<insert_record> method:
997 $sysno = $sess->insert_record(data => $rec1,
998 recordType => 'grs.perl.pod',
999 groupName => "demo1");
1001 In this case, sysno will be -1, if the record could not be added, because there was already one in the database, with the same record identifier (generated according to the I<recordId> setting).
1003 =head1 DATABASE SELECTION
1005 Within a zebra repository you can define logical databases. You can either do this by record groups, or by providing the databaseName argument for update methods. For each record the database name it belongs to is stored.
1007 For searching, you can select databases by calling:
1009 $sess->databases('db1','db2');
1011 This will not do anything if the given and only the given databases are already selected. You can get the list of the actually selected databases, by calling:
1013 @dblist = $sess->databases();
1017 It's nice to be able to store data in your repository... But it's useful to reach it as well. So this is how to do searching:
1019 $rs = $sess->search(databases => [qw(demo1,demo2)], # optional
1020 pqf => '@attr 1=4 computer');
1022 This is going to execute a search in databases demo1 and demo2, for title 'com,puter'. This is a PQF (Prefix Query Format) search, see YAZ documentation for details. The database selection is optional: if it's provided, the given list of databases is selected for this particular search, then the original selection is restored.
1024 =head2 CCL searching
1026 Not all users enjoy typing in prefix query structures and numerical attribute values, even in a minimalistic test client. In the library world, the more intuitive Common Command Language (or ISO 8777) has enjoyed some popularity - especially before the widespread availability of graphical interfaces. It is still useful in applications where you for some reason or other need to provide a symbolic language for expressing boolean query structures.
1028 The CCL searching is not currently supported by this API.
1030 =head2 CQL searching
1032 CQL - Common Query Language - was defined for the SRW protocol. In many ways CQL has a similar syntax to CCL. The objective of CQL is different. Where CCL aims to be an end-user language, CQL is the protocol query language for SRW.
1034 In order to map CQL queries to Zebra internal search structures, you have to define a mapping, the way it is described in YAZ documentation: I<Specification of CQL to RPN mapping>. The mapping is interpreted by the method:
1036 $sess->cqlmap($mapfile);
1038 Or, you can directly provide the I<mapfile> parameter for the search:
1040 $rs = $sess->search(cqlmap => 'demo/cql.map',
1041 cql => 'dc.title=IDZebra');
1043 As you see, CQL searching is so simple: just give the query in the I<cql> parameter.
1047 If you'd like the search results to be sorted, use the I<sort> parameter:
1049 $rs = $sess->search(cql => 'IDZebra',
1052 Note, that B<currently> this is (almost) equivalent to
1054 $rs = $sess->search(cql => 'IDZebra');
1055 $rs->sort('1=4 ia');
1057 but in the further versions of Zebra and this API a single phase search and sort will take place, optimizing performance. For more details on sorting, see I<IDZebra::ResultSet> manpage.
1061 As you have seen, the result of the search request is a I<Resultset> object.
1062 It contains number of hits, and search status, and can be used to sort and retrieve the resulting records.
1064 $count = $rs->count;
1066 printf ("RS Status is %d (%s)\n", $rs->errCode, $rs->errString);
1068 I<$rs-E<gt>errCode> is 0, if there were no errors during search. Read the I<IDZebra::Resultset> manpage for more details.
1072 Zebra supports scanning index values. The result of the
1074 $sl = $sess->scan(expression => "a");
1076 call is an I<IDZebra::ScanList> object, what you can use to list the values. The scan expression has to be provided in a PQF like format. Examples:
1078 B< a> (scan trough words of "default", "Any" indexes)
1081 B< @attr 1=1016 a> (same effect)
1084 B< @attr 1=4 @attr 6=2 a> (scan trough titles as phrases)
1086 An illegal scan expression will cause your code to die. If you'd like to select databases just for the scan call, you can optionally use the I<databases> parameter:
1088 $sl = $sess->scan(expression => "a",
1089 databases => [qw(demo1 demo2)]);
1091 You can use the I<IDZebra::ScanList> object returned by the i<scan> method, to reach the result. Check I<IDZebra::ScanList> manpage for more details.
1093 =head1 SESSION STATUS AND ERRORS
1095 Most of the API calls causes die, if an error occures. You avoid this, by using eval {} blocks. The following methods are available to get the status of Zebra service:
1101 The Zebra provided error code... (for the result of the last call);
1105 Error string corresponding to the message
1109 Additional information for the status
1113 This functionality may change, see TODO.
1115 =head1 LOGGING AND MISC. FUNCTIONS
1117 Zebra provides logging facility for the internal events, and also for application developers trough the API. See manpage I<IDZebra::Logger> for details.
1121 =item B<IDZebra::LogFile($filename)>
1123 Will set the output file for logging. By default it's STDERR;
1125 =item B<IDZebra::LogLevel(15)>
1127 Set log level. 0 for no logs. See IDZebra::Logger for usable flags.
1131 Some other functions
1135 =item B<$sess-E<gt>init>
1137 Initialize, and clean registers. This will remove all data!
1139 =item B<$sess-E<gt>compact>
1141 Compact the registers (? does this work)
1143 =item B<$sess-E<gt>show>
1145 Doesn't have too much meaning. Don't try :)
1153 =item B<Clean up error handling>
1155 By default all zebra errors should cause die. (such situations could be avoided by using eval {}), and then check for errCode, errString... An optional flag or package variable should be introduced to override this, and skip zebra errors, to let the user decide what to do.
1157 =item B<Make the package self-distributable>
1159 Build and link with installed header and library files
1163 Test shadow system, unicode...
1167 Cleanup, arrange, remove redundancy
1177 Peter Popovics, pop@technomat.hu
1181 Zebra documentation, Zebra::ResultSet, Zebra::ScanList, Zebra::Logger manpages