1 package WebPAC::Output::Estraier;
6 use base qw/WebPAC::Common/;
16 WebPAC::Output::Estraier - Create Hyper Estraier full text index
24 our $VERSION = '0.05';
28 Create full text index using Hyper Estraier index from data with
35 Connect to Hyper Estraier index using HTTP
37 my $est = new WebPAC::Output::Estraier(
38 masterurl => 'http://localhost:1978/',
42 encoding => 'iso-8859-2',
51 URI to C<estmaster> node
55 C<estmaster> user with sufficient rights
63 name of database from which data comes
67 character encoding of C<data_structure> if it's differenet than C<ISO-8859-2>
68 (and it probably is). This encoding will be converted to C<UTF-8> for
73 Name of database will be used to form URI of documents in index.
82 my $log = $self->_get_logger;
84 #$log->debug("self: ", sub { Dumper($self) });
86 foreach my $p (qw/masterurl user passwd database/) {
87 $log->logdie("need $p") unless ($self->{$p});
90 my $url = $self->{masterurl} . '/node/' . $self->{database};
93 $log->info("opening Hyper Estraier index $self->{url}");
95 my $nodes = $self->master( action => 'nodelist' );
97 $log->debug("nodes found: $nodes");
99 if ($nodes !~ m/^$self->{database}\t/sm) {
100 $log->info("creating index $url");
103 name => $self->{database},
104 label => "WebPAC $self->{database}",
105 ) || $log->logdie("can't create Hyper Estraier node $self->{database}");
108 $self->{'db'} = HyperEstraier::Node->new($self->{url});
109 $self->{'db'}->set_auth($self->{'user'}, $self->{passwd});
111 my $encoding = $self->{'encoding'} || 'ISO-8859-2';
112 $log->info("using encoding $encoding");
114 $self->{'iconv'} = new Text::Iconv($encoding, 'UTF-8') or
115 $log->logdie("can't create conversion from $encoding to UTF-8");
117 $self ? return $self : return undef;
123 Adds one entry to database.
129 text => 'optional text from which snippet is created',
132 This function will create entries in index using following URI format:
134 C<file:///type/database%20name/000>
136 Each tag in C<data_structure> with specified C<type> will create one
137 attribute and corresponding hidden text (used for search).
146 my $log = $self->_get_logger;
148 my $database = $self->{'database'} || $log->logconfess('no database in $self');
149 $log->logconfess('need db in object') unless ($self->{'db'});
151 foreach my $p (qw/id ds type/) {
152 $log->logdie("need $p") unless ($args->{$p});
155 my $type = $args->{'type'};
156 my $id = $args->{'id'};
158 my $uri = "file:///$type/$database/$id";
159 $log->debug("creating $uri");
161 my $doc = HyperEstraier::Document->new;
162 $doc->add_attr('@uri', $self->{'iconv'}->convert($uri) );
164 $log->debug("ds = ", sub { Dumper($args->{'ds'}) } );
166 # filter all tags which have type defined
168 ref($args->{'ds'}->{$_}) eq 'HASH' && defined( $args->{'ds'}->{$_}->{$type} )
169 } keys %{ $args->{'ds'} };
171 $log->debug("tags = ", join(",", @tags));
173 return unless (@tags);
175 foreach my $tag (@tags) {
177 my $vals = join(" ", @{ $args->{'ds'}->{$tag}->{$type} });
179 $log->logconfess("no values for $tag/$type") unless ($vals);
181 $vals = $self->{'iconv'}->convert( $vals ) or
182 $log->logdie("can't convert '$vals' to UTF-8");
184 $doc->add_attr( $tag, $vals );
185 $doc->add_hidden_text( $vals );
188 my $text = $args->{'text'};
190 $text = $self->{'iconv'}->convert( $text ) or
191 $log->logdie("can't convert '$text' to UTF-8");
192 $doc->add_text( $text );
195 $log->debug("adding ", sub { $doc->dump_draft } );
196 $self->{'db'}->put_doc($doc) || $log->logdie("can't add document $uri to node " . $self->{url} . " status: " . $self->{db}->status());
202 # REST parametars validation data
205 my $estraier_rest = {
207 userdel => [ qw/name/ ],
209 nodeadd => [ qw/name label/ ],
210 nodedel => [ qw/name/ ],
213 _set_link => [ qw/url label credit/ ],
219 Issue administrative commands to C<estmaster> process and receive response
222 my $nodelist = $est->master( action => 'nodelist' );
230 my $log = $self->_get_logger;
232 my $action = $args->{action} || $log->logconfess("no action specified");
234 $log->logdie("action '$action' isn't supported") unless ($estraier_rest->{master}->{$action});
236 $log->debug("master action: $action");
238 return $self->estcall(
239 validate => 'master',
240 rest_url => $self->{masterurl} . '/master?action=' . $action ,
259 my $log = $self->_get_logger;
261 my @labels = $self->master( action => 'nodelist' );
263 $log->debug("got labels: ", join("|", @labels));
265 @labels = grep(/^$args->{to}/, @labels);
267 my (undef,$label) = split(/\t/, shift @labels);
270 $log->warn("can't find label for $args->{to}, skipping link creaton");
274 $log->debug("using label $label for $args->{to}");
276 return $self->estcall(
278 action => '_set_link',
279 rest_url => $self->{masterurl} . '/node/' . $args->{from} . '/_set_link' ,
280 url => $self->{masterurl} . '/node/' . $args->{to},
282 credit => $args->{credit},
288 Workhourse which does actual calls to Hyper Estraier
291 rest_url => '/master?action=' . $action,
292 validate => 'master',
296 C<rest_url> is relative URL to C<estmaster> and C<validate> is entry into
297 internal hash which will check if all parametars are available before
305 my $log = $self->_get_logger;
307 $log->debug("estcall: ",Dumper($args));
309 foreach my $p (qw/rest_url validate action/) {
310 $log->die("ectcall needs $p parametar") unless ($args->{$p});
313 my $url = $args->{rest_url};
315 $del = '&' if ($url =~ m#\?#);
319 foreach my $arg (@{ $estraier_rest->{ $args->{validate} }->{ $args->{action} } }) {
320 $log->logdie("missing parametar $arg for action $args->{action}") unless ($args->{$arg});
321 $url_args .= $del . $arg . '=' . uri_escape( $args->{$arg} );
325 $url_args =~ s#^\&#?# if ($url =~ m#\?#);
328 $log->debug("calling $url");
330 my $res = $self->est_ua()->get($url);
332 if ($res->is_success) {
333 #$log->debug( $res->content );
334 return split(/\n/, $res->content) if wantarray;
335 return $res->content || 0E0;
337 $log->warn("unable to call $url: " . $res->status_line);
345 This is helper function to create C<LWP::UserAgent> object with Super User
348 my $ua = $self->est_ua( user => 'admin', passwd => 'admin' );
357 return $self->{_master_ua} if ($self->{_master_ua});
360 package AdminUserAgent;
361 use base qw/LWP::UserAgent/;
363 my $self = LWP::UserAgent::new(@_);
364 $self->agent("webpac/$VERSION");
367 sub get_basic_credentials {
368 my($self, $realm, $uri) = @_;
369 return ($self->{user}, $self->{passwd});
371 sub set_basic_credentials {
372 my ($self, $user, $passwd) = @_;
373 $self->{user} = $user;
374 $self->{passwd} = $passwd;
378 $self->{_master_ua} = AdminUserAgent->new( ) || sub {
379 my $log = $self->_get_logger;
380 $log->logdie("can't create LWP::UserAgent: $!");
383 $self->{_master_ua}->set_basic_credentials($self->{user}, $self->{passwd});
385 return $self->{_master_ua};
390 Dobrica Pavlinusic, C<< <dpavlin@rot13.org> >>
392 =head1 COPYRIGHT & LICENSE
394 Copyright 2005 Dobrica Pavlinusic, All Rights Reserved.
396 This program is free software; you can redistribute it and/or modify it
397 under the same terms as Perl itself.
401 1; # End of WebPAC::Output::Estraier